Tribology of Composite Materials [1 ed.] 9781613247723, 9781616683191

Citation preview

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved. Tribology of Composite Materials, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved. Tribology of Composite Materials, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

MATERIALS AND MANUFACTURING TECHNOLOGY

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

TRIBOLOGY OF COMPOSITE MATERIALS

No part of this digital document may be reproduced, stored in a retrieval system or transmitted in any form or by any means. The publisher has taken reasonable care in the preparation of this digital document, but makes no expressed or implied warranty of any kind and assumes no responsibility for any errors or omissions. No liability is assumed for incidental or consequential damages in connection with or arising out of information contained herein. This digital document is sold with the clear understanding that the publisher is not engaged in rendering legal, medical or any other professional services.

Tribology of Composite Materials, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

MATERIALS AND MANUFACTURING TECHNOLOGY J. PAULO DAVIM - SERIES EDITOR UNIVERSITY OF AVEIRO AVEIRO, PORTUGAL Drilling of Composite Materials J. Paulo Davim (Editor) 2009. ISBN: 978-1-60741-163-5 (Hardcover) 2009. ISBN: 978-1-60876-584-3 (e-book)

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Metal Cutting: Research Advances J. Paulo Davim (Editor) 2010. ISBN: 978-1-60876-207-1 (Hardcover) 2010. ISBN: 978-1-61122-573-0 (e-book)

Tribology Research Advances J. Paulo Davim (Editor) 2011. ISBN: 978-1-60692-885-1 (Hardcover) Medical Device Manufacturing Mark J. Jackson and J. Paulo Davim (Editors) 2011. ISBN: 978-1-61209-715-2 (Hardcover) Metal Matrix Composites J. Paulo Davim (Editor) 2011. ISBN: 978-1-61209-771-8 (Hardcover)

Artificial Intelligence in Manufacturing Research J. Paulo Davim (Editor) 2010. ISBN: 978-1-60876-214-9 (Hardcover) 2011. ISBN: 978-1-61761-564-1 (e-book) Micro and Nanomanufacturing Research J. Paulo Davim (Editor) 2010. ISBN: 978-1-61668-488-4 (Hardcover) 2012. ISBN: 978-1-61942-003-8 (Softcover) 2010. ISBN: 978-1-61324-366-4 (e-book)

Biomedical Tribology J. Paulo Davim (Editor) 2011. ISBN: 978-1-61470-056-2 (Hardcover) 2011. ISBN: 978-1-61470-153-8 (e-book) Tribology of Composite Materials J. Paulo Davim (Editor) 2010. ISBN: 978-1-61668-319-1 (Hardcover) 2012. ISBN: 978-1-62100-999-3 (Softcover) 2010. ISBN: 978-1-61324-772-3 (e-book)

Tribology of Composite Materials, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

MATERIALS AND MANUFACTURING TECHNOLOGY

TRIBOLOGY OF COMPOSITE MATERIALS

J. PAULO DAVIM

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

EDITOR

Nova Science Publishers, Inc. New York

Tribology of Composite Materials, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

Copyright © 2012 by Nova Science Publishers, Inc. All rights reserved. No part of this book may be reproduced, stored in a retrieval system or transmitted in any form or by any means: electronic, electrostatic, magnetic, tape, mechanical photocopying, recording or otherwise without the written permission of the Publisher. For permission to use material from this book please contact us: Telephone 631-231-7269; Fax 631-231-8175 Web Site: http://www.novapublishers.com NOTICE TO THE READER The Publisher has taken reasonable care in the preparation of this book, but makes no expressed or implied warranty of any kind and assumes no responsibility for any errors or omissions. No liability is assumed for incidental or consequential damages in connection with or arising out of information contained in this book. The Publisher shall not be liable for any special, consequential, or exemplary damages resulting, in whole or in part, from the readers’ use of, or reliance upon, this material.

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Independent verification should be sought for any data, advice or recommendations contained in this book. In addition, no responsibility is assumed by the publisher for any injury and/or damage to persons or property arising from any methods, products, instructions, ideas or otherwise contained in this publication. This publication is designed to provide accurate and authoritative information with regard to the subject matter covered herein. It is sold with the clear understanding that the Publisher is not engaged in rendering legal or any other professional services. If legal or any other expert assistance is required, the services of a competent person should be sought. FROM A DECLARATION OF PARTICIPANTS JOINTLY ADOPTED BY A COMMITTEE OF THE AMERICAN BAR ASSOCIATION AND A COMMITTEE OF PUBLISHERS. Additional color graphics may be available in the e-book version of this book. LIBRARY OF CONGRESS CATALOGING-IN-PUBLICATION DATA Tribology of composite materials / editor, J. Paulo Davim. p. cm. Includes bibliographical references and index. ISBN:  (eBook) 1. Composite materials--Mechanical properties. 2. Tribology. I. Davim, J. Paulo. TA418.9.C6T728 2010 620.1'1892--dc22 2010015634

Published by Nova Science Publishers, Inc. † New York Tribology of Composite Materials, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

CONTENTS   vii 

Preface Chapter 1

1 

Chapter 2

Tribology of Injection Molded Thermoplastic Nanocomposites Carmine Lucignano and Fabrizio Quadrini 

35 

Chapter 3

Tribology of Composite Materials with Inorganic Lubricants Kunhong Hu, Xianguo Hu and Ralph Stengler 

55 

Chapter 4

Mechanical and Tribological Behaviors of Nanometer Al2O3 and SiO2 Reinforced PEEK Composites Guo Qiang and Pan Guoliang 

87 

Effects of Matrix Crystalline Structure and Molecular Weight on the Tribological Behavior of PEEK-Based Materials G. Zhang and M. Schehl 

123 

Chapter 5 Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Tribology of Functionalised Carbon Nanofillers Based Polymer Composites M. Fahim and J. Paulo Davim 

Chapter 6

Friction and Wear of Al2O3-Ni Composite Jinjun Lu, Junhu Meng, Bin Liu, Jingbo Wang and Shengrong Yang 

Chapter 7

Modeling and Analysis on Wear Behaviour of Metal Matrix Composites K. Palanikumar, T. Rajasekaran and J. Paulo Davim 

Chapter 8

Tribology of Glass-Ceramic Bonded Composite Materials M. J. Jackson 

Index

Tribology of Composite Materials, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

143 

157  175  207 

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved. Tribology of Composite Materials, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

PREFACE Recently, the use of composite materials has increased in various areas of science and technology due to their special properties, with applications in biomedical, aircraft, automotive, defence and aerospace, as well other advanced industries. Tribology is defined as the “science and technology of interacting surfaces in relative motion”. It includes the research and application of principles of friction, wear and lubrication. Recently, nanocomposites had been gaining ground. Frictional interactions in micro and nanocomponents are becoming increasingly important for the development of new products in several industries. This book aims to provide the research and review studies on tribology of composite materials. The first chapter provide information on tribology of functionalised carbon nanofillers based polymer composites. Chapter 2 is focused on tribology of injection molded thermoplastic nanocomposites. Chapter 3 discuss tribology of composite materials with inorganic lubricants. Chapter 4 is focused on mechanical and tribological behaviors of nanometer Al2O3 and SiO2 reinforced PEEK composites. Subsequently, the chapter 5 deal with the effects of matrix crystalline structure and molecular weight on the tribological behaviour of PEEK-based materials. Chapter 6 is focused on friction and wear of Al2O3-Ni composite. Chapter 7 discuss modelling and analysis on wear behaviour of metal matrix composites. Finally, the last chapter of this research book is focused on some aspects of tribology of glass-ceramic bonded composite materials. The present research book can be used for final undergraduate engineering course (for example, materials, mechanical, physics, etc) or as a subject on tribology of composite materials at the postgraduate level. Also, this book can serve as a useful reference for academics, researchers, materials, mechanical and physics engineers, professional in related industries with composite materials. The Editor acknowledges gratitude to Nova Publishers for this opportunity and for their professional support. Finally, I would like to thank all the chapter authors for their availability for this work.

Tribology of Composite Materials, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

J. Paulo Davim Aveiro, Portugal November 2009

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved. Tribology of Composite Materials, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

In: Tribology of Composite Materials Editor: J. Paulo Davim

ISBN 978-1-62100-999-3 © 2012 Nova Science Publishers, Inc.

Chapter 1

TRIBOLOGY OF FUNCTIONALISED CARBON NANOFILLERS BASED POLYMER COMPOSITES M. Fahim1∗ and J. Paulo Davim2 1

Department of Physics, Zakir Husain College, University of Delhi, Delhi, India 2 Department of Mechanical Engineering, University of Aveiro, Campus Santiago, Portugal

ABSTRACT Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

The multidisciplinary science of interacting surfaces in relative motion is now identified distinctly as tribology (tribos meaning rubbing). It has grown immensely over the years and evolved to an extent that studies on surface related aspects such as generation and transmission of force, dissipation of energy (friction), dissipation of mass (wear) and lubrication between two bodies in relative motion have now narrowed down to the fundamental understanding of genesis of friction at atomic level. The subject has received greater attention in view of the ever growing demands to design low friction, low wear materials that are lightweight; corrosion, fatigue and radiation resistant; chemically inert, possess high load bearing capacity and can serve in severe operating conditions of speed, pressure, temperature and erosive-corrosive environment. Such socalled tribo-efficient materials can enhance components’ life, reduce fuel and power consumption in a machine and cause minimum damage to environment that is the driving force of current global research. A steady but fast pace progress in high performance selflubricating polymer composites has given a much needed impetus in this area of research. These polymer composites developed using a judicious choice of polymer matrices (thermosets, thermoplastics), fibrous/fabric reinforcement (synthetic; glass, carbon, aramid, as well as natural plant fibers sisal, hemp, flax) and a range of performance enhancing fillers (solid lubricants, electrically conductive, heat dissipative) offer tremendous opportunities to tailor and design specific properties based on stronger filler∗

Corresponding author: Department of Physics, Zakir Husain College, University of Delhi, Delhi, INDIA 110 007 E-mail: [email protected].

Tribology of Composite Materials, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

2

M. Fahim and J. Paulo Davim

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

matrix interface in tribo-materials. Polymer nanocomposites in which nanosized fillers interact at the level of inter-molecular spacing in a polymer chain have opened up new vistas of understanding the genesis of friction. Carbonaceous fillers (carbon fibers, fabric, graphite, nanotubes, fullerenes) have immense tribo-potential to enhance mechanical performance, reduce friction and wear and transfer as a lubricating film during operation. More recently high aspect ratio (~103) carbon nanotubes (CNTs) and their frictional anisotropy has established them as a multifunctional filler for polymer composites in which large reduction in wear and friction has been achieved through loading as small as 1 wt%. Functionalization (grafting of specific polar functional groups chemically onto the filler surface) of these carbon nanotubes by acidic treatment is an area in which substantial systematic results have started being reported. However, achieving uniform dispersion and stronger interfacial adhesion between CNTs and the matrix polymers to transfer the superior properties of the former to the latter is still a bigger challenge more so because nanofillers offer a large interfacial surface area per unit volume of filler. Several strategies have been reported such as ultrasound mixing, high energy melt mixing, twin screw extrusion, in-situ polymerization, functionalization of fillers to address such issues with various degrees of success. The pertinent questions which yet remain to be answered are the measurement of interface strength and percolation threshold needed to realise a strong interface without compromising on the mechanical integrity of composites. Conflicting data exist in the literature on the transfer of selflubricating films on counterface and positive synergism between functionalized carbon nanotubes and polymer so far as tribo-performance is concerned may be due to the wide difference in test configuration and processing techniques. Secondly, friction and wear mechanisms could not be understood completely due to lack of sufficient characterization data of nanocomposites particularly based on the high performance, heat resistant specialty polymers such as polyetherimide, polyetheretherke-tone, polyamideimide, polybismaleimides, polyphenylene sulphide, polyethersulphone; polyblends and bidirectional composites. The sole objective of this chapter is to address such critical issues and highlight the need to perform systematic quantitative studies that will guide the development and application of carbon-polymer nanocomposites as tribo-material on industrial scale.

1. INTRODUCTION TO TRIBOLOGY AND TRIBO-MATERIALS Tribology, derived from the Greek word tribos which means rubbing, is a science that deals with the interfacial phenomena between two contacting surfaces in relative motion. These phenomena which arise due to generation and transmission of forces at surfaces lead to friction and wear. Lubrication is a means to reduce both friction as well as wear of materials. Thus, in broader perspectives, tribology covers all aspects of friction, wear and lubrication. Although practiced by mankind since prehistoric time, and present in innumerable situations, it was recognized as a separate entity in the beginning of the 1960s when a committee appointed by the British Government and headed by H. Peter Jost studied the impact of friction and wear on machines. The committee in its report submitted in 1966 suggested to bring all the activities involving friction, wear and lubrication under tribology [1]. The evolution of tribology is summarized in Table 1. An entertaining history of tribology can be found elsewhere [2].

Tribology of Composite Materials, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

Tribology of Functionalised Carbon Nanofillers Based Polymer Composites

3

Table 1. Evolution of Tribology as a separate entity

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Year Leonardo da Vinci (1452-1519) Amontons (1699) Charles-Augustin de Coulomb (17361806). Euler (1750) Charles Hatchett (1760- 1820) Isaac Newton and O. Reynolds (19th century) Stribeck (beginning of the 20th century) 1966

Significance Laws of friction were introduced Laws of friction rediscovered Laws of friction were verified (friction is proportional to load and independent of the area of skidding surfaces) Verification of friction laws First reliable test on frictional wear Hydrodynamic lubrication principles

Stribeck curves; variation of coefficient of friction with bearing characteristics no. ηN/P. Committee headed by H. Peter Jost submitted report and the word Tribology (science of wear, friction, lubrication and design) introduced as separate branch of science

Figure 1. Classification of tribo-materials and their applications.

Friction is generally defined as the ratio of two forces acting, respectively, perpendicular and parallel to an interface between two bodies under relative motion or impending relative motion and is measured by a dimensionless quantity known as coefficient of friction (µ). Friction between two surfaces in relative motion causes wear (progressive loss of material from the interacting surfaces) and energy dissipation that leads eventually to catastrophic failures in machines that comprise numerous moving parts such as gears, bearings, seals, pistons, valves, cams, clutches, sliders etc. Smooth running of machine means there should be minimum friction and wear between two moving surfaces so as to prevent any dimensional

Tribology of Composite Materials, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

4

M. Fahim and J. Paulo Davim

changes even at a micron level. The control of wear and friction has been an indispensable area of research since the pre-historic times when simple machines like pulley, lever, wheel etc. were invented. Lubrication, the intervening layer between contact surfaces, is an effective means of controlling wear and reducing friction. Since friction, wear and lubrication are tribological phenomena, the subject of tribology has assumed greater significance in the current times. The most common tribological problem that has a larger bearing on human lives is the case of human joints which are subjected to lubrication and wear. Serious efforts are in progress to replace degraded human joints by self-lubricated artificial joints that have an efficiency equivalent to natural joints. Hence, the main thrust of tribology nowadays is on studying the characteristics of intervening layers between contacting bodies and the consequences of film transfer, film failure or absence of a film which results in severe friction and wear of materials.

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

1.1. Tribo-efficient Materials Metals, alloys, ceramics, polymers; composites based on them and carbon-carbon composites, all these materials have been used traditionally in tribological applications as anti-friction material (low wear, low friction) such as in gears, bearings, seals, bushes, artificial joints, sliders and friction materials (low wear, moderately high friction) as in brake pads, clutch plates, tyres etc. [3]. Friction and wear are not inherent material properties but depend largely on the operating conditions, test configuration and parameters. Tribo-efficient materials are defined as those materials which deliver desired friction and wear performance at high pressure-velocity (PV) limits and high operating temperatures. A simple collection of tribo-materials and their performance dependent applications is given in Figure 1. Tribomaterials have their own advantages and limitations and hence are chosen for specific application with great care [4]. From Figure 1 it becomes obvious that the current demand is for lightweight, wear and corrosion resistant tribo-materials which are expected to increase the life of machine components despite operating in severe operating conditions, and at the same time would be easy on fuel and power consumption. These two prime objectives drive the current research across the world.

1.1.1.

Polymeric Tribo-composites

Engineering polymers appear to be an obvious choice for developing tribo-materials because of their excellent property profile such as lightweight, wear, corrosion and radiation resistance, solvent resistance, self-lubrication, quiet operation and easy mouldability and machining [3]. However, pure polymers are seldom used in tribo-applications because of their inherent weaknesses such as poor mechanical strength, low thermal stability, low thermal conductivity, low dissipativity and high thermal expansion which limit their tribological performance at high loads, speeds and temperature. It appears to be a blessing in disguise because it provides ample opportunities to modify polymers by a judicious choice of fibrous/fabric reinforcement (synthetic: glass, carbon, aramid and natural: plant fibres), and a range of performance enhancing fillers such as solid lubricants (MoS2, PTFE, graphite), thermal and electrical conductivity enhancing fillers (such as metal particles), fire retardants and cost reducing fillers [4]. The most remarkable feature about polymer based tribo-efficient

Tribology of Composite Materials, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Tribology of Functionalised Carbon Nanofillers Based Polymer Composites

5

materials is their self-lubricity. They can be used in tribo-related situations where liquid lubrication is difficult either due to high temperatures or possibility of contamination leading to erosive-corrosive wear and most importantly in situations where hydrodynamic lubrication could not be established due to small oscillatory motion or frequent starts and stops [5]. Thus, heat resistant, high performance self-lubricated or solid-lubricated fiber reinforced polymer composites have been in greater demand. Subsequently concerted research has been done extensively over the years to explore their tribological performance as can be found in the literature [6-10]. It has been observed that polymer composites based tribo-materials offer a range of properties and a careful selection is needed for specific applications because different polymer composites/blends behave differently in different wear situations. Likewise fillers and fibrous reinforcement have not always improved the wear performance. Hence, the performance of polymer composite based tribo-material should not be predicted a priori. Friction behaviour of polymer composites is different from that of other materials because they do not obey Amonton’s and Coulomb’s basic laws of friction. Friction coefficient (µ) is a function of load and sliding speed [4]. With increasing load, µ necessarily decreases because polymers are viscoelastic plastic with low moduli and low melting points. With increase in temperature or speed, µ does not show a fixed pattern for all the polymers and it is unpredictable too. It may show a peak at typical speed followed by a decrease which is mainly due to the dependence of their viscoelastic nature on temperature. Counterface material and their roughness also influence µ significantly. Generally µ is the lowest in the range 0.1 – 0.2 for most of the polymers. Similarly wear behaviour of polymer composites depends on the type of fibre and matrix concentration, dispersion, aspect ratio (length/radius of fibre), alignment and filler-matrix adhesion. The higher the aspect ratio (A) more load will be transferred from the matrix to the fibre and more is the wear resistance following the equation [11], σf = τA + σm, where σf is the contact stress, σm is the compressive stress of the matrix in the composite loaded against counterface under a load W, τ is the tangential stress produced because of the difference in the moduli of matrix and fibre. Table 2. Historical development of carbonaceous materials [13] Natural diamond reported Graphite as a solid lubricant First carbon fibers on record Chemical vapor deposition (CVD) of carbon Patented Industrial production of pyrolytic graphite Industrial production of carbon fibers from Rayon Development of carbon fibers Discovery of the fullerene molecules Industrial production of CVD diamond Industrial production of DLC Discovery of arc-grown carbon nanotubes Large-scale production of carbon nanotubes

2700 years ago The Middle Ages 1879 1880 1950s 1950s Early 1960s 1985 1990s 1990s 1991 Late 1990s

It is however, not true that with increase in the concentration of fibers, the wear resistance increases continuously. In fact either it deteriorates or becomes constant beyond a typical optimum concentration.

Tribology of Composite Materials, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

6

M. Fahim and J. Paulo Davim

For polymer composites, wear performance or wear resistance (reciprocal of specific wear rate, ko) is sensitive to PV values, it is customary to express it along with the details of PV values [12]. While wearing under harsh conditions of either P or V, or both, lot of frictional heat is generated and because polymers are low melting solids, their strength and surface properties deteriorate drastically at a point where they fail in the tribo-test. This value is so called PV limit and highlights range of utility of a typical material for extended PV conditions. Higher the value, more useful will be the material.

2. TRADITIONAL CARBON FILLED POLYMER COMPOSITES IN TRIBOLOGY: SIGNIFICANCE AND APPLICATIONS A historical perspective of carbon-derived materials and the important years in the development of carbon technology, from a tribological point of view is given in Table 2 [13]. Carbon and graphite fibers are effective in reducing both wear and friction due to their lubricating nature and heat dissipating action. Only few high performance specialty polymers can be used in severe operating conditions of pressure, speed, temperature, environment, radiation. Hence, they show improved tribo-performance when reinforced with carbon fibers and solid lubricated by MoS2, graphite or PTFE. Such polymers are PEEK, PI, PEI, LCP, POM, PPS, PES, PAI, PBI etc. Tribo performance of these polymers and their carbon reinforced composites are discussed in the following section.

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

2.1. Carbon Fiber Polymer Composites The friction and wear behaviour of carbon filled PTFE composites for bearing applications have been found to be excellent even at elevated temperature tests. A low friction and wear factor were measured for carbon/graphite filled PTFE at 260oC [14]. The addition of carbon fibre reduced the wear rate because of the difficulty in material removal as observed through microscopy [15]. Tevruz [16] performed tribological tests on the journal bearings made up of 35% carbon filled PTFE composites arguing that component (bearings) testing reveal the performance of material in more realistic way rather than testing on pin-on-disc (POD) type machine. He reported that friction and wear were strongly influenced by the thickness and composition of the PTFE/carbon composite films depending upon the adhesion between steel and composite surfaces, the cohesive properties of the base matrix, pressure and the sliding distance. He concluded that effect of load and velocity on wear can be reduced considerably by making the bearing temperature as low as possible. Bijwe and co-workers [17] tribo-evaluated PTFE+25% CF composite sliding against mild steel disc on a fabricated machine suitable for high load, speed and temperature testing. Rotating metal counterface was designed not to dissipate heat freely by proper insulation to make the tribo-sliding possible under more harsh conditions. Under selected conditions, though friction coefficient of PTFE rose from 0.18 to 0.38 due to incorporation of CF, wear resistance remarkably increased by more than two orders of magnitude. The presence of solid lubricants such as PTFE and graphite was critical for enhancing the tribological properties of CF reinforced polyimides. Incorporation of solid lubricant enhanced

Tribology of Composite Materials, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

Tribology of Functionalised Carbon Nanofillers Based Polymer Composites

7

the tribo-behaviour of PAI significantly because of improved thick and uniform film transfer on the counterface. SEM analyses showed evidence of fatigue cracking. CF reinforcement reduced wear of PTFE by four times without influencing friction coefficient significantly [18]. In the case of PI, incorporation of PTFE as a lubricant was very much beneficial for reducing friction and wear performance. CF reinforcement without PTFE did not improve friction and wear behaviour, in fact it deteriorated it [19]. Tribology of PEEK composites shows that 30% short CF reinforced PEEK composites with a heat deflection temperature of 315oC were most suitable for use at selected high temperatures [14]. The CF reinforced PEEK exhibited the highest PV values at 204 and 260oC among all the materials including PTFE, PFA, PVDF, PES, PEI, PA, and PPS. Polyertheretherketone (PEEK) composites with 15% CF led to a minimum wear rate at elevated temperatures. At 230oC graphite powder lubricated CF reinforced PEEK showed behaviour similar to PTFE lubricated composite. However, at 260oC, the graphite lubricated composites demonstrated 40% more wear than the PTFE lubricated composite. The combination of graphite and PTFE in CF reinforced PEEK proved more beneficial in reducing wear than individual ones at RT and 260oC.

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

2.2. Carbon Fabric Reinforced Polymer Composites Fabric/textile reinforced polymer composites exhibit very good mechanical strength properties in both longitudinal as well as transverse directions. The unique advantage of fabrics as a reinforcement lies in their ability to drape or conform to curved surfaces without wrinkling. In the case of polymer matrix composites, carbon, graphite, glass, aramid fabrics, etc., are the most commonly used fabrics that have immense potential in tribo-applications in automotive and aircraft industry as well [20]. Carbon fabric not only offers maximum extent of strength enhancement, it also increases the thermal conductivity of the composite which is very important from tribo-point of view. The rapid dissipation of frictional heat produced at the asperity contacts protects the matrix from excessive degradation and helps in the retention of performance properties to a greater extent. Apart from this, incorporation of nanofillers in carbon fabric reinforced composites also influence the tribological behaviour immensely. For instance, in the case of dry sliding wear against steel counterface fillers like polyfluo-150 wax (PFW), nano-ZnO and nano-SiC contributed significantly in increasing the wear resistance of CF reinforced phenolic composites [21]. Nano-SiC particles, however, increased μ. The improvement in the wear resistance (WR) was attributed to the stronger interface bonding among the carbon fabric, polymer and the particles, and the transfer film formation on the counterface. The WR of the composites decreased at elevated temperatures. The authors attributed this effect to the degradation and decomposition of the adhesive resin at excessively elevated temperatures. For composites containing carbon fabric filled with nano SiO2, TiO2, CaCO3 via dip coating techniques in phenolic resin, it has been reported that nano CaCO3 contribute significantly in improving the wear resistance while nano SiO2 was the most effective in increasing the friction reducing ability and mechanical properties [21a]. Nano particulates enhance the bonding strength between the carbon fabric and the resin. The wear rate at elevated temperature (above 180oC) was much larger due to the degradation and decomposition of the adhesive resin.

Tribology of Composite Materials, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

8

M. Fahim and J. Paulo Davim

2.2.1. Influence of Type of Carbon Fabric Weave

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

It has been found that the weave of carbon fabric influenced both strength and tribological performance of polyetherimide composites significantly [22]. However, influence in different wear modes was different. For instance a twill weave proved to be the best choice for achieving highest possible strength properties, viz. tensile modulus, tensile strength, flexural strength, flexural modulus and interlaminar shear strength, but not toughness. Operating load affected wear performance substantially. In adhesive [22] and low amplitude oscillating wear modes [23], specific wear rate increased with load while in the case of abrasive wear mode [23a] it decreased.

Adhesive Wear Mode In the case of fabric reinforced composites, wear depends mainly on severe fiber damage by various processes such as micro-cracking, micro-cutting and pulverization resulting in generation of debris and removal of debris from the surface resulting in “positive” wear [22]. The removal of debris depends on how easily it can be peeled off or pulled out from the matrix, which in turn depends on the fiber–matrix bonding, which again depends on the operating parameters. Severe operating conditions lead to deterioration in the fiber–matrix interface leading to more damage to fibers and removal of debris as a vicious cycle. The type of weave contributes to these major mechanisms in various ways as follows. First two contribute in negative way leading to higher wear while the last two act for wear protection. Net wear is a resultant of these mechanisms. Other factors that influence wear are rigidity of weave, crimp, debris retention and amount of resin transfer on the disc as a result of weave structure. If resin transfer is high, fibers will be less supported and this would lead to more fiber damage and less retention of debris. Among these four factors, wear rate would be the minimum if weave is least rigid with least number of crimp points, maximum number of tight pockets to trap debris and least tendency for resin transfer on the disc. Thus, the plain weave is least favorable from fiber damage point of view (first two points) and most appropriate from debris retention and minimal resin transfer point of view (last two points). For satin weave, the situation is exactly reverse. Twill weave composite is always moderate in these four mechanisms working in opposite directions. This could be the reason wear rate of composite CT was lowest. CP performed second best because weightage of last two mechanisms is somewhat higher than the first two. Fretting Wear Mode Wear behavior in low amplitude oscillating wear (LAOW) mode is very different from that in sliding wear mode. Due to reciprocating motion in the former, debris formed get trapped in the contact zone and continue shearing till “wear thinned” particles can manage to escape from the zone and contribute towards “positive” wear. Among these mechanisms, the escape of wear debris is minimal if the slip amplitude is very small. The debris generated and trapped into contact zone form a third body interface by separating original two sliding surfaces. Thus, third body abrasion is a predominant mechanism in this wear situation. The wear process of fiber-reinforced composites involves various mechanisms such as fiber– matrix de-bonding followed by fiber breakage as a result of micro-cracking, micro-cutting and micro-pulverization due to reciprocating shearing stresses [23]. The wear also depends on

Tribology of Composite Materials, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

Tribology of Functionalised Carbon Nanofillers Based Polymer Composites

9

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

the amount of fiber debris produced and its extent of escape from the contact zone. Amongst all weaves, plain weave is most tight (least flexible) followed by twill and satin. This would result in maximum fiber damage to the plain weave composite. However, the same weave is also responsible for the retention of fiber debris in the contact zone. Thus, fretting wear of fabric reinforced composites is a complex phenomenon that comprises multiple mechanisms which are operative successively or simultaneously. The satin weave is the least compact amongst three followed by twill and plain. Hence, the tendency of retention of debris beneath the crimp and pockets is minimum for satin and maximum for plain weave composite, a fact that supports the highest and lowest wear rates for the former and latter, respectively. Thus, the initial fiber breakage process though is maximum in plain weave composite, the further damage processes are minimal and retention processes are maximum. Satin weave being loose, on other hand, fiber breakage is minimal because of highest length of fiber, and hence maximum flexibility, between crossover points. Thus, CP showed highest wear resistance despite maximum fiber pulverization, mainly because of its highest tendency of retention of wear debris in the pockets and beneath the crimp point. The maximum tightness of the plain weave and maximum crimp points were responsible for retention of debris leading to lowest wear of this composite.

Abrasive Wear Mode In the case of abrasive wear the basic mechanisms are drastically different from adhesive and fretting wear modes [23a]. The shearing forces being very severe during abrasion tend to cut the fibers at first instance. Whether they will be cut or not definitely depends on how rigidly they are held between crossover points. Secondly wear debris being quite large as compared to other wear modes, entrapment of wear debris in the pockets or beneath the crimp points is not possible. Debris if produced, get removed from the surface contributing to “positive” wear. Thus, the abrasive wear of such composites is mainly controlled by the ease with which fibers are broken which, in turn, depends on how tightly they are held between the crossover points. Fibers under or over crossover points are under more tension and are more vulnerable to breakage. The wear of CS was lowest at lowest load because of arrangement of strands of fibers in satin weave. Satin weave has maximum fiber length between the two consecutive crossover points followed by twill weave while plain weave has the shortest because of its alternate pattern. During abrasion, fibers in CS could get more elongated though on microscale, as compared to the fibers in CT and CP. Fibers in CS offer maximum possible resistance to elongation on microscale before breaking. These are the reasons why CS showed lowest wear and CP showed highest wear at lower loads. At increased loads, however, the phenomenon of micro-elongation before breaking, leading to wear protection, is not effective and response to the shearing forces by grits in the form of fiber cutting is more or less similar at higher loads. 2.2.2. Influence of Carbon Fabric Content Adhesive Wear Mode Dry adhesive wear studies on polyacrylonitrile (PAN) based high strength carbon fabric (plain weave) reinforced polyetherimide (PEI) composites showed that a moderate CF content (75, 65 and 55 vol.%) proved to be the most effective in increasing the mechanical strength of

Tribology of Composite Materials, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

10

M. Fahim and J. Paulo Davim

PEI and secondly, the composites with fabric in the direction normal to sliding plane led to very high coefficient of friction (μ) [22]. When fabric was parallel to the sliding plane, significant improvement in the tribo-properties of PEI in terms of very high tribo-utility (up to 600 N), appreciably low μ and enhanced wear resistance (WR) (in the range of 10−16 m3/N m) was achieved. A fairly good correlation was obtained between WR and combination of mechanical properties such as ultimate tensile strength (S), and interlaminar shear strength (ILSS).

Fretting Wear Mode Fretting wear studies of same composites showed that CF inclusion resulted in reduction in friction coefficient of PEI (0.4–0.3) and wear rate significantly (by more than an order). Very low wear rates in the range of 10-15 m3/Nm were shown by three composites with 55, 65 and 75 vol% of CF [23]. Same composites showed best range of friction coefficient and mechanical properties also. However, the enhancement in performance (both mechanical and tribological) did not improve with further increase in percentage of fibers. Fiber amount, too low or too high proved less beneficial. Optimum amount for highest strength and triboproperties was in the range of 55–65 vol%. If the ratio of fiber and matrix is optimum, matrix can hold the fibers strongly so that fibers wear preferentially by wear thinning. Otherwise fiber pulverization results leading to comparatively high wear.

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Abrasive Wear Mode In the case of dry abrasive wear performance of the same composites, the one reinforced with 65 vol% CF exhibited the best tensile and shear strength. When abraded against silicon carbide paper it showed highest wear resistance (WR) and lowest friction coefficient among all composites [23a]. The composites containing CF content higher and lower than 65 vol% showed results other than the desired performance. 2.2.3. Influence of Carbon Fabric Treatment Surface functionalization of carbon fibers is required to compensate for the small active specific surface area, low surface energy, and surface lipophobicity of carbon fibers that result in weak cohesive force between carbon fibers and the matrix. The functional groups attached on the surface of the fiber can enhance its wettability, dispersibility, and surface reactivity [24]. Chemical method, electrochemical method, plasma treatment etc., have been developed to increase the quantity of surface functional groups and thus enhance the ability to establish strong interactions between fibers and matrix [Cross references in 24]. Nano-SiO2 has been found to be one of the most important fillers for the improvements in the friction and wear behavior of fabric composites [25,26]. Hence it was thought worthwhile to coat nanosized silicon dioxide on the surfaces of carbon fabric. The adhesion between the fibers and phenolic matrix was improved greatly after surface treatment, hence the strong interface could transmit the load from the matrix to fibers efficiently. It further prevented the formation of the third body by preventing the peeling off of CF and reduced the detachment of CF from phenolic matrix [27]. Thus, adhesive wear mechanisms were more dominant than abrasive wear mechanisms in the friction process. The latter was found to be detrimental in the case of untreated CF phenolic composite. Worn surfaces and the transfer films of the

Tribology of Composite Materials, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

Tribology of Functionalised Carbon Nanofillers Based Polymer Composites

11

CFRP composites showed severe fibers cutting and lots of debris of fibers and phenolic matrix. In a different work, carbon fabric was modified with strong HNO3 etching, plasma bombardment and anodic oxidation respectively [28]. Modified carbon fabric reinforced phenolic composites exhibited improved mechanical and tribological properties. HNO3 etched carbon fabric reinforced composite showed best improvement while the anodic oxidation treated carbon fabric reinforced composite showed the least improvement in the properties. Table 3 Use of nanofillers in polymer composites Common nanofillers Carbon nanotubes (SWCNT, MWCNT)

Properties enhanced Electrical, mechanical, tribological

Exfoliated clay

Compatibiliser for polymer blends, flame resistance Thermal, Glass transition temperature Tribological properties, viscosity modification Anti-microbial UV adsorption, lubricity Charge transport Electrical conductivity, charge transport

Polyhedral oligomeric silsesquioxane (POSS) Nano silica

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Nano silver Nano zinc oxide CdSe, CdTe Graphene

Applications Food packaging, aerospace, aircraft, electrical, electronics Polymer composites

Polymer nanocomposites Medical UV screens Photovoltaic cells Electrical, electronic

3. CARBON NANOFILLERS BASED POLYMER NANOCOMPOSITES As discussed in the preceding sections, in traditional polymer composites, micron sized fibrous reinforcement, incorporation of fillers and fabric reinforcement has indeed improved the load carrying capacity of polymer. However, there is always a possibility that poor fibrematrix interaction and inhomogeneous dispersion of particles may become detrimental to achieve complete reinforcement efficiency. In order to overcome this problem, nanosized fillers have been introduced in the polymer. The main advantage with nanosized fillers is that they have large surface area and can be wet completely by the polymer matrix and thus provide stronger particle-polymer interaction. Nanofillers have proved to enhance mechanical, electrical and tribological properties (Table 3)[29]. However, nanofillers have not always led to an improvement in the properties which greatly depend on the strong bonding between the polymer matrix and nanofillers. When carbon atoms on the surface of carbon nanofibers covalently bond to the matrix it will result in significant increase in the properties. However, when carbon nanofiller concentration is increased beyond an optimum limit, the properties are adversely affected mainly because of the increase in physical interaction between fiber-fiber and due to the rheological limitations which restrict the dispersion of nanofillers in the matrix [30]. Notwithstanding this fact, potential of polymer nanocomposites have been greatly utilised in food packaging, aerospace, automotive,

Tribology of Composite Materials, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

12

M. Fahim and J. Paulo Davim

biomedical and engineering sectors mainly due to a remarkable improvement in material properties such as high storage modulus, increased tensile and flexural properties, and thermal properties [31,32]. The progress in the development of nanocomposites has got a boost with the use of carbon nanofibers and nanotubes because of their extremely high aspect ratio. For example, single-walled nanotubes are reported to have 100 times the strength of steel and less than one-sixth of the weight, which would lead to nanocomposites of exceptional properties [33,34]. In the following section carbonaceous nanofillers used for developing polymer composites are discussed in detail.

3.1. Novel Carbonaceous Nanofillers for Polymer Composites There are several ways that carbon atoms bond together to form carbon-derived materials: nanotubes, diamonds, graphite, fullerenes, and other less common forms, each having a distinct crystalline or molecular form. In diamond, each carbon atom is attached to four others in a three-dimensional lattice, which gives diamond its strength. On the other hand, in graphite, each carbon atom is attached to three others in a plane and form a hexagonal lattice, whilst the remaining bond is used to hold the planes above and below. The bonds in the plane are stronger than in diamond, but the interplanar bonds are relatively weak and enable the planes to slide. Therefore, whereas diamond is isotropic, graphite is anisotropic.

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

3.1.1. Carbon Nanotubes Iijima in 1991 discovered the carbon nanotube [35], in the soot at the negative electrode of an arc discharge. Little tubes mixed with a large amount of other forms of carbon were found. Such multi-walled carbon nanotubes (MWNT) contained 2–50 concentric graphite cylinders with a diameter of 3–10 nm and a length of up to 1 micron. Single-walled carbon nanotubes (SWNT) were developed much later. Because of adhesive forces nanotubes often bunch to form ropes. The tubes can either be open-ended or have caps formed from half a C60 molecule at either end. The structure of a nanotube is similar to that of graphite, with the difference that the sheets are closed to form a tube. Ideally, a carbon nanotube consists of either one cylindrical graphite sheet (single-walled nanotube) or several nested cylinders (multi-walled nanotube) with an interlayer spacing of 0.34–0.36 nm, that is close to the typical atomic spacing of graphite. The C–C bonds have a length of 0.14 nm, which is shorter than the bonds in diamond, indicating that the material is stronger than diamond [36]. The rolling-up of the hexagonal lattice can be performed in different ways. The sheet can be rolled-up along one of the symmetry axes, producing, either a zigzag or an armchair tube. The properties of carbon nanotubes are listed in Table 4.

Synthesis of Nanotubes The three ways to make soot that contains a reasonably high yield of nanotubes are electric arc discharge (EAD), laser ablation (LA) and chemical vapour deposition (CVD).

Tribology of Composite Materials, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

Tribology of Functionalised Carbon Nanofillers Based Polymer Composites

13

Table 4. Mechanical, electrical and thermal properties of CNTs Properties of nanotubes Young’s modulus Tensile strength Density of bundled nanotubes Band gap Electrical conductivity of bundle of nanotubes Thermal stability In vacuum In air Thermal conductivity

Value 1.8 TPa. 11 to 63 GPa 1.33–1.40 g/cm3 0.4–1 eV 1 x 109 A/cm2

up to 2800 ◦C, up to 750 ◦C. 6000W/mK at room temperature.

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Electric Arc Discharge Method In this process, the carbon electrodes are placed a few millimetres apart and the current of approximately 100 amperes vaporises the carbon into a hot plasma, some of which recondenses in the form of nanotubes which form only where the current flows, i.e. on the larger negative electrode. The voltage of about 20V, maintains a high temperature of 2000– 3000 ◦C [37]. The typical yield of nanotubes is up to 30% by weight. The tubes have diameters between 2 and 20 nm and tend to be short (50 micron or less), deposited in random sizes and directions; the typical rate of deposit is about 1 mm/min. An addition of a small amount of catalysts such as transition metal powder, like cobalt, nickel or iron to the rods, favours the growth of single-walled nanotubes. These catalysts also allow the reduction of temperature which is required to prevent the nanotubes to coalesce and merge rapidly into disorder [38]. Laser Ablation Nanotubes have been prepared by laser vaporisation of a carbon target in a furnace at 1100–1200◦C. A cobalt–nickel catalyst assists the growth of the nanotubes, presumably because it prevents the ends from being “capped” during synthesis. By using two laser pulses, growth conditions can be maintained over a larger volume and for a longer time. This scheme provides more uniform vaporisation and better control of the growth conditions. The diameter range of the tubes can be controlled by varying the reaction temperature. A flow of argon or nitrogen gas sweeps the nanotubes from the furnace to a water-cooled copper collector placed just outside of the furnace. Chemical Vapour Deposition CVD method is capable of controlling growth direction on a substrate and synthesising a large quantity of nanotubes. In this process a mixture of hydrocarbon gas, acetylene, methane or ethylene and nitrogen is introduced into the reaction chamber. During the reaction, nanotubes are formed on the substrate by the decomposition of the hydrocarbon at temperatures 700–900 ◦C and atmospheric pressure [39]. In this process, nanotubes are obtained at much lower temperature albeit at the cost of lower quality, and the catalyst can be deposited on a substrate, which allows for the formation of novel structures. The as grown nanotubes can be purified using an ultrasonic bath [40]. The larger contaminants can also be

Tribology of Composite Materials, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

14

M. Fahim and J. Paulo Davim

removed by dispersing the powder in a solvent and subsequent centrifugation. Oxidative treatment, either by heating the powder in air at 650◦C or by a liquid phase treatment in acidic environment is also used for purification. For SWNTs, standard methods are used to eliminate catalyst particles and amorphous carbon such as re-fluxing the raw material in acid followed by centrifugation or cross-flow filtration. For MWNTs, a purification method that uses the properties of colloidal suspensions has been developed. Smaller objects remain dispersed while larger particles form aggregates that are deposited as a sediment after a few hours. The size-exclusion chromatography has also been used for the purification and size selection for MWNTs and SWNTs [36,41].

LCVD Laser-assisted CVD is another method which is promising for the development of highly localized surface-bound growth of carbon nanotubes and nanofibers especially on the micrometer scale. This is a technique in which selective heating of catalytic metal nanoparticles is done in the presence of a hydrocarbon precursor. With this technique synthesis of SWNTs can be restrained to a circular region 5 μm in diameter [42]. However, synthesis of carbon nanofibers or nanotubes with a precise control over diameter, length, shape, structure, alignment and position is still to be achieved with this laser-assisted technique.

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

3.1.2. Carbon Nanofibers Carbon nanofibers are composed of various arrangements of graphene sheets. Their structure is characterized by a multitude of short-range well ordered crystalline domains. Nanofiber synthesis by catalytic chemical vapor deposition (CCVD) and its variants such as catalytic plasma-enhanced chemical vapor deposition (C-PECVD) are highly controllable [43]. Nanofibers are now used as functional components in devices either as electrochemical probes, electrodes or field-emission electron sources [44].

3.1.2.1. Fullerenes C60 Fullerenes and functionalized fullerenes have immense potential as filler in improving the properties of polymer composites. However, they are not soluble in common solvents and need to be chemically modified [45,46]. Apart from this, the wettability of fullerenes and interfacial interaction with composite materials is also a cause of concern. Fullerenes have a unique spherical shape with cage diameter of 0.71 nm and possess high load-bearing capacity, low surface energy, high chemical stability, weak intermolecular and strong intramolecular bonding by virtue of which it has been extensively investigated to explore its tribological potential as a lubricant [47-51]. 3.1.3. Carbon Nanorods/Nanowhisker The nanoparticles/whiskers are produced by catalytic chemical vapor deposition (CCVD) process. One of the unique features of these carbon nanoparticles/whiskers is that the degree of agglomeration is surprisingly small which enables their easier miscibility with the matrix polymers.

Tribology of Composite Materials, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

Tribology of Functionalised Carbon Nanofillers Based Polymer Composites

15

3.1.4. Amorphous Carbon Amorphous carbon (a-C) coatings, containing high sp3 bond could improve the friction and abrasion properties in ambient air [52-55] and could deteriorate in vacuum or during the air-to-vacuum transition because the amorphous carbon coatings failed within a short sliding distance [56]. Amorphous carbon nanorods can be prepared controllably by catalytic chemical vapor deposition (CCVD). The aligned films of amorphous carbon nanorods promise a low friction coefficient and better wear resistance under dry conditions [57,58].

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

3.2. Functionalisation of Carbon Nanofillers Nanotubes tend to agglomerate or self-assemble into bundles and need to be separated if they are to be used as a filler to make polymer nanocomposites. They are generally separated using both non-covalent as well as covalent modifications [59 and references therein]. Noncovalent modifications are based upon physisorption of molecules and polymers on CNT surfaces. The covalent modifications involve chemical functionalization by the attachment of various functional groups to the ends or sidewalls of the nanotubes through covalent bonds. The covalent modification also alters the structural and electronic properties of the nanotubes. Since CNTs can react with various classes of compounds they can be integrated into inorganic, organic, and biological systems [60]. For instance, grafting polymers to CNTs can be performed by generation of active polymeric species (radicals [61] or anions [62,63]) in a suspension of the CNTs in an organic solvent. It can be realized by using ready-made polymers (in the case of radical initiation), or in situ polymerization of monomers in both types of initiation. Various chemical modification methods tried for functionalization of CNTs include oxidation, hydrogenation, fluorination, addition of free radicals and other reactive molecules. Fluorine and other functional groups on the side walls of SWNTs surface not only alter their physical properties such as conductivity and solubility they also enhance their chemical reactivity. The functional groups chemically attached to the nanotubes deagglomerate the nanotube bundles and help in uniform dispersion of CNTs in polymers and ceramics. The functional groups also provide multiple sites for covalent bonding of nanotubes to polymer matrices. The fluorine modified SWNTs possess an ultra-low friction coefficients (0.002– 0.07), making them attractive candidates for solid lubricant applications [64]. Treatments of SWNTs by sonication in HNO3 or H2SO4/ HNO3 and H2SO4/H2O2 mixtures have also showed generation of carboxyl groups and other oxofunctionalities on nanotube open ends. Thermally stable (up to 400oC) hydrogenated carbon nanotubes (HCNTs) have been prepared from purified CNTs through a reduction process, utilizing Li and methanol in liquid ammonia. The direct fluorination of the SWNTs with elementary fluorine results in the sidewall-fluorinated SWNTs (fluoronanotubes). They were prepared by fluorination of both L-SWNTs [65,66] and nanotubes produced by a high-pressure disproportionation of CO (HiPco-SWNTs) [67,68]. The in situ covalent bonding of fluoronanotubes to a polyethylene matrix during melt processing has also been observed in some cases [69]. SWCNTs have also been debundled and solubilised in a poly(N-vinyl carbazole) (PVK) matrix [70]. Solvent-free CNT functionalization by polymers involves reaction with aryl diazonium species [71]. Such

Tribology of Composite Materials, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

16

M. Fahim and J. Paulo Davim

functionalization techniques remove the need for any solvent during the functionalization step. This delivers functionalized CNTs with increased solubility in organic solvents and processibility in polymeric blends. The second methodology involves the functionalization of CNTs that are first dispersed as individual tubes in surfactants in aqueous media. Functionalized nanotubes have no tendency to re-rope. They remain as individuals in organic solvents, giving enormous increases in solubility.

3.3. Processing Techniques The traditional methods to prepare carbonaceous filler based nanocomposites are discussed in the following section. The two most critical factors in developing a polymer nanocomposites using carbon nanofillers discussed in the preceding section are dispersion of nanofillers and better structure-property correlation achieved based on the strength of fillermatrix interface. Subsequently various techniques used so far for developing nanocomposites have been used with precaution keeping in mind the above issues. Some common processing techniques used for nanocomposites are discussed in the following section.

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

3.3.1. Melt Mixing/Compounding Mixing of the nanosized second phase filler particles in polymer is the most critical issue in developing a polymer nanocomposite. Better mixing between these two ensures higher strength and stiffness. When CNTs are thoroughly dispersed in the polymeric material, an interconnecting network of CNTs is formed which provides a pathway for electrical charge to flow and axial properties. Numerous techniques have been used to achieve uniform dispersion of CNTs in the polymer. The most popular technique being the preparation of the CNTpolymer composites under sonication [72] and in some cases using chemicals such as alkoxysilane terminated amide acid oligomer to disperse the CNTs [73]. The major aim is to incorporate an optimum concentration of MWCNTs required for property enhancement without severely reducing the melt flow properties of the polymeric resin. Dry mixing technique can be used to incorporate MWCNTs, however, in this technique CNTs tend to agglomerate at high loading levels and the viscosity becomes too high for processing. Application of shear during mixing can avoid the agglomeration. Solution mixing, melt blending and melt spinning techniques used to develop polymer nanocomposites have their own advantages and disadvantages. No use of organic solvents and compatibility with industrial processes, such as extrusion, injection and blow molding, and other polymer processing techniques make these techniques desirable. However, it is usually difficult to achieve homogeneous dispersion of CNTs throughout polymer matrices by melt-blending method.

Sonication Ultrasound sonication is one of the promising approaches to disperse the nanoparticles into the base material thoroughly. High-intensity ultrasonic waves generate nonlinear effects in the liquids such as transient cavitation and acoustic streaming [74-77]. Liquid medium is necessary because sonochemistry is driven by acoustic cavitation that only occurs in liquids. Acoustic cavitation involves the formation, growth, pulsating and collapsing of tiny bubbles, producing transient (in the order of microseconds) micro-hot spots that can reach

Tribology of Composite Materials, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

Tribology of Functionalised Carbon Nanofillers Based Polymer Composites

17

temperatures of about 5000◦C, pressures of about 1000 atm, and heating and cooling rates above 1010 K/s [78]. The strong impact coupling with local high temperatures can also enhance the wettability between polymer and nanoparticles; thus can break the agglomerating bodies by damaging the Coulomb and van der Waals forces between the particles and make them disperse homogeneously in the liquid medium. However, process parameters and base materials properties affect the mixing process in sonication. Furthermore, sonication has an optimum time limit which varies with sonicator power, wt% of nanoparticles and foam amount.

Ball Milling Ball-milling is a mechanical process that leads to local generation of high pressure as a result of collisions throughout the grinding media [79]. This method has been used to obtain nano-barrels from cup-stacked carbon nanotubes, transform nanotubes into nanoparticles (ellipsoidal and spherical), generate nanostructures from graphite, and shorten the lengths of nanotubes [80 and references therein].

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

3.3.2. Extrusion MWCNTs can also be dispersed using a micro-scale twin-screw extruder in which uniform dispersion is achieved through melt mixing of molten polymer and CNTs [81,82]. The high shear mixing necessary to disentangle and uniformly disperse MWCNTs in the matrix can be achieved with the extruder. Extrusion of nanocomposite through a die and subsequent drawing yields continuous ribbons of nanocomposites with aligned nanotubes that can be further processed into laminates having improved elastic modulus and yield strength compared to randomly oriented nanocomposites [82]. For thermoplastic polyetherimide reinforced with SWCNTs, some improvement in mechanical properties could be achieved, however, it has been observed that melt processing is desirable for thermoplastics rather than thermosets since melt stability can be achieved in the required processing zone [83].

3.3.3. In Situ Polymerisation In situ polymerization is an important technique for making polymer/CNT nanocomposites because it is claimed to achieve fine dispersion of CNTs [84-86]. PMMA was produced in a process of addition polymerization. In this process, the free radical initiator, benzoyl peroxide (BPO), was added into the MMA at the reaction temperature (85– 90oC). A BPO molecule formed two free radicals during the reaction. With the initiation of these free radicals, the CjC double bonds in MMA molecules would be opened, and then linked with each other to form long chain of PMMA molecules [86]. After a prolonged reaction (almost 1 h) CNTs were added into reacting mixtures and the resulting PMMACNTs viscous mixtures were mixed ultrasonically and stirred to ensure better dispersion of CNT in polymer matrix [85].

3.3.4. Thin Film Deposition Nanocomposite thick films can be solution cast by dissolving extrudate rods in a solvent and deposited on a substrate in a glove box. Films can then be air dried at room temperature

Tribology of Composite Materials, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

18

M. Fahim and J. Paulo Davim

and then detached from the substrate [87]. In another technique, the solution was subjected to homogenization using a motorized homogenizer with an attached facility of continuous cooling [88]. This solution was used to coat the polymer film onto the piranha treated Si surface using spin coating. Multi-walled VACNT films have also been grown on a silicon substrate using a DC-PECVD method [89].

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

4. TRIBOLOGY OF CARBON NANOFIBRE BASED POLYMER NANOCOMPOSITES Addition of even a very low amount of rigid nanoparticles/fibers in polymers signifycantly improves the base properties of polymer especially stiffness and strength that are mainly attributed to the availability of large numbers of nanoparticles with huge interfacial areas. This is one reason why nanosized reinforcements are preferred over macro- and microsized reinforcements. From the tribological viewpoint, the small size of nanoparticles with homogenous dispersion in the matrix and good interfacial adhesion between nanoparticles and matrix are the most essential requirements for a polymer nanocomposite. Wear or progressive loss of material is assumed to be low since the nano-particles have almost same size as that of surrounding polymer chains. The tribological properties of carbon nanofibers reinforced nanocomposites become sensitive to operating parameters and environment conditions because the nanosized fillers and their removal as lubricating wear debris at the contact surface greatly influences the surface interactions. Tribological evaluation of such composites thus assumes greater significance in view of understanding the genesis of friction and dissipation of mass at low dimensions of nanofillers. Polyetheretherketone (PEEK) is one of the most tribo-efficient polymer because it is thermally stable, exhibits high wear and friction performance even at high PV limits. Various carbon fibre reinforced and solid lubricated variations of PEEK has been developed over the years which are used for developing tribo-components. Hence, the effect of carbon nanofibers on the tribology of PEEK has always been a subject of great interest. PEEK based nanocomposites containing vapour-grown carbon nanofibre have shown a linear increase in tensile stiffness and strength with nanofibre loading fractions up to 15 wt% while matrix ductility was maintained up to 10 wt% [90]. A homogeneous dispersion and alignment of nanofibres was claimed by the authors. Based on DSC thermograms an interaction between matrix and the nanoscale filler was suggested. The influence of same vapour-grown carbon nanofibres (CNF), of average diameter 150 nm, on the wear behaviour of semicrystalline poly(etherether ketone) (PEEK) was investigated later [91]. Carbon nanofibres were found to reduce the wear rate of PEEK significantly. A comprehensive review on the comparative tribological performance of carbon nanofillers filled PEEK in dry sliding against 440C in various gas environments can be found in Ref [91a].

4.1. Tribology of Functionalized Carbon Nanofiber Polymer Composites The surface treatment of carbon fibers (CF) and its effect on the tribological properties of CF reinforced polytetrafluoroethylene (PTFE) composites has revealed that RE treated CF reinforced PTFE (CF/PTFE) composite shows lowest friction coefficient and wear under

Tribology of Composite Materials, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Tribology of Functionalised Carbon Nanofillers Based Polymer Composites

19

various applied loads and sliding speeds as compared to untreated and air-oxidated composites [92]. X-ray photoelectron spectroscopy (XPS) study of carbon fiber surface showed that, after RE treatment, oxygen concentration increased obviously, and the amount of oxygen-containing groups on CF surfaces were largely increased. The increase in the amount of oxygen-containing groups enhanced interfacial adhesion between CF and PTFE matrix. With strong interfacial adhesion of the composite, stress could be effectively transmitted to carbon fibers that enhanced the tribological performance of the composite. In another work, carbon nanofiber (CNF) was treated with HNO3 and a coupling agent [93]. This surface modification was reported to decrease the friction coefficient as well as wear loss of CNF/PTFE composites slightly. Anti-wear property of the composite was achieved when CNF was treated with HNO3 followed by coupling agent treatment. In the case of polyurethane, which exhibits best abrasion properties, when it is filled with a diisocyanate treated CF, effective improvement of the interfacial adhesion between the CF and polyurethane matrix was achieved which further enhanced the tribological properties of PU coating [94]. Effects of untreated and pretreated carbon nanofibers (CNFs) on the crystallization behavior, friction behavior, and mechanical properties of ultra high molecular weight polyethylene (UHMWPE)/high density polyethylene (HDPE) nanocomposites prepared by a twin-screw extrusion have showed that the addition of CNFs though affected the temperature of crystallization it did not alter the crystalline structure of the UHMWPE/HDPE blend [94a]. The degree of crystallinity, and the tensile strength and modulus of the UHMWPE/HDPE systems increased initially with addition of CNFs but decreased with further increase in CNF loading. With the increase of untreated CNF content, the friction coefficient of UHMWPE/HDPE decreased albeit with no change in the friction process. The degree of crystallinity of the nanocomposites with the pretreated CNFs exhibited a decrease due to the better interface adhesion compared to that in the nanocomposites with the same loading of untreated CNFs. The enhancement in tensile strength of nanocomposites containing 0.5 wt% treated CNFs was found to be four times higher (32%) than that of the nanocomposites containing untreated CNFs (8%) over that of the pure polymer.

4.2. Tribological Characterization of Functionalised Carbon Nanotubes Polymer Composites The carbon nanotubes are still in the early stages of tribological investigation and it has been found single-wall carbon nanotubes (SWNTs), multiwall carbon nanotubes (MWNTs), graphitized MWNTs, fluorinated SWNTs, carbon nano-onions, and carbon nitride spheres have superior friction properties and endurance lives in air or ultrahigh vacuum indicating their potential use as lubricant in aerospace applications in microelectromechanical systems (MEMS) and micromachines. All nanocarbons can dramatically improve the stiction (or adhesion) and friction between contacting surfaces under dry conditions, a major issue for MEMS and micromachines. In particular, the coefficient of friction for reduced SWNTs and nascent SWNTs is 1/50th and 1/10th, respectively, of that for conventional graphite lubricants in air. The coefficient of friction for graphitized MWNTs is one fifth that of MoS2 in ultrahigh vacuum [13]. Among nanofillers, carbon nanotubes (CNTs) have fancied the imagination of scientists mainly because of its exceptionally high modulus (1–1.8 TPa),

Tribology of Composite Materials, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

20

M. Fahim and J. Paulo Davim

stiffness, axial strength and high aspect ratio (~103) which is expected to improve the mechanical strength of polymers manifold [95]. Polymer composites reinforced by CNTs have been extensively researched in the last few years for their strength and stiffness properties [96]. However, it has been observed that a poor CNT polymer interface can lead to poor intertube load transfer in agglomerates and between CNT and polymer matrix resulting in reduced performance. There are three major aspects which need to be addressed to achieve complete reinforcement efficiency; better CNT-polymer interface to transfer superior properties of CNT to polymer, functionalization of CNTs, optimum concentration of CNT to realise maximum benefits and controlling the viscosity of CNT-polymer mixture so that CNT concentration can be increased. These three aspects will be discussed in the subsequent section vis-à-vis tribological performance. A comprehensive review of the polymeric nanocomposites for tribological applications can be found in Ref. 97. Table 5 shows a compilation of the friction coefficients and specific wear rates of neat and carbon nanotubes filled polymer nanocomposites characterised over the years since the discovery of nanotubes in early 90s. Polytetrafluoroethylene (PTFE) is one of the most important heat resistant tribologically significant polymer and is used in almost all tribological applications as solid lubricants, seals and wear reducing fillers. Though it has the lowest friction coefficient in sliding wear mode it is seldom used because of its high wear. Various attempts have been made to modify its properties using fibrous reinforcement and filler addition so that it can be used in different forms for tribological applications. Since carbon nanotubes have proved to be a potential reinforcement for polymers so far as improvement in mechanical properties is concerned it was expected that it will do wonders for PTFE. Expectedly, when reinforced with carbon nanotubes, PTFE exhibited considerable improvement in the wear and friction properties as shown in Table 5. Unfortunately not much data followed in the subsequent years to supplement the author’s claim. It has been reported that CNTs filled PTFE composites exhibit friction coefficient that decreases with increasing CNT content probably due to self-lubricating nature of CNT [98]. It was also claimed that CNTs strengthen the structure of PTFE and effectively reduce its adhesive and ploughing wear, thereby significantly improve the wear resistance of CNT/PTFE composites. A CNT/PTFE composite containing 20 vol% of CNT exhibited the lowest wear rate which was attributed to the strength and high aspect ratio of CNTs. CNTs were released from the composite during sliding and transferred to the counterface. The transfer film prevented direct contact and reduced both wear rate and friction coefficient. Multifunctionality of SWCNT-PTFE nanocomposites and PTFE/Cotton fabric composites filled with MWCNT in improving tribological performance of PTFE has also been reported recently [98a, 98b]. In contrast to PTFE, lot of work has been done on another tribologically significant polymer UHMWPE which exhibits the best abrasion resistance and equally good properties in other wear modes [99-101]. However, it is seldom used in high temperature conditions because of its poor thermal stability. When reinforced with MWCNTs, an increase in hardness as well as wear resistance was observed [99]. It was thought that the variation in the microstructure and load carrying capacity of the composites might have been responsible for the enhancement. However, it was found that CNT addition makes no change to the internal structure of UHMWPE. Instead applied stresses and frictional stresses were transferred to the

Tribology of Composite Materials, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Table 5. Tribological properties of the CNT filled polymer composites Ref

Tribo-material

112

Pure PMMA + 0.05wt%MWCNT + 0.10wt%MWCNT + 0.25wt%MWCNT + 0.75wt%MWCNT + 1.00wt%MWCNT + 1.50wt%MWCNT + 2.50wt%MWCNT UHMWPE/HDPE Cop (80/20) + 0.2wt%MWCNTf + 0.5wt%MWCNTf + 1.0wt%MWCNTf + 2.0wt%MWCNTf + 0.2wt%MWCNTu + 0.5wt%MWCNTu + 1.0wt%MWCNTu + 2.0wt%MWCNTu Pure UHMWPE + 0.1wt%MWCNT + 0.2wt%MWCNT + 0.5wt%MWCNT Pure UHMWPE + 1.0wt%MWCNT

101

100

99

+ 5.0wt%MWCNT

102

Pure HDPE + 1.0wt%MWCNT + 3.0wt%MWCNT + 5.0wt%MWCNT

Specific wear rate, (mm3/Nm) 2.90 x 10-4 2.60 x 10-4 2.40 x 10-4 1.80 x 10-4 1.25 x 10-4 0.90 x 10-4 1.00 x 10-4 1.25 x 10-4 44.3 x 10-4

Friction coefficient 0.45 0.44 0.41 0.37 0.36 0.33 0.32 0.32

33.0 x 10-4 20.0 x 10-4 22.5 x 10-4 19.0 x 10-4 17.5 x 10-4 18.0 x 10-4 16.0 x 10-4 13.0 x 10-4 0.35g 0.25g 0.12g .04g 8nm 17nm 6nm 9 Nm 4nm 6 Nm Dr30 Dr10 0.1546 mm/106 cy 0.0840 0.0798 0.0637

0.10 – 0.12

0.05 0.06 0.10 0.11 0.27 0.25

0.31 0.29

0.22

0.27

Test conditions sliding wear,block-on-ring v= 0.431m/s;L=50N, t=1h,RT

Modified four ball tester; L=92.3N on 100Cr6 and X5CrNi18-10 steel

5.1 5.4 5.5 5.8

Dr30 Dr10 0.0506

0.03 0.04

Hardness HV 26 27 28 31 35 37 36 36

sliding wear, ball-on-disc v= 0.3m/s;L=5N, t=2h,RT

L=75µN;v=5µm/s; scratch length= 5µm; nanowear

sliding wear, block-on-ring v= 200rpm; L=45kg, 500,000 cycles,40C, water lubricant

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Table 5. Tribological properties of the CNT filled polymer composites (Continued) Ref

Tribo-material

113 .

Pure PS + 0.05wt%CNT + 0.10wt%CNT + 0.25wt%CNT + 0.75wt%CNT + 1.00wt%CNT + 1.50wt%CNT + 2.50wt%CNT + 4.0wt%CNT Pure polyamide 6 + 2wt%CNT(CaCO3) + 2wt%CNT(CaCO3)f + 2wt%CNT(MCM)

109

104

114

108

Pure Polyimide + 1.0vol%CNT + 2.5vol%CNT + 5.0vol%CNT + 8.0vol%CNT + 15vol%CNT + 25vol%CNT + 30vol%CNT PMMA/PS + 0.10wt%MWCNT + 0.50wt%MWCNT + 1.0wt%MWCNT + 1.5wt%MWCNT + 2.0wt%MWCNT + 3.0wt%MWCNT Polyimide on Si PI+SWCNT on Si

Specific wear rate, (mm3/Nm) 13.0 x10-5 12.0 x10-5 9.0 x10-5 7.0 x10-5 4.0 x10-5 3.0 x10-5 1.0 x10-5 1.5 x10-5 2.0 x10-5 280 30-40 200 25-30 260 60-90 180 20-25 Rp / Rh/ µm µm 0.5 mm3 ≤ 0.5

2.30 x10-4 2.10 x10-4 1.70 x10-4 1.45 x10-4 1.32 x10-4 1.38 x10-4 1.39 x10-4 3000 cycles 7200 cycles

Friction coefficient 0.42 0.41 0.39 0.38 0.36 0.33 0.31 0.31 0.30

Hardness HV 29 30 31 32 34 37 40 39 38

Test conditions Block-on-ring (HRC 48-50) v= 0.431m/s;L=50N, t=1h,RT

Scratch test, L=15N; scratch length= 5 mm; v= 5mm/min, RT; conical indentor

0.300 0.260 0.240 0.242 0.245 0.244 0.240 0.220 0.48 0.46 0.41 0.38 0.36 0.37 0.36 0.10 0.12

27 36 38 42 45 44 43 44 24 25 28 34 35 35 35 0.43GPa 0.72GPa

Sliding wear; block-on-ring; L=50N; v= 0.431 m/s; t=1.5h; RT

Block-on-ring (HRC 48-50) v= 0.428m/s;L=50N, t=1h,RT

Ball-on-disk, v = 0.042 m/s; L = 7g = 370 MPa; RT

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Table 5. Tribological properties of the CNT filled polymer composites (Continued) ref

Tribo-material

98

Pure PTFE + 2.5vol%CNT + 5.0vol%CNT + 10.0vol%CNT + 15.0vol%CNT + 20.0vol%CNT + 25.0vol%CNT + 30.0vol%CNT Pure BMI + 1.0wt%MWCNT + 2.5wt%MWCNT + 5.0wt%MWCNT + 7.5wt%MWCNT + 1.0wt%MWCNTf + 2.5wt%MWCNTf + 5.0wt%MWCNTf + 7.5wt%MWCNTf Pure Epoxy

111

116

+ Fullerene(10wt%) + MWCNT(0.5wt%) + Predispersed MWCNT (0.5wt%) +aligned CNT(0.3wt%)

115

Pure epoxy + 0.20wt%MWCNTf + 0.50wt%MWCNTf + 1.0wt%MWCNTf + 2.0wt%MWCNTf + 4.0wt%MWCNTf + 0.2wt%MWCNTu + 0.5wt%MWCNTu

Specific wear rate, (mm3/Nm) 900 x10-6 20 x10-6 9.0 x10-6 4.0 x10-6 3.0 x10-6 2.0 x10-6 2.5 x10-6 2.6 x10-6 2.80 mg/m 1.90 1.70 1.75 1.79 1.85 1.40 1.45 1.50 100-240nm (5-15cycles)

Friction coefficient 0.200 0.190 0.186 0.184 0.179 0.175 0.177 0.172 0.80 0.70 0.65 0.55 0.50 0.20 0.19 0.19 0.18 0.553

Hardness HV

Test conditions Sliding wear, L=200N; 200 rpm; t=2h;Cr18Ni9Ti

32 33 35 32 34 45 66 55 58

Ball-on-plate, v = 0.84 m/s; L = 500N, t=1h;RT

Scanning probe microscopy; v= 5µm/s; L= 50µN; t=15s

0.408 0.391 0.389 70-130nm (5-15cycles) Wear depth 9.0 x10-6 7.5 x10-6 7.0 x10-6 4.5 x10-6 5.5 x10-6 6.5 x10-6 12.5 x10-6 9.0 x10-6

0.342

Ball-on-prism, Dead wt=30N; L=21.2N; v=28.2 mm/s; t=60h=6000m pv = < 0.76 MPa.m/s

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Table 5. Tribological properties of the CNT filled polymer composites (Continued) Ref

118

117

Tribo-material + 1.0wt%MWCNTu + 2.0wt%MWCNTu + 4.0wt%MWCNTu EP/CNT(1%)+PTFE(10%) EP/CNT(1%)+MoS2(10%) EP+ MoS2(10%) EP/CNT(1%)+Grp(10%) EP+Grp(10%) HNBR/FKM(75/25) + MWCNT HNBR/FKM(50/50) + MWCNT HNBR+ MWCNT(10%) HNBR Pure HNBR

Specific wear rate, (mm3/Nm) 20.0 x10-6 9.0 x10-6 12.0 x10-6 6.0 x10-6 1.0 x10-6 0.9 x10-6 3.5 x10-6 2.0 x10-6 0.3 x10-4 0.1 x10-4 0.9 x10-4 0.1 x10-4 2.5 x10-4 2.9 x10-4 10-1 10-1

0.050 0.052 0.053 0.055 0.035 0.040 1.1

2.9

+ MWCNT(10phr)

10-2

10-3

1.4

2.6

+ MWCNT(30phr)

10-3 POP

10-3 ROP

1.4 POP

2.3 ROP

Friction coefficient

Hardness HV

Test conditions

Orbital RBOP; ball-on-sheet; 100Cr6 steel; L=90N. rotational speed=280 rpm; t=3h

100Cr6 Steel Pin-on-rubber plate (POP), L=2N; v=250 mm/s, t=1.5h Roller(steel)-on-plate (rubber) (ROP); 100Cr6 steel100Cr6 Steel Pin-on-rubber plate (POP), L=2N; v=250 mm/s, t=1.5h

f stands for functionalized, u for non-functionalised, MWCNT: Multiwalled carbon nanotubes; specific wear rate is in mm3/Nm unless specified, all values are approximate and obtained from experimental curves.

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Tribology of Functionalised Carbon Nanofillers Based Polymer Composites

25

nanotubes during indentation. CNT addition increased both the local compressive strength as well as the shear strength. Since UHMWPE is an abrasion resistant polymer a lower debris generation was expected when contact surfaces with higher strength of UHMWPE were exposed during sliding. A slight increase in the coefficient of friction was attributed to the increase in shear strength and surface roughness by CNT addition because the mechanically strong CNTs might have caused abrasion of counterface. Supporting data is available in a study done on CNT/UHMWPE composites at nanoscale [100]. The images of the nanoscratch tests done on MWCNT/UHMWPE films and UHMWPE films showed that more material deposits were piled up near the scratches on the UHMWPE film. Moreover, on the surface of the MWCNT/UHMWPE film a microstructure different from that of UHMWPE film was observed. This observation suggested that the microstructure was the interface between the CNTs and UHMWPE. Several strips dragged along the scratch also suggested that the same micostructure had a strong bonding between CNTs and UHMWPE and was responsible for the enhancement of the chain mobility of MWCNT/UHMWPE. Due to this chain mobility, the fracture cracks could not propagate in the MWCNT/UHMWPE composite. Therefore, the nanotube-reinforced composite exhibited better wear resistance than UHMWPE. The authors refrained from attributing the low friction coefficient of the MWCNT/UHMWPE composite to self-lubrication of CNTs transferred on the counterface because they could not obtain any supporting evidence when AFM tip was used to scratch the composite and then examined through Raman spectroscopy. Johnson et.al. [102] also noticed a fall in friction coefficient with the addition of CNT for CNT/HDPE composite. However, they claimed that since friction mechanism is complicated and there are probably competing factors associated with it the low friction could not be attributed alone to the worn off CNTs debris acting as lubricant, much like powder graphite lubricants. However, the authors cautioned that since there work was in water-lubricated condition it cannot be compared with dry sliding conditions. Apart from this result they also reported that the addition of CNTs to HDPE improved certain material properties including material stiffness, maximum load to failure and work to failure. The wear tests showed that the addition of 5 wt% CNT decreased the overall wear rate by up to 50% and friction coefficient decreased by atleast 12%. Based on these results the authors suggested that since UHMWPE is a very good tribo-material for artificial joint replacements, CNT should be reinforced in UHMWPE to achieve the desired result. However, the drawback with UHMWPE is its inherent weaknesses such as high creep compared to metal and bone. Since HDPE exhibits a better creep resistance it could be blended with UHMWPE and then reinforced with CNTs to achieve desired results. That way the poor wear resistance of HDPE will also be compensated for. Tribological characterization of UHMWPE/HDPE blends reinforced with CNTs showed that though pure UHMWPE exhibited a pronounced running-in phase characterised by a steep slope and followed by an almost constant wear rate, the UHMWPE/HDPE blend and the CNT reinforced composites exhibited more or less linear wear curves [101]. This indicated the formation of self-lubricating transfer film on the steel counterpart which took a longer time in the case of the pure UHMWPE than in the case of the composites. There was also a significant reduction in wear rate caused by CNT addition and was correlated with the increase in young’s modulus with increasing CNT content.

Tribology of Composite Materials, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

26

M. Fahim and J. Paulo Davim

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

4.2.1. Influence of Functionalization of CNT on Friction and Wear Behaviour A noticeable factor observed in the materials discussed in the preceding paragraph was that the composites reinforced with untreated CNTs had a better wear performance than the functionalised CNTs reinforced composite. It was in quite contradiction to the belief that functionalization improves the filler-matrix adhesion [103]. The authors claimed that the purpose of functionalization is to graft the functional polar groups onto the CNT surface to promote filler/polymer interaction. However, for a polymer such as polyethylenes, the functional groups may have been less important because PEs are non-polar. Secondly, the boiling acid treatment must have cracked and peeled the MWCNTs thereby adversely affecting its aspect ratio and hence the reinforcing efficiency was badly affected. This may be the reason untreated CNTs based composite returned better wear resistance than acid treated CNTs based composites. The authors also claimed that apart from functionalization, dispersal of CNTs was a critical factor, an issue discussed at length in later sections. The benefit of CNT functionalization could not have been significant for PEs, but with high temperature polymers such as polyimides, polyetherimides and bismaleimides, a clear advantage has been reported [104-108]. For instance, with unfunctionalized CNTs, PI/CNT nanocomposite showed a sharp decrease in wear rate as well as friction coefficient in dry sliding against the steel counterpart when the CNT content was below 8.0% [104]. The enhancement was attributed to increased microhardness and bending strength. CNT as a reinforcing agent also contributed in restraining the adhesion and scuffing of the PI matrix and helped in the formation of the better quality transfer films on the counterface. This, together with the increased load carrying capacity due to CNT reinforcement contributed to improved friction and wear performance. Encouraged by this result, it was expected that a better CNT-matrix adhesion following the functionalization of CNTs will definitely result in better results. Functionalised CNTs have already been reported to have improved the mechanical properties of PI nanocomposite films [105]. The design of the amine terminated PI to share the same structural units with the polyimide ensured the full compatibility between the functionalised CNTs and the matrix. In the case of polyetherimides, it was observed that the mechanical properties of functionalised MWCNT/PEI composites increased significantly with addition of a very small amount of MWCNT (≤ 1 wt%) [106]. The high performance of the nanocomposites was attributed to the homogeneous dispersion of CNTs in the matrix as well as the strong interfacial adhesion between the two due to nitric acid treatment which created COOH and OH groups. The carboxylic groups increased the anchoring sites along the nanotubes with polymer matrix thus favouring the stress transfer from the matrix to CNTs. The OH groups on the other hand helped in improving the interaction with PEI which possess polar -CONCO- groups along the polymer chains. This result is supported by the data published in a paper on MWCNTs grafted chemically with PEI [107]. The grafting occurred via both amide as well as imide linkages. It was reported that the tensile strength and modulus of the PEI films grafted with carboxylic acid functionalised MWCNT increased with CNT concentration (0.14 – 0.38 wt%). In the case of SWCNT/PI film, the authors claimed that an improvement in hardness and elastic modulus resulted in the improved load-bearing capacity of the PI films [108]. In addition to this the resulting textured microstructure reduced the real area of contact which restricted the wear particle generation during sliding. The vertical alignment of nanotubes (as observed through AFM) probably prevented the breaking and pull-out of these tubes due to their high elongational tensile properties which contributed in

Tribology of Composite Materials, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Tribology of Functionalised Carbon Nanofillers Based Polymer Composites

27

enhancing the wear durability. However, the increase in friction coefficient despite no wear to the film was attributed to the vertical alignment of CNTs induced during sliding and the continuous increase in the contact area. The benefit of CNT functionalization was also studied in the case of polyamides 6 nanocomposites [109]. It was observed that wear penetration depths were considerably reduced with the addition of MWCNTs. However, the reduction was larger for CNTs functionalised with Fe/MCM41 catalyst. Functionalised CNTs reduced only slightly the penetration depth of PA6 as compared to non-functionalized CNTs. It was claimed that interfacial adhesion between the filler and polymer was indeed improved by the functionalization. However, the melt mixing process suppressed the free movement of the nanotubes due to its high viscosity that impeded the adequate dispersion of the CNTs in the matrix. It was also suggested that the restricted movement due to the van der Waals forces and hydrogen bonds between the functionalised nanotubes might have caused tubes to agglomerate that reduced the stress transfer from the polymer to the filler. Secondly, in those regions of polymer in which the CNTs were well dispersed and interacted with the polymeric chains, possibly by means of the amide groups in the PA6 and hydroxyl groups in the tube, might have reduced the chains movement that impeded the recovery of the polymer. The authors claimed that acidic functionalization of CNTs did not considerably increased the scratch resistance and attributed it to the inadequate dispersion and agglomeration of CNTs. Similar results have been reported for polyamide 6 reinforced with non-functionalised CNTs [109a]. It was observed that non-functionalised CNTs improved the wear resistance and friction performance under dry as well as water lubricated sliding conditions probably due to the effective reinforcing and self-lubricating effects of CNTs. This result was contradicted for another tribologically active and heat resistant polymer, bismaleimides [110-111a]. The authors used two kinds of original MWCNTs with different diameters, and a COOH-MWCNT to prepare three different CNT/BMI composites [110]. They reported that the addition of MWCNTs in BMI decreased the friction coefficient irrespective of the kind of MWCNTs used. However, functionalization of CNTs changed the wear mechanism from adhesive wear (for pure BMI resin) to abrasive attrition by changing the self-lubricating property of the worn surface. For MWCNTs-DBA/BMI, the improved microhardness as well as wear and friction performance was attributed to the improved interface between functionalised MWCNT and BMI which was reflected in wear mechanisms [111]. The worn surface of pure BMI showed a rough microstructure displaying the features of fatigue wear while the worn surface of MWCNT-DBA/BMI composite displayed a smooth texture and features of adhesive wear. Using a different functionalization; MWCNT-EDA, authors observed that aminofunctionali-zation triggered defects in the MWCNTs which thus reduced the self-lubricating effects of CNTs and increased the friction coefficients of functionalised CNT/BMI composites [111a]. On the contrary, it strengthened the interface between the MWCNT and BMI and enhanced the dispersive state of the MWCNTs in the BMI matrix.

4.2.2. Threshold Concentration of CNT The threshold concentration for the composite discussed in the preceding paragraph was found to be around 2.5 wt%. Several other reports are also available for other commodity polymers such as PMMA, PS and their blends in which authors have found a threshold

Tribology of Composite Materials, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

28

M. Fahim and J. Paulo Davim

concentration of CNTs [112-114]. For PMMA, around 1 wt% of CNT results in maximum wear and friction performance [112] while for polyacrylonitrile –MMA 1.5 wt% CNT reinforcement produces best wear and friction performance [112a]. Similarly, for PS and PMMA/PS blend around 1.5 wt% of CNT is the threshold concentration [113-114]. It has also been reported that 1 wt% of CNT was optimum to achieve better wear and friction performance in epoxy/CNT nanocomposite [115]. To this composite, solid lubricants like PTFE, MoS2 and graphite were added individually to see if there was any further reduction in wear. It was observed that CNTs and graphite/MoS2 mutually aid one another and there was a positive synergism which further increased the wear resistance of composite. However, this composite performed badly against 100Cr6 bearing steel compared to its incredible performance against X5CrNi18-10 stainless steel. In a previous paper it was reported that wear properties of CNT-Epoxy composites depend greatly on the surface coverage area of CNTs. If surface coverage area is more than 25% then the wear rate can be reduced considerably [115a]. In the case of nanowear tests performed on aligned CNT-epoxy composite, it was observed that the wear depth increased linearly with the wear cycles [116]. However, the wear depth was much smaller for aligned CNT-epoxy composite than for neat epoxy. The same composite also exhibited a lower friction coefficient than (i) epoxy, (ii) MWCNT reinforced epoxy and (iii) fullerene epoxy composite. It was attributed to the increased strength and stiffness of aligned CNT-epoxy composite. However, authors claimed that before ruling out the positive effects of fullerene more composites need to be prepared using different polymers. The influence of CNT reinforcement on the sliding and rolling friction and wear behaviour of elastomers such as HNBR and fluororubber has also been reported in the literature [117-118]. For such composites 30 phr of MWCNT has been found to give maximum friction and wear performance. It was found that the crosslinking density, hardness, tensile and tear strengths of rubber increased with increasing MWCNT content [117]. The coefficient of friction as well as wear resistance increased with the addition of MWCNT. However, no correlation could be attempted between the increase in mechanical strength and tribological performance in lack of substantial data. In the case of HNBR/Flurorubber blend filled with CNTs, the reinforcement enhanced the friction and wear performance but again it could not be related to the change in the contact angles of the composites [118].

4.2.3. Influence of Mixing and Sonication on CNT Dispersion Most of the authors have reported that CNT/functionalised CNT addition improves the friction and wear performance of polymer composites. However, very few papers have reported on the factors that can improve the homogeneous dispersion of CNTs in matrix. Jacobs et.al [115] have dealt in detail with this issue. According to them, if functionalised CNTs could be mixed with a high energy mixing then agglomerates can be crushed. Though the use of dual asymmetric centrifuge renders the use of the functionalization using acid unnecessary but it leads to high viscosity mixture suitable for press moulding but less applicable for casting. In contrast, functionalization results in low viscosity mixture. Sonication of the mixture improves the dispersion but the effect was found to be insignificant. Similar results have been found in another work in which prolonged sonication of MWCNTs in ultrasonic bath have resulted in introducing defects in nanotubes. Sonication can break them up causing a change in the aspect ratio thereby the entire purpose of utilising its high

Tribology of Composite Materials, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

Tribology of Functionalised Carbon Nanofillers Based Polymer Composites

29

aspect ratio is defeated [119]. Melt processing/extrusion of CNT reinforced polymer composites has been found successful in achieving a reasonably good dispersion of CNTs [120-125]. In a very recent paper ultrasound assisted twin screw extrusion of MWCNT reinforced polyetherimide composite has shown better dispersion results [126]. The homogeneous dispersion of CNTs was verified indirectly using rheological measurements. An increase in complex viscosity and storage modulus of PEI was observed. The results were supported by HRSEM studies which showed that ultrasound helped in the dispersion of CNTs by disintegrating the CNT bundles/agglomerates.

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

CONCLUSION Carbonaceous nanofillers, in particular, carbon nanotubes by virtue of its high aspect ratio have always been thought of as promising filler for polymer composites. In fact it has been found true for most of the polymer composites so far as the enhancement in mechanical properties is concerned. However, to be successful as filler in improving the friction and wear performance of polymer composites leaves much to be desired. Only a few polymers have been studied so far and most of the high performance engineering polymers have been ignored. Even for the polymers studied so far, contradictory results appear in the data regarding the positive synergism between CNT, wear/friction performance enhancing fillers and functionalisation of CNTs. Self-lubricating CNT transfer film on counterface is also debatable [127, 128]. MWCNTs show consistently high friction when aligned normal to the contact plane and very low friction when aligned parallel to the contact surface. The orientation of nanotubes in the films thus affects the magnitude of friction coefficient [129]. Similarly, other than sliding wear mode, the tribological performance of CNT/polymer composites has not been determined in other wear modes. A great deal of research work needs to be done to derive maximum tribological potential of CNTs as effective fillers by resolving the three major issues that still pose a bigger challenge to scientists; firstly, the homogeneous dispersion of CNTs in polymer matrix; secondly optimising the percolation threshold; and finally the functionalisation of CNTs albeit causing minimum harm to environment. The potential of CNTs as filler in fabric reinforced polymer composites is another area which needs to be seriously focussed.

ACKNOWLEDGMENT One of the authors (MF) wish to acknowledge the facilities provided at CIT and MTM, Katholieke Universiteit, Leuven, Belgium for the above work under European commission Erasmus Mundus Exchange program (2009).

REFERENCES [1]

Peter Jost H., Lubrication (Tribology) Education and Research. UK Department of Education and Science, HMSO (1966)

Tribology of Composite Materials, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

30

M. Fahim and J. Paulo Davim

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

[2]

Dowson, D., History of Tribology, Second Edition. Professional Engineering Publishing, (1997) [3] Zum Gahr, K Microstructure and wear of materials, Tribology series 10, Elsevier Science publishers (1987) [4] Hager A.M. and Davies M., in Advances in Composite tribology, Composites Materials Series 8, K. Friedrich (ed) chap 4, Elsevier, Amsterdam (1993) [5] Clauss, F.J., Solid lubricants and self-lubricating solids, Chap 3, Academic press, New York, (1972) [6] Bijwe J. and Fahim M., in Handbook on Advanced functional molecules and polymers, ed H.S. Nalwa, vol. 4, chap 8, Gordon and Breach Science publishers, (2001) [7] Friedrich K., (ed) Friction and wear of polymer composites, Composites Materials Series, 1 (Elsevier, Amsterdam) (1986) [8] Friedrich K., (ed) Advances in Composite tribology, Composites Materials Series 8, Elsevier, Amsterdam (1993) [9] Anderson J.C. and Williamson, P.K., in Polymer Wear and Its Control, ACS Symp. Series, 287 (ed. L.H. Lee), ACS Washington, 315 (1985) [10] Zhang S.W., Tribol. Intl., 31(1-3) (1998) 49 [11] Lancaster, J.K., In: Friction and wear of polymer composites, Composites Materials Series, 1 (ed) K. Friedrich Ch 11, (Elsevier, Amsterdam) (1986) [12] Bhushan B and Gupta B.K., Handbook of Tribology, McGraw Hill, NY (1991) [13] Novel carbons in tribology, Tribology International 37 (2004) 865 [14] Wolverton, M.P., Theberge J.E and McCadden, M.L., Machine Design, 3 (1983) 55 [15] Lhymn C, Wear, 107 (1986) 95 [16] Tevruz T, Wear, 221 (1998) 61 [17] Bijwe J., Neje S, Indumathi J and Fahim M., J. Reinf. Plast. and Comp., 21 (13) (2002) 1221 [18] Jain V.K. and Bahadur S., Wear 123 (1988) 143 [19] Rehbein P and Wallaschek J, Wear 216 (1998) 97 [20] Soutis C, Mater.Sci. Eng. A 412 (2005) 171 [21] Zhao-Zhu Zhang, Feng-Hua Su, Kun Wang, Wei Jiang, Xue-hu Men, Wei-Min Liu Mater. Sci. Engg: A, 404(1-2)(2005) 251 [21a] F.H. Su, Z.Z. Zhang, W.M. Liu, Wear 260 (2006) 861 [22] Rattan R, Bijwe J and Fahim M, Wear, 264, (1-2), (2008) 96; Bijwe J, Rattan R, Wear 262 (5-6) (2007) 749 [23] Rattan R, Bijwe J and Fahim M, Wear, 262 (5-6), (2007) 727; Bijwe J, Rattan R, Tribol. Letts., 23(3) (2006) 223 [23a] Bijwe J, Rattan R and Fahim M, Tribol. Intl., 40 (2007) 844; Rattan R, Bijwe J, Tribol. Letts., 22(1) (2006) 105 [24] Zhao J.G., Liu L, Guo Q.G, Shi J.L. et.al, Carbon 46 (2008) 380 [25] Zhang X, Xianqiang Pei, Zhang J, Qihua Wang, Colloids and Surfaces A: Physicochem. Eng. Aspects 339 (2009) 7 [26] Qi hua Wang, Xin Rui Zhang, Xian Qiang Pei, Mat.and Design, 31(3) (2010) 1403 [27] Zhang X.R., Pei X.Q, Wang Q.H,, Eur. Polym. J. 44 (2008) 2551 [28] Feng-hua Su, Zhao-zhu Zhang, Kun Wang, Wei Jiang and Wei-min Liu, Composites Part A: Applied Science and Manufacturing, 36 (12), (2005), 1601 [29] Paul D.R. and Robeson L.M., Polymer 49 (2008) 3187

Tribology of Composite Materials, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

Tribology of Functionalised Carbon Nanofillers Based Polymer Composites [30] [31] [32] [33] [34] [35] [36] [37] [38] [39] [40] [41] [42]

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

[43] [44] [45] [46] [47] [48] [49] [50] [51] [52] [53] [54] [55] [56] [57] [58] [59] [60] [61] [62] [63] [64] [65] [66] [67]

31

Kuriger R.J., Alam M.K, Anderson D.P, Jacobsen R.L., Comp. Part A, 33, (2002) 53 Schmidt D, Shah D.,Giannelis E.P., Cuur. Opin. Solid State Mater. Sci., 6 (2002) 205 Alexandre M, Dubois P, Mater. Sci. Eng. 28 (2000) 1 Carbon Nanotechnologies, Plastics Addit. Compounds (2003) 12 Dalton A.B., Collins S, Munoz E, Razal J.M. et.al Nature, 423 (2003) 703 Iijima S. Nature (1991) 354(6348), 56. Bonard JM, Kind H, Stöckli T, Nilsson LO. Solid-State Electronics 45(6), (2001) 893 Yakobson BI, Smalley RE. Am. Sci. 85, (1997) 324 Collins P.G., Avouris P. Scientific American; December (2000) 38 Xie S, Li W, Pan Z, Chang B, Sun L. Carbon nanotube arrays. Mater. Sci. Eng. A 2000;286(1):11 Ebbesen TW, Ajayan PM. Nature 1992;358:220 Mamalis A.G., Vogtländer L.O.G., Markopoulos A., Precision Engg. 28 (2004) 16 Fujiwara Y, Maehashi K, Ohno Y, Inoue K, Matsumoto K, Jpn. J. Appl.Phys. 44 (2005) 1581. Melechko A.V., Merkulov V.I., McKnight T.E, Guillorn M.A, Klein K.L., Lowndes D.H., Simpson M.L., J. Appl. Phys. 97 (2005) 041301. Merkulov V.I., Melechko A.V, Guillorn M.A. et.al., Nanotechnology 3 (2003) 66. Jiang G.C, Zheng Q.X, Yang D.Z, J. Appl. Polym. Sci. 99 (2006) 2874 Braun T, Buvari A, Barcza L, Solid State Ionics 74 (1994) 47. Kratschmer W, Lamb L.D, Fostiropoaulos K, Nature 347 (1990) 354. Kroto H.W., Heath J.R, O’Brien S.C, Nature 318 (1985) 162. Feng B, Lubr. Sci. 9 (2) (1997) 181 Bhushan B, Gupta B.K., Van Cleef G.W, Tribol. Trans. 36 (1993) 573. Blau P.J, Haberlin C.E, Thin Solid Films 219 (1992) 129. Mohrbacher H, Celis J.P, Diam. Relat. Mater. 4 (1995) 1267 Yan P.X, Song Q.M, Mater. Chem. Phys. 58 (1999) 195 Novikov V.P., Dymont V.P., Appl. Phys. Lett. 70 (2) (1997) 200 Lettington A.H, Carbon 36 (1998) 555 Krumpiegl T, Meerkamm H, Fruth W, Schaufler C, Erkens G, Bohner H, Surf. Coat. Technol.120–121 (1999) 555 Tu J.P., Zhu L.P, Hou K, Guo S.Y, Carbon 41 (2003) 1257 Tu J.P., Jiang C.X, Guo S.Y, Fu M.F., Mater. Lett. 58 (10) (2004) 1646 Khabashesku V.N. and Pulikkathara M.X, Mendeleev Commun.16(2) (2006) 61 Tasis D, Tagmatarchis N, Bianco A, Prato M. Chem. Rev. 106 (2006) 1105. Koshio A, Yudasaka M, Zhang M, Iijima S. Nano. Lett.1 (2001) 361 Wu W, Zhang S, Li Y, Li J, Liu L, Qin Y, et al. Macromolecules 36 (2003) 6286. Blake R, Gunko YK, Coleman J, Cadek M, Fonseca A, Nagy JB, et al. J. Am. Chem. Soc. 126 (2004) 10226 Vander Wal R.L., Miyoshi K, Street K.W, Tomasek A.J, Peng H, Liu Y, Margrave J.L. and Khabashesku V.N., Wear, 2005, 259, 738. Rinzler A.G, Liu J, Dai H, Nikolaev P, et.al., Appl. Phys. A, (1998), 67, 29 Thess A, Lee R, Nikolaev P, Dai H, Petit P, Robert J, et.al Science, 273 (1996) 483. Nikolaev P, Bronikowski M.J, Bradley R.K, Rohmund F, Colbert D.T, Smith K.A and Smalley R.E., Chem. Phys. Lett., 313 (1999) 91.

Tribology of Composite Materials, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

32

M. Fahim and J. Paulo Davim

[68] Bronikowski M.J., Willis P.A, Colbert D.T, Smith K.A and Smalley R.E, J. Vac. Sci. Technol. A, 19 (2001) 1800. [69] Shofner M.L., Khabashesku V.N. and Barrera E.V, Chem. Mater.,18 (2006) 906. [70] Wang W, Lin Y, Sun Y-P. Polymer 46 (2005) 8634 [71] Tour JM, Dyke CA., Chem. Eur. J. 10, (2004), 812 [72] Park C, Ounaies Z, Watson K.A., Crooks R.E. et.al, Chem. Phys. Letts, 364 (2002) 303 [73] Smith Jr J.G. Watson K.A., Thompson C.M., Cornell J.W., Sci. Adv Mater. Process Eng. Tech. Conf. Ser, 34 (2002) 365 [74] Nyborg W.L.M , Acoustic Streaming, in: W.P. Mason (Ed.), Physical Acoustics, IIB, Academic Press, New York, 1965, p. 355. [75] Noltingk B.E., Neppiras E.A, Proc. Phys. Soc. 63B (1950) 674. [76] Suslick K.S. (Ed.), Ultrasound: It’s Chemical, Physical, and Biological Effects, VCH, New York, 1988. [77] Abramov O, Ultrasound in Liquid and Solid Metals, CRC Press, Boca Raton, FL, 1994, p. 289. [78] Suslick K.S., Didenko Y, Fang M.M., Hyeon T, Kolbeck K.J, Mcnamara W.B., Midleleni M.M., Wong M, Phil. Trans. R. Soc. Lond. A 357 (1999) 335 [79] Konya Z, Zhu J., Niesz K, Mehn D, Kiricsi I., Carbon, 42 (2004) 2001 [80] Ghose S, Watson K.A, Sun K.K., Criss J M., Sochi E. J. and Connell J.W, Comp. Sci Technol., 66 (2006) 1995 [81] Potschke P., Fornes T.D., Paul D.R., Polymer, 43 (2002) 3247 [82] Thostenson E.T., Chou T. J. Phys D: App. Phys. 35(16) (2002), L77 [83] Siochi E, Park C., Lillehei P.T., Rouse J.H. Topping C.T. et.al, Composites Part B, 35 (2004) 439 [84] Huang J.E, Li X.H., Xu J.C. and Li H.L, Carbon, 41(14) (2003) 2731 [85] Yang Z, Dong B, Huang Y, Liu L. et.al, Mater. Letts, 59 (2005) 2128 [86] Jia Z.J., Wang Z.Y. Xu C.L., Liang B.Q. et.al, Mater. Sci. Engg. A, 271 (1999) 395 [87] Emilie J. Siochi, Dennis C., Cheol Park et.al., Composites: Part B 35 (2004) 439 [88] Satyanarayana N, Rajan K.S., Sinha S.K., Shen L, Tribol. Lett. 27 (2007) 181 [89] Ler J.G.Q, Yufeng Hao, J.T.L. Thong, Carbon 45 (2007) 2737 [90] Sandler J, Werner P, Milo S.P. Shaffer, Vitaly Demchuk, Volker Altsta¨dt,Alan H. Windle, Composites: Part A 33 (2002) 1033 [91] Werner P, Altstadt V,. Jaskulka R, Jacobs O, Sandler J.K., Shaffer M.S, Windle A.H, Wear 257 (2004) 1006 [91a] McCook N.L., Hamilton M.A, Burris D.L and Sawyer W.G, Wear, 262 (2007) 1511 [92] Shangguan Qianqian, Cheng Xianhua, J. Rare Earths, 26 (4) (2008) 584 [93] Yijun Shi, Xin Feng, Huaiyuan Wang, Xiaohua Lu, Wear 264 (2008) 934 [94] Zhao-Zhu Zhang, Hao-Jie Song, Xue-Hu Men, Zhuang-Zhu Luo, Wear 264 (2008) 599. [94a] Sui G, Zhong W.H, Ren X, Wang X.Q., Yang X.P, Materials Chemistry and Physics 115 (2009) 404 [95] Wong E.W., Sheehan P.E, Science, 277, (1997) 1971; Daniel Wagner H and Vaia R.A, Materials Today, 2004, 38; Thostenson E.T, Li C, Chou T.S, Comp. Sci. Technol. 65 (2005) 491 [96] Ajayan P.M., Shadler L.S., Giannaris C., and Rubio A., Adv. Mater., 12, (2000), 750; Shadler L.S, Giannaris S.C and Ajayan P.M., Appl. Phys. Lett., 73, (1998), 3842

Tribology of Composite Materials, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Tribology of Functionalised Carbon Nanofillers Based Polymer Composites

33

[97] Burris D.L, Boesl B, Bourne G.R and Sawyer W.G, Macromol. Mater.Eng, 292 (2007), 387; Dasari A, Yu Z.Z., Mai Y.W , Mat. Sci. Engg. R, 63 (2003) 31 [98] Chen W.X, Li F, Han G, Xia J.B, Wang L.Y, Tu J.P and Xu Z.D, Tribol. Letts., 15, (2003) 275 [99] Zoo Y, J. An, D. Lim and Dae-Soon Lim, Tribol. Letts. 16, (2004) 305 [100] Wei Z, Zhao Y, Ruan S.L, Gao P and Yu T.X, Surface and Interface Analyses, 38, (2006), 883 [101] Xue Y, Wu W, Jacobs O, Schadel B, Polym. Testing, 25, (2006) 221 [102] Johnson B.B, Santare M.H, Novotny J.E, Advani S.G., Int. J. Mech. Mater., (2009) doi: 10.1016/j.mechmart.2009.04.003 [103] Kumar S, Rath T, Mahaling R.N, Reddy C.S. et.al., Mater. Sci. Engg. B, 141 (2007) 61 [104] Cai H, Yan F and Xue Q, Mater. Sci. Engg. A364 (2004) 94 [105] LW Q, Lin Y, Hill D.E., Zhou B, Wang W, Sun XF, et.al., Macromolecules, 37(16), (2004) 6055 [106] Liu T, Tong Y, Zhang W, Comp. Sci. Technol. 67, (2007) 406 [107] Ge J.J, Zhang D, Li Q, Hou H, Graham M.J, Dai L, Harris F.W. and Cheng S.Z.D, J. Am. Chem. Soc., 127 (2005), 9984 [108] Satyanarayana N, Rajan K.S.S., Sinha S.K. and Shen L, Tribol. Letts., 27 (2007) 181 [109] Giraldo L.F, Lopez B.L., Brostow W, Polym. Sci. Engg., 49 (2009) 896 [110] Liu L.N, Gu A.J, Fang Z.P, Tong L.F, Xu Z.B, Comp. Part A, 38, (2007), 1957 [111] Fang Z, Liu L, Gu A, Wang X and Guo Z, Polym. Adv. Technol., 20(11) (2009) 849 [111a] Liu L, Fang Z, Gu A and Guo Z, J. Appl. Polym. Sci., 113(6) (2009) 3484 [112] Yang Z, Dong B, Huang Y, Liu L, Yan F and Li H, Mater. Lett., 59 (2005), 2128 [113] Dong B, Wang C, He B.L., Li H.L, J. Appl. Polym. Sci., 108, (2008) 1675 [114] Yang Z, Dong B, Huang Y, Liu L, Yan F and Li H, Mater. Chem. Phys., 94 (2005) 109 [115] Jacobs O, Xu W, Schadel B, and Wu W, Tribol. Letts., 23 (2006) 65 [116] Wang H, Feng J, Hu X, Ng K.M, Polym. Engg Sci., 48 (2008) 1467 [117] Felhos D, Karger-Kocsis J, Xu D, J. Appl. Polym. Sci., 108, (2008) 2840 [118] Xu D, Karger-Kocsis J,. Major Z,. Thomann R, J. Appl. Polym. Sci., 112 (2009) 1461 [119] Lu K.L., Lago R.M, Chen Y.K, et. al, Carbon, 34(6), (1996), 814 [120] Jin Z, Pramoda K.P, Xu G, Goh S.H, Chem. Phys. Lett., 337 (2001) 43 [121] Potschke P, Fornes T.D, Paul D.R., Polymer, 43 (2002) 3247 [122] Potschke P, Bhattacharya A.R., Janke A, Europ. Polym. J, 40 (2004) 137 [123] McNally T, Potschke P, Halley P, Murphy M, Martin D, et.al, Polymer, 46 (2005), 8222 [124] Chen G.X, Kim H.S, Park B.H, Yoon J.S, Polymer, 47 (2006), 4760 [125] Thostenson E.T, Chou T.J., J Phys; D Appl Phys, 35(16) (2002) L77 [126] Isayev A.I, Kumar R, Lewis T.M, Polymer, 50 (2009) 250 [127] Hu X, Men, Z. Zhang Z, Song H.J., Wang K, Jiang W, Comp. Sci. Technol, 68 (2008) 1042 [128] Fahim M et.al, unpublished results (2009) [129] Dickrell P.L, Sinnot S.B, Hahn D.W, Raravikar N.R et.al, Tribol. Letts, 18(1) (2005) 59

Tribology of Composite Materials, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved. Tribology of Composite Materials, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

In: Tribology of Composite Materials Editor: J. Paulo Davim

ISBN 978-1-62100-999-3 © 2012 Nova Science Publishers, Inc.

Chapter 2

TRIBOLOGY OF INJECTION MOLDED THERMOPLASTIC NANOCOMPOSITES Carmine Lucignano and Fabrizio Quadrini∗ Department of Mechanical Engineering, University of Rome Tor Vergata, Italy

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

ABSTRACT Injection molding of high performance polymers is an attractive technology for producing engineering parts. New high performance polymers can be produced by melt mixing of common tribological polymers with nano-fillers. In fact by using nano-fillers, a large reduction of the weight content of the filler is expected, with relevant effects on material performances and weight. Morevoer, thermoplastic nanocomposites may be injection molded as the unfilled polymers. In this study, several nanocomposites were injection molded as well as the unfilled matrices, and several mechanical, thermal and wear tests were performed. There is no strict correlation between the nanocomposite mechanical properties and the tribological behavior; in particular, the wear mechanism is fundamental for defining the wear performances, despite of the nanocomposite strength. In such cases, a very small amount (1 wt%) is sufficient to a remarkable improvement both in mechanical and wear strength.

1. INTRODUCTION Nowadays, the development of advanced materials for tribological purposes is becoming a pressing demand of manufacturing industries for long time applications, without lubrication, and in conditions of cryogenic and elevated temperatures. Engineering plastics are very attractive materials for many applications under oil-less conditions, therefore tribological properties are becoming the limiting factor for the use of high performance polymers in a lot of industrial applications. Many studies have been conducted to evaluate the wear resistance of high performance polymers, as well as to predict their tribological behavior under real ∗ Corresponding author, E-mail: [email protected].

Tribology of Composite Materials, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

36

Carmine Lucignano and Fabrizio Quadrini

operative conditions. Plastic parts can be produced by different processes like extrusion, injection molding, compression molding, depending on the material to work, the expected performance and the part geometry. Moreover, plastic materials are a very large material class, which allows one to find a suitable polymer for nearly every application. Gears, cams, bearings and seals, are typically produced by injection molding of thermoplastic matrix composites (the so called “tribological plastics”). Injection molding is a flexible production technique for the manufacture of complex shaped, thin walled plastic parts with near neat shape. During the process, a melted polymer is forced to flow into a cavity of desired shape and then allowed to solidify under high holding pressure. The molded parts generally show high material and shape modification because of many factors, such as orientation, degradation, thermal and flow stresses, weld lines, crystallization distribution and thermal shrinkage. Stresses are induced by polymer flow during the filling stage and by temperature change principally during the cooling stage. The stresses generally cause the warpage of the molded part and have also an effect on its appearance and properties. Moreover, a large amount of energy and time is lost to initially melt and subsequently flow and cool the material. Despite the disadvantages, the ease and economics of manufacturing complex parts by injection molding are well recognized. Even if it is known that production processes always affect properties of processed materials compared with the unprocessed ones, a bad production process could completely damage all the expected properties. In tribological plastics, micro-fillers are added in a medium or high weight content (up to 40 %wt). By using nano-fillers, a large reduction of the weight content of the filler is expected, with relevant effects on the material performances and weight. Even if nanocomposites have the potential to substitute typical micro-filled thermoplastic materials also for bulk components, their use is still limited by the difficulty in their processing. Thermoplastic nanocomposites may be injection molded as typical reinforced thermoplastics but the effect of the injection molding process on the bulk properties of this class of materials is still under investigation.

2. RECENT RESEARCH DEVELOPMENTS Recent studies have focused on the evaluation of the tribological behavior of several thermoplastic nanocomposites. Chang et al. deepened the sliding wear of nanoparticle filled polyamide 66 composites [1]. They performed sliding tests by means of a pin-on-disk apparatus under different contact pressures, p, and sliding velocities, v. They observed that the addition of 5 vol% of TiO2 nanoparticles in a 5 vol% graphite + 15 vol% short carbon fiber reinforced PA66 resulted in a reduction of the friction coefficient and wear rate. The friction coefficient of the nanocomposite ranged between 0.2 and 0.4 depending on the pv term. Srinath and Gnanamoorthy studied the effect of organoclay on the wear performances of polyamide 6 nanocomposites [2]. Also in this case a pin on disk tribometer was used but no other filler was added to the thermoplastic matrix. The nanoparticle content ranged between 1 and 5 wt% and all the nanocomposites investigated exhibited a low abrasive wear resistance compared with neat PA6. The friction coefficient ranged between 0.3 and 0.4 depending on the nanofiller content whereas the friction coefficient of the neat PA6 was close

Tribology of Composite Materials, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

Tribology of Injection Molded Thermoplastic Nanocomposites

37

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

to 0.5. By increasing the nanofiller content, the friction coefficient decreased. In a further study, Srinath and Gnanamoorthy also investigated the wear performance of this class of nanocomposites in water [3]. In the same year, McCook et al. discussed the tribological results of PEEK nanocomposites in dry sliding in various gas environments. They used a wide variety of available nanoparticles and microparticles (graphite powder, carbon nanoonions, single-walled carbon nanotubes, tungsten disulphide powder, tungsten disulphide fullerenes, alumina nanoparticles, and PTFE nanopowder): as a result the friction coefficients ranged from below 0.05 to above 0.60, whereas the wear rates varied from 10−4 to 10−7 mm3/(Nm). Okhlopkova et al. discussed the tribological and physico-mechanical properties of composites on the base of polytetrafluoroethylene and different grades of nanosize aluminum oxide powders as dependent on the filler content and process conditions [4]. Contemporarily, Carrión et al. studied the physical and tribological properties of a polycarbonate-organoclay nanocomposite [5]. Acetals (i.e. polyoxymethylene, POM) are also interesting tribological materials; in fact POM is widely used for the fabrication of sliding parts or gears in un-lubricated systems, as well as polyamide (PA). Kukureka et al. studied the wear mechanism of POM by means of a twin disc machine: a POM disc ran against a disc of the same material [6]. In similar testing conditions, Rao et al. evaluated the effect of a 20 wt% PTFE in a PA and POM matrix [7]. They inferred that PTFE, by reducing friction, would inhibits crack formation for components such as gears, but no tests were carried out on real gears. Benabdallah studied the friction and wear behavior of POM-based composites using reciprocating, line contact tests against two types of corrosion-protected steel plates [8]. Dealing with POM-steel coupling, Mergler et al. focused on the transfer of POM on steel surface during sliding [9]. In fact, the mechanism of the material transfer onto steel governs the development of friction and wear in the course of time. Latest studies on POM, preferentially deal with material modifications to enhance the tribological behavior (for example by defining new blends [10]).

3. SPECIMEN MOLDING In this study, thermoplastic nanocomposites were produced by melt mixing in a twinscrew extruder, and subsequent injection molding of pellets. High performance polymers, generally used in tribological applications, were chosen for the experimentation: polyamide 6 (Durethan B30S), polyamide 66 (Durethan A31), and polyoxymethylene (Delrin 500). The thermoplastic matrices were mixed with different weight contents of several nano-fillers (silica, alumina, titanium dioxide, and carbon nanotube, CNT). Dog-bone specimens were injection molded by means of an electric press (Fanuc Roboshot S-2000i 50B), and used for thermal analyses, tensile tests, macro-indentation tests, and wear tests. The specimens were 4 mm thick, 10 mm width with a gage-length of 100 mm. The molding parameters were chosen according to the recommended values of the manufacturers: both nanocomposites and unfilled thermoplastics were molded. Not all the combinations of matrices and nano-fillers were considered as only 4 different nanocomposites were prepared by melt mixing: 2 with the PA66 matrix (1 wt% nano-titania filled, 3 wt% CNT filled), and only one for each other matrix (1 wt% nano-alumina filled PA6 and 1 wt% nano-silica filled POM).

Tribology of Composite Materials, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

38

Carmine Lucignano and Fabrizio Quadrini

Figure 1 shows the density of the injection molded specimens together with the appearance of one of them. As the filler density is generally higher than the matrix density (2.6 g/cm3 for silica, 4.2 g/cm3 for titanium dioxide, 3.6 g/cm3 for alumina, and 1.4 g/cm3 for CNT), a density increase is observed for the nanocomposite in comparison with the unfilled polymer. However this increase is very small due to the low filler content: the higher density increase is related to the CNT filled PA66 because of the higher filler content (3 wt%).

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Figure 1. Density of injection molded nanocomposites and unfilled matrices.

Dynamic Mechanical Analysis For thermoplastic nanocomposites, it is very difficult to measure the filler exfoliation and/or dispersion into the organic matrix, whereas it is easier to evaluate the related effect on the composite properties in terms of strength, rigidity or functional properties. For example, viscoelastic properties are strongly dependent on the nano-filler content and dispersion as nano-fillers affect the polymer molecular mobility. In DMA tests, the effect of the nano-filler content on the nanocomposite properties is more evident in terms of storage modulus increase and loss factor decrease. Figures 2 and 3 show the results of a dynamic mechanical analysis (by means of Netzsch DMA 242 C) for POM and PA66, respectively. DMA tests were carried out in a three-point bending mode with the oscillatory frequency of 10 Hz, from room temperature to 150 °C. Samples with the size of 50x10x4 mm3 were extracted from the dog-bone specimen gage length. In Figures 2 and 3, a comparison is shown between the unfilled matrices and their nanocomposite, in terms of storage modulus and loss factor. The effect of the nano-silica filler in the POM matrix is mainly visible in the higher storage modulus at room temperature and in a small change of the loss factor curve: a small peak is visible in the loss factor curve of the unfilled POM at 90 °C whereas a similar peak is observed at higher temperature in the nanocomposite. However, the effect of the nano-filler for POM seems to be very slight. In fact, in the case of the 3 wt% CNT filled PA66 (Figure

Tribology of Composite Materials, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

Tribology of Injection Molded Thermoplastic Nanocomposites

39

2), the effect of the CNT content is quite stronger. The storage modulus of the nanocomposite is higher than the matrix storage modulus in all the temperature range, and also the loss factor is generally lower. The shift of the loss factor peak toward higher temperatures is evident. In a temperature scan, the maximum of the loss factor corresponds to the inflection point of the storage modulus and the related temperature is the glass transition temperature. As the nanofiller content strongly limits the molecular mobility, this effect is mainly observable during the polymer glass transition, and a peak shift occurs.

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Figure 2. Dynamic mechanical analyses of unfilled POM and its nanocomposite.

Figure 3. Dynamic mechanical analyses of unfilled PA66 and its CNT filled nanocomposite.

Tribology of Composite Materials, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

40

Carmine Lucignano and Fabrizio Quadrini

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Differential Scanning Calorimetry A sample was also extracted from the molded specimens to perform differential scanning calorimetry (DSC). The tests were performed by DSC Netzsch 200 PC, from room temperature to above 300 °C at 10 °C/min. For each specimen, one sample was extracted from the same position, near the centre of the dog-bone. Figures 4, 5, and 6 show the effect of the nano-filler on the DSC thermogram of POM, PA6 and PA66 respectively. Apart from the filler typology, nanocomposites exhibit an irregular melting peak. In the case of the nanosilica filled POM (Figure 4), a double peak is clearly visible with a great reduction of the melting heat. For the nano-alumina filled PA6 (Figure 5) many more peaks are visible but the extension of the melting range seems to be comparable with the matrix alone. For PA66, the shape of the melting peak is influenced by the nano-filler typology in terms of peak shape and extension of the melting range (Figure 6). Figure 7 shows the melting heats extracted from the scans of Figures 4, 5 and 6. As expected, for all the nanocomposites, a decrease of the melting heat is always observed as a consequence of the nano-filler addition. In fact nano-fillers are inert and cannot melt during the scan. However, a small reduction was expected because of the small quantity of fillers. Instead, the addition of 1 wt% of nano-silica in a POM matrix results in a 20% reduction of the melting heat in comparison with the unfilled POM. The same amount of nano-alumina determines a 5% reduction of the melting heat of PA6. In the case of PA66, the melting heat reduction is 15 and 25% for the 1 wt% titanium dioxide and the 3 wt% CNT filled nanocomposite, respectively. As the nano-fillers strongly limit the molecular mobility of the matrix, a lower crystallization degree is achieved in the material during cooling. In fact, for crystalline polymers, increasing the mechanical properties by reducing the molecular mobility leads to a loss of the crystal content. Moreover, the mixing stage is fundamental to have the maximum effect from the nano-filler addition. Lower filler contents are more easily spread into the resin matrix whereas at high filler content the nano-fillers tend to aggregate.

Figure 4. DSC scan of unfilled POM and its nanocomposite. Tribology of Composite Materials, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

Tribology of Injection Molded Thermoplastic Nanocomposites

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Figure 5. DSC scan of unfilled PA6 and its nanocomposite.

Figure 6. DSC scan of unfilled PA66 and its nanocomposites.

Tribology of Composite Materials, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

41

42

Carmine Lucignano and Fabrizio Quadrini

Figure 7. Melting heat of unfilled polymers and nanocomposites.

4. MECHANICAL TESTING

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

DSC curves are strongly influenced by the nano-filler content but it is not obvious that a similar effect may be observed also in terms of mechanical properties. Apart from the thermal analyses, it is important to evaluate if an improvement in the mechanical and tribological properties was achieved as well.

Tensile Tests Tensile tests were carried out by means of a material testing machine (MTS Alliance RT/50) at the rate of 10 mm/min. Figure 8, 9, and 10 show the tensile curves of unfilled and filled thermoplastics in the case of POM, PA6 and PA66 matrix respectively. The main effect of the nano-silica content on the tensile behavior of POM is a slight increase of the material rigidity and strength (Figure 8). As a consequence, the material ductility decreases. A similar effect is observed for the nano-alumina filled PA6 (Figure 9). Dealing with the PA66 matrix (Figure 10), the addition of the nano-titania mainly affects the material ductility whereas the yield stress is only minimally influenced. Also CNTs strongly affect the material ductility as well as the related strength: in this case, the reduction of the strain at break is very high in comparison with the unfilled PA66. Tensile strength was extracted from the first maximum of the tensile curves, after the elastic stage, and reported in Figure 11.

Tribology of Composite Materials, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

Tribology of Injection Molded Thermoplastic Nanocomposites

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Figure 8. Tensile test of unfilled POM and its nanocomposite.

Figure 9. Tensile test of unfilled PA6 and its nanocomposite.

Tribology of Composite Materials, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

43

44

Carmine Lucignano and Fabrizio Quadrini

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Figure 10. Tensile test of unfilled PA66 and its nanocomposites.

Figure 11. Tensile strength of unfilled polymers and nanocomposites.

In terms of bulk properties, the effect of the nano-filler content is very poor: nanocomposites show a similar strength and a lower strain at break in comparison with the unfilled matrix. A small increase, about 1%, in the strength was observed for the nano-silica filled POM and the nano-titania filled PA66 whereas a larger increase was measured for the nano-alumina filled PA6, about 8%. Instead, for the 3 wt% CNT filled PA66 the tensile strength exhibited a very large decrease (15%), if considering the small CNT content. An increase in the strain at break was never observed in any nanocomposite. Therefore, at best

Tribology of Composite Materials, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

Tribology of Injection Molded Thermoplastic Nanocomposites

45

case, the effect of nano-filler content on the polymer bulk properties is generally negligible for all the studied combinations of filler and matrix. In fact, tensile properties of semicrystalline polymers strongly depend on the crystallization degree, and all the used matrices (POM, PA6, PA66) are highly crystalline. As the nano-fillers inhibit the polymer crystallization, a negative effect on the tensile properties was generally observed. Particularly, the material ductility was mainly damaged as the mechanism of the crystal orientation during the tensile test is responsible for the high strain at break of crystalline polymers.

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Indentation Tests In order to deepen the effect of the nano-filler content on the nanocomposite mechanical properties, also indentation tests were performed. Indentation tests are able to measure local mechanical properties and are strongly influenced by the polymer molecular mobility [12]. Tests were performed by means of the same material testing machine which was equipped with a WC indenter, having a tip diameter of 2 mm. The test rate was 5 mm/min and the preload 200 N. In order to make a comparison, the indentation pressure at the penetration depth of 0.4 mm was extracted: 8 tests were performed on each specimen and the average of the indentation pressure was acquired. Typical indentation curves are reported in Figure 12 in the case of PA66 and its nanocomposites. The effect of the nano-filler is clearly visible as higher pressures are necessary to indent the CNT filled PA66 in comparison with the unfilled polymer, whereas the nano-titania filled PA66 seems to be easier to indent. Figure 13 summarizes the indentation results in terms of average indentation pressure at the penetration depth of 0.4 mm. For PA6 and POM, an increase of the indentation pressure was observed for the nanocomposites in comparison with the unfilled polymers. The extent of this increase is very low (1-2%), and Figure 13 seems to be very similar to Figure 11 where the nanocomposite tensile strength is reported.

Figure 12. Typical indentation tests of unfilled PA66 and its nanocomposites.

Tribology of Composite Materials, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

46

Carmine Lucignano and Fabrizio Quadrini

Figure 13. Indentation pressure of unfilled polymers and nanocomposites.

A very different behavior was observed for PA66, as the indentation pressure of the nano-titania filled PA66 is lower than the unfilled PA66 (-2%), whereas a large increase was measured for the CNT filled PA66 (+11%). The results for PA66 are opposite to the discussed values of strength. However, apart form the CNT filled PA66, indentation data confirm that nano-fillers provide only negligible changes of the material mechanical properties.

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

5. TRIBOLOGICAL PROPERTIES Nano-fillers may potentially improve functional properties of polymeric materials. Tribological performances of thermoplastics are enhanced by the addition of a small amount of nano-filler. However, the production of bulk nanocomposites does not seem to be the best technological solution because only the part surface is interested to sliding. Moreover, due to the injection stage, a strong difference is present between the skin and the core structure of the molded part, with highly oriented molecules (by extensional flow) in the former and slightly oriented molecules (by shear) in the latter. Finally, the presence of the nano-filler in the polymer bulk may strongly affect the material mechanical properties. Because of all these limitations, the accurate evaluation of tribological performances of injection molded nanocomposites is an important goal to avoid time loss and costs. For example, in this study, tribological polymers (such as POM, PA6, and PA66) were used. These materials are already widely used for the production of sliding parts, and their characteristics are tailored to lead to the best performance in terms of mechanical and wear strength. Particularly, POM, PA6, and PA66 are all semicrystalline polymers and their crystalline nature is an important factor for the definition of their final properties. Unfortunately, nano-fillers reduce the material crystallization degree, resulting in an unwanted loss of performances. This way, if poor tribological properties would be also measured, the production of such nanocomposites by injection molding should be avoided.

Tribology of Composite Materials, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

Tribology of Injection Molded Thermoplastic Nanocomposites

47

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Friction and wear tests were performed by means of a ball-on-block reciprocating tribometer (by CSM Instrument), at room temperature under dry conditions. The samples were cut from the center of the dog-bone specimens. A steel ball was used with a diameter of 6 mm, the reciprocating friction stroke was 6 mm and tests were carried out at the reciprocating sliding frequency of 2.65 Hz, with a correspondent maximum linear velocity of 5 cm/s. Two values for the normal spring-driven load were chosen, 5 and 10 N. The maximum sliding distance was 200 m for the former normal force, and 1 km for the latter. However, all the samples passed the test at the lower normal force with negligible surface alterations, whereas at the higher normal force some tests were interrupted for the excess in the friction force, which means a strong alteration of the sliding surface. Particularly, at 10 N of normal force, the tests of the unfilled PA66, the CNT filled PA66, and the nano-alumina filled PA6 stopped after 200 m; the test of the nano-titania filled PA66 stopped after only 100 m. PA6, POM and its nano-silica filled nanocomposite arrived to the final distance of 1 km. Before and after the wear test at 10 N of normal force, the surface topography of the samples was acquired by means of a 3D surface profiling system (Taylor Hobson Talysurf CLI) equipped with an inductive gauge, and the surface roughness was extracted. Figure 14 shows some typical friction curves in the case of PA66 and its nanocomposites with the normal load of 5 N. It is evident that the material behavior under sliding is quite complex, as some oscillations and sudden rises occur.

Figure 14. Comparison between the friction curves of PA66 and its nanocomposites at 5 N of normal force.

For all these curves, a single value for the friction coefficient was extracted from the average of the friction coefficient curve before the final rise. Oscillations may be related to small surface irregularities which are flattened during sliding, or to some debris which are detached from and re-attached onto the surface. Friction coefficients from all the friction tests are reported in Figure 15. All the samples arrived to the final distance of 200 m, even if in such cases (for example the CNT filled PA66) a strong variation of the friction force occurred near the end. At the lower normal force, a similar friction coefficient, about 0.1, was observed for all the unfilled polymers and nanocomposites. The lowest friction coefficient was

Tribology of Composite Materials, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

48

Carmine Lucignano and Fabrizio Quadrini

observed in PA6 (0.069) whereas its nanocomposite was about 0.075. PA66 and its nanocomposite exhibited higher coefficients, but the highest friction coefficients were measured for the POM (0.21) and its nanocomposite (0.16). However, at the lower friction load, the sliding surfaces remained unaltered and no wear was measured. Therefore, the normal load was increased up to 10 N. By increasing the normal load, the friction coefficient strongly increases for all the tested materials, from an average of 0.1 to 0.4. In fact, due to the higher deformation of the material under sliding, the surface contact between the steel ball and the samples increases and the friction force increases as well. However, different behaviors were observed in dependence of the thermoplastic matrix.

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Figure 15. Friction coefficients from all the friction tests.

Figure 16. Surface aspect after wear test for POM and its nanocomposite.

Tribology of Composite Materials, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

Tribology of Injection Molded Thermoplastic Nanocomposites

49

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Figure 17. Surface aspect after wear test for PA6 and its nanocomposite.

At the higher normal load, PA66 and its CNT filled nanocomposite behaved in the same way, exhibiting the same friction coefficient (0.52) and the same maximum sliding distance (200 m). Instead, at lower normal force, PA66 showed a friction coefficient lower than its nanocomposites. At higher normal force, the lowest friction coefficient was measured for the nano-titania filled PA66 (0.45) but a serious surface damaging occurred in the first 100 m. These results suggest that for thermoplastic nanocomposites low friction coefficients are not necessarily related to low wear. For PA66 samples, the lowest friction coefficient was measured for the nano-titania filled sample which failed very soon under dry sliding at 10 N of normal force. Best results were measured for PA6 which exhibited the lowest friction coefficient both at the lower and at the higher normal load, moreover the test arrived to the final distance of 1 km. Instead, the nano-alumina filled PA6 showed a good behavior under low normal force, but its performance collapses at higher load: the highest friction coefficient was measured between all the samples and the test stopped after 200 m. Good results were also obtained with the POM and its nano-silica filled nanocomposite. In both cases, the maximum sliding distance of 1 km was reached. Even if they showed the highest friction coefficients at low normal force, in terms of the percentage increment, they exhibited the lowest increase by increasing the normal force (about 67% for POM and 88% for nano-silica POM, whereas it was about 250% for PA6). Having a constant friction coefficient under different loads can be considered a very important feature for tribological materials under working conditions. However, the friction coefficient is not sufficient alone to evaluate the quality of a tribological material: Figures 16, 17, and 18 show the aspect of the worn surface at the end of the sliding test with the higher normal load, for POM, PA6, and PA66 samples, respectively. For unfilled POM and nano-silica filled POM (Figure 16), the appearance of the wear scar appears to be smooth and regular. In fact, in both cases the tests never stopped until the maximum sliding distance of 1 km. That is the case of unfilled PA6 (Figure 17) where a small

Tribology of Composite Materials, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

50

Carmine Lucignano and Fabrizio Quadrini

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

narrow track appears on the sliding line. Instead, the wear test of the nano-alumina filled PA6 stopped after 200 m and the surface is evidently damaged by the ball sliding. Material protrusion is visible, probably occurring in the last stages of the sliding test. Making a comparison between PA6 and POM, it is possible that a very good interaction between the POM matrix and the nano-silica leads to good mechanical and wear properties of the nanocomposite, despite of the loss in cristallinity. In fact, an increase in tensile strength and indentation pressure was measured, as well. Also for nano-silica filled PA6 better mechanical properties were measured, but the affinity between matrix and fillers seems to be poor in terms of tribology as an higher friction coefficient was measured. As a result, the surface of the nano-silica filled PA6 is completely damaged by the ball sliding. Probably, a different wear mechanism makes the difference between PA6 and POM nanocomposites: PA6 is a very good tribological material and its nano-silica nanocomposite is worse, POM is worse than PA6 but its nanocomposite is better than the unfilled POM. Therefore, adding a small amount of nano-silica to the POM matrix can be a good way to improve its tribological and mechanical properties.

Figure 18. Surface aspect after wear test for PA66 and its nanocomposite.

Tribology of Composite Materials, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

Tribology of Injection Molded Thermoplastic Nanocomposites

51

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Dealing with PA66, mechanical testing already suggested that PA66 nanocomposites do not lead to a clear improvement in performance. The nano-titania filled PA66 had an slightly higher tensile strength of the unfilled matrix and a slightly lower indentation pressure, whereas the CNT filled PA66 showed opposite trends. In both cases, a lower friction coefficient was never observed. The appearance of the worn surface (Figure 18) confirms that a similar wear behavior was observed: all the three samples were strongly damaged by the test at the higher normal load. The test was stopped after 200 m for PA66 and its CNT filled nanocomposite, whereas the nano-titania filled PA66 ended the test after only 100 m. A strong material protrusion is present together with a periodic profile of the scar, particularly evident in the nanocomposite samples. Despite of the better mechanical properties, PA66 is worse than PA6 in terms of tribological behavior, and unfortunately the proposed nanocomposites were not able to fill the gap. In order to quantify the differences between the worn samples, the wear loss was extracted from the surface maps of Figures 16, 17, and 18 in terms of volume loss (Figure 19). The very good behavior of PA6 is confirmed by the lowest wear loss, as well as the good behavior of the nano-silica filled POM in comparison with the unfilled matrix. The poor behavior of PA66, its nanocomposites, and nano-alumina filled PA6 is confirmed as well. In terms of wear loss, the CNT filled PA66 is better than the unfilled PA66, as a consistent reduction was obtained (about 50%).

Figure 19. Wear loss of unfilled polymers and nanocomposites.

This way, CNTs can be used to improve the durability of PA66 under sliding, if the related loss in ductility is accepted. The reduction of the wear loss of the nano-titania filled PA66, in comparison with the unfilled PA66, cannot be taken into account because of the difference in the maximum sliding distance of the tests. It is interesting to observe that wear loss data cannot be correlated with mechanical data, in terms both of tensile strength (Figure 11) and indentation pressure (Figure 13). It was observed that an increase in tensile strength was not related to an increase in indentation pressure for all the materials, because of the different behavior of the tested material under traction and indentation. An increase in strength or indentation pressure is not related to an increase in wear resistance, as well,

Tribology of Composite Materials, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

52

Carmine Lucignano and Fabrizio Quadrini

because of the importance of the different wear mechanism for the performance of the material under sliding.

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Figure 20. Surface roughness of unfilled polymers and nanocomposites.

As a last remark, the surface roughness was extracted from the worn surface of the maps of Figures 16, 17, and 18, and compared with the surface roughness of the molded specimens. Figure 20 shows the surface roughness data in terms of Sa. As expected, all the molded specimens showed a similar surface roughness, about 0.8 µm. Smaller values (about 0.7 µm) were observed for POM and its nanocomposite. The highest value was measured for the nano-alumina filled PA6 (about 1 µm). Probably the addition of nano-alumina to the PA6 matrix produces surface irregularities which affects the friction coefficient (Figure 15) and the wear resistance (Figure 19). After the test, the surface roughness increases for all the samples. Higher increments were observed in PA66 samples because of the discussed appearance of the worn surface. Lower increments were observed for unfilled POM and PA6. The nanosilica filled POM shows a roughness lower than the unfilled matrix, according to the lower wear loss and friction coefficient. A stronger increase is measured for the nano-alumina filled PA6.

REFERENCES [1]

[2]

[3]

Chang, L., Zhang, Z., Zhang, H. and Schlarb, A.K. (2006), On the sliding wear of nanoparticle filled Polyamide 66 composites, Composites science and technology, 66, 3188-3198. Srinath, G. and Gnanamoorthy, R. (2007), Effect of organoclay addiction on the twobody abrasive wear characteristics of polyamide 6 nanocomposites, Journal of Material Science, 42, 8326-8333. Srinath, G. and Gnanamoorthy, R. (2007), Sliding wear performance of Polyamide 6clay nanocomposite in water, Composites science and technology, 67, 399-405.

Tribology of Composite Materials, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

Tribology of Injection Molded Thermoplastic Nanocomposites

McCook, N.L., Hamilton, M.A., Burris, D.L. and Sawyer, W.G. (2007), Tribological results of PEEK nanocomposites in dry sliding against 440C in various gas environments, Wear, 263, 1511-1515. [5] Okhlopkova, A.A, Sleptsova, S.A., Parnikova, A.G., Ul’yanova, T.M. and Kalmychkova, O.Yu. (2008), Triboengineering and Physicolmechanical properties of nanocomposites based on PTFE and Aluminum Oxide, Journal of friction and wear, 29(6), 466-469. [6] Carrion, F.J., Arribas, A., Bermudez, M.D. and Guillamon, A. (2008), Physical and tribological properties of a new polycarbonate-organoclay nanocomposite, European polymer journal, 44, 968-977. [7] Kukureka, S.N., Chen, Y.K., Hooke, C.J., Liao, P. (1995), The wear mechanism of acetal in unlubricated rolling-sliding contact, Wear, 185, 1-8. [8] Rao, M., Hooke, C.J., Kukureka, S.N., Liao, P., Chen, Y.K. (1998), The effect of PTFE on the friction and wear behavior of polymers in rolling-sliding contact, Polymer Engineering and Science, 38(12), 1946-1958. [9] Benabdallah, H. (2003), Friction and wear of blended polyoxymethylene sliding against coated steel plates, Wear, 254, 1239-1246. [10] Mergler, Y.J., Schaake, R.P. (2004), Huis in’t Veld, A.J., Material transfer of POM in sliding contact, Wear, 256, 294-301. [11] Chen, J., Cao, Y., Li, H. (2006), Investigation of the friction and wear behaviors of polyoxymethylene/ linear low-density polyethylene /ethylene-acrylic-acid blends, Wear, 260, 1342-1348. [12] Guglielmotti, A., Quadrini, F. And Squeo, E.A. (2008), “Macroindentation of polymers”, Polymer engineering and science, 48( 7), 1279-1288.

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

[4]

53

Tribology of Composite Materials, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved. Tribology of Composite Materials, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

In: Tribology of Composite Materials Editor: J. Paulo Davim

ISBN 978-1-62100-999-3 © 2012 Nova Science Publishers, Inc.

Chapter 3

TRIBOLOGY OF COMPOSITE MATERIALS WITH INORGANIC LUBRICANTS Kunhong Hu1,2, Xianguo Hu1,∗ and Ralph Stengler3

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

1. Institute of Tribology, Hefei University of Technology, Hefei, P.R. China 2. Department of Chemistry and Materials Engineering, Hefei University, Hefei, P.R. China 3. Hochschule Darmstadt, University of Applied Sciences, Darmstadt, Germany

ABSTRACT In this chapter, the preparation and tribological properties of the composite materials with inorganic lubricants were reviewed. The so composite materials can be prepared using mechanical mixing, chemical synthesis and technology of coating and film. There have been a lot of studies concerning the tribology of composite materials, especially nanocomposites. According to the results reviewed in the section, the composite materials are generally of better tribological properties than their original materials. The addition of traditional solid lubricants such as MoS2 into the composites can generally reduce both the friction and wear. With carbon nano-tubes it is possible to reduce the frictional forces on polymeric surfaces. The anti-wear capacity of the composites can also be enhanced by their inorganic high-hardness or high-intensity components such as metal, ceramic particles and carbon fiber. Moreover, some tribological mechanisms has been put forward to interpret the tribological behavior of the lubrication composites and their components, such as chemical inertness, rolling friction, elastic deformation, exfoliation, transferring film, debris assembling and so on. However, there is still a lack of systemic tribological mechanisms to illuminate the tribological behavior of the composites, especially in the aspects of micro-tribology and nano-tribology. ∗ Corresponding author, E-mail: [email protected].

Tribology of Composite Materials, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

56

Kunhong Hu, Xianguo Hu and Ralph Stengler

1. INTRODUCTION The research on the composites materials is an important field in materials science and technology. The composite materials is composed of two or more components which are of different chemical and physical properties. The components in the composite materials may compensate for their defects and enhance their merits mutually. The composite materials usually exhibit better physical and chemical characteristics than their any component. As a result, the composite materials have wide applications in optics, magnetism, catalysis and so on. In recent years, the importance of composite materials in tribology has also been paid so much attention, and a serial of methods were developed to prepare the tribological composite materials including mechanical mixing, chemical synthesis and technology of coating and film. The components in the tribological composite materials may offset their defects and enhance their lubrication and antiwear properties mutually. For example, their antiwear capacity can be enhanced by their inorganic high-hardness components such as metal [1,2] and ceramic particles [3]. In addition, some soft solid lubricants are often used to improve their lubrication and antifriction properties [4]. Herein, several selected features concerning the tribological composite materials were surveyed. In the first part, the preparation of composite materials for tribological application were reviewed. The second will focus on the tribological properties of the composite materials, especially nanocomposites.

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

2. PREPARATION OF COMPOSITE MATERIALS FOR TRIBOLOGICAL APPLICATION 2.1. Mechanical Mixing Mixing two or more components is the most simple method to prepare the composite materials for tribological application. A lot of organic and inorganic materials can be mixed mutually with proper treatment process. The most typical is to add inorganic fillers into organic matrix materials, especially plastic materials. The inorganic fillers mainly include high strength ceramic particles [3], carbon fiber [5] and solid lubricants such as MS2 (Mo, W) [4]. The desired composite materials can be obtained usually via a mechanical mixture and thermal process respectively, and a typical mixing method was described as follows.

The mechanical mixture can not generally destroy the structure of the ingredients in the obtained composites. For example, the XRD pattern of polyoxymethylene/molybdenum disulfide nano-balls (POM/MoS2 nano-balls) composites is still of the XRD lines both POM and nano-MoS2 (Figure 1a).

Tribology of Composite Materials, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

Tribology of Composite Materials with Inorganic Lubricants

57

Adapted from Ref. [6]. Figure 1. XRD pattern (a) and TEM image (b) of POM/MoS2 nano-balls composites prepared by mechanical mixture.

Figure 1b is the TEM image of the POM/MoS2 nano-balls composite. As shown in this figure, the structure of MoS2 nano-balls is remained in the nanocomposite and the POM polymer surrounds on the nano-balls. Though the mechanical mixture is a very simple and practical method to prepare the composite materials, it still has the insurmountable shortcoming. The ingredients in the composites are distributed mutually only via the mechanical mixing, which is very difficult to obtain the uniform composites and disadvantages their performances obviously. Thus, some chemical synthesis routes was developed.

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

2.2. CHEMICAL SYNTHESIS 2.2.1. Intercalation Intercalation is a peculiar chemical technique to obtain the composite materials of the layered inorganic materials and organic materials. The typical layered compound is of the strong covalent bond inside layers and the weak Van der Waals force between layers. The relatively large space and weak Van der Waals force between the adjacent layers lead to the feasibility to insert guests into the layered materials. The intercalation hosts include graphite, clay, transition metal sulfide and so on [7-9]. The graphite and MS2 intercalation compounds show potential applications in tribology. There are mainly two graphite intercalation compounds (GIC), i.e. acceptor GIC and donor GIC. The donor GIC includes alkali metals GIC, alkali earth metal GIC, and transition metal GIC, while the Acceptor GIC bronst acid GIC, halide GIC and so on. The GIC can now be prepared by some chemical methods such as Vapor Phase Transportation [10] and electrochemical method [11-18]. The differences of intercalation behavior between MS2 and other transition metals includes the coordination properties of M atoms and the ligand field defined in the interlaminar spaces of the compound [19]. At present, the most effective approach to obtain MS2 intercalation compound is through the exfoliation-restacking method, which may also be described as monomolecular layer technology [20]. In the method, an intercalation compound

Tribology of Composite Materials, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

58

Kunhong Hu, Xianguo Hu and Ralph Stengler

of lithium into MS2 (LixMS2) was first prepared from n-butyl lithium and MS2 under N2. And then, a stabile suspension of monolayer MS2 layers may be formed by adding reagents such as water into LixMS2, which is described as the exfoliation process. The obtained MS2 monolayers in suspension can be restacked by acidification or high-speed centrifugation. Adding different inorganic or organic groups into the monolayer suspension can lead to a number of MS2 intercalation compounds (MS2-IC) in the restacked process [21-24]. The intercalation process of guests into MS2 is described in Figure 2.

Figure 2. Schematic illustration of the intercalation of guest into MS2. +

δ/H − LiC 4 H 9 H 2O MS2 ⎯n⎯ ⎯⎯ ⎯→ MS2-IC. ⎯⎯→ LixMS2 ⎯⎯ ⎯→ [MS2] + H2 + LiOH ⎯Guest

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

The crystal structure and the layer space of MS2 are commonly changed after the intercalation of guests into the Van der Waals gap, leading to the (001) line shift to (002) line in its XRD pattern. For example, the XRD pattern of the intercalation compound of POM/MoS2 shows the shift (See Figure 3a,b). The (001) and (002) lines are at about 8° (dspacing value = 1.118 nm) and at d-spacing value=0.613 nm respectively, which confirms that the spaces of MoS2 gallery was expanded by 0.505 nm with respect to pristine MoS2. Figure 3c represents the microscopy of POM/MoS2 nanocomposites.

Figure 3. XRD pattern of POM/MoS2 (b) and the amplification of (a), and TEM image of (c) POM/MoS2 (and b) [25].

As shown in this figure, MoS2 dispersed in the polymer matrix were still in a layered structure, which is the basis of using for the solid lubricant. The aggregate MoS2 layers with a average thickness of 20 nm could be found in the figure, which was thinner than those of original bulk micro-MoS2. It is also indicated that the restacked MoS2 has a good dispersion

Tribology of Composite Materials, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

Tribology of Composite Materials with Inorganic Lubricants

59

in POM. So it can be affirmed that the in-situ intercalation/polymerization is a good way to avoid the agglomeration of inorganic nanoparticles. The intercalation of polyimide into MoS2 show also similar results. The XRD pattern of polyimide/MoS2 (PI/MoS2) nanocomposite also has the so shift. The main MoS2 line is located at 2θ=14.44°(d=0.613 nm), while the PI/MoS2 intercalation nanocomposite at 2θ=8.306°(d=1.064 nm) (Figure 4). The space expansion in PI/MoS2 intercalation nanocomposite is 0.451 nm.

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Figure 4. XRD patterns of MoS2 (a) PI/MoS2 nanocomposite (b) [26].

Considering structural differences between nano-sized MoS2 particles and bulk microsized MoS2, they possibly has differences in the exfoliation-restacking behavior. The so problem has been involved in the work of Hu et al [27]. MoS2 particles may be arranged in two groups including the closed layered MoS2 and the opened layered MoS2 according to their structures. The closed layered MoS2 such as MoS2 nano-balls is very difficult to be intercalated, and the main diffraction lines of nano-MoS2 such as (002), (100) and (110) is remained in the XRD pattern of its exfoliating resultant, see Figure 5a. The TEM micrograph of exfoliated MoS2 nano-balls in Figure 5c confirms that MoS2 nano-balls were destroyed only partly during the intercalation-exfoliation treatment. This is because the close ball-like structure is of better chemical stability than normal 2H-MoS2 nano-platelets. However, the opened layered MoS2 such as MoS2 nano-platelets can be exfoliated completely. Its exfoliated suspension only represents (hk0) peaks, see Figure 5b, and MoS2 nano-platelets were almost disappeared via the intercalation-exfoliation treatment, see Figure 5d. The exfoliation of MoS2 nano-platelets represents a similar process with bulk micro-MoS2 reported by Gee et al [20,28-30]. n − LiC H

4 9 2 ⎯→ Li/nano-MoS2 ⎯⎯ ⎯ → [nano-MoS2]δ- + H2 + LiOH nano-MoS2 ⎯ ⎯ ⎯

H O

Like bulk MoS2, The exfoliated MoS2 nano-platelets may also be restacked by adding H+ into the monolayer suspension to remove the negative charge on the surface of MoS2 monolayer. However, the restacked product of MoS2 nano-platelets is different from that of the bulk MoS2. The restacked nano-MoS2 maintained the structure of monolayer and did not form layered structures, i.e. nano-MoS2 monolayer restacked irregularly along c axis.

Tribology of Composite Materials, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

60

Kunhong Hu, Xianguo Hu and Ralph Stengler

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Figure 5. XRD patterns of exfoliated suspension of MoS2 nano-balls (a), nano-platelets (b), and TEM micrographs of exfoliated suspension of MoS2 nano-balls (c) and nano-platelets (d) [27]).

Figure 6. XRD patterns of restacked nano-MoS2 (a) and micro-MoS2 (b), SEM image of restacked nano-MoS2 (c), HRTEM image of restacked nano-MoS2 (d), SAED pattern of restacked nano-MoS2 (e) [27].

Therefore, the XRD pattern of the restacked nano-MoS2 is similar to that of the exfoliated nano-MoS2 suspension, as shown in Figure 5b and Figure 6a, in which all (00l) lines and mixed (hkl) lines were absent. However, the bulk micro-MoS2 monolayer restack along c axis and represent obvious (00l) line (Figure 6b). The irregular restacking of nano-MoS2 monolayer may lead to a cellular structure (Figure 6c). The irregular restacking can also be observed in its high-resolution TEM micrograph

Tribology of Composite Materials, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

Tribology of Composite Materials with Inorganic Lubricants

61

(Figure 6d). The selected-area electron diffraction (SAED in Figure 6e represents no (00l) lines and mixed (hkl) lines, which is consistent with the XRD result. The reason was discussed in Ref. [27]. Due to very small sizes, nano-MoS2 monolayer has active-high centers in every direction. The attracting force, which led to the restacking of nano-MoS2 monolayer, is disordered and non-directional leading to cellular structures, which is simulated in Figure 7.

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Figure 7. Schematic demonstration of the restacked processes of exfoliated bulk micro-MoS2 (a) exfoliated nano-MoS2 (b) [27].

The intercalation composites of nano-MoS2 may also be formed by the exfoliationrestacking technology. For example, the inorganic Ni2+ and organic cetyl trimethyl ammonium ion have been intercalated into nano-MoS2. The XRD (001) line concerning the Ni2+/nano-MoS2 composite is located at 2θ = 7.48° (d =1.181 nm), while cetyl trimethyl ammonium ion at 2θ = 6.54° (d =1.350 nm) (Figure 8). The layer space was expanded by 0.566 nm and 0.735 nm with Ni2+ and cetyl trimethyl ammonium ion intercalation respectively. Due the organic cetyl trimethyl ammonium ion is of more size than inorganic Ni2+, the expansion of layer space is larger in the cetyl trimethyl ammonium ion/nano-MoS2 composite than Ni2+/nano-MoS2 composite.

Figure 8. XRD patterns of the intercalation compounds of Ni2+ (a) and Cetyl trimethyl ammonium bromide (b) into nano-MoS2 [31].

Tribology of Composite Materials, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

Kunhong Hu, Xianguo Hu and Ralph Stengler

62

HC

O

O O

O

O

CH2 O

.

O

.

O

HC

CH2 O

O

C

C

.

C

C

n

HC

CH2

PS C60

C

O

.

C

C

C

n

C60

PS

C60

O

Figure 9. Schematic demonstration of the preparation reaction for PS/C60 nanocomposite [37].

2.2.2. Chemical Modification

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Organic and inorganic composite materials with excellent tribological properties can also synthesized by the chemical surface modification of inorganic materials. The composites concerning SiO2, MoS2, PbO, ZnS, TiO2, and LaF3 modified by organic compounds were synthesized respectively [32-36]. The modified inorganic composites showed satisfactory tribological properties, and a typical surface modification in situ was adapted from Ref. [32] and shown as follows.

2H5OH Si(OC2H5 )4 + 2H2O+ nRCH= CH2 ⎯C⎯ ⎯ ⎯→4C2H5OH+

Inorganic materials can also be modified by the polymerization of monomers on their surfaces. The polymerization of styrene on the surface of C60 has been reported, and the result showed that the Lubrication film from the modified C60 was of very low friction coefficient [37]. The polymerization reaction may be described as follows.

2.3. Technology of Coating and Film The methods to form inorganic coatings and films on matrix materials include sputtering [38-40], chemical vapor deposition [41], electrochemical deposition [42-44], and spraying [45-47]. The sputtering technology is a very effective physical vapor deposition (PVD) method for preparation of lubrication film, which main includes radio frequency sputtering [48], magnetron sputtering [49-52], reactive sputtering [53], and ion beam assistant deposition

Tribology of Composite Materials, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

Tribology off Composite Materials M withh Inorganic Luubricants

63

[554]. At presen nt, researches mainly focuss on the MoS S2 and WS2 composite c film ms such as Ti/MoS2 [55], Cr/MoS2 [56], W/MoS2 annd Mo/MoS2 [57], Zr/MoS S2 [58], TiN/M MoS2 [59], C CrN/MoS 0], WC/MoS2 [61], WS2/M MoS2 [48], Graphite/MoS G MoS2 [63], 2 [60 2 [62], PbO/M A Ag/WS a so on. Fiigure 10 show w a typical equipment e forr the preparattion of the 2 [64] and Zr/MoS2 comp posite coatingss, and the miccrostructure of the so obtaiined Zr/MoS2 composite cooatings also prrovided in Figgure 11.

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Fiigure 10. Schem matic diagram of o the depositionn system for thee MoS2-Zr compposite coatings [58].

Fiigure 11. SEM images of the composite c coatinng of Zr/moS2 on o matrix prepaared via the meddiumfrrequency magneetron sputteringg [58].

The thermaal processing method includding hot moullding and hot rolling methood is proper too prepare the plastic-basedd composites materials succh as the thrree-layer self--lubrication coomposite mateerial[65,66] Thhe three-layerr materials is composed c of steel s matrix layyer, copper poowder adhesiv ve and plastic self-lubricatioon layer (Figuure 12), whichh can be prepaared via the foollowing proceess.

Tribology of Composite Materials, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

64

Kunhong Hu, Xianguo Hu and Ralph Stengler

Figure 12. Schematic demonstration of the three-layer composite materials [65].

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Figure 13. Models of different MS2 structures: nano-ball (a); bulk particle (b); nano-slice (c).

3. TRIBOLOGY OF COMPOSITE MATERIALS 3.1. Tribology of Composites with MS2 as Fillers MS2 (M=Mo or W) is one of important fillers in the lubrication materials. The usable MS2 includes bulk micro-MoS2, spherical micro-MS2 [67], ball-like [68,69], fiber-like [70], inorganic fullerene-like [71], tube-like [72] and slice-like nano-MS2 [73]. The fullerene-like MS2 have been used as the fillers in polymers, and the tribological properties of the obtained composite materials have been investigated as well [4]. The close-structural MoS2 nano-balls without the rim-edge surfaces (Figure 13a) are a proper filler in POM. Though the bulk micro-MoS2 is of the rim-edge surface, it has a very low catalytic activity in POM degradation because of its large sizes and is also a proper filler in POM (See Figure 13b). Though the two MoS2 samples, i.e. the structure-closed MoS2 nanoparticles and the structure-opened bulk MoS2, are the proper fillers in POM, they represents different lubrication effect on the POM matrix. The proper added amount of MoS2 in POM is not more than 1.0 wt.% for nano-balls while 0.5 wt% for bulk mciro-MoS2 [66]. As shown in Figure 14a, the friction reduction concerning 1% nano-MoS2 represented a advantage over these for

Tribology of Composite Materials, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

Tribology of Composite Materials with Inorganic Lubricants

65

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

micro-MoS2. According to Figure 14b, the POM plastic layers with nano-MoS2 represented lower wear volumes than these with micro-MoS2. This confirms that adding proper nanoMoS2 balls into the POM plastic can increase the wear resistance and the friction reduction of POM. According to the results from SEM provided in the literature, the change of the worn manner of the POM plastic layer with increasing nano-MoS2 content may be described as follows: deep plough wear → slender plough wear →adhesion wear → acute adhesion and plough wear. However, the wear manner of the POM plastic layer only presented a simpler changing process with increasing micro-MoS2 content: deep plough wear → slender plough wear → being broken. It has been well known that the lubrication mechanism of layered 2H-MoS2 is associated with the shearing and sliding of the MoS2 molecular layers with the weak Van der Waals bonds (See Figure 15). The structure-closed MoS2 nano-balls, like the IF-nanoparticles [7476], the excellent tribological performances may be ascribed to its chemical inertness, rolling friction, deformation, exfoliation and delivery of MoS2 sheets in the contact area (Figure 16 and Figure 17). However, the TEM images (Figure 18) of worn debris only represent part exfoliation of nano-balls without obvious traces of rolling and deformation. Due to POM matrix possessed good intension and hardness, the downward part of nano-ball was still confined in the POM matrix. As a result, the exfoliation of nano-sheets from MoS2 nano-balls in the contact area became a main anti-wear and lubrication manner. The anti-wear and lubrication manner of MoS2 as fillers in POM is described in Figure 19.

Figure 14. Average friction coefficient versus time (a) and the average wear volume (b) [66].

Figure 15. Possible lubrication mechanism of bulk layered 2H-MoS2.

The friction coefficient of POM/nano-MoS2 composites presented a small-range wave in long-time rubbing process. The wave possibly resulted from the alternate forming and peeling of lubrication film come from POM and nano-MoS2. Additionally, according to our Tribology of Composite Materials, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

66

Kunhong Hu, Xianguo Hu and Ralph Stengler

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

experiments, other mechanism of debris assembling for the regular wave of the friction coefficient was also suggested. Figure 20 shows the debris assembling lubrication processes [66]. The Stage I passed through a very short period about 1 minute with a remarkable decline in friction coefficient. Under the selected end-face rubbing manner, the resultant debris could not be removed in time, and so they were able to assemble on the friction surface to form clusters of separation (Stage I). The obtained separation clusters decreased the actual contact area of the friction pairs, and consequently reduced the friction coefficient of the POM/nanoMoS2 sample. After Stage I, the main friction and wear happened on the surface of the separation clusters. As a result, the separation clusters was increasingly worn out in Stage II. Due to these clusters suffered from high loads, they were rubbed into a very flat surface. When the assembling separation clusters were destroyed completely by rubbing against mated steel, the friction coefficient of POM/nano-MoS2 presented an abrupt elevation within a very short period (Stage III).

Figure 16. Structure-closed hollow MoS2 nano-balls and its possible lubrication mechanism.

Figure 17. Structure-closed solid MoS2 small nano-balls and its possible lubrication mechanism. Tribology of Composite Materials, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

Tribology of Composite Materials with Inorganic Lubricants

67

Figure 18. TEM image of debris with worn MoS2 nano-balls under 480 N at 0.8 m/s for 30 min. [66].

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Figure 19. Schematic illustration of the lubrication mechanism of micro-MoS2 (a) and prepared nanoMoS2 (b) [66].

Figure 20. Schematic illustration of the mechanism of debris’ assembling. [66]. Tribology of Composite Materials, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

68

Kunhong Hu, Xianguo Hu and Ralph Stengler

.

Figure 21. TEM image of debris with worn MoS2 nano-balls under 480 N at 0.8 m/s for 30 min.

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

The open layered structure in MoS2 nano-slices [77] is similar with that of bulk MoS2 and also has the shearing and sliding lubrication function (Figure 13c). However, MoS2 nanoslices with the high BET area and active dangling bonds possibly have a degradation effect on the stability of POM polymer.

Figure 22. Friction coefficients under vacuum of POM samples [78].

As a result, the wear mechanism of MoS2 nano-slices as a filler in POM should be the catalytic degradation of POM on the nano-slices (Figure 21). Actually, due to the degradation effect, MS2 nano-slices are difficult to be added to plastics. For example, MoS2 nano-slices can degrade polyoxymethylene (POM) [66] at the thermal process leading to poisonous formaldehyde in the thermal process. This implies that MoS2 nano-slices are not a proper fillers in POM. Moreover, the POM/MoS2 nano-balls composite also shows excellent tribological properties under vacuum environment [78]. Figure 22 provides the results of tribological tests under ≤ 10-3 Pa vacuum. As shown in this figure, the POM with MoS2 nano-balls represented better friction reduction than both pure POM and POM with micro-MoS2 in the friction tests.

Tribology of Composite Materials, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

Tribology of Composite Materials with Inorganic Lubricants

69

3.2. Tribology of Composites with Carbon Nanotube 3.2.1. Introduction of Carbon Nanotube The tribological behavior of plastic parts is one of the most important properties in applications. One major goal of material development is to design tailor-made polymers with specific frictional coefficients. Therefore micro- and nanoparticles are used to influence the surface properties like wear or scratch resistance and frictional behavior [79]. The influences of Multiwall Carbon Nanotubes as fillers for different polymers were investigated in this section. A significant influence on the mentioned macroscopic surface functionalities could be shown. Table 1. Typical Properties of fibers Property Young’s Modulus (GPa) Tear strength (GPa) Length/Diameter ratio

carbon fiber 230 4 1000

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Figure 23. Carbon Nanotube.

Micro- and nanoparticles are used for reinforcement of polymers. But in a lot of cases the bulk properties like modulus or shear strength are not so important. Surface properties or functionalities like wear or scratch resistance are of more importance in some applications. For example Daimler-Chrysler is using a new developed nanopaint to improve scratch resistance of their automobiles [80]. The new 40 µm thick lacquer finish includes ceramic particles with a diameter of 20 nm. Those polymeric composite materials are of increasing importance for many products. It is now possible to produce tailor-made materials for special applications. A broad range of mechanical properties can be achieved. In a lot of cases carbon is used as the appropriate filler. Carbon black is used for improvement of rubbers whereas carbon fibers are used for optimizing mechanical properties like the young’s modulus, tear strength and wear resistance. Even the dynamical properties of composite materials are better. Carbon exists in a number of variations due to the involved atomic bonds. A sp1-bond leads to a linear structure of the carbon, a sp2-bond to planar structures like graphite and the sp3-bond leads to tetrahedral structures like diamond. Fibers are building up networks and give better results in the mechanics. By downsizing the filler to the nanoscale it is possible to reach a new range of influence. This is mainly due to the increased active surface, respectively the increased surface to volume ratio. Carbon shows a whole family of unusual sp²-hybridized modifications like fullerenes and carbon Nanotubes. Carbon Nanotubes show extremely high stability also at

Tribology of Composite Materials, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

70

Kunhong Hu, Xianguo Hu and Ralph Stengler

high temperatures. A lot of interesting applications are possible by using Nanotubes for reinforcement [81]. Table 1 shows the properties of typical fibers in comparison to carbon Nanotubes [82]. Since the discovery of Fullerenes [83] and Carbon Nanotubes [84], a lot of investigations are done with these unusual sp2 carbon modifications. Nanotubes are generally hollow cylindrical structures made up of carbon atoms. They exist in various shapes - cylinders, spheres, cones, tubes and also complicated shapes. Single Walled Nanotubes (SWNT) can be considered as long graphene sheets. Nanotubes generally have a length to diameter ratio of 1000 so they can be approximated as one-dimensional structures. Multi Walled Nanotubes (MWNT) can be considered as a collection of concentric SWNTs with different diameters, lengths, and properties. Carbon Nanotubes are produced by several methods. One possibility is the Arc discharge. In this synthesis technique, vapor is created by an arc discharge between two carbon electrodes with or without catalyst. Nano-tubes self-assemble from the resulting carbon vapor. This method generally produces large quantities of impure material. Another possibility is the production by Laser ablation. A graphite target gets vaporized by a NdYAG laser generating small carbon molecules and atoms which condense to form single walled Nanotubes held together by Van Der Waals forces. SWNT are produced if the target consists of graphite mixed with cobalt, nickel or iron. Although a chemical vapor deposition (CVD) is used to produce MWNTs. Figure 23 shows a Carbon Nanotube [85].

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

3.2.2. Measurement of Surface Functionalities All measurements are done with the Universal Surface Tester (UST®). First it scans along a defined path X on the material’s surface and determines continuously the vertical deflection Z and also the surface profile. Then the same path is once again scanned with the same tip which is loaded additionally with a defined load. The resulting surface profile G represents the local total deformation. At last the surface is scanned once again with the same load as in step 1.

Figure 24. The total deformation G and its elastic and plastic parts E and P could be calculated from the profiles c and e measured with minimum load (0.7 mN) and d with a load range between 1 and 100 mN. Tribology of Composite Materials, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

Tribology of Composite Materials with Inorganic Lubricants

71

The elastic part of the total deformation E is now recovered. Thus the surface profile is established only by the remaining deformation. The difference between the measured profiles results from the determined surface deformations – total deformation G, elastic deformation E and plastic deformation P (Figure 24). Load FN

Measurement of the topology

Sample

Movement and measurement of the frictional forces Figure 25. Experimental setup.

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

3.2.3. Microtribology During the standard measurement the load on the surface is controlled. With a new module it is now possible to measure also the tangential forces during the movement of the tip over the surface. With this additional information of the frictional force it is possible to calculate the coefficient of friction µr. Figure 25 shows the experimental setup for micro friction.

3.2.4. Nanocomposites One of the most critical problems is the dispersion of nanoparticles into the polymer matrix. Most micro- and nanoparticles tend to agglomerate. To prevent this the filler must be chemically functionalized. Nevertheless to change surface properties it is necessary that those particles are close to the surface. Sample production by injection molding lead to very poor results in surface property changes. This may be due to the fact that the high shear forces generate an orientation parallel to the surface. It is known from glass fiber reinforcement, that injection molded parts are covered by a nearly fiber free zone. Using high pressure and temperature to produce test specimen lead to good results. In the case of micro particle fillers high percentages are necessary to see an effect on the surface. Figure 26 shows the reduction

Tribology of Composite Materials, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

72

Kunhong Hu, Xianguo Hu and Ralph Stengler

of frictional force measured by a 5 mm steel ball with 100 mN load. More than 50% of micro particles are necessary to see the reduction of frictional forces. Using nanosize particles only a few percent of filler are necessary to change the surface properties. Quite a lot of particles appear at the surface. Figure 27 shows a comparison of the surface profile measured with an AFM.

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Figure 26. Frictional forces of a micro particle filled epoxy resin.

Figure 27. Epoxy resin without and with nanoparticles.

3.2.5. Experimental Results Three samples of MWNT from Zyvex Cooperation processed in SU-8 epoxy resin with 2, 5 and 10 wt% were investigated. The samples were prepared on glass substrate with a thickness of 1mm. They were cured for 2 days at 60°C. The Carbon Nanotubes can be used to reduce the friction of the surface. An increasing content of the MWNT filler lead to an increase of the surface roughness. Therefore the effective surface contact is reduced. As mentioned above normally a high amount of filler is necessary to modify the coefficient of friction [86]. In the observed case even small amounts of MWNT lead to a decrease in frictional forces. Figure 28 shows the results.

Tribology of Composite Materials, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

Tribology of Composite Materials with Inorganic Lubricants

73

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Figure 28. Frictional forces between the epoxy resin and a 5 mm steel ball.

Figure 29. Frictional force measured along a 10mm path on the surface of the sample.

The same influence could be seen by the investigation of a polyamide filled with up to 10 weight percent of MWCNT. The samples were prepared by compounding and pressing a plate with 3 cm diameter. Figure 29 shows the measured frictional force of a 5 mm steel ball with a 70 mN load over a 10 mm path on the surface. Figure 30 shows the result of average frictional forces between the different samples. In summary, it is also possible to reduce the frictional forces on polymeric surfaces with Carbon Nanotubes. Even a few percent of filler show a significant influence. This may be caused by the high aspect ratio of the MWNT which makes it easier to disturb the ideal polymeric surface. In a next step the effect on wear resistance will be investigated.

Tribology of Composite Materials, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

74

Kunhong Hu, Xianguo Hu and Ralph Stengler

Figure 30. change of frictional forces with MWCNT content.

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

3.3. Tribology of Composites with other Fillers Zinc compounds such as ZnS and ZnO can be used as fillers in lubrication composites. Li et al studied the friction and wear characteristics of nano-ZnO filled polytetrafluoroethylene (PTFE) [87]. The filling of nano-ZnO to PTFE could greatly reduce the wear rate of the nanocomposite and the best anti-wear property was obtained with the nanocomposite containing 15 vol.% nanometer ZnO (Figure 31). The nano-ZnO filler could prevent the destruction of PTFE banded structures during friction process, and a uniform and tenacious transfer film was formed on the friction surface, which promised an excellent anti-wear property of this composite.

Figure 31. Variations of the friction coefficient (a) and wear volume (b) of the PTFE/nano-ZnO composite with sliding duration [87].

Tribology of Composite Materials, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

Tribology of Composite Materials with Inorganic Lubricants

75

The ZnS-filled polyelectrolyte multilayers (PEMs) was synthesized and their tribological properties were investigated on a UMT-2 against stainless steel ball by Yang et al [88]. The ZnS-filled PEMs had higher antiwear lives than unfilled ones (Figure 32). The paper suggested that the ZnS nanoparticles were formed in the PEMs to enhance the load-carrying capacity. An optimum amount of ZnS nanoparticles within PEMs to improve the tribological performances was also found in the paper.

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Figure 32. The variations of the friction coefficients with the sliding time for different thin films sliding against stainless steel ball at 100 mm/min under normal load of (a) 1.0N and (b) 2.0 N [88].

Figure 33. Wear loss versus load of unfilled and graphite filled C-E composites at 3m/s [89].

Graphite is another important fillers in the lubrication composites. Suresha et al studied the wear behavior of graphite filled carbon fabric reinforced epoxy (CE) composites using a pin-on-disc wear tester under dry contact condition [89]. The results show that the excellent wear characteristics were obtained with carbon-epoxy containing graphite as filler. A graphite surface lubrication film was formed on the counter friction surface, which was confirmed to be effective in improving the wear characteristics of the composite. A microcracking and

Tribology of Composite Materials, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

766

Kunhhong Hu, Xiannguo Hu and Ralph R Stenglerr

fiber fracture wear w mechanissms of the caarbon-epoxy composite c wass found in thee literature. king mechanissm was causedd by the progrressive surface damage. Hoowever, the The microcrack w wear of the caarbon-epoxy composite c waas reduced to a greater exxtent by addittion of the grraphite filler (Figure 33), in which weear was domiinated by miccroplowing/m microcutting m mechanisms insstead of microocracking.

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Fiigure 34. Effectt of load on the frictional coeffficient (a) and wear w rate (b) of PTFE and its coomposite w ultrafine diaamond (UFD) [995]. with

Some high h-hardness inoorganic chemiicals are also often used as a fillers to ennhance the anntiwear of com mposites suchh as SiO2, Si3N4, SiC, Al2O3, Cr3C2, ZrO O2 and diamonnd [90-94]. D to holding high hardnesss, the so inorgganic chemicalls can generally not reduce the friction Due obbviously, but they can decrrease the wearr rate remarkaably. This wass confirmed by b Lai et al [995] who investigated that thhe friction andd wear propertties of polytetrrafluoroethyleene (PTFE) filled with ultrrafine diamonnd (UFD) werre studied in detail on a block-onring b w wear tester n significant change in unnder dry slidiing conditions. The resultss showed thatt there was no frriction coefficient, but the wear w rate of thhe PTFE com mposite was redduced consideerably with thhe UFD which h had effects of o loading-carrry and increasiing formation of transfer fillms (Figure 344). The relativ ve wear mechhanism was suuggested in thhe literature thhat UFD particles had a fuunction of rollling bearing in i frictional innterface, and resulted r in chhange of PTFE E frictional foorm from sing gle macromollecular slidingg friction to a mixed form m of sliding and a rolling frriction. Yu et al co ompared the tribological t prroperties of foour high-hardnness ceramic particles p as fillers in polyph henylene (PPS S) including Si S 3N4, SiC, All2O3 and Cr3C2 [93]. It was found that PS. They all thhe used ceramiic particles exxcept Si3N4 couuld increase thhe friction coeefficient of PP deecreased the wear w rate of PPS P within 20 km sliding distance, d and only o Al2O3 inccreased the w wear rate of PP PS after 20 km m sliding. Mooreover, the PPS P with Si3N4 showed exccellent both anntifriction and d antiwear propperties (Figuree 35). Carbon fib ber is also onne of importaant fillers, whhich is often used to enfoorce plastic coomposites. Th hough the adddition of carboon fiber can not n decrease thhe friction coeefficient of thhe plastic com mposite, the anntiwear properrties of the plaastic compositte can be imprroved [96].

Tribology of Composite Materials, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

Tribology of Composite Materials with Inorganic Lubricants

77

The tribological properties of carbon fabric/epoxy (CF/EP) composites under condition of ring-on-ring dry sliding were studied [97]. CF/EP composites which prepared with semi-dry method with 40 vol. % EP exhibited a steady friction coefficient and low wear (Figure 36). The main wear mechanism of pure CF/EP composites was adhesive. Friction-reducing and anti-wear properties of carbon fabric composites can be greatly improved by fillingMoS2 and graphite.

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Figure 35. Effect of load on the frictional coefficient (a) and wear rate (b) of PPS and its composite with ceramic particles [93].

Figure 36. Friction and wear behavior of composite vs. CF cohesive material content [97].

3.4. Tribology of Intercalation Composites T. M. Wang et al reported tribological behavior of Polyimide/MoS2 intercalation composite as an additive in the commercial lithium grease [26]. The intercalation composite could improve the antiwear and antifriction properties of the lithium grease obviously. The

Tribology of Composite Materials, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

78

Kunhong Hu, Xianguo Hu and Ralph Stengler

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

XPS results confirm that the excellent lubrication effect of the intercalation composite was attributed to the formation of a surface protective film composed of FeSO4, MoO3 and iron oxides on the worn steel surface. However, there was a lack of a comparison between the intercalation composite and non-intercalation composite. J. Wang et al [25] reported the tribological behavior of the POM/MoS2 intercalation composite. Figure 37a adapt from this literature showed the relationship between the friction coefficient and loads within 30 minutes friction time. The friction coefficient of the samples with pure POM or POM/MoS2-IC became lower when the load turned to 900N from 800 N. The too high load such as 1000 N leaded to a remarkable increase in the friction coefficient of the sample with pure POM, which indicated the plastic layer had been broken.

Figure 37. Friction coefficients (a) and wear depths (b) of POM and POM/MoS2 intercalation composite [25].

However, the sample with POM/MoS2-IC presented the lowest friction coefficient under 1000 N. This implied the POM/MoS2-IC layer remained the excellent capability of friction reduction under high loads. Figure 37b showed that the wear scar depth of the POM/MoS2-IC layer was smaller than that of the pure POM layer. This indicated that the POM/MoS2-IC layer had better capability of wear resistance. However, This literature did not also take the comparison between the intercalation and non-intercalation composites into consideration. The comparison between intercalation and non-intercalation composite was done in a forthcoming literature [98], which investigated the micro-tribological behavior of POM-based composites with MoS2. Figure 38 adapted from this literature shows that the non-intercalation POM/MoS2 nano-balls had the lowest friction coefficient, while the POM/MoS2-IC the highest one. In addition, the non-intercalation POM/micro-MoS2 also represented a lower friction coefficient than the POM/MoS2-IC. As shown in Figure 38b, the best wear resistance was also found in the POM/MoS2 nano-balls composite.

Tribology of Composite Materials, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

Tribology of Composite Materials with Inorganic Lubricants

79

Figure 38. Friction coefficients (a) and wear resistance (b) of POM/MoS2 composites [98].

There was a similar antiwear resistance between POM/MoS2-IC and POM/micro-MoS2 composites. This indicates that the intercalation compound is not of an observable advantage over the non-intercalation compound, which was ascribed to the transformation of 2H-MoS2 into 1T-MoS2 via intercalation reaction. The transformation enables the intercalation composite to lose the 2H structure, which is of high lubrication function necessary in the MoS2-based lubrication composites.

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Intercalation MoS2(2H, D3H) ⎯⎯⎯⎯ → POM/MoS2-IC(1T, Oh)

3.5. Tribology of Surface-Modified Composites The surface-modified composites have excellent dispersion in oils, and as a result, they may improve the tribological properties of the base oils to a satisfactory extent. Chen and Liu et al investigated the tribological behavior of the oleic acid (OA) coated PbO nanoparticles [99]. The results show that the liquid paraffin (LP) containing the OA-modified PbO nanoparticles have better antifriction and antiwear properties than both the pure (LP) and the lead OA (LOA) compound. The tribological advantage of the OA modified nano-PbO over LP and LOA may be further magnified with the increased load (Figure 39). The corresponding lubrication mechanism may be ascribed to the formation of a deposited boundary lubrication film of PbO on the rubbing surface. Liu and Chen still studied the tribological properties of the ZnS nanoparticles modified by di-n-hexadecyldithiophosphate (DDP) as an additive in liquid paraffin using a four-ball machine [100]. The results show that the DDP-coated ZnS nanoparticles as additive in liquid paraffin is capable to reduce wear of steel even at an extremely low concentration. The lubrication mechanism may be ascribed to the formation of a physically adsorbing and tribochemically reacted film with ZnS, ZnO, FeS, Fe2O3, SO42− and PO43−. The properties of the DDP-capped PbS nanoparticles as an antiwear additive were also investigated by Liu et al [101].

Tribology of Composite Materials, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

800

Kunhhong Hu, Xiannguo Hu and Ralph R Stenglerr

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Fiigure 39. Variattion of the frictiion coefficient (a) ( and wear scar diameter (b)oof the steel balll with load luubricated by LP, LP containingg OA-coated nanno-PbO and LP P containing LO OA respectively [99].

Fiigure 40. The prroperties of the surface-modifiied nano-SiO2 particles p as addiitive in lithium grease [1102].

The results indicate thaat the DDP-P PbS nanopartticles were well-distributed w d with the avverage diametter of 3-5 nm,, which can be well disperssed in minerall oil. The four-ball wear teests show that they can remaarkably improove the antiweear ability of liiquid paraff inn even with 0..05 wt.%. Reecently, Zhangg et al investtigated the anntiwear and looad-carrying capacity c of litthium grease with the surfface-modifiedd nano-SiO2 particles p as addditive [102]. The loadcaarrying capacity of lithium m grease can be b increased by ~75% whhen it containss 0.5 wt.% m modified nano--SiO2, and meanwhile, the wear w amount may m be decreaased by 31% (F Figure 40). The nano-SiO2 particles cann deposit in the boundaryy film and plaay the role of o the selfreeplenish in thee worn surfacee. The dispersion of the suurface-modifieed inorganic particles may be b better than that of the noon-modified particles p in organic oils. Theoretically, T they should have better tribological t prroperties than n their originaal non-modifiied particles. However, theere are still less studies cooncerning the comparison between b the moodified particlles and their original o particlles.

Tribology of Composite Materials, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

Tribology of Composite Materials with Inorganic Lubricants

81

3.6. Tribology of Lubrication Composite Coating and Film

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

The lubrication coating and film have important applications in the self-lubrication composite materials. The composite coating and film are generally composed of selflubrication substances including MoS2, graphite, and WS2 and antiwear components such as Cr, Zr and DLC. Kao found that the properties of MoS2 coatings can be significantly improved through the co-deposition of an appropriate amount of chromium [103]. The literature demonstrated that The MoS2–Cr coatings exhibit improved adhesion strength, microhardness and tribological properties. The formation of stable transfer layers plays a fundamental role in determining the wear mechanisms of the wear pairs. The transfer layer formed from the MoS2–Cr coating contains solid lubricant materials of Mo and S elements, and provides a protective effect which reduces wear and friction. Other study was concerning the tribological properties of the MoS2-Zr coating on YT15 cemented carbide substrates reported by Zhao et al [58]. The MoS2-Zr coating showed a better antifriction and antiwear properties than both pure MoS2 coating (Figure 41). This indicated that suitable level of Zr doping in MoS2 coatings can effectively improve the mechanical and tribological properties. The MoS2-Zr coating, similar the MoS2–Cr coatings, form a transfer film which determine the wear behavior of MoS2-Zr coatings.

Figure 41. Friction coefficients (a) and wear rate (b) of the pure MoS2 coating and the MoS2-Zr nanocomposite coating [58].

The other research focus is concerning the tribological properties of the DLC-based nanocomposite coating, which consist of an unhydrogenated diamond-like carbon (DLC) matrix, hard (WC) and soft (WS2) nanoparticles. The DLC/WC/WS2 nanocomposite coatings can have very low friction over a wide range of environments including humid air, vacuum, nitrogen and cycling conditions [104-106]. Tribological characteristics of DLC-based nanocomposite coatings were also studied by Wu et al [107]. The literature suggested that the formation of lubricious surface tribolayers is responsible for the low frictional behavior of WCS coatings in different sliding environments. The tribolayers are dominated by mixed components including graphitic carbon, WS2 and/or WO3 in varying amounts. The synergistic effect among these mixed components provides low friction during the sliding of WCS coatings in different environments.

Tribology of Composite Materials, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

822

Kunhhong Hu, Xiannguo Hu and Ralph R Stenglerr

Moreover, there are alsoo some researrches concernning the tribollogical properrties of the coomposite coatting consistinng of several soft lubricatiion substances such MoS2, WS2 and grraphite. Yin et e al reported the microstruucture and tribbological propperties of the MoS2/WS2 coomposite lubriication film prrepared by maagnetron co-spputtering methhod (Figure 422) [108]. As coompared with h pure MoS2 film, the MoS M 2/WS2 com mposite film revealed bettter friction sttability, wear durability and load-carryinng capacity, and a lower fricction coefficieent (Figure 433). Other com mposite lubricaation film consisting of MoS2 and graphiite was studiedd by Liu et all [109]. The results show that the fricttion coefficiennt is decreaseed and the wear w rate is inncreased with increasing thee normal loadd. The wear mechanism m of MoS2 /graphiite coatings inn vacuum is fatigue fa under the t lower loadds, while the wear is controolled by a“plaatelet”wear duue to plastic deformation d unnder the higher loads.

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Fiigure 42. The micro-structure m o the MoS2/WS of S2 composite fillms [108].

Fiigure 43. Frictio on coefficient for fo MoS2 film annd MoS2/WS2 composite c filmss at different loaads of (a) 15 N and (b) 25 N [108].

Tribology of Composite Materials, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

Tribology of Composite Materials with Inorganic Lubricants

83

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

CONCLUSION The composite lubrication materials is an important field in the materials science and technology. They can be prepared using mechanical mixing, chemical synthesis and technology of coating and film. The mechanical mixing is the most simple method to prepare the composite materials for tribological application, while the method can not enable the ingredients to distribute uniformly in the composites. The chemical routes are able to overcome the shortcoming, while they are usually more complex. The technology of coating and film may lead to a lot of self-lubrication materials. However, the method need special equipments, which restrains its practical application in common fields. The more efforts should be put into the preparation process of the composite lubrication materials to reduce the complexity and the cost. There have been a lot of studies concerning the tribology of composites, especially nanocomposites. According to the results reviewed in the section, The composite materials are generally of better tribological properties than their original materials. The addition of traditional solid lubricants such as MoS2 and WS2 into the composites can generally reduce both the friction and wear. With Carbon Nanotubes it is possible to reduce the frictional forces on polymeric surfaces. Even a few percent of filler show a significant influence. This may be caused by the high aspect ratio of the MWNT which makes it easier to disturb the ideal polymeric surface. The antiwear capacity of the composites can also be enhanced by their inorganic high-hardness or high-intensity components such as metal, ceramic particles and carbon fiber. Moreover, some tribological mechanisms were put forward to interpret the tribological behavior of the lubrication composites and their components, such as chemical inertness, rolling friction, elastic deformation, exfoliation, transferring film, debris assembling and so on. However, there is a lack of better systemic tribological mechanism to illuminate the tribology of the composites, especially micro-tribology and nano-tribology.

ACKNOWLEDGMENT We wish to thank National Natural Science Foundations of China (Grant Nos. 50905054 and 50475071) for financial support.

REFERENCES [1] [2] [3] [4] [5] [6] [7] [8]

Renevier N M, Fox V C, Teer D G, et al. Surf. Coat. Technol. 127 (2000) 24. Zhou H, Wen Q P, Hao H, et al. Tribology 26 (2006) 183. Wang Q, Xue Q, Shen W, Zhang J. J. Appl. Polym. Sci. 69 (1998) 135. Rapoport L, Nepomnyashchy O, Verdyan A, et al. Adv. Eng. Mater. 6(2004) 44. Shi Y J, Feng X, Wang H Y, Lu X H. Wear 264 (2008) 934. Hu K H, Sun X J, Xu Y F, et al. J. Hefei Univ. Technol. 32 (2009) 615. Akelah A, Moet A, J. mater. Sci. 31(1996) 3589. Whittingham M S, Gamble Jr F R. Mater. Res. Bull. 10 (1975) 363.

Tribology of Composite Materials, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

84 [9] [10] [11] [12] [13] [14] [15] [16] [17] [18] [19] [20] [21] [22] [23] [24] [25] [26] [27] [28] [29] [30] [31] [32] [33] [34] [35] [36] [37] [38] [39] [40] [41] [42] [43] [44] [45] [46] [47] [48] [49] [50] [51] [52] [53] [54]

Kunhong Hu, Xianguo Hu and Ralph Stengler Dines M B. Mat. Res. Bull. 10 (1975) 287. Dresselhaust M S, Dresselhaust G. Adv. Phys. 30 (1981) 139 Shioyama H, Crespin M, Seron A, et al. Carbon 31 (1993) 223. Frackowiak E, Tatsumi K, Shioyama H, et al. Synth. Met, 73 (1995) 27. Kang F, Leng Y, Zhang T Y. Carbon, 34 (1996) 889. Izumi I, Sato J, Iwashita N, et al. Synth. Met. 75 (1995) 75. Horn D, Boehm H P. Mater. Sci. Eng. 31 (1977) 87. Liu D Z, Lan G X, Hu S F, et al. Carbon 30 (1992) 251. Jobert A, Touzain P, Bonnetain L. Carbon 19 (1981) 193. Boehm H P, Heje W, Ruisinger B. Synth. Met. 23 (1988) 395. Benavente E, Santa Ana M A, Mendizábal F, González G. 224 (2002) 87. Gee M A, Frindt R F, Joensen P, Morrison S R. Mater. Res. Bull. 21 (1986) 543. Golub A S, Shumilova I B, Novikov Y N, et al. Solid State Ionics 91 (1996) 307. Kosidowski L, Powell A V. Chem. Comm. 1998, 2201. Powell A V, Kosidowski L, McDowall A. J. Mater. Chem. 11 (2001) 1086. Nakagaki S, Mangrich A S, Wypych F. Inorganica Chimica Acta 254 (1997) 213. Wang J, Hu K H, Xu Y F, Hu X G. J. Appl. Polym. Sci. 110 (2008) 91. Wang T M, Shao X, Wang Q H, Liu W M. Tribology 25 (2005) 322. Hu K H, Hu X G. Mater. Sci. Technol. 25 (2009) 407. Yang D, Sandoval J S, Divigalpitiya W M R, et al. Phys. Rev. B, 43 (1991) 12053. Morrison S R, Frindt R F, Joensen P, et al. US Patent 4822590, 1989. Tagaya H, hashimoto T, Karasu M, et al. Chem. Lett. 12 (1991) 2113. Hu K H, Chen H, Wang Q, Han C L. J. Hefei Univ. 18 (2008), 45. Li X H, Li Q H, Zhang Z J, et al. Tribology 25 (2005), 499. Chen S, Liu W M. Mater. Res. Bull. 36 (2001) 137. Zhang P Y, Qi S K, Xue Q J, Liu W M. Tribology 20 (2000) 211. Xue Q J , Liu W M , Zhang Z J. Wear, 213 (1997) 29. Zhou J F, Wu Z S, Zhang Z J, et al. Wear, 249 (2001) 333. Shi B, Luo J B, Chen D R, Huang L. J. Tsinghua. Univ. (Sci and Tech) 40(2000) 28. Regula M, Ballif C, Moser J H, et al. Thin Solid Films 280 (1996) 67. Noshiro U ,Watanabe S ,Sakurai T, et al. Surf. Coat. Technol. 200 (2006) 5849. Kim S K, Cha B C. Surf. Coat. Technol. 188/189 (2004) 174. Carmalt C J, Parkin I P, Peters E S. Polyhedron 22 (2003) 1499. Yan X B, Xu T, Yang S R, Xue Q J. Tribology 23 (2003) 169. Namba Y J. Vac. Sci. Technol. A 10 (1992) 3368. Wang H, Shen M R, Ning Z Y, et al. Appl. Phys. Lett. 69 (1996) 1074. Gell M, Jordan E H, Sohn Y H, et al. Surf. Coat. Technol. 146-147 (2001) 48. Ye H, Li Q, Wei Z Y, Zhao X J. Tribology 29 (2009) 246. Goberman D, Sohn Y H, Shaw L, et al. Acta Materialia. 50 (2002) 1141. Watanabe S, Noshiro J, Miyake S. Surf. Coat. Technol. 183 (2004) 347. Arslan E, Bulbul F, Alsaran A, et. al. Wear, 259 (2005) 814. Wang X, Xing Y M, Ma S L, et. al. Surf. Coat. Technol. 201 (2007) 5290. Haider J, Rahman M, Corcoran B, et al. Surf. Coat. Technol. 200 (2005) 1080. Fenker M, Balzer M, Kappl H, et al. Surf. Coat. Technol. 201 (2006) 4099. Weiss V, Bohne W, Rohrich J, et al. J. Appl. Phys. 95 (2004) 7665. Zhuang D M, Liu J J, Zhu B L, et al. Wear 210 (1997) 45.

Tribology of Composite Materials, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

Tribology of Composite Materials with Inorganic Lubricants

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

[55] [56] [57] [58] [59] [60] [61] [62] [63] [64] [65] [66] [67] [68] [69] [70] [71] [72] [73] [74] [75] [76] [77] [78] [79] [80] [81] [82] [83] [84] [85] [86] [87] [88] [89] [90] [91] [92] [93] [94] [95] [96]

85

Zhang X H, Liu D X, Gao G R, et al. Tribology 28 (2008): 219. Su Y L, Kao W H. Tribol. Int. 36 (2003) 11. Renevier N M, Fox V C, Teer D G, et. al. Surf. Coat. Technol. 127 (2000) 24. Zhao J L, Deng J X, Song W L. Tribology 29 (2009) 238. Haider J, Rahman M, Corcoran B, et al. Surf. Coat. Technol. 200 (2005) 1080. Kim S K, Cha B C. Surf. Coat. Technol. 188 (2004) 174. Onate J I, Brizuela M, Viviente J L, et al. Trans. Inst. Met. Finish. 2007, 85 (2007) 75. Liu Y, Luo C T, Ye Z Y, et al. Lubrication Engineering, 32 (2007) 131. Han J S, Liu W M, Lu J J, et al. J. Mater. Sci. Eng. 26 (2008) 118. Wang Q, Tu J P, Zhang S C, et al. Surf. Coat. Technol. 201 (2006) 1666. Huang L, Xu D Y, Xu C P, Acta Materiae Compositae Sinica, 16 (1999) 30. Hu K H, Wang J, Schraube S, et al. Wear, 266 (2009) 1198. Zhou X D, Fu X, Shi H Q, Hu Z S. lubr. Sci. 19 (2007) 71. Hu K H, Wang Y R, Hu X G, Wo H Z. Mater. Sci. Technol. 23 (2007) 242. Hu K H, Hu X G, Sun X J, et al. Key Eng. Mater. 353-358 (2007) 2107. Zelenski C M, Dorhout P K. J. Am. Chem. Soc. 120 (1998) 734. Feldman Y, Wasserman E, Srolovitz D J, et al. Science, 267 (1995) 222. Nath M, Govindaraj A, Rao C N R. Adv. Mater. 13 (2001) 283. Hu K H, Hu X G, Mater. Sci. Technol. 25 (2009) 407. Rapoport L, Feldman Y, Homyonfer M, et al. Wear 225-229 (1999) 975. Rapoport L, Bilik Y, Feldman Y, et al. Nature, 387 (1997) 791. Chhowalla M, Amaratunga G A J. Nature, 407 (2000) 164. Hu K H, Hu X G, Jiang P, Int. J. Mater. Prod. Technol. (in press). H K H, Sun X J, Xu Y f, et al. J. Hefei Univ. Technol. 32 (2009) 615. May B, Hartwich M R, Stengler R, et al. Determination of surface functionalities of micro- and nanoparticles modified polymers, ChinaNano, Beijing 2005. N.N. Mikrokosmos auf der Außenhaut, Automobil-Entwicklung 3 (2004) 22 -23. Lahr B. Mechanical Properties of Carbon Nanotube Polymer Composites, Ph.D. Thesis, Trinity College Dublin, Nov 2001. Eitan A. Nanotube – Polymer Composites: Tailoring the interface for improved mechanical properties, Renssellaer Polytechnic Institute, Troy NY April 2002. Kroto H W, Heath J R, O'Brien S C, et al. Nature 318 (1985) 162. Iijima S. Nature 354 (1991) 56. Dresselhaus M S, Dresselhaus G, Saito R. Carbon 33 (1995) 883. Stengler R K, Weinhold W P. New method to determine the influence of micro/nanoparticles on surface and functional properties, ECCM11, Rhodos. Li F, Hu K A, Li J L, Zhao B Y. Wear 249 (2002) 877. Yang G B, Ma H X, Wu Z S, Zhang P Y, Wear, 262 (2007) 471. Suresha B, Siddaramaiah, Kishore, et al. Wear 267 (2009) 1405. Wang Q H, Xue Q J, Shen W C. Tribol. Int. 30 (1997), 193. Wang Q H, Xu J F, Shen W C, Liu W M, Wear 196 (1996) 82. Wang Y, Lim S, Luo J L, Xu Z H, Wear 260 (2006) 976. Yu L G , Bahadur S, Xue Q J. Wear, 214 (1998) 54. Wang Q, Xue Q, Shen W, Zhang J. J. Appl. Polym. Sci. 69 (1998) 135. Lai S Q, Yue L, Li T S, Hu Z M, Wear 260 (2006) 462. Wang H T, Liu W M, Yang S R, Xue Q J, Polym. Mater. Sci. Eng. 13 (1997) 126.

Tribology of Composite Materials, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

86

Kunhong Hu, Xianguo Hu and Ralph Stengler

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

[97] Zhou X H, Sun Y S, Wang W S, Tribology, 28 (2008) 56. [98] Hu K H, Schraube S, Xu Y F, et al. Tribology, (submitted). [99] Chen S, Liu W M, Ou Z W, Xu B S. Tribology 21 (2001), 344. [100] Liu W M, Chen S, Wear 238 (2000) 120. [101] Chen S, Liu W M, Tribology, 17 (1997), 260. [102] Cao Z, Li X H, Zhang Z J, et al. Tribology 25 (2005) 390. [103] Kao W H. Wear 258 (2005) 812. [104] Voevodin A A, O’Neill J P, Zabinski J S, Surf. Coat. Technol. 116-119 (1999) 36. [105] Wu J H, Sanghavi M, Sanders J H, et al. Wear 255 (2003) 859. [106] Wu J H, Rigney D A, Falk M L. et al. Surf. Coat. Technol. 188–189 (2004) 605. [107] Wu J H, Karthikeyan S, Falk M L., Rigney D A. Wear 259 (2005) 744. [108] Yin G L, Huang P H, Yu Z, et al. Tribology, 27 (2007) 41. [109] Liu Y, Luo C T, Ye Z Y, et al. Lubr. Eng. 32 (2007) 131.

Tribology of Composite Materials, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

In: Tribology of Composite Materials Editor: J. Paulo Davim

ISBN 978-1-62100-999-3 © 2012 Nova Science Publishers, Inc.

Chapter 4

MECHANICAL AND TRIBOLOGICAL BEHAVIORS OF NANOMETER AL2O3 AND SIO2 REINFORCED PEEK COMPOSITES Guo Qiang∗ and Pan Guoliang School of Materials Science and Engineering, Shanghai University, Shanghai, China

ABSTRACT

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Polyetheretherketone (PEEK) is a high performance engineering plastics with high mechanical strength, modulus, toughness and wear resistance. For the application in some more severe work conditions, it is necessary to further improve the mechanical properties and wear resistance of PEEK. Nanoparticles reinforced PEEK has attracted more and more attentions and actually plays an important role in the field of modified PEEK plastic due to its unique properties resulting from the nano-scale structure of composites. This chapter reviewed a serial of previous works in our laboratory and focused on the mechanical and tribological properties of nanometer Al2O3 or SiO2 particles filled PEEK composites during which the research on the fretting properties of PEEK composites was novel. Results showed that the incorporation of nanometer Al2O3 or SiO2 with certain diameter and content improved the wear resistance of PEEK greatly under different testing conditions of sliding friction or fretting. Furthermore, some mechanical strength and modulus also increased. By comparison, Al2O3 always performed superior in mechanical properties and tribological properties than SiO2 especially in the fretting wear resistance. It should be noted that the modification of some coupling agents on the surface of nanometer inorganic particles would improve also the dispersion of the particles in PEEK matrix and correspondingly reduce the particles agglomeration, resulted in increase mechanical performance. Therefore, the key factors influencing the mechanical properties and tribological properties were all discussed detail in this chapter, such as the nanometer particle content, diameter, type, material preparation and modification method. In addition, the fracture mechanism and wear mechanism were both investigated through the observation of fractograph and wear surface by SEM with a view to make clear the relations between structure and performance and to provide a instruction for the nanoparticles reinforcing PEEK composites. ∗

Corresponding author, E-mail: [email protected].

Tribology of Composite Materials, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

88

Guo Qiang and Pan Guoliang

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

1. INTRODUCTION The high performance poly(ether-ether-ketone) (PEEK) polymer was first prepared by Bonner in 1962. It is a derivative of poly(aryl-ether-ketones). PEEK is chemically recognized as a linear poly(aryl-ether-ketone) and is a melt processable aromatic polymer; the melting point Tm lies between 330 and 385ºC, depending on the relative proportion of ether-ketone groups linking the phenylene rings. The excellent mechanical properties of PEEK such as high strength, modulus, toughness and wear resistance together with its good thermal stability and chemical inertness make it one of the highest-performance thermoplastics and a potential candidate for dry friction units under severe conditions. In recent years, particulates reinforced polymers attract more and more attentions due to their unique properties resulting from a special rigid-tough structure. It is worth noting that the inclusion of micrometer sized particulates into polymers, high filler content is generally required to bring the above stated positive effects into play. This would detrimentally affect some important properties of the matrix polymers such as processability, appearance, density and aging performance. Therefore, composites with improved performance and low particle contents are highly desired. With this concern, the newly developed nanocomposites, i.e., polymers or metals reinforced by nano scaled fillers would come into the competitive candidates. The extremely high surface area is one of the most attractive characteristics of nanoparticles because it facilitates creating a great amount of interphase in a composite and thereby, a strong interaction between the fillers and the matrix at a rather low nano-filler loading. The incorporation of nanometer particulate fillers into polymer matrix has been proved to be an effective way for improving the mechanical properties of the matrix, as well as wear resistance. The properties of the resulting polymer composites depend on the characteristics, dimensions, and shapes of the inorganic fillers and also on the interfacial bonding strength. It was proposed that the decreasing filler dimension and increasing filler content will significantly improve the specific area of the filler, and in turn it would greatly and effectively improve the transfer of the load between the fillers and the polymer matrix [1]. Generally, the filling modification on PEEK plastic to improve its mechanical and tribological properties involved a lot of different kinds of particulates as follows: Si [2], copper [3], SiO2 [4], SiC [5, 6], Al2O3 [1], MoS2 [7], CuS [8], ZrO2 [9], Al3N4 [10], La2O3 [11], CaCO3 [12], hydroxyapatite [13, 14], potassium titanate [15], PTFE [16], graphite [17], carbon fiber [18-20], poly(ether imide) [21], etc. Among all the normal fabrication methods for polymer nanocomposites, the sol–gel method appears to be the most promising one. The nanoparticles were first dispersed, and then mixed with the polymer gel at the molecular or near molecular level. However, it is well known that PEEK is of good resistance to most organic solvents except concentrated sulfuric acid (95–98 %) and methyl sulfonic acid (CH3SO3H). Accordingly, it is highly unlikely or impossible to fabricate commercially the nanoparticle-filled PEEK composites by means of the sol–gel method. Therefore, it is more feasible to fabricate its nanocomposites through the compression molding technique or injection molding technique instead of the sol–gel method. However, a homogeneous dispersion in a polymeric matrix is a difficult task due to the strong agglomerating tendency of the nanoparticles. The nano-sized silica or alumina fillers

Tribology of Composite Materials, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Mechanical and Tribological Behaviors of Nanometer Al2O3 and SiO2 …

89

reinforced PEEK composites appear to be somewhat agglomeration or clustering in the nanofiller phase. Therefore, how to disperse homogeneously nanoparticles into polymer matrix is a challenge for enlarging the application of nanocomposites [22]. It is well known that the dispersion of nanofillers in the polymer matrix can be improved with the aids of surface modification by chemical reaction or non-reactive modifier. In order to obtain the better dispersion characteristics of the silica nanofillers in the PEEK matrix, it was inspired that the surface modification on the nanofillers should be carried out to impart the uniform filler dispersion and the smaller agglomerates domain [23]. Alkyl silane coupling agents have been applied as the chemically reactive modifiers. Applying the alkyl silane coupling agent, the nanosized silica content can reach as high as 40–50 wt.% with less agglomerates or apparent silica domain in the polymer matrix. Except for alkyl silane coupling agents, alkoxysilane coupling agents can also approach the same work. The coupling agents, such as alkyl silane and alkoxysilane coupling agents, were mostly applied for the sol–gel method to fabricate the inorganic nanoparticle reinforced polymer composites. On the other side, stearic acid has been proposed as a non-interacting surface modifier. It can be considered that the absorbed stearic acid on the surface of the nano sized silica fillers could reduce the interaction between the silica nanofillers, and also lower the size of agglomerates with increasing filler content. The dispersion of nanometer particles in PEEK matrix examined by AFM using a Nanoscope E Atomic force microscope (Digital instrument, USA) at room temperature was reported [24]. Figure 1(a) and (b) show the surface micrographs of specimen with 5 wt.% 15nm Al2O3 particles and specimen with 10 wt.% 15nm Al2O3 particles, respectively. As is shown in Figure 1(a), the white spots in figure are nanometer Al2O3 particles which almost keeping uniform dispersion without any visible aggregating, and the maximal scale of it is about 40 nm. While the maximal scale of Al2O3 particles in Figure 1(b) approaches to 90 nm. Therefore low content of Al2O3 performed superior in the dispersing homogenization than high content of Al2O3 when the Al2O3 content was less than 10 wt.%. As the mechanical properties of particulates reinforced PEEK composite are concerned, it seems that the incorporation of fillers always results in the improvement of some mechanical properties. However, the comprehensive improvement in all mechanical properties is hardly achieved. The filling of hydroxyapatite always results in increase in Young’s modulus while the decrease in tensile strength [25]. When potassium titanate whiskers (PTW) was filled into PEEK, Zhang etc. [15] found the tensile strength and tensile modulus of PEEK composites increased with the increasing PTW content within the used loading range till 30 wt%. In addition, it is revealed that the compounding processes exhibited great influence on the reinforcement efficiency of PTW. The composites pre-compounded with the rheometer possessed higher mechanical performance than those pre-compounded with the extruder. As for some polymer fillers, the filling of ekonol can improve the compressive strength and hardness while at the cost of the decrease of bending strength and impact strength [26]. The filling of PTFE even deteriorated all mechanical properties (except impact strength) [16]. Therefore, to obtain the optimal comprehensive mechanical performance, it is much deserved to study the effects of kind, diameter, content, surface modification and process methods etc. on mechanical properties.

Tribology of Composite Materials, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

90

Guo Qiang and Pan Guoliang

Figure 1. AFM micrograph of dispersion of 15nm Al2O3 particles in the PEEK composites: (a) specimen with 5 wt.% Al2O3; (b) specimen with 10 wt.% Al2O3.

Literatures on the wear resistance of PEEK composite under sliding friction are abundant during which PEEK composites are prepared by incorporating various particulates into PEEK matrix. Certain fillers were selected to meet the requirement of different applications. If the requirements of high wear resistance play an important role in applications, some rigid particles is preferred such as Si, SiC, SiO2, Al2O3 and CuO etc. however, It seems the filling of graphite, MoS2 and PTFE can result in low friction and ultra-low wear at the same time [7, 27-29]. In addition, the effects of operating temperature on the dry sliding friction and wear performance of PEEK composite are also great. As test temperatures increased from below to

Tribology of Composite Materials, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Mechanical and Tribological Behaviors of Nanometer Al2O3 and SiO2 …

91

above the glass transition temperature (Tg) of the composite, wear and friction transitions were seen to occur [30]. As for the tribological properties of PEEK composites under fretting wear, the relative literatures are seldom. Bijwe has reported the fretting wear performance of five injectionmoulded blends of PEEK with PTFE which were evaluated on a pin-on-disc configuration on an SRV Optimol Tester. It was observed that inclusion of PTFE affected the adhesive wear and low amplitude oscillating wear (LAOW) in a beneficial way. With an increase in PTFE contents, coefficient of friction in both the wear modes (adhesive and low amplitude oscillating) decreased but the trends in wear performance differed [16]. Jacobs also reported the wear behaviour of continuous-fiber-reinforced plastics under oscillatory sliding against aluminium counterparts. The amplitude, the frequency and the contact pressure were found to have a critical influence on the fretting wear rate of composites [31]. To investigate the mechanism of high wear resistance of PEEK composites, literatures reported the microstructure of fracture surface and transfer film on steel counterparts were investigated by Scanning Electron Microscope(SEM), Atomic Force Microscope(AFM), Xray Photoelectron Spectroscopy(XPS) and energy dispersive spectrum(EDS), etc. [32-34]. Voort considered it was a mechanical process in which the fragments of the material removed were locked into the crevices of counterface asperities [35]. Guo [36] and Friedrich [37] considered that transfer film layer plays an important role in the load transmission, and it therefore also affects the wear process. The cracking and delamination of the transfer film will also be affected by the bonding of transfer film to the counterface [8]. Lin etc. considered a thin, uniform and tenacious transfer film closely related to tribological behavior contribute to the better wear resistance of the filled PEEK composite [38]. However, it has been argued that the ability of these fillers to reduce wear depends upon their ability to form transfer films of the composites on the counterface which are thin and uniform and are strongly bonded to the substrate [3, 35, 37, 39]. The inclusion of much cheaper (in comparison with CNF or carbon nanotubes CNT) nano SiO2 or Al2O3 particles (with diameters about 15–30 nm) into PEEK is of basic interest for the purposes of processability and mechanical enhancement [40]. The PEEK polymer filled with nano-sized silica or alumina particles measuring 15–30 nm has demonstrated an improvement of elastic modulus and tensile strength by 20–50 % [1]. In addition, the promising high performance nanocomposites reveal a significant improvement in the tribological characteristics, resulting in considerably decreased frictional coefficient and wear rate. It is reported that the incorporation of nano-SiO2 (13 nm) leads to a significant improvement in PEEK matrix stiffness. Nanoparticle-induced molecular or morphological immobilization can be responsible for the enhanced stiffness. After incorporating nano-SiO2, the wear rate of PEEK is significantly decreased and at very low nano-SiO2 content, the composite presents the lowest wear rate. The worn surfaces of the nanocomposites were much smoother than that of pure PEEK. The reductions of perpendicular deformation of PEEK matrix and tangential plastic flow of the surface layer involved into friction were supposed to be important for the modification of the tribological behavior after incorporating nano-SiO2 [4, 41]. The purpose of this chapter is to describe the mechanical and wear properties of nanometer SiO2 and/or Al2O3 filled PEEK composites with different filler proportions. Furthermore, the friction and wear properties of PEEK composites were evaluated under different friction conditions of sliding and fretting. The effects of nanoparticles properties on

Tribology of Composite Materials, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

92

Guo Qiang and Pan Guoliang

the mechanical performance and tribological characteristics of composite were discussed in detail as well as the wear mechanism. In addition, the effects of different coupling agents on the dispersion of particles and structure of composite were also described. It was believed that this work would be helpful for the understanding of the function mechanism of nanometer SiO2 and/or Al2O3 as fillers in PEEK and for providing the guidance of the application of PEEK in tribological field.

2. EXPERIMENTAL METHOD

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

2.1. Material Specimen Preparation Generally, there are two kinds of preparation methods for composite specimens. One is the injection molding and another is compression molding. To prepare a specimen by injection molding, the compounding of composites were first carried out in an internal mixer above the melt point of matrix resin. The mixtures were then granulated and dried prior to the next processing stage. Finally injection molding was performed using an injection-molding machine. To prepare a specimen by compression molding, the compounding of composites were first mixed by mechanical or ultrasonic method and then compressed using a flat vulcanizing machine. Prior to mechanical testing, the composite blocks were cut into specimens of a certain size to meet the experimental request. The fracture surfaces of tensile and impact test specimens were coated with a thin layer of gold and then observed using a Scanning Electron Microscope (SEM) [25]. In this chapter, the PEEK composites were filled with SiO2 and/or Al2O3 and prepared by compression molding method. The type, diameter and content of inorganic nanometer particles, coupling agents and dispersing methods are all listed in Table 1. The preparation process of PEEK composite materials was as follows [24]. Inorganic nanometer fillers were firstly mixed with coupling agents in absolute ethanol, and then blended with PEEK powder, finally prepared by heat compression moulding. Four kinds of dispersing methods were used to blend the mixture with PEEK powder. The first one is dry powder direct mechanical mixing method (DMM), which is to dry the mixed solution of fillers by heating at 110ºC for 3 h, and mixed it with dry PEEK powder by mechanical method. The second one is liquid-solid mechanical dispersing method (LSMD), which is firstly to blend the mixed solution with PEEK powder by mechanical dispersion directly, then to filtrate and dry it at 110ºC for 8h. The third one is ultrasonic dispersing method (UD), which is the same with LSMD except that the mixed solution and PEEK powder was dispersed by ultrasonic for 1.5 h. Absolute ethanol was selected as liquid medium, and the mixture was stirred by a glass stick continuously during the ultrasonic dispersion. The last one is ball milling dispersing method (BMD), which is also the same with LSMD except that it was blended in a ball mill for 1.5 h. Some Si3N4 balls of 2-6 mm in diameter were set in the ball mill and absolute ethanol was also used as liquid medium. During the molding process, the material was heated to 340ºC under the pressure of 20 MPa and kept for 30 min, then heated to 365ºC, followed by cooling to 100ºC in the mould while keeping the pressure, finally opened the mould and cooled to room temperature.

Tribology of Composite Materials, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

Mechanical and Tribological Behaviors of Nanometer Al2O3 and SiO2 …

93

Table 1. Composition of nanometer Al2O3 or SiO2 / PEEK composites and dispersing methods

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Particle PEEK PEEK Al2O3 Al2O3 SiO2 Al2O3 Prescription PTFE, Dispersing surface Φ250μm, Φ50μm, Φ15nm, Φ90nm, Φ500nm, Φ12nm, code wt.% method treating wt.% wt.% wt.% wt.% wt.% wt.% agent D000

100

―

―

―

―

―

―

―

―

D121

95

―

5

―

―

―

―

DMM

Titanate

D221

95

―

―

5

―

―

―

DMM

Titanate

D321

95

―

―

―

5

―

―

DMM

Titanate

L120

95

―

5

―

―

―

―

LSMD

―

L121

95

―

5

―

―

―

―

LSMD

Titanate

L122

95

―

5

―

―

―

―

LSMD

L123

95

―

5

―

―

―

―

LSMD

L131

90

―

10

―

―

―

―

LSMD

Stearic acid Sodium stearate Titanate

L424

95

―

―

―

―

5

―

LSMD

Silane

L434

90

―

―

―

―

10

―

LSMD

Silane

L141

―

95

5

―

―

―

―

LSMD

Titanate

U141

―

95

5

―

―

―

―

UD

Titanate

B141

―

95

5

―

―

―

―

BMD

Titanate

D005

90

―

―

―

―

―

10

DMM

―

D125

85

―

5

―

―

―

10

DMM

Titanate

D225

85

―

―

5

―

―

10

DMM

Titanate

―

5

―

10

DMM

Titanate

―

―

―

10

DMM

Titanate

D325

85

―

―

D135

80

―

10

The first letter in prescription code, such as D, L, U and B refers to various dispersing methods of DMM, LSMD, UD and MD, respectively. Four kinds of coupling agents, such as titanate, stearic acid, sodium and silane, were used to modify the surface characteristic of particle fillers.

2.2. Methods for Mechanical Properties Testing To investigate the effects of the type, size and content of fillers, dispersing methods and coupling agents on the mechanical properties of composites, specimens of D000, D121, D221, L120, L121, L122, L123, L131, L424, L434, L141, U141, B141 were selected to be tested. The tested mechanical properties of composites include tensile strength, flexural strength, compressive strength, bending strength and impact strength, as well the corres-

Tribology of Composite Materials, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

94

Guo Qiang and Pan Guoliang

ponding modulus. In addition, ball indentation hardness and elongation at break are also important mechanical properties of composites. Tensile and compressive tests were carried out on a universal testing machine (CSS441000) under ambient condition at a nominal strain rate of 1.5 mm·min-1 and 1 mm·min-1, respectively. The dimension of tensile specimens meets the requirement of Chinese National Standard Testing Methods GB/T1041-92. Impact strength of the samples was measured with XCJ-4 Impact Tester at room temperature according to GB/T16420-1996. Ball indentation hardness of the samples was measured with PHBI-625 Plastic Ball Indentation Hardness Tester according to GB2298-82. Flexural tests were carried out on a universal material testing machine (INSTRON 1195, INSTRON Corporation, UK) according to GB/T 16419-1996, and the specimens were tested with a crosshead speed of 1.0 mm/min. The flexural strength was calculated as Eq. 1 [24]. σ f = 1 . 5 L 0 p ( BH 2 ) − 1

(Eq. 1)

where σf is flexural strength, L0 is the span between two acting points on specimen, P is the force loaded on specimen according to a given flexibility, B is the width of specimen and H is the thickness of specimen. A JXA-840A scanning electron microscope (SEM) was used to evaluate the fractograph of tensile testing specimens. The fractured surfaces were cleaned with pure alcohol to eliminate impurities and coated with a thin evaporated layer of gold to improve conductivity before examination.

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

2.3. Methods for Sliding Friction Testing As far as methods testing the sliding friction characteristics of polymer matrix composites are concerned, three methods were always used in literatures, i.e. pin-on-disc tribometer [16], ball-on-disc tribometer [42]and ring-on-block tribometer [26]. When pin-on-disc tribometer was used, the polymer pin was abraded in a single-pass condition against the water-proof silicon carbide (SiC) abrasive paper. The paper was fixed on the disc rotating with a constant speed. Various loads were applied on the pin keeping other operating parameters such as speed and abrading distance constant. The specific wear rate K0 was calculated from the following equation [43]. k 0 (m

3

/ Nm ) =

Δm

ρ Ld

(Eq. 2)

where △m was the weight loss in kg, ρ the density in kg/m3, L the load in N and d the distance abraded in m.

Tribology of Composite Materials, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

Mechanical and Tribological Behaviors of Nanometer Al2O3 and SiO2 …

95

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Figure 2. The schematic diagram of a frictional couple during sliding friction and wear testing.

As for the ball-on-disc tribometer, the ball was loaded on a rotating aluminum disc through a steel stick, where the polymer specimens were deposited on the aluminum discs. Through modify the diameter of disc, different sliding speed can be obtained. It was used to systematically investigate the effects of the applied load and sliding velocity on the friction and wear behavior of the coatings. The friction force data were simply divided by the applied loads to give the friction coefficients. The wear rates represent the worn volumes per unit of the applied load and of the sliding distance. The cross-section areas can be obtained from the section profile of the wear tracks measured using a profilometer, and then multiplied by the length (perimeter) of the wear tracks to give the total worn volumes. To study the effects of the type and size of fillers on the tribological properties of composites under sliding friction, specimens of D000, D121, D221, D321, D005, D125, D225, D325 were tested on an Amsler model friction and wear tester, as shown in Figure 2 [43]. Sliding was performed under ambient conditions over a period of 2 h at a sliding speed of 0.42ms-1 and under a load of 196 N. The ambient temperature was around 18ºC and the relative humidity was 50 %. Before each test, the AISI 1045 carbon steel ring (hardness of HRC50-55) and PEEK or its composite blocks were abraded with no. 900 water abrasive paper, and then both steel rings and the PEEK composite blocks were cleaned with cotton dipped in acetone followed by drying. At the end of each test, the blocks were cleaned in the acetone and weighed. For each experimental condition, three tests were carried out and the average result reported. The widths of the wear traces were measured using a 15 J model microscope. The volume loss was calculated from Eq. 3 [43].

Tribology of Composite Materials, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

96

Guo Qiang and Pan Guoliang V s = 7 ⋅ [( π r 2 180 ) ⋅ arcsin( B 2 r ) − B

4r 2 − B 2 4]

(Eq. 3)

where Vs is the volume loss in cubic millimeters, B is the width of the wear traces in millimeters, R is the 20 mm radius of the steel ring and “7” is the width of the specimen in millimeters. The wear coefficient w was calculated from Eq. 4 [43]. Ws = Vs L ⋅ F

(Eq. 4)

where Vs is the volume loss in cubic millimeters, L is the sliding distance in meters and F is the applied load in Newton. The friction coefficient m was calculated from Eq. 5 [43].

μ = M /( r ⋅ F )

(Eq. 5)

where M is the friction moment in Nms, r is the 0.02 m radius of the steel ring and F is the applied load in Newton. The morphologies of the wear traces were observed using a JXA840A model scanning electron microscopy. The chemical state of the elements in the transfer film was analyzed on a PHOENIX model energy-dispersive spectrometry made by EDS.

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

2.4. Methods for Fretting Testing Fretting wear is a kind of reciprocating sliding wear, sometimes called Low amplitude oscillating wear (LAOW) [44]. Generally, during the fretting testing a polymer pin oscillated against the counterface of mild steel disc. Interrupted mass measurements are used to quantify wear rather than dimensional measurements of pin height because of sample creep and thermal expansion. These wear measurements are made periodically during each test. The uncertainty intervals on wear rate data represent the experimental uncertainty in the measurement while the confidence intervals on friction coefficient data represent the standard deviation of the friction coefficient for the entire test [27]. In the case of fretting wear, all selected parameters including amplitude are set before the experiment starts. Once the oscillation starts, generally, amplitude falls below the set value and it is reset with the control panel. However, when load is very large and shearing forces are high, amplitude cannot be maintained even after applying maximum range provided in the control unit. This shows that the limiting load has reached. Sometimes amplitude reduces because of trapping of large particles of metal or polymer and falls to a low value for a short time. It regains once the particle either is thrown off or gets “wear thinned”. However, if the value falls below 80 % of original value for a very long time even after applying full possible force through control unit, it is realized that the material has reached the “failure limit” and cannot withstand the selected load. The experiment is abandoned and lower load is selected for the next experiment. In present work, specimens of D000, D121, D221, D321, L121, L131, L424, L434, D005, D125, D225, D325, D135 were selected and their fretting wear rates were determined with a view to investigate the effects of the type, size and content of fillers on the tribological properties of composites under fretting [45].

Tribology of Composite Materials, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

Mechanical and Tribological Behaviors of Nanometer Al2O3 and SiO2 …

97

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

The experimental apparatus for fretting wear is outlined in Figure 3, which consisted essentially of an eccentric wheel through which the controlled amplitude could be produced [45]. The upper specimen is an Φ 10 mm steel ball made of AISI 1045 treated in a vacuum furnace to a surface hardness of 20HRC, and is fixed on a holder. The normal load on steel ball is 19.6 N applied by a load system with standard weights. The lower specimen is a block with dimensions of 10 mm × 10 mm × 12 mm made of the filled PEEK composites, which is processed by grinding, polishing, cleaning, and then drying in a desiccator for 30 days prior to testing, finally, is mounted on a vibration platform to deliver reciprocating movement at a given amplitude. The work frequency is 27 Hz with testing period of 2 × 105 times.

Figure 3. Fretting wear tester. (1) weight loading; (2) friction coefficient strain gage; (3) vibration platform; (4) exciter; (5) eccentric wheel; (6) lower specimen; (7) upper specimen (steel ball).

The amplitude is set at 500 μm through adjusting the phase difference of two eccentric wheels. Experiments were conducted with no lubricant in laboratory air at room temperature and relative humidity of about 80 %. Due to the shape of fretting scars approximates to an ellipse, the scar areas on the specimen and steel ball, S1 and S2, respectively, can be calculated through measuring the major axis and minor axis of the ellipsoidal scar using an optical microscope with micro scale. For each experiment, three parallel tests were performed and the average result reported. SEM and EDS were also used for the relevant examination and the specimen was coated with a thin evaporated layer of gold to improve conductivity prior to examination. According to the result of earlier paper [43], the filling of Al2O3 powder in PEEK has little influence on the friction coefficient under dry sliding, the variation of friction coefficient was omitted in this work but the wear loss during fretting was discussed.

Tribology of Composite Materials, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

98

Guo Qiang and Pan Guoliang

3. MECHANICAL PROPERTIES OF PEEK COMPOSITE According to previous literatures, the mechanical properties of composites were always closely related to their wear characteristics. Through investigation on short fiber reinforced PEEK composites, Sinmazcelik etc. [46] found that the tribological properties can be improved by enhanced mechanical properties of polymers and reduced adhesion in the contact area. The incorporation of particles and fiber can always reinforce some strength and modulus of composite while deteriorated another. Therefore a comprehensive improvement of mechanical performance is difficult to obtain. To prepare a product met the special requirement, it is necessary to study the effect of diameter, content, preparation method and modification/treating method on mechanical performance. As for PEEK composites filled with Al2O3 or SiO2, the mechanical performance and relative factors were discussed in next several paragraphs.

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

3.1. Effect of Filler Nanoparticle Size on Mechanical Properties and Fracture Mechanism As is shown in Figure 4(a) and (b), specimens filled with 15 nm Al2O3 particles and dispersed by DMM possesses maximal tensile and impact strength, which increased by 5 % and 5.6 times than neat PEEK, respectively [24]. However, compressive and flexural strength would increase as the diameter of Al2O3 particles rises. Compressive and flexural strength of specimen filled with 90 nm Al2O3 increased by 21 % and 17 % than that of neat PEEK, respectively. Nanometer particles possess large specific surface area, high surface activity and better interactivity with the polymer chain segment in comparison with normal size particles, so the filling of it could improve the toughness, rigidity and strength of composites. Meanwhile the rigid inorganic particles in polymer would lead to the concentration of stress, then easily resulting in more microcracks and more absorption to impact energy. In addition, inorganic particles could interrupt and delay the spread of microcracks or stop its transformation to the fracture crack. With the increase of diameter, the specific surface area of inorganic particles would decrease, then lead to the weakening of interaction between inorganic particles and polymer, finally would result in the decrease of tensile and impact strength. However, compressive and flexural strength would increase with the improvement of material rigidity. Representative SEM micrographs of the tensile fracture surface of D000, D121 and D221 at various magnifications are all shown in Figure 5 [24]. Dimples are distinctly visible in the micrograph of D000 (see Figure 5(a)). The high magnification micrograph of one of the dimples in Figure 5(a) is shown in Figure 5(b). The bottom of the dimple appears smooth but some distinctly visible radial strias can be found on the edge (see Figure 5(b) and (c)). The fractograph of D121 appears rough multilayer structure with distinct edges like mica (see Figure 5(d) and (e)). The multilayer structure could increase the area on which the tensile force acts, which could account for the raise of tensile strength. And one of the layers is composed of abundant of smooth grains in high magnification micrograph as shown in Figure 5(f). The micrograph of D221 also appears multilayer structure analogous to that of D121, but the number of layers is fewer (see Figure

Tribology of Composite Materials, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

Mechanical and Tribological Behaviors of Nanometer Al2O3 and SiO2 …

99

5(g) and (h)), which indicates that tensile strength would not increase as the diameter of Al2O3 particles rises.

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

a.

b. Figure 4. Effect of diameter of nanometer Al2O3 particles on the mechanical properties. (a) tensile, compressive and flexural strength; (b) impact strength and ball indentation hardness.

Tribology of Composite Materials, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

Guo Qiang and Pan Guoliang

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

100

Figure 5. SEM micrographs with different magnification of tensile fracture surface for neat PEEK and composites filled with various diameters Al2O3 particles. (a), (b) and (c) D000; (d), (e) and (f) D121; (g) and (h) D221. Tribology of Composite Materials, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Mechanical and Tribological Behaviors of Nanometer Al2O3 and SiO2 …

Figure 6. Effect of content of various nanometer particles on the mechanical properties. (a) tensile, compressive and flexural strength; (b) impact strength and ball indentation hardness.

Tribology of Composite Materials, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

101

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

102

Guo Qiang and Pan Guoliang

Figure 7. SEM micrographs with different magnification of tensile fracture surface for PEEK composites filled with various contents Al2O3 or SiO2. (a), (b) and (c) L121; (d), (e) and (f) L131; (g) L424.

Tribology of Composite Materials, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

Mechanical and Tribological Behaviors of Nanometer Al2O3 and SiO2 …

103

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

3.2. Effect of Filler Content on Mechanical Properties and Fracture Mechanism Tensile, impact, compressive, flexural strength and ball indentation hardness as the function of content of inorganic nanometer particles are all shown in Figure 6 (a) and (b) [24]. All mechanical properties of specimens L121 and L131 which dispersed by LSMD are superior to that of D000, except that tensile strength of L131 is lower than that of D000. Furthermore, tensile, compressive and impact strength of L121 are optimal of all and the impact strength of L121 is even eight times that of D000. But the higher flexural strength and ball indentation hardness of L131 are also presented compared with L121. In addition, all the mechanical properties of nanometer SiO2 reinforced PEEK are superior to that of D000 except for the tensile strength of L424 and L434, but the improvement of it is worse than that of nanometer Al2O3 reinforced PEEK. Analogous with specimens filled with nanometer Al2O3 particles, tensile, compressive and impact strength of L424 are superior to that of L434, except that other two mechanical properties of L434 are better than that of L424, even the ball indentation hardness of specimens L434 is highest of all. Representative SEM micrographs of the tensile fracture surface of L121, L131 and L424 at various magnifications are all shown in Figure 7, respectively [24]. The micrograph of L121 also appears multilayer structure but the drop in level is greater than that of D121, which would account for the higher strength in tensile test (see Figure 7 (a)). The edge of one of the layer shown in Figure 7 (a) is composed of smooth grains, which are less than 100nm in diameter (see Figure 7 (b) and (c)). However, there is a big smooth area in the middle of the micrograph of L131 as shown in Figure 7 (d), which indicates more brittle fracture characteristic than other specimens. Clear plastic deformation in the edge of the smooth area could be found at higher magnification as shown in Figure 7 (e) and (f). And lots of holes could be found in the micrograph of L424 as shown in Figure 7 (g), which indicates that the compatibility of PEEK matrix and SiO2 is not good and account for its lowest tensile strength of all specimens.

3.3. Effect of Coupling Agents for Nanoparticle Treatment on Mechanical Properties As is shown in the Figure 8(a) and (b), the tensile, flexural and impact strength of specimens which filled with coupling agents treated nanometer Al2O3 particles are higher than that of neat PEEK, which implies the well interaction between the nanometer Al2O3 particles and polymer segment chain [24]. Furthermore, the impact strength of specimens filled with coupling agents treated Al2O3 is higher than that of L120 filled with untreated Al2O3. It is possible that the generation of ductile interface layer between the surface of particles and coupling agents improves the absorption to impact energy. But there is no analogous trend in the others mechanical properties, which is possible due to the weakening of strength and rigidity of composites for the generation of ductile interface.

Tribology of Composite Materials, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

Guo Qiang and Pan Guoliang

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

104

Figure 8. Mechanical properties of composites filled with Al2O3 particles treated by various coupling agents. (a) tensile and flexural strength; (b) impact strength and ball indentation hardness.

Tribology of Composite Materials, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

105

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Mechanical and Tribological Behaviors of Nanometer Al2O3 and SiO2 …

Figure 9. Variation of mechanical properties of PEEK composites with different dispersing methods. (a) tensile, compressive and flexural strength; (b) impact strength and ball indentation hardness.

Tribology of Composite Materials, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

106

Guo Qiang and Pan Guoliang

3.4. Effect of Dispersing Methods of Nanoparticles in PEEK on Mechanical Properties Four kind of dispersing methods: DMM and LSMD were used to disperse fillers in PEEK matrix, which were important factors that would influence the mechanical properties of reinforced PEEK. Figure 9(a) and (b) give the variation of mechanical properties with different dispersing methods [24]. It is seen that the compressive, flexural and impact strength of L121 have improved by 14.8 %, 10.4 % and 15.7 %, respectively, compared with that of D121. But the tensile strength and ball indentation hardness of L121 are slightly greater than that of D121. As far as three methods of LSMD, UD and BMD are concerned, the variation of mechanical properties with dispersing methods is slightly. Among the three kinds of specimens, the tensile and compressive strength of B141 are superior, while the flexural strength and ball indentation hardness of U141 are best.

4. SLIDING FRICTION PROPERTIES OF PEEK COMPOSITE

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Sliding friction is the most common condition in many applications of PEEK composites. Therefore it is important to reveal the sliding friction properties of various PEEK composites and the relative influence factors. In this section, the friction and wear properties of PEEK composite filled with Al2O3 and/or PTFE were described, as well as the effect of diameter on friction coefficient and wear rate were investigated. Furthermore, the wear mechanism were discussed in detail through the investigation on the wear trace and transfer film.

4.1. The Friction and Wear Properties of Filled PEEK in Dry Sliding Conditions The effect of particle size on the wear coefficient and friction coefficient of the filled PEEK are shown in Figs. 10 and 11 [43]. It is seen that when filled with Al2O3 particles and without PTFE, the composite exhibited a decreased wear coefficient in comparison with the unfilled one. In particular, the lowest wear coefficient in this work was obtained by composite filled with 15 nm Al2O3. When filled with 90 and 500 nm Al2O3, the wear rates of specimens D221 and D321 reached at a little more than twice the value of specimen D121 which was filled with 15 nm Al2O3. When filled with Al2O3 particles and PTFE, the composite showed an increased wear rate, especially, the wear rate of specimen D225 filled with 90 nm Al2O3 is sharply increased. When filled with PTFE alone, specimen D005 exhibited a decreased wear rate in comparison with the unfilled composite of specimen D000. The friction coefficient of the composite filled with PTFE is much lower than that without PTFE. Composites filled with 15 nm Al2O3 shows higher friction coefficient compared with those filled with other particles, both with and without PTFE. Thus it can be concluded that nanometer or micron Al2O3 particles as fillers in PEEK does not change the friction coefficient and wear coefficient of the filled composites simultaneously.

Tribology of Composite Materials, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

Mechanical and Tribological Behaviors of Nanometer Al2O3 and SiO2 …

107

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Figure 10. Effect of particle size of Al2O3 on the wear coefficient of the filled PEEK. (load 196 N, sliding velocity 0.42 ms-1, test duration 120 min).

Figure 11. Effect of particle size of Al2O3 on the friction coefficient of the filled PEEK. (load 196 N, sliding velocity 0.42 ms-1, test duration 120 min).

Tribology of Composite Materials, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

108

Guo Qiang and Pan Guoliang

4.2. SEM Observation of the Wear Traces and Transfer Films

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Figure 12 shows SEM micrographs of the worn surfaces of unfilled PEEK and filled PEEK with 5 wt. % Al2O3 particles [43]. It can be seen that the plucked and ploughed marks appeared on the wear scar of unfilled PEEK specimen D000, while the scuffing on the surface of specimen D121 filled with 15 nm Al2O3 was obviously abated.

Figure 12. SEM micrographs of worn surfaces of unfilled PEEK and filled PEEK with 5 wt. % Al2O3. (load 196 N, sliding velocity 0.42 ms-1, test duration 120 min). (a) D000, (b) D121, (c) D221, (d) D321.

Like the wear scar on the surface of unfilled PEEK block, there are some plucked and ploughed marks on the wear scar of specimen D221 filled with 90 nm Al2O3, some pits formed by the abscission of the particles are also found. The obvious scuffing on specimen D221 shows that abrasive wear plays an important role when the particle size is up to 90 nm. Flakes appeared on the wear scar of specimen D321 filled with 500 nm Al2O3 indicates that adhesive wear plays an important role when the particle size is up to 500 nm. It can also be inferred that the morphologies of the wear traces are relevant to the wear coefficients of Al2O3 filled PEEK. The transfer films formed on the steel ring surfaces running against the filled PEEK with 5 wt. % Al2O3 particles are shown in Figure 13 [43]. It can be seen that the thin, uniform and coherent transfer film was formed by running the steel ring against the specimen D121, while ploughed marks, as well as thick, lumpy and incoherent transfer film was formed by running the steel ring against specimen D221.

Tribology of Composite Materials, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

Mechanical and Tribological Behaviors of Nanometer Al2O3 and SiO2 …

109

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Figure 13. SEM micrographs of transfer films formed by running the steel ring against the filled PEEK with 5 wt. % Al2O3 particles. (load 196 N, sliding velocity 0.42 ms-1, test duration 120 min). (a) D121, (b) D221.

Figure 14. SEM micrographs of worn surfaces of specimen D125 filled with 5 wt. % 15 nm Al2O3 and 10 wt. % PTFE (a) and transfer films formed by running the steel ring against the composite (b) (load 196 N, sliding velocity 0.42 ms-1, test duration 120 min).

There exists an obvious difference between the morphologies of the wear traces on the composites with and without PTFE. SEM micrographs of the worn surfaces of specimen D125 filled with 5 wt. % 15 nm Al2O3 and 10 wt. % PTFE and transfer films formed by running the steel ring against the composite are shown in Figure 14 [43]. Worn debris of multi-layer laminar film is found on the wear scar of specimen D125, which indicates that transfer films were formed and transferred, repetitively. A thick, lumpy and incoherent transfer film is found on the steel ring surface, indicating that PTFE transferred prior to PEEK, and consequently a lowered friction coefficient was reached. The repetitive formation and destruction of the film occurred at a high rate and resulted in higher wear coefficient compared with specimen D121 filled with 5 wt. % 15 nm Al2O3.

Tribology of Composite Materials, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

110

Guo Qiang and Pan Guoliang

Figure 15. EDS spectra of transfer films formed by running the steel ring against the specimen D121 filled with 5 wt. % 15 nm Al2O3. (load 196 N, sliding velocity 0.42 ms-1, test duration 120 min).

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

4.3. EDS Analysis EDS analysis of the transfer films formed by running the steel ring against the specimen D121 filled with 5 wt. % 15 nm Al2O3 shows that some proportion of Al element exists in the transfer film, besides elements the steel substrate contain, such as Fe, Mn and C. This indicates that Al2O3 particles transferred to the counterpart ring surface together with PEEK, forming the transfer films during the rubbing process (Figure 15) [43].

5. FRETTING PROPERTIES OF PEEK COMPOSITES Some literatures have reported the fretting properties of PEEK composites filled with PTFE. It was concluded from the investigations that the inclusion of PTFE in PEEK definitely and significantly improved the performance of PEEK. The blends did not show any scuffing problems. A 30 times improvement in wear rate and five times in friction coefficient was observed due to inclusion of PTFE. The influence of PTFE in LAOW did not match completely with that in adhesive or abrasive wear mode, though LAOW/fretting wear has a common wear mechanism with both the wear modes. In the case of abrasive wear, hardness and tensile strength proved to be dominating wear-controlling material properties [16]. Burris etc. [27] found that the PEEK/PTFE composite has a wear rate lower than unfilled PTFE and PEEK for every sample tested. This composite was 900 times as wear resistant as the unfilled PEEK and 260,000 times as wear resistant as the unfilled PTFE. In addition, this composite material has a friction coefficient lower than unfilled PTFE and PEEK for every sample tested. However, for PEEK composites filled with inorganic particles, research on their

Tribology of Composite Materials, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

Mechanical and Tribological Behaviors of Nanometer Al2O3 and SiO2 …

111

fretting wear characteristics is seldom. Therefore the effects of relative factors on the fretting wear characteristics of PEEK composites are deserved to study and described in detail as follows.

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

5.1. The Influence of Al2O3 Diameter on Fretting Wear Resistance of PEEK Composites The results of fretting wear testing for PEEK composites filled with 5 wt.% Al2O3 powder with various particle diameters are as shown in Figure 16 [45]. It can be seen that, the area of wear scar on composite specimen decreases after filling Al2O3 in comparison with pure PEEK, and reaches the lowest value of 0.503 mm2 at Al2O3 diameter of 15 nm. At Al2O3 diameter of 90 nm, the wear scar area of specimen D221 is about 1.5 times that of specimen D121. However, when Al2O3 diameter is 500 nm, the wear scar area of specimen D321 is lower than that of D221. Figure 17 shows the results of the fretting wear testing for PEEK composites filled with both PTFE and Al2O3 fine powder [45]. It can be found that, the area of wear scar on specimen D005 filled with 10 wt.% PTFE super fine powder is the lowest of all specimens, which reveals that the filling of PTFE can improve the fretting wear resistance of PEEK composite effectively. However, when both PTFE and Al2O3 powder are filled into PEEK, the wear scar areas of all specimens increases slightly and that of specimen D225 filled with 90 nm Al2O3 is the highest. The variation of Al2O3 diameter has smaller influence on fretting wear resistance in comparison with PEEK composites only filled with Al2O3. During the fretting wear, the friction counterpart of steel ball also suffers from wear. For only Al2O3 filled PEEK composite, the area of wear scar on steel ball fretting against specimen D121 is 0.360 mm2, which is the lowest of all friction counterparts. For PEEK composite filled with PTFE, the wear scar area of steel ball fretting against specimen D005 is the lowest but that against specimen D225 is the highest.

Figure 16. The area of the wear scars on Al2O3/PEEK composite and the corresponding steel ball, S1 and S2, respectively, plotted as functions of diameter of Al2O3 particles after fretting test. Tribology of Composite Materials, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

112

Guo Qiang and Pan Guoliang

Figure 17. The area of the wear scars on Al2O3/PEEK composite filled with PTFE and the corresponding steel ball, S1 and S2, respectively, plotted as functions of diameter of Al2O3 particles after fretting test.

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

5.2. The Influence of Different Content of the Fillers on Fretting Wear Resistance of PEEK Composites Figure 18 shows the variations of the wear scar area of 15 nm Al2O3 filled PEEK composite and that of the corresponding steel ball with content of Al2O3 particles after fretting test [45]. It can be seen that the filling of 15 nm Al2O3 can obviously improve the fretting wear resistance. The area of wear scar decreases to 60 % that of pure PEEK when 5 % Al2O3 powder is filled, and then increases with the increase of Al2O3 content. Figure 19 shows the variations of the wear scar areas of 12 nm SiO2 filled PEEK composites and that of the corresponding steel balls with content of SiO2 particles after fretting test [45]. It can be found that the wear scar area of SiO2 filled PEEK composite is greater than that of Al2O3 filled PEEK composite, but both the curves of the variations with content of fillers are similar. The wear scar area of specimen L424 filled with 5 wt. % SiO2 is 1.5 times that of specimen L121. It should be noted that the wear scar area of specimen L434 filled with 10 wt. % SiO2 is even greater than that of pure PEEK. Figure 20 shows the variations of the wear scar area of Al2O3/PEEK composite filled with PTFE and that of the corresponding steel ball with content of Al2O3 particles after fretting test [45]. It can be seen that, the area of wear scar on PEEK composite increases monotonically with the increase of Al2O3 content, which indicates that the filling of both PTFE and Al2O3 would result in a more severe wear on PEEK composites, in comparison with the filling of only Al2O3. From Figure 18, 19 and 20, it can be also found that, the variations of the wear scar areas of composite specimens with filler content are very similar to that of steel balls with filler content. The maximal value of the wear scar areas of steel balls is 0.853 mm2 which belongs to the steel ball fretting against specimen L424.

Tribology of Composite Materials, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

Mechanical and Tribological Behaviors of Nanometer Al2O3 and SiO2 …

113

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Figure 18. The area of the wear scars on Al2O3/PEEK composite and the corresponding steel ball, S1 and S2, respectively, plotted as functions of Al2O3 particles wt. % after fretting test.

Figure 19. The area of the wear scars on SiO2/PEEK composite and the corresponding steel ball, S1 and S2, respectively, plotted as functions of SiO2 particles contents after fretting test.

Tribology of Composite Materials, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

114

Guo Qiang and Pan Guoliang

Figure 20. The areas of the wear scars on Al2O3/PEEK composite filled with PTFE and the corresponding steel ball, S1 and S2, respectively, plotted as functions of Al2O3 particles content.

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

5.3. SEM and EDS Analysis on the Worn Surfaces Figure 21(a) and (b) shows the SEM micrographs of worn surface of specimen D000 and the corresponding abrasive debris, respectively [45]. It can be observed from Figure 21(a) that, the basically smooth worn surface has multilayer plate-like abrasive debris with small debris particles distributing in the wear scar area. As is shown in Figure 21(b), there is obvious fiber form debris around the wear scar. The debris is formed step by step due to the cyclic extrusion initiated by fretting.

Figure 21. SEM micrographs of the worn surface on specimen D000 (a) and the corresponding abrasive debris (b) after fretting test.

Tribology of Composite Materials, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

Mechanical and Tribological Behaviors of Nanometer Al2O3 and SiO2 …

115

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Figure 22. EDS spectrogram of the worn surface on specimen D121 after fretting test.

EDS spectrogram of the worn surface on specimen D121 after fretting wear testing is as shown in Figure 22 [45]. It can be found the characteristic peak of chemical element Fe except for that of chemical elements of C, O and Al originated from specimens, which indicates the occurrence of chemical erosion on the surface of steel ball and the transfer of generated ferric oxide to the surface of specimen during fretting wear testing. SEM micrographs of the worn surface on specimens filled with Al2O3 or SiO2 after fretting wear testing are as shown in Figure 23 [45]. From Figure 23(a), it can be observed that, the surface of wear scar is basically smooth with little trace of local adhesive transfer but no obvious debris, which indicates a slight wear of specimen D121 filled with 15 nm Al2O3. At Al2O3 diameter of 90 nm, as is shown in Figure 23(b), the surface of wear scar is rough with lots of abrasive debris and has dense and deep stria-like grooves parallel to the direction of fretting. However, when Al2O3 diameter is 500 nm, as shown in Figure 23(c), the surface of wear scar is smoother than that of D221 and also has shallow stria-like grooves parallel to the direction of fretting. Furthermore, it can be found that some multilayer plate-like abrasive debris adhere to the surface of wear scar and some grooves are obvious on it, which indicates a moderate adhesion wear. As is shown in Figure 23(d), the surface of wear scar is basically smooth with some obvious long cracks locally but no debris, which indicates a deterioration of toughness due to the increase of Al2O3 content. As is shown in Figure 23(e), the surface of wear scar of specimen L424 filled with SiO2 is rougher than that of D221 filled with Al2O3 and has a mass of dense and deeper stria-like grooves which are parallel to the direction of fretting and arrange regularly, indicates a more severe abrasive wear than other specimens.

Tribology of Composite Materials, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

116

Guo Qiang and Pan Guoliang

Figure 23. SEM micrographs of the worn surface on specimens filled with Al2O3 or SiO2 after fretting test: (a) D121; (b) D221; (c) D321; (d) L131; (e) L424.

Figure 24 shows the SEM micrographs of the worn surface on specimens filled with both Al2O3 and PTFE after fretting wear testing [45]. As shown in Figure 24(a), the worn surface of specimen D125 is smooth and has some plate-like abrasive debris with small cracks on them. When Al2O3 diameter increases to 90 nm, as shown in Figure 24(b), some broad and stria-like grooves parallel to the direction of fretting appear on the worn surface or even on the surface of multilayer abrasive debris adhesion to the wear scar, which reveals a strong adhesive transfer between the abrasive debris and substrate material. When Al2O3 content increases to 10 wt. %, as is shown in Figure 24(c), the surface of wear scar is smooth as that of D125, but the plate-like abrasive debris is bigger.

Tribology of Composite Materials, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

Mechanical and Tribological Behaviors of Nanometer Al2O3 and SiO2 …

117

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Figure 24. SEM micrographs of the worn surface on specimens filled with both Al2O3 and PTFE after fretting test. (a) D125; (b) D225; (c) D135.

Figure 25 shows the EDS spectrogram of the worn surface on steel ball fretting against specimen D125 after fretting wear testing [45]. It can be observed that the characteristic peaks of elements Al and F are obvious, which proves the transfer of Al2O3 and PTFE from specimen to steel ball during fretting wear testing. It should be noted that the formation of a moderated transferred film would be beneficial to the fretting wear resistance for PEEK composites. Figure 26 shows the SEM micrographs of the worn surface on the steel ball fretting against specimens filled with Al2O3 and PTFE together [45]. It can be found that the variations of Al2O3 diameter and content result in the variations of the topographies of transfer film or wear scar on steel balls, and a mass of particulate debris adhering to the surfaces of wear scars for all the four specimens. From Figure 26(a) and (b), it can be seen that both the worn surfaces have clear wear contour. The worn surface on the steel ball fretting against specimen D125 is smooth and a little polymer materials adhere to the surface of wear scar with homogeneous distribution. However, the worn surface on the steel ball fretting against specimen D225 has thick transfer film which covers the whole worn surface. From Figure 26(c), when Al2O3 diameter increases to 500 nm, polymer materials adhering to the worn surface of D325 are fewer than that of D225 and the striae with bright-dark interphase appear on the worn surface. For steel ball fretting against specimen D135 filled with 10 wt.% Al2O3, there are fewer adherent polymer material but more particulate debris with partly agglomeration on the smooth surface of wear scar in comparison with D125, as is shown in Figure 26(d).

Tribology of Composite Materials, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

118

Guo Qiangg and Pan Guooliang

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Fiigure 25. EDS spectrogram s of the worn surfacce on the steel ball b fretted against specimen D125. D

Fretting wear differs froom other form ms of wear in two aspects, i.e. very loow relative veelocity of the contacting suurfaces and thhat most partss of the contaacting surfacess are never brrought out of contact. c Frettiing produces considerable c a amount of debrris, known as third body, w which does not easily escappe from the contact and pllays an imporrtant role in thhe wear of m materials. As th he friction paair of polymerr and steel balll is concerned, thermal efffect always pllays an imporrtant role durring fretting for f a repeatedd contact streess in small area a during frretting and low w heat conducctivity of polym mer. Frictionaal heat produced during frettting would ennhance the adhesion a betw ween the steeel ball and polymer. p Theerefore, the teemperature reesistance of po olymer materiaal would affecct the wear deggree on surfacce of fretting scar. Due to thhe filling of Al A 2O3 in PEEK K improved thhe glass transition temperatuure of PEEK composite, thhereby the weaar resistance of o it during fretting was improved accordiingly. With the variation v of filler diameter, content c and kind, k the degreee and mode of o wear are diifferent. Speciimen D121 haas high frettinng wear resisttance and makkes a slight wear w on the suurface of steell ball for its good g temperatuure resistance, adhesion weear resistance and plastic deeformation ressistance. Weaar of the steel ball fretting against a specim men D221 is more m severe thhan that againsst specimen D121. D The weaar mechanism m of D221 is mainly m abrasivve wear due too the embeddin ng and ploughhing action off ferric oxide debris d generateed in the interrface, while m minor adhesion n wear. The wear w resistancce of specimeen D321 is beetween that off specimen D D121 and D221, and the weear mechanism m appears stroong adhesion wear and weaak abrasive w wear. The moree Al2O3 powder results in thhe increase off friction tempeerature and thhe falling of abbrasive debris with sheets. Furthermore, F t cooling off abrasive debrris with high teemperature the w would result in n more cracks as shown in Figure F 23(d). The T wear of sppecimen filledd with SiO2 iss more severe than that filleed with Al2O3 and the wear mechanism appears a strongger abrasive w for the existence of SiO wear O2 particles andd ferric oxide debris.

Tribology of Composite Materials, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

Mechanical and Tribological Behaviors of Nanometer Al2O3 and SiO2 …

119

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Figure 26. SEM micrographs of the worn surface on the steel ball fretted against specimen filled with both Al2O3 and PTFE. (a) D125; (b) D225; (c) D325; (d) D135.

The molecular chain structure in PTFE is regular and symmetrical, carrying the property of a tendency to crystallize, having a fairly high melting point and good thermal stability. However, the filling of PTFE would deteriorate the mechanical strength of PEEK composites and result in the formation and falling of abrasive debris during fretting. As is shown in Figure 24, all the wear scars of specimen D125, D225 and D135 appear strong adhesion wear. As the sheet like abrasive debris is not easy to be got out from contact field of wear, the mutual transfer of composite material between the surface of specimen and steel ball is more frequently. Therefore, the filling of PTFE makes a more severe wear in comparison with Al2O3/PEEK composites unfilled PTFE. Specimen D125 filled with 15 nm Al2O3 has moderate mechanical performance and adhesion resistance. Therefore, moderated polymer materials adhere to the surface of steel ball which form a good polymer lubrication film and result in a slight wear on the surface of steel ball. And then the most severe wear was obtained when 90 nm Al2O3 particles was filled. However, when Al2O3 diameter increases to 500 nm, the wear of steel ball is slighter than D225. It is probably due to that the existence of big filler particle decrease the adhesion of polymer material to steel ball. For the adhesion wear dominated work condition, when the effect of the big particle size on the adhesion wear decreasing is greater than that on the abrasive wear increasing, the slighter wear could be obtained.

Tribology of Composite Materials, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

120

Guo Qiang and Pan Guoliang

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

CONCLUSION Tensile strength, ball indentation hardness and impact strength of PEEK filled with 15 nm Al2O3 particles are better than those of PEEK filled with 90 nm Al2O3 particles, however, the latter exhibits better performance in flexural and compressive strength but has lowest value of ball indentation hardness of all. PEEK filled with 5 wt.% Al2O3 represents better performance than that filled with 10 wt.% Al2O3 in tensile, compressive and impact strength , but the latter exhibits better performance in flexural strength and ball indentation hardness. In addition, all mechanical properties of PEEK filled with nanometer SiO2 particles are worse than that of PEEK filled with Al2O3 particles except for the ball indentation hardness. The treatment of coupling agent results in an increase in toughness, but decrease in hardness. And impact strength of PEEK filled with titanate treated nanometer Al2O3 particles is about eight times that of neat PEEK or 2 times that of PEEK filled with untreated Al2O3. Nanometer and micron Al2O3 particles as fillers in PEEK can also reduce the wear coefficient, but not the friction coefficient of PEEK. The lowest wear rate was obtained with the composite filled with 5 wt. % 15 nm Al2O3. With 15 nm Al2O3 filled PEEK, a thin, uniform and tenacious transfer film was formed on the counterpart steel surface during the friction process. The decreased wear coefficient of the filled PEEK is attributed to the abated scuffing between the transfer film and the composite surface. With 90 and 500 nm Al2O3 filled PEEK, the increased wear coefficients of the filled PEEK composites are attributed to abrasive wear and adhesive wear respectively, on the PEEK composite surface, compared with 15 nm Al2O3 filled PEEK. The incorporation of 10 wt. % PTFE into unfilled PEEK got a decreased friction coefficient and wear coefficient simultaneously, while the incorporation of 10 wt. % PTFE into PEEK composites caused a lower friction coefficient and a higher wear coefficient, indicating no synergistic effect with nanometer Al2O3 particles. For PTFE transferred prior to PEEK, a lowered friction coefficient was reached. The filling of Al2O3 powder improves the fretting wear resistance of PEEK composite. With the increase of Al2O3 diameter from 15 nm to 500 nm, the area of wear scar on specimen increases first and decreases afterward. However, when both the Al2O3 and PTFE are filled into PEEK, the wear of all specimens is more severe while the variation of Al2O3 diameter has little influence on fretting wear resistance in comparison with the only Al2O3 filled PEEK composite. It should be noted that the area of wear scar of specimen D005 filled with 10 wt.% PTFE super fine powder is the lowest of all specimens, indicates no synergistic effect of Al2O3 and PTFE in PEEK composite. For PEEK composite filled with Al2O3 with/without PTFE, the area of wear scar increases monotonically with the increase of Al2O3 content. For SiO2 filled PEEK composite, the area of wear scar is greater than that of Al2O3 filled PEEK composite, but both the curves of variations with content of fillers are similar. Meanwhile, the more severe wear for the PEEK composite specimen the more severe wear for the corresponding steel ball. For the friction counterparts of PEEK composite paired against a steel ball, abrasive wear and adhesive wear dominate the fretting wear mechanism during fretting. The embedding and plough actions of fillers and ferric oxide debris generated in the interface of friction counterparts during fretting would intensify the wear on the surface of steel ball. The filling of PTFE in PEEK composite enhances the adhesion and coherence force of polymer material and increases the frequency of mutual transfer of composite material between the friction pair of composite specimen and steel ball, thus makes a more severe wear in comparison with PEEK composites unfilled PTFE.

Tribology of Composite Materials, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

Mechanical and Tribological Behaviors of Nanometer Al2O3 and SiO2 …

121

REFERENCES [1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] [16] [17] [18] [19] [20]

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

[21] [22] [23] [24] [25] [26] [27] [28] [29] [30] [31] [32] [33] [34] [35] [36] [37] [38]

Kuo M. C.; Huang J. C.; Chen M. Mater. Chem. Phys. 2006, 99(2-3), 258-268. Lhymn C.; Lhymn Y. O. Adv. Polym. Tech. 1988, 8(4), 417-430. Yu L.; Yang S.; Liu W.; et al. J. Appl. Polym. Sci. 2000, 76(2), 179-184. Zhang G.; Chang L.; Schlarb A. K. Compos. Sci. Technol. 2009, 69(7-8), 1029-1035. Wang Q. H.; Xue Q.; Shen W. Gongneng Cailiao/J. Funct. Mat. (Chinese). 1998, 29(5), 558-560. Wang Q. H.; Xue Q. J.; Liu W. M.; et al. J. Appl. Polym. Sci. 2000, 78(3), 609-614. Long C. G.; Wang X. Y. J. Mater. Sci. 2004, 39(4), 1499-1501. Vande V. J.; Bahadur S. Wear. 1995, 181-183(1), 212-221. Wang Q. H.; Xue Q.; Shen W.; et al. J. Appl. Polym. Sci. 1998, 69(1), 135-141. Goyal R. K.; Negi Y. S.; Tiwari A. N. Eur. Polym. J. 2005, 41(9), 2034-2044. Wang H. Y.; Feng X.; Shi Y.; et al. J. Reinf. Plast. Comp. 2009, 28(6), 645-655. Zhou B.; Ji X.; Sheng Y.; et al. Eur. Polym. J. 2004, 40(10), 2357-2363. Converse G. L.; Yue W.; Roeder R. K. Biomaterials 2007, 28(6), 927-935. Tang S. M.; Cheang P.; Abubakar M. S.; et al. Int. J. Fatigue 2004, 26(1), 49-57. Zhang G. S.; Sui G. X.; Meng H.; et al. Compos. Sci. Technol. 2007, 67(6), 1172-1181. Bijwe J.; Sen S.; Ghosh A. Wear 2005, 258(10), 1536-1542. Donaldson S. L. Composites 1985, 16(2), 103-112. Gao S. L.; Kim J. K. Compos Part A-Appl S. 2001, 32(6), 775-785. Fracasso R.; Rink M.; Pavan A.; et al. Compos. Sci. Technol. 2001, 61(1), 57-63. Lamontagne C. G.; Manuelpillai G. N.; Kerr J. H.; et al. Int. J. Impact. Eng. 2001, 26(110), 381-398. Stuart B. H.; Briscoe B. J. High Perform Polym. 1996, 8(2), 275-280. Zhang G.; Schlarb A. K.; Tria S.; et al. Compos. Sci. Technol. 2008, 68(15-16), 30733080. Lai Y. H.; Kuo M. C.; Huang J. C.; et al. Mat. Sci. Eng. A. 2007, 458(1-2), 158-169. Pan G. L.; Guo Q.; Tian A. G.; et al. Mat. Sci. Eng. A. 2008, 492(1-2), 383-391. Abu B. M.; Cheang P.; Khor K. A. Compos. Sci. Technol. 2003, 63(3-4), 421-425. Long C. G.; Wang X. Y. J. Reinf. Plast. Comp. 2004, 23(15), 1575-1582. Burris D. L.; Sawyer W. G. Wear 2006, 261(3-4), 410-418. Mody P. B.; Chou T. W.; Friedrich K. J. Mater. Sci. 1988, 23(12), 4319-4330. Tripathy B. S.; Furey M. J. Wear 1993, 162-164(Part 1), 385-396. Hanchi J.; Eiss N. S. Wear 1997, 203-204, 380-386. Jacobs O.; Friedrich K.; Marom G.; et al. Wear 1990, 135(2), 207-216. Bahadur S.; Gong D.; Anderegg J. W. Wear 1993, 160(1), 131-138. Bahadur S.; Gong D. Wear 1992, 154(1), 151-165. Feng X.; Zhang R. Cailiao Yanjiu Xuebao/ J. Mat. Res. (Chinese). 1999, 13(6), 645649. Voort J. V.; Bahadur S. Wear 1995, 181-183(1), 212-221. Guo Q.; Friedrich K. Journal of Synthetic Lubrication 1993, 10(3), 213-224. Friedrich K.; Flock J.; Varadi K.; et al. Wear 2001, 251(1-12), 1202-1212. Lin Y. X.; Gao C. H.; Li Z. F. Cailiao Rechuli Xuebao/T Mat. Heat Treat (Chinese). 2006, 27(4), 20-23.

Tribology of Composite Materials, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

122

Higgs C. F.; Wornyoh E. Wear 2008, 264(1-2), 131-138. Kuo M. C.; Tsai C. M.; Huang J. C.; et al. Mater. Chem. Phys. 2005, 90(1), 185-195. Wang Q. H.; Xue Q. J.; Shen W. C. Tribol. Int. 1997, 30(3), 193-197. Li J.; Liao H.; Coddet C. Wear 2002, 252(9-10), 824-831. Qiao H. B.; Guo Q.; Tian A. G.; et al. Tribol. Int. 2007, 40(1), 105-110. Braza J. F.; Furst R. E. Wear 1993, 162-164(Part 2), 748-756. Pan G. L.; Guo Q.; Zhang W. D.; et al. Wear 2009, 266(11-12), 1208-1215. Sinmazcelik T.; Yilmaz T. Mater. Design 2007, 28(2), 641-648.

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

[39] [40] [41] [42] [43] [44] [45] [46]

Guo Qiang and Pan Guoliang

Tribology of Composite Materials, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

In: Tribology of Composite Materials Editor: J. Paulo Davim

ISBN 978-1-62100-999-3 © 2012 Nova Science Publishers, Inc.

Chapter 5

EFFECTS OF MATRIX CRYSTALLINE STRUCTURE AND MOLECULAR WEIGHTON THE TRIBOLOGICAL BEHAVIOR OF PEEK-BASED MATERIALS G. Zhang∗ and M. Schehl Institute for Composite Materials, University of Kaiserslautern, Kaiserslautern, Germany

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

ABSTRACT Polyetheretherketone (PEEK) materials are being widely used as tribomaterials. The crystalline structure and molecular weight of PEEK were proven to exert significant influences on its mechanical profiles. The effects of matrix crystalline structure and molecular weight on the tribological properties of PEEK-based materials were studied in this work. The tribological properties were correlated with its mechanical properties. For pure PEEK, the tribological behaviour is closely related to its mechanical properties, e.g. stiffness and ductility. Therefore, the crystalline structure and molecular weight play significant roles on the tribological behaviors. For PEEK composites, however, the matrix/filler interactions are of great importance. The interfacial bonding and stress transfer between the matrix and the fillers are important factors influencing the tribological performances of the composites.

1. INTRODUCTION PEEK (polyetheretherketone) becomes one of the most attractive semi-crystalline thermoplastics due to the combination of a high toughness and a high stiffness. Because of its excellent tribological performance, PEEK based materials are widely used as tribomaterials

∗

Corresponding author, Email: [email protected].

Tribology of Composite Materials, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

124

G. Zhang and M. Schehl

[1-3]. In the last two decades, lots of researches were conducted to investigate the tribological behaviors of PEEK-based materials. The selection of filler types (including the surface modification of certain fillers) and the optimization of their dimensions, contents, and dispersing states constitute most of the efforts on the tribology of PEEK in the last two decades. Many delicate formulations of PEEK composites [4-11], especially the ones with multiple fillers [4,5] were proposed. The crystalline structure of PEEK is significantly influenced by the processing conditions, i.e. cooling speed from its melting state and a post-annealing treatment. A low cooling speed or a post-annealing treatment at a temperature higher than the cold crystallization temperature can result in a high crystallinity of PEEK. With a fast cooling speed, an amorphous structure is achievable. With a low cooling speed, however, a semi-crystalline structure can be obtained. For a semi-crystalline PEEK, the molecular network is represented by crystallites acting in an amorphous matrix as physical cross-links. A previous study shows that the microhardness of PEEK presents a strong dependence on the crystalline structure [12]. With increasing the crystallinity, the hardness of PEEK is increased. The frictional processes of polymeric materials are very complicated. Generally, for a single-phase material, the combination of a low surface energy, a high stiffness, and a high toughness results in a good tribological performance, i.e. low friction coefficient and wear rate. The structure of poly-meric materials, e.g. molecular weight and crystalline structure, exerts important roles on their properties including tribological performances [12,13]. The previous work [12] indicates that a semi-crystalline PEEK exhibits a much higher wear resistance than an amorphous one. The higher wear resistance of semi-crystalline PEEK was mainly attributed to the higher stiffness than the amorphous PEEK. Compared with a single-phase polymer, the polymer composite presents a more complicated structure-tribology relationship. Generally, the roles of fillers can be summarized into the three following aspects: lubricating effect, improving mechanical properties, e.g. compressive strength and stiffness, and promoting the formation of a homogeneous transfer film. The internal lubricants refer to the materials with a low surface energy, e.g. polytetrafluoroethylene (PTFE), and the layer-structural materials in which the layers are linked by weak Van der Waals bonds, e.g. graphite, MoS2 etc. A pioneering work was carried out by Voss and Friedrich [14] on short glass and carbon fibers reinforced PEEK composites. The fibers were incorporated into the matrix to increase its creep resistance and compressive strength. Their results indicated that short carbon fiber (SCF) improves the wear resistance of PEEK more effectively than glass fiber. Polymeric material filled with SCF/graphite/PTFE is a successful tribomaterial formulation [4]. The multiple fillers play synergetic roles on improving the tribological performance of polymer. Moreover, such fillers like nano-sized particles and internal lubricants might promote the formation of a homogenous transfer film on counterpart surface. A homogenous transfer film was assumed to benefit the tribological performance of the material [5,7,9]. In the composite system, the mechanical properties of the matrix can influence the interaction between the fillers and the matrix. The crystalline structure of the matrix affects its mechanical properties. Therefore, it can be reasonably expected that the interaction between the fillers and the matrix can be influenced by the matrix crystalline structure. Taking into account the fact that the matrix crystalline structure is closely related to the processing parameters, e.g. cooling speed and annealing treatment, it’s of interest to know the influence of matrix crystalline structure on the tribological profile of a composite. The first part of this

Tribology of Composite Materials, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

Effects of Matrix Crystalline Structure and Molecular Weight ...

125

work focuses on the effect of matrix crystalline structure on the tribological behaviors of PEEK based composites. Two composites, i.e. graphite (10 wt%) filled amorphous PEEK and graphite (10 wt%) filled semi-crystalline PEEK, were studied. The two composites were referenced hereafter as AMC (amorphous composite) and SCC (semi-crystalline composite). Graphite is chosen in this work as filler because it is one widely used lubricant in polymeric composites. In order to simplify the system, only graphite particles were incorporated into the matrix. Effort on fundamental understanding of tribology is always crucial for the formulation of tribomaterials. Surely, it is of interest to understand how PEEK’s mechanical properties affect its tribological behaviors. The second part of this work relies on the correlation of the mechanical and the tribological properties of three pure PEEKs and two SCF/graphite/PTFE filled PEEK composites with different matrix molecular weights. The tribological behaviors of the PEEK materials were examined and correlated with their mechanical properties. The objective of this part of work is to describe a comprehensive effort to correlate the tribological behaviors of PEEK materials with their mechanical properties. The three pure PEEKs were referenced as MA (material A), MB, and MC and two SCF/graphite/PTFE filled PEEK composites were referenced as MB FC30 and MC FC30. MB FC30 and MC FC30 were compounded respectively from MB and MC with each 10 wt% SCF (9.1 vol%), graphite and PTFE. The molecular weights of the three PEEKs follow the order: MA < MB < MC.

2. EXPERIMENTAL

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

2.1. Sample Preparation Amorphous bulk PEEK material can not be prepared easily because of its rapid crystallization speed. However, the cooling speed can be fast enough if the material is in a thin film form. In order to get a composite with an amorphous matrix (AMC), the materials used in this work were prepared using a printing process for obtaining coating forms. PEEK powders having a mean diameter of 10 µm, supplied by ICI (Victrex Ltd., UK), were used in the present work. The graphite particles as shown in Figure 1 were supplied by Sigma– Aldrich Ltd., USA. For beginning, PEEK powders mixed with 10 wt% graphite particles were put into an aqueous solution to form slurry. The slurry was continuously stirred for 30 min and then subjected to an ultrasonic bath for 20 min until the system became fully dispersed. Consequently, the slurry was applied evenly on degreased substrates. The substrates were 60 mm diameter and 5 mm thick aluminum discs. After being dried in air, the substrate-coating system was heated to 400 oC and held at this temperature for 5 min. In order to get SCC, the coating-substrate systems were cooled naturally in air (~20 oC). In order to get AMC, the coating-substrate system was rapidly quenched from melting state into water at room temperature. The coatings have a thickness of about 60 µm. More details on coating procedure were described in a previous work [15]. From previous works [12,15,16], the quenched PEEK composite has an amorphous matrix while the air-cooled PEEK material has a semi-crystalline matrix. The semi-crystalline PEEK exhibits a much higher stiffness than the amorphous PEEK [12].

Tribology of Composite Materials, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

126

G. Zhang and M. Schehl

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Figure 1. Morphologies of graphite particles.

Reprinted with permission from Elsevier. Figure 2. Basic configuration of the measuring part of the BoD tribometer [16].

Plates of MA, MB, MC, MB FC30 and MC FC30 were compression molded at 400 oC and slowly cooled to room temperature in the mold. All samples were prepared under the same compression and cooling conditions. Surely, the PEEK matrix in these five materials has semi-crystalline structures. It should be noted that using the same processing parameters, the PEEKs with different molecular weights have slightly different crystallinities: a high molecular weight corresponding to a slightly lower crystallinity [17]. Therefore, the difference of mechanical and tribological properties between the PEEKs in this work should be considered as a result of synthetic effects of molecular weight and crystallinity [18].

Tribology of Composite Materials, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

Effects of Matrix Crystalline Structure and Molecular Weight ...

127

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

2.2. Friction Tests The tribological tests for AMC and SCC were conducted on a Ball-on-Disc (BoD) tribometer (CSEM, Switzerland) at room temperature. The counterpart was a 6 mm diameter 100Cr6 steel ball with a mirror finished surface (Ra: 0.02 µm). The hardness of the 100Cr6 ball is 62 HRC. The friction force was measured with a Linear Variable Differential Transformer (LVDT) sensor and dynamically recorded into a computer. The basic configuration of the measuring part comprised in the tribometer is illustrated in Figure 2. The core and coils of the LVDT sensor are fixed with cantilever I and cantilever II, respectively. As shown in the schematic, the deformation of cantilever I occurs due to the shearing force produced during sliding process. As a consequence, the core moves in the central bore of the coils. The frictional force was measured as a function of the position of the core in the LVDT sensor. The applied load was fixed at 9 N and the sliding velocities ranged from 0.2 to 1.4 m/s. The sliding distance was 1000 m. The friction data recorded during the period 500-1000 m were averaged for calculating the mean friction coefficient. The wear rate was defined as the worn volume per unit of applied load and sliding distance. The cross-section of the worn track was obtained using a Taylor-Hobson Surtronic 3P profilometer (Rank Taylor Hobson Ltd., UK) after the friction test. The cross-section area of the wear tracks timing the perimeter permitted to obtain the total worn volume. The friction coefficients and wear rates presented in this paper are the mean values of at least three experimental data. The specimens, MA, MB, MC, MB FC30 and MC FC30, having a dimension of 4 x 4 x 12 mm3, were cut from compression molded plates. The tribological tests were performed using a block-on-ring apparatus. The counterpart was a 100Cr6 steel ring with a 60 mm diameter and a mean roughness, Ra, 0.2 µm. In order to reduce the running-in process, the specimens were “pre-worn” with grinding paper before the test (firstly P 800 and then P 1200) to an arc outline to match the configuration of the counterpart. During the test friction force was measured and dynamically recorded into a computer. The ratio between the friction force and load equals friction coefficient. For each test the friction coefficient refers to the mean value of the data recorded after running-in process. All the tests in this work were conducted for 20 hours under dry conditions at room temperature. Specimen’s mass loss, Δm , was measured after frictional test and the specific wear rate wS of the material was calculated using the equation:

wS =

Δm (mm3/Nm) ρFL

Eq. (1)

where ρ is the density of the specimen, F is the normal load applied on the specimen during sliding, and L is the total sliding distance. The inverse of wear rate is usually considered as material’s wear resistance. The sliding velocity was controlled by the rotating speed of the ring and was fixed at 1 m/s in this work. As suggested in a previous work [19], the applied load can play an important role on the tribological behavior of PEEK by influencing the depth of the surface layer involved in the friction process. Therefore, to investigate the pressure dependence of PEEK’s tribological behaviors is of importance for a deep understanding of the sliding process. In this work the apparent pressure ranged from 1

Tribology of Composite Materials, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

128

G. Zhang and M. Schehl

MPa to 4 MPa. The friction coefficients and wear rates given in this paper are mean values of at least three experimental data.

3. RESULTS AND DISCUSSIONS 3.1. Effects of Crystalline Structure on the Tribological Behavior of Graphite Particles Filled PEEK

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Figure 3a and Figure 3b respectively show the friction coefficients and wear rates of AMC and SCC as a function of sliding velocity. It is clearly seen that SCC presents lower friction coefficients and wear rates than AMC. The tribological characteristics of the two materials are sensitive to the variation in sliding velocity. In the studied range, for both AMC and SCC the increase in sliding velocity results in a decrease of friction coefficients. In contrast to the friction coefficients, the wear rates of AMC and SCC are increased when the sliding velocity is increased.

Figure 3. Effect of sliding velocity on the friction coefficients (a) and wear rates (b) of AMC and SCC.

Tribology of Composite Materials, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Effects of Matrix Crystalline Structure and Molecular Weight ...

Figure 4. Worn surface of AMC produced at 0.2 m/s with (a) low and (b) high magnifications; (c) Transfer film on the counterpart after sliding against AMC at 0.2 m/s. Tribology of Composite Materials, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

129

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

130

G. Zhang and M. Schehl

Figure 5. Worn surface of AMC produced at 0.8 m/s with (a) low and (b) high magnifications.

Figure 4a and Figure 4b show the worn track of AMC produced at 0.2 m/s at low and high magnifications, respectively. The sliding direction is indicated in the figures. At a low sliding velocity, ploughs constitute the mean feature of the worn surface. Figure 4c shows the transfer film adhering on the counterpart surface. Clearly, the polymer material was transferred to the counterpart surface during the friction process. When the asperities or protruding parts on the transfer film are pushed forward, a strain hardening of PEEK ahead of the asperities occurs [20]. As a result, the material is continually displaced sideways to form ridge adjacent to the developing plough. With increasing the sliding velocity, the worn surface presents a distinct morphology. Figure 5a and Figure 5b show the worn surface of AMC produced at 0.8 m/s.

Tribology of Composite Materials, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

Effects of Matrix Crystalline Structure and Molecular Weight ...

131

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Figure 6. Worn surface of AMC produced at 1.4 m/s.

Some cracks as indicated in the figures by arrows were noticed near graphite/matrix interfaces. These cracks are caused by the separation of graphite particles and the matrix. Figure 6 shows the AMC worn surface produced at 1.4 m/s. With increasing the sliding velocity, more cracks occur during the sliding process. Due to the higher modulus of the graphite particles than the amorphous matrix, even the graphite particles are progressively ground layer-by-layer during the sliding process, stress concentration can occur near the graphite particles in the surface layer. With repeated effects, the interfacial adhesion can be destroyed and cracks occur near the filler/matrix interface. Figure 7a, Figure 7b, and Figure 7c show the worn surfaces of SCC obtained at 0.2, 0.8, and 1.4 m/s, respectively. At low velocities, ploughs are clearly observed on the worn surface. At high velocities, compared with the worn surface of AMC, much less cracks are noticed on the worn surface of SCC. Even at 1.4 m/s, the interfacial crack hardly shows up in the worn surface (cf. Figure 7c). Moreover, as is clearly noticed, the surface becomes smoother with increasing the sliding velocity from 0.2 to 1.4 m/s. For SCC, the high matrix stiffness can improve the stress transfer between the matrix and the filler. Therefore, the stress concentration occurring near the graphite particle can be reduced. Moreover, the high matrix stiffness can also reduce the relative motion of the fillers. These two factors seem to reduce the destruction of the filler/graphite adhesion. Accordingly, the occurrence of the interfacial crack is significantly reduced. Figure 8a and Figure 8b show the transfer films on the counterparts after sliding against SCC at 0.2 and 1.4 m/s, respectively. It is clearly seen that the increase in sliding velocity significantly improves the homogeneity of the transfer film. The worn surface morphology is closely related to the transfer film morphology: a homogenous transfer film corresponds to a smooth worn surface. However, it should be noted that a homogeneous transfer film does not necessarily correspond to a low wear rate of SCC here. The high wear rate obtained by increasing the sliding velocity could be related to the high strain rate occurring in the surface layer involved in friction process.

Tribology of Composite Materials, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

132

G. Zhang and M. Schehl

Figure 7. Worn surfaces of SCC produced at (a) 0.2 m/s, (b) 0.8 m/s, and (c) 1.4 m/s.

Tribology of Composite Materials, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

Effects of Matrix Crystalline Structure and Molecular Weight ...

133

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Figure 8. Transfer films on the counterparts after sliding against SCC at (a) 0.2 m/s and (b) 1.4 m/s.

At a low strain rate a viscoelastic behavior prevails in the surface layer while at a high strain rate an elastic behavior prevails in the surface layer [19,20]. Moreover, the increase in sliding velocity provokes an increase in contact temperature. This can also increase the wear rate by decreasing material’s mechanical strength [21].

3.2. Correlation of the Tribological Behaviors with the Mechanical Properties of PEEK Materials with Different Matrix Molecular Weights 3.2.1. Mechanical Properties Figure 9a and Figure 9b show respectively the Young’s modulus and elongations at break of the MA, MB, MC, MB FC30 and MC FC 30. Compared with MA, MB presents a slightly lower Young’s modulus but a slightly higher elongation rate. MC presents the lowest Young’s modulus but the highest elongation rate. Figure 9c shows the hardness of the three pure PEEKs. As is seen, MC exhibits the lowest hardness and MA presents the highest hardness. Summarized, an increase in PEEK molecular weight corresponds to a decrease in material stiffness but an increase in material ductility.

Tribology of Composite Materials, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

134

G. Zhang and M. Schehl

Reprinted with permission from Elsevier. Figure 9. (a) Young’s modulus; (b) elongations at break; and (c) universal hardness of studied PEEK based materials [18].

Tribology of Composite Materials, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Effects of Matrix Crystalline Structure and Molecular Weight ...

135

Reprinted with permission from Elsevier. Figure 10. (a) Overview of MA fracture surface; (b) overview of the fracture surface of MC; (c) magnified observation of zone I indicated in b; (d) fiber/matrix interfaces in the fracture surfaces of MB FC30 and (e) MC FC30 [18].

Figure 10a shows the overview of the tensile fracture surface of MA. Two distinct zones, indicated respectively as zone I and zone II, were noticed on the fracture surface. Clearly, zone I corresponds to fracture initiation and zone II corresponds to fracture propagation. Figure 10b shows the overview of the fracture surface of MC and Figure 10c illustrates the zone corresponding to fracture initiation. The failures initiate from randomly distributed impurities acting as stress concentration sites at the early stage of tensile tests. The drawn

Tribology of Composite Materials, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

136

G. Zhang and M. Schehl

nature of the morphology in zone I suggests ductile deformation rather than stable crack growth. Once fracture is initiated, a fast crack takes place in zone II. Compared with MA, MC exhibits a higher ductility and therefore a longer fibration process occurs. Being filled with the multiple fillers, MB FC30 and MC FC30 present a significantly improved tensile modulus. However, the ductility of PEEK is much decreased. Figure 10d and Figure 10e show representative PEEK/SCF interfaces observed on the fracture surfaces of MB FC30 and MC FC30, respectively. The arrows in the images indicate the carbon fibers. It is clearly seen that the matrix/fiber adhesion is stronger in MC FC30 than in MB FC30.

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

3.2.2. Tribological Behaviors Friction coefficients and wear rates were summarized in Table 1. For three pure PEEKs, in studied range, the increase in apparent pressure leads to slightly higher friction coefficients. The friction coefficient seems to increase slightly, if any, with increasing molecular weight. Under low apparent pressures (1 MPa and 2 MPa), the increase in molecular weight corresponds to a lower wear resistance. Under 4 MPa, however, the increase in molecular weight corresponds to a higher wear rate. After incorporating SCF/graphite/ PTFE, the wear rates of the PEEKs are significantly decreased by 93.0% - 96.3%. Under 1 MPa and 2 MPa, incorporating the multiple fillers does not change significantly the friction coefficients of PEEKs. Under 4 MPa, however, especially for MC FC30, the incorporation of the fillers distinctly decreases the friction coefficient. The wear rates of the composites are increased by raising the apparent pressure from 1 MPa to 4 MPa. Figure 11a shows the worn surface of MA obtained under 1 MPa. The sliding direction in Figure 11 is downward. The keen-edged grooves suggest that a microcutting effect constitutes the main wear mechanism. The sliding is essentially governed by the dynamic process occurring in the surface layer involved in the friction process [22]. In the frictional layer, stress distributes in both normal and parallel directions to the sliding direction. Under a low pressure, the thickness of the PEEK surface layer involved in the frictional process is small. In this case, for a polymer with high stiffness, elastic behavior can prevail in the surface layer. Accordingly, a microcutting effect exerted by protruding regions of the counterpart constitutes the main wear mechanism. When material’s ductility is increased, the material loss due to the microcutting effect tends to decrease. Figure 11b shows the worn surface of MC. Compared with MA, the grooves on the surface of MC are alleviated. This can be the reason why the wear rate of MC is lower than that of MA. However, besides the grooves, short ripple-like deformations perpendicular to the sliding direction are noticed on the surface of MC. This morphological feature is more obvious on the worn surface of MC produced under 4 MPa (Figure 11e). These ripple-like deformations are assumed to be caused by a stick-slip motion of the counterpart. Due to the adhesion between the two sliding pairs, the PEEK surface presents a larger tangential deformation along the sliding direction than the subsurface material. Accordingly, plastic flow of PEEK surface occurs. Figure 11c shows the worn surface of MB FC30. During the frictional process, carbon fibers support most load and internal lubricants, i.e. graphite and PTFE, reduce the adhesion between the sliding pairs. Under a low apparent pressure, the fibers were ground progressively and therefore the thinning of fibers mainly determines the wear rate.

Tribology of Composite Materials, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

Effects of Matrix Crystalline Structure and Molecular Weight ...

137

Table 1. Friction coefficients and wear rates of PEEKs and their composites

Material type

MA

MB

MC

MB FC30

MC FC30

Sliding condition

Friction coefficient

Relative error (%)

6

Wear rate (10mm3/Nm)

Relative error (%)

1 MPa, 1m/s 2 MPa, 1m/s 4 MPa, 1m/s 1 MPa, 1m/s 2 MPa, 1m/s 4 MPa, 1m/s 1 MPa, 1m/s 2 MPa, 1m/s

0.34 0.38 0.41 0.37 0.39 0.39 0.37 0.41

3.28 1.36 5.63 16.10 1.18 9.49 13.14 3.44

19.69 25.85 18.10 12.41 16.37 14.14 11.59 13.00

13.87 21.57 23.18 33.93 21.60 1.70 21.63 35.62

4 MPa, 1m/s

0.42

5.55

22.30

15.62

1 MPa, 1m/s 2 MPa, 1m/s 4 MPa, 1m/s 1 MPa, 1m/s 2 MPa, 1m/s 4 MPa, 1m/s

0.35 0.41 0.37 0.36 0.41 0.27

11.93 3.30 18.87 12.05 10.14 1.80

0.51 0.75 0.99 0.60 0.75 0.83

7.64 6.53 13.36 13.16 5.24 3.36

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Reprinted with permission from Elsevier.

With increasing pressure, the thickness of the frictional layer becomes larger. Figure 11d and Figure 11e respectively show the worn surfaces of MA and MC produced under 4 MPa. With comparing the worn surfaces of the three pure PEEKs produced under 4 MPa, one can notice that the plastic flow occurring in the frictional layer is more evident for materials with a lower stiffness and a higher ductility. As is seen from Figure 11e, periodic and long ripplelike deformations are clearly observed on the worn surface of MC. For most polymers, Van der Waals and hydrogen bonds are typical factors for the junctions occurring between the two counterparts [20]. Formation and rupture of these junctions control the adhesion component. Increasing the pressure, the adhesion between the sliding pairs is increased. Due to the adhesion force, the PEEK surface presents a larger tangential deformation than the subsurface material. Once the stress applied on the polymer surface exceeds the critical stress [23,24], a slip stage initiates and runs until the stress decreases to below the critical stress, when the sliding pairs stick again [23]. Under a high pressure cracking and debonding of fibers become important factors contributing to material loss. Figure 11f shows the worn surface of MB FC30 produced under 4 MPa. Compared with the worn surface produced under 1 MPa (Figure 11c) more scratch traces are noticed on the worn surface. The scratch of the surface is caused by the cracked fibers. Under a high pressure, fractures occur in fibers where they are mostly loaded. When the broken parts are removed from the matrix, they scratch the matrix in the following sliding process. Moreover, when the broken parts collide with other fibers, they can cause further failures of fibers. Figure 11g shows the worn surface of MC FC30. Compared with MB FC30, removal of fibers is reduced and less scratch traces are observed on the worn surface of MC FC30. The better matrix/SCF adhesion can reduces the removal of fibers and thereby the wear of the material.

Tribology of Composite Materials, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

G. Zhang and M. Schehl

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

138

Tribology of Composite Materials, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Effects of Matrix Crystalline Structure and Molecular Weight ...

Tribology of Composite Materials, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

139

140

G. Zhang and M. Schehl

Reprinted with permission from Elsevier. Figure 11. Worn surfaces of MA (a), MC (b) and MB FC30 (c) produced under 1 MPa; worn surfaces of MA (d), MC (e) and MB FC30 (f) and MC FC30 (g) produced under 4 MPa [18].

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

CONCLUSION The effect of matrix crystalline structure on the tribological performance of graphite particles filled PEEK coating was studied. Two composites with an amorphous matrix (AMC) and a semi-crystalline matrix (SCC) were used. The velocity dependence of the tribological behaviors of the two composites was considered. In addition, the correlation between the mechanical properties and the tribological performances of PEEK-based materials with matrix having different molecular weights was investigated. Following conclusions can be drawn: 1. In the studied velocity range, i.e. 0.2 m/s to 1.4 m/s, SCC presents lower friction coefficients and wear rates than AMC. With increasing the velocity, the friction coefficients of AMC and SCC were decreased but the wear rates of these two materials were increased. 2. At low velocities, ploughs and material transferring can be dominant factors determining the friction processes of AMC and SCC. For AMC, at high velocities, cracks occur near graphite/matrix interfaces due to the stress concentration occurring on the graphite particles in the surface layer. As to SCC in which the matrix has a higher stiffness, the stress concentration occurring near graphite particle can be reduced. Therefore, the interface is better maintained during the sliding process. 3. Under the same compression molding and cooling conditions, the increase in molecular weight corresponds to a higher material ductility and a lower material stiffness. The tribological mechanism of PEEK is closely related to its mechanical properties. Under a low pressure, the microcutting effect exerted by the protruding region of the counterpart constitutes the main wear mechanism. In this case, an increase in the material’s ductility decreases the wear rate. Under a high pressure,

Tribology of Composite Materials, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

Effects of Matrix Crystalline Structure and Molecular Weight ...

141

however, plastic flow occurring in PEEK surface layer can be an important factor contributing to material loss. In this case, a high stiffness seems to benefit the material’s wear resistance. 4. Being filled with SCF/graphite/PTFE, PEEK exhibits a much improved wear resistance. The increase in apparent pressure increases the wear rates of the composites. Under a low pressure, thinning of fibers dominates the wear of the composites. Under a high pressure, fiber failures are important factors contributing to material loss. A high interfacial bonding between the PEEK matrix and fibers seems to be an important factor for getting a high wear resistance.

REFERENCES [1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11]

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

[12] [13] [14] [15] [16] [17] [18] [19] [20] [21] [22] [23] [24]

Bahadur S, Gong DL, Anderegg JW. Wear 1993;160:131-8. Davim JP, Marques N, Baptista AM. Wear 2001;251:1100-4. Wang QH, Xue QJ, Liu HW, Shen WC, Xu JF. Wear 1996;198:216-9. Friedrich K, Zhang Z, Schlarb AK. Compos. Sci. Technol. 2005;65: 2329-43. Voort J, Bahadur S. Wear 1995;181-183:212-21. Yamamoto Y, Takashima T. Wear 2002;253:820-6. Wang QH, Xue QJ, Liu WM, Chen JM. Wear 2000;243:140-6. Eiss Jr NS, Hanchi J. Wear 1996;200:105-21. Xue Q, Wang Q. Wear 1997;213:54-8. Hanchi J, Eiss Jr NS. Wear 1997;203-204:380-6. Schelling A, Kausch HH. In: Friedrich K, Editor. Advances in composites tribology. Amsterdam: Elsevier; 1993. p. 65. Zhang G, Liao H, Yu H, Ji V, Huang W, Mhaisalkar SG, Coddet C. Surf. Coat Technol. 2006;200:6690-5. Lu ZP, Friedrich K. Wear 1995;181-183:624-31. Voss H, Friedrich K. Wear 1987;116:1-18. Zhang G, Li W-Y, Cherigui M, Zhang C, Liao H, Bordes J-M, Coddet C. Prog. Org. Coat 2007;60:39-44. Zhang G, Yu H, Zhang C, Liao H, Coddet C. Acta Mater. 2008;56:2182-90. Zhang G. Effect of molecular weight on the crystallization kinetics of PEEK. unpublished results. Zhang G, Schlarb AK. Wear 2009;266:337-44. Zhang G, Zhang C, Nardin P, Li WY, Liao H, Coddet C. Tribol. Int 2008;41:79-86. Myshkin NK, Petrokovets MI, Kovalev AV. Tribol. Int. 2005;38:910-21. Li J, Liao H, Coddet C. Wear 2002;252:824-31. Briscoe BJ. In: Friedrich K, Editor. Friction and Wear of Polymer Composites, Amsterdam: Elsevier; 1986, p. 25. Hadal RS, Misra RDK. Mater. Sci. Eng., A 2005;398:252-61. Misra RDK, Hadal RS, Duncan SJ. Acta Mater. 2004;52:4363-76.

Tribology of Composite Materials, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved. Tribology of Composite Materials, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

In: Tribology of Composite Materials Editor: J. Paulo Davim

ISBN 978-1-62100-999-3 © 2012 Nova Science Publishers, Inc.

Chapter 6

FRICTION AND WEAR OF AL2O3-NI COMPOSITE Jinjun Lu1,∗, Junhu Meng1, Bin Liu2, Jingbo Wang1 and Shengrong Yang1 1

State Key Laboratory of Solid Lubrication, Lanzhou Institute of Chemical Physics, Chinese Academy of Sciences, Lanzhou, P.R. China 2 School of Stomatology, Lanzhou University, Lanzhou, PR China

ABSTRACT Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Al2O3-Ni powders with different volume fraction of Ni were prepared by a three-step reduction method and then hot-pressed into Al2O3-Ni composites at elevated temperatures. Scanning electron microscopy observation on the microstructure of Al2O3Ni composites indicated inter-type and intra-type of Ni particles. Considerable growth of Al2O3 grain and Ni grain was also observed. The addition of Ni resulted in reduced hardness while increased bending strength. Very limited improvement on the fracture toughness can be obtained only for composite with 15 vol.% Ni. Friction and wear of monolithic Al2O3 and Al2O3-Ni composites in sliding against a Si3N4 ball were investigated at room temperature in air. Results indicated that a transition from mild wear to severe wear was found for Al2O3/Si3N4 tribo-couple. There was no such a transition for Al2O3-Ni composites in sliding against a Si3N4 ball. At high loads, Al2O3-Ni composites exhibited much better wear resistance than that of monolithic Al2O3. Wear mechanism for monolithic Al2O3 in mild wear regime was asperity-scale failure while it was related to the formation and detachment of tribo-layer in severe wear regime. Worn surfaces of Al2O3-Ni composites at 3 N and 5 N were very smooth. At high loads, the formation and detachment of tribo-layer on the worn surfaces of Al2O3-Ni composites played very important roles in affecting the tribological behavior and wear mechanism.

∗

Corresponding author, E-mail: [email protected].

Tribology of Composite Materials, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

144

Jinjun Lu, Junhu Meng, Bin Liu et al.

1. INTRODUCTION Alumina (Al2O3) has been the focus of research and development in the past decades as tribo-material. Its tribological behaviors, e.g. friction and wear in sliding, erosion, fluids and high temperature [1-4] were extensively investigated. Study on effect of second phase on the mechanical property of Al2O3 was active in recent years. In addition, there were some interesting papers concerning the effect of second phase on tribological behavior [5]. Al2O3-based nanocomposites, e.g. Al2O3-Ni composite [6] and Al2O3-SiC composite [78] have attracted many attentions because of their excellent mechanical properties. In addition, the tribological behavior of Al2O3-SiC composite was investigated [8]. It is interesting to investigate the tribological behavior of Al2O3-metal composites [9-10], especially for composite reinforced by sub-micron or even nano metal particles. In this chapter, Al2O3 ceramics reinforced by submicron Ni particles were prepared. Their microstructure, mechanical strength and tribological behavior were investigated. Finally, wear mechanisms were discussed.

2. MATERIALS AND EXPERIMENTAL DETAILS 2.1. Materials

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

2.1.1. Preparation of Al2O3-Ni Composite Powder The Ni/Al2O3 powders with different amounts of Ni (5%, 10% and 15% in volume fraction unless otherwise stated) were prepared by a three-step reduction method as shown in Figure 1. This strategy for preparation of Al2O3-Ni composite powder in this chapter has been widely adopted for the synthesis of oxide-supported metal catalyst, which enables successful synthesis of γ-Al2O3/Ni powder with Ni nanoparticles well-dispersed on the γ-Al2O3 support. Firstly, α-Al2O3 powder (commercially available from China Building Materials Academy) with an average particle size of 0.36 μm and Ni(NO3)2⋅6H2O powder (commercially available from Baiyin Chemical Reagent Factory, China) were mixed and ballmilled in the presence of ethanol for 24 h with agate balls (step 1). And then, the mixture was calcined at 450 °C in air in an oven and followed by another ball-milling process in ethanol for 24 h to break the agglomerate (step 2). The powder was reduced in dry hydrogen at 600 °C for 1 h (step 3). The phases of products in each step were α-Al2O3 and Ni(NO3)2⋅6H2O in Figure 2a (step 1), α-Al2O3 and NiO in Figure 2b (step 2), and α-Al2O3 and Ni in Figure 2c (step 3). The following reactions were proposed for steps 2 and 3, respectively: 2Ni(NO3)2⋅6H2O(s)→2NiO(s)+4NO2(g)+O2(g)+6H2O(g)

(1)

NiO(s)+H2(g)→Ni(s)+H2O(g)

(2)

Tribology of Composite Materials, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

Friction and Wear of Al2O3-Ni Composite

145

Spherical Ni particles (10 to 40 nm in diameter) were found to be well-dispersed on αAl2O3 particles (Figure 3). The phase composition of the sintered body (Figure 2d) was the same as that of Al2O3/Ni composite powder in step 3 (Figure 2c).

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Figure 1. Schematic illustration of routine to prepare Al2O3-Ni composite powder and Al2O3-Ni composite.

Figure 2. XRD patterns of products of each step in Figure 1.

Tribology of Composite Materials, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

146

Jinjun Lu, Junhu Meng, Bin Liu et al.

Figure 3. TEM micrograph of Al2O3-5%Ni composite powder.

2.1.2. Preparation and Microstructure of Al2O3-Ni Composite After being crushed and sieved, the Al2O3-Ni composite powder was hot-pressed at 1400 °C to 1550 °C for 60 min under a pressure of 30 MPa in an argon atmosphere. The sintering temperature and schedule were optimized for Al2O3-Ni composites but is not the topic of this chapter. Monolithic Al2O3, as a reference sample, was hot-pressed at 1500 °C for 30 min. The average grain size d50 of monolithic Al2O3 was 10 μm.

Tribology of Composite Materials, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

Friction and Wear of Al2O3-Ni Composite

147

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

The average size of Ni particles in Al2O3-Ni composite increased considerably to several hundred nanometers (Figure 4) compared with that of Ni particles in Al2O3-Ni composite powders (Figure 3) . Inter-type Ni particles as well as intra-type particles were found. Intertype refers to particles located at the Al2O3/Al2O3 grain boundaries and the triple junctions while intra-type means particles located within Al2O3 grains. Grain size d50 of Al2O3 in Figure 4a was ca. 2 μm. As the content of Ni increased, the grain sizes d50 of Al2O3 with 10% and 15% Ni decreased to be less than 1 μm. Therefore, it can be deduced that Ni particles effectively inhibited the growth of Al2O3 grains.

Figure 4. Fractured surfaces of (a) Al2O3-5%Ni composite, (b,c) Al2O3-10%Ni composite, and (d) Al2O3-15%Ni composite.

Tribology of Composite Materials, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

148

Jinjun Lu, Junhu Meng, Bin Liu et al. Table 1. Grain size and mechanical strengths of monolithic Al2O3 and Al2O3-Ni composite Composition

Al2O3

Al2O3-5%Ni

Al2O3-10%Ni

Al2O3-15%Ni

Grain size of Al2O3 d50, μm

10

2

0.8

0.8

Hardness, GPa Three-point bending strength, MPa Fracture toughness, MPa m1/2

17.40±0.85

16.00±0.53

15.53±0.44

13.84±0.22

194.4±19.9

378.8±36.5

585.00±106.4

528.6±55.0

3.60±0.30

3.37±0.16

3.37±0.13

4.38±0.16

Table 2. The physical and mechanical properties of Si3N4 ball from supplier Density, g/cm3

Vickers hardness, GPa

Fracture toughness, MPa m1/2

Bending strength, MPa

Elasticity modulus, GPa

3.20~3.30

13~16

5.0~7.0

600~1000

300~320

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

2.1.3. Mechanical Property of Al2O3-Ni Composite The hot-pressed samples were machined and polished for mechanical and tribological tests. The density of the samples was determined by Archimeds’ principle using distilled water. All the samples were sintered to nearly full density. Rectangular beam samples (3×4×40 mm3) were used to measure the three-point bending strength with a span of 20 mm at a crosshead speed of 0.05 mm/min. Vickers hardness was measured using a load of 98 N and a dwell time of 5 s. Fracture toughness (KIC) was obtained by the indentation method using the equation:

K IC = P(π .b)

−3

2

(tgβ )−1

(1)

where P is the load, b is the crack length and β is 68o. Mechanical strengths of monolithic Al2O3 and Al2O3-Ni composite were listed in Table 1. The addition of Ni led to a reduction in hardness despite of much smaller grain size in Al2O3-Ni composites, as shown in Figure 4. According to Hall-Petch relationship, the hardness of fine-grain material is higher than that of coarse-grain material. This relationship has been experimentally supported in many studies. Ni is ‘softer’ than alumina and more easily plastically deformed. Then it is not surprising that the reduction in hardness can be attributed to Ni phase in the composite. As seen in Table 1, the bending strength of Al2O3 can be markedly improved with the addition of Ni. Composite with 10% Ni had the highest bending strength. The initial idea of introducing Ni particle is to prompt the toughness of Al2O3. However, very limited improvement on the fracture toughness can be obtained only for composite with 15% Ni. There were no improvement in cases of Al2O3-5% Ni and Al2O3-10% Ni composites.

Tribology of Composite Materials, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

Friction and Wear of Al2O3-Ni Composite

149

According to Figure 4, Al2O3-Ni composites showed the mixed mode of transgranular and intergranular fractures.

2.2. Friction and Wear Test The tribological tests were conducted on a reciprocating tribometer (UMT-2MT, USA) with a ball-on-disk configuration. The upper specimen was a commercially available Si3N4 ball with 3 mm in diameter (Shanghai Research Institute of Materials), and the lower specimen was an Al2O3-Ni composite disk with a size of 25 mm in diameter and 8 mm in thickness. An Al2O3 disk was used as a reference. The surface roughness (Ra) of Si3N4, Al2O3 and Al2O3-Ni composites was about 0.020 μm. The mechanical properties of Si3N4 balls are listed in Table 2. Prior to commencing a tribological test, the specimens were ultrasonically cleaned in an alcohol bath and allowed to dry. The test condition was as follow: a sliding speed of 1 m/s, a stroke of 5 mm and a sliding distance of 5400 m under different normal loads in the range of 3 to 20 N The measured relative humidity and ambient temperature were 30 to 40% and 25 °C, respectively. Wear volumes were calculated by measuring the worn volume of the ball and the area of wear track cross section on the disk, using an optical microscopy and a surface profilometry. The friction coefficients were recorded continuously during the test by a computer. The worn surfaces of monolithic Al2O3, Al2O3-Ni composite and wear debris were analyzed on JSM-5600LV scanning electron microscopy (SEM) equipped with energy dispersive spectroscopy (EDS).

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

3. RESULTS AND DISCUSSION 3.1. Friction Coefficient Figure 5 shows the friction coefficient of monolithic Al2O3 and Al2O3-Ni composites in sliding against Si3N4 under different loads. Friction coefficients of monolithic Al2O3 sliding against Si3N4 were in the range of 0.26 to 0.55. In case of Al2O3-5%Ni against Si3N4, friction coefficients were 0.60 at 3 N and around 0.40 at loads of 5 to 20 N. Friction coefficients of Al2O3-10% Ni and Al2O3-15% Ni against Si3N4 at 5 N and 10 N were higher than that of Al2O3-5% Ni composite.

3.2. Wear Rate The wear rates of monolithic Al2O3 in Figure 6a indicated a transition from mild wear to severe wear as a function of load . The critical load for this transition was between 5 N and 10 N. Likewise, similar transition of wear was found for the counterpart Si3N4 (Figure 6b). The wear rates of the three Al2O3-Ni composites at loads of 3 N and 5 N were higher than that of monolithic Al2O3 while the wear rates at loads of 10 N, 15 N and 20 N were much

Tribology of Composite Materials, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

150

Jinjun Lu, Junhu Meng, Bin Liu et al.

lower than that of monolithic Al2O3, see Figure 6. No transition from mild wear to severe wear was found for Al2O3-Ni composites at given loads. In summary, Al2O3-Ni composites exhibited better wear resistance and higher critical load for transition from mild wear to severe wear than that of monolithic Al2O3.

3.3. Wear Mechanisms 3.3.1. Monolithic Al2O3

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Wear mechanisms for monolithic Al2O3 sliding against Si3N4 in mild wear regime and severe wear regime will be discussed on two aspects, i.e. topography and chemical composition of worn surface of monolithic Al2O3 as well as topography of wear debris. Figure 7 shows SEM micrographs of the worn surface of monolithic Al2O3 and corresponding wear debris at 3 N. The wear of monolithic Al2O3 at 3 N was so low that the cavities, which were the result of grain pulloff during grinding and polishing, still can be seen on the worn surface (Figs. 7a and 7b).

Figure 5. Friction coefficient of monolithic Al2O3 and Al2O3-Ni composites in sliding against Si3N4 under different loads.

Tribology of Composite Materials, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

Friction and Wear of Al2O3-Ni Composite

151

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

The wear debris was agglomerates up to several hundreds of microns (Figure 7c) and actually very fine particle (Figure 7d). Most of them were less than 1 μm in size, which was much lower than the grain size of Al2O3 . This means that grains were gradually removed in asperity-scale and corresponding wear was low. EDS result showed that no transfer and mechanical mixing occurred.

Figure 6. Wear rates of (a) monolithic Al2O3 and Al2O3-Ni composites and (b) Si3N4 as a function of normal load.

3.3.2. Al2O3-Ni Composite The worn surface of Al2O3-5%Ni composite at 3 N was very smooth and free of cavities (Figure 9a). For Al2O3-Ni composites, this was the typical characteristic of the worn surfaces under loads of 3 N and 5 N. At 10 N, fracture can be found on the worn surface of Al2O35%Ni composites (Figure 9b) and was enhanced at loads of 15 N (Figure 9c) and 20 N (Figure 9d). The critical loads for fractures on the worn surfaces of Al2O3-10%Ni and Al2O315%Ni composites were 10 N and 15 N, respectively. Under loads higher than the critical load, smooth area (Figure 9e) and rough area (Figure 9f) can be found on the worn surface, which was similar to that in Figure 8b. However, the number of cracks in the smooth area in Figure 9e was much less than that in Figure 8b. The typical wear debris at 20 N was plate-like (Figure 10a) and their typical size was several tens of microns, which was much larger than the typical grain size of Al2O35%Ni composite. The structure of some wear debris revealed that they were detached by subcritical propagation, Figure 10b.

Tribology of Composite Materials, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

152

Jinjun Lu, Junhu Meng, Bin Liu et al.

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Figure 7. SEM micrographs of (a,b) worn surfaces of monolithic Al2O3 and (c,d) wear debris at 3 N. Black line in Figure 7a is the boundary for wear track and non-wear track.

The topographies of the worn surfaces of monolithic Al2O3 at loads of 10 N (Figure 8a) and 15 N (Figure 8b) were totally different to that at 3 N and 5 N. The worn surfaces were characterized by two distinct areas: smooth area (tribo-layer, Figure 8c) and rough area (Figure 8d). The tribo-layer was on top of the rough area. The area fraction of the tribo-layer to the total worn surface increased from 10 N (Figure 8a) to 15 N (Figure 8b). Both elements Al and Si were found in the tribo-layer and considered as mechanically mixed layer (EDS result). The rough area was characterized by sharp grains and fine fragments (Figure 8d). The fragments generated from both Al2O3 and Si3N4 were compacted and sintered at high local pressure and speed. This might be the formation mechanism of tribo-layer. The detachment of wear debris (platelike, Figures 8e and 8f) was attributed to the nucleation and propagation of cracks in the tribolayer (Figure 8c). There were two types of wear debris in Figures 8e and 8f, i.e. fine particles similar to that in Figures 7c and 7d, and large plate-like particles up to several tens of microns. Apparently, the generation of wear debris was time-dependent. Wear mechanisms of monolithic Al2O3 and Al2O3-Ni composites in sliding against Si3N4 under high loads were related to formation and failure of tribo-layer. The composition, microstructure, mechanical properties of tribo-layer varied from monolithic Al2O3 to Al2O3Ni composites. They also depended on content of Ni in the composites. As such, it is understandable that various modes of detachment of tribo-layer in Figures 8, 9 and 11. To reveal the composition, microstructure, mechanical properties of tribo-layer, a lot of work should be done in the future.

Tribology of Composite Materials, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Friction and Wear of Al2O3-Ni Composite

Figure 8. SEM micrographs of worn surfaces of monolithic Al2O3 at (a) 10 N, (b,c,d) 15 N and (e,f) wear debris at 15 N. Tribology of Composite Materials, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

153

Jinjun Lu, Junhu Meng, Bin Liu et al.

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

154

Figure 9. SEM micrographs of worn surfaces of Al2O3-5%Ni composite at (a) 3 N, (b) 10 N, (c) 15 N and (d,e,f) 20 N.

Tribology of Composite Materials, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

Friction and Wear of Al2O3-Ni Composite

155

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Figure 10. SEM micrographs of wear debris of Al2O3-5%Ni composite sliding against Si3N4 at 20 N.

Figure 11. SEM micrographs of worn surfaces of Al2O3-15%Ni composite at (a,b) 15 N and (c,d) 20 N.

Tribology of Composite Materials, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

156

Jinjun Lu, Junhu Meng, Bin Liu et al.

CONCLUSION Microstructure of Al2O3-Ni composites was characterized by inter-type and intra-type of Ni particles in Al2O3 matrix. The addition of Ni resulted in reduced hardness while increased bending strength. A transition from mild wear to severe wear with increased load was found for Al2O3/Si3N4 tribo-couple. There was no such a transition for all Al2O3-Ni composites in sliding against a Si3N4 ball. At high loads, Al2O3-Ni composites exhibited much better wear resistance than that of monolithic Al2O3. The formation and detachment of tribo-layer were the key factors influencing the tribological behavior and wear mechanisms of monolithic Al2O3 and Al2O3-Ni composites.

REFERENCES Jahanmir S. Friction and wear of ceramics. Marcel Dekker, Inc: New York, US, 1994. Tomlinson W.J.; Matthews S.J.; Ceram. Inter. 1994, vol20, 201-209. Hsu S.M.; Shen M. Wear 2004, vol256, 867-878. Adachi K.; Kato K.; Chen N. Wear 1997, vol203-204, 291-301. Pasaribu H.R.; Sloetjes J.W.; Schipper D.J. Wear 2003, vol255, 699-707. Sekino T.; Nakajima T.; Ueda S.; Niihara K. J. Am. Ceram. Soc. 1997, vol80, 11391148. [7] Niihara K. J. Ceram. Soc. Jpn. 1991, vol99, 974-982. [8] Chen J.; Rainforth W.M.; Lee W.E. Scrip. Mater. 2000, vol42, 555-560. [9] Portu de G.; Guicciardi S.; Melandri C.; Monteverde F. Wear 2007, vol262, 1346-1352. [10] Scheppokat S.; Hannink R.; Janseen R.; Portu de G.; Claussen N. J. Euro. Ceram. Soc. 2005, vol25, 837-845.

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

[1] [2] [3] [4] [5] [6]

Tribology of Composite Materials, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

In: Tribology of Composite Materials Editor: J. Paulo Davim

ISBN 978-1-62100-999-3 © 2012 Nova Science Publishers, Inc.

Chapter 7

MODELING AND ANALYSIS ON WEAR BEHAVIOUR OF METAL MATRIX COMPOSITES K. Palanikumar∗, T. Rajasekaran and J. Paulo Davim 1

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Sri Sairam Institute of Technology, Chennai, India 2 Department of Mechanical Engineering, S.R.M. University, Kattankulathur, Chennai, India 3 Department of Mechanical Engineering, University of Aveiro, Campus Santiago, Aveiro, Portugal

ABSTRACT This chapter presents the experimental investigation on wear loss of fabricated silicon carbide particle reinforced aluminium metal matrix composites. Experiments were conducted on pin-on-roller wear tester at various conditions. The parameters considered for the experiments were %volume fraction of SiC, speed, and load. The response considered for the analysis was wear loss. An empirical model has been developed for predicting the wear loss of Al/SiC MMC composites. Response surface regression and analysis of variance (ANOVA) are used in order to study the effects of tribological parameters. The influences of different parameter in wear loss of Al/SiC particulate composite have been analyzed in detail and presented in this study.

1. INTRODUCTION Metal matrix composites (MMC) are attractive materials because of their high specific strength, stiffness and wear resistance. Among modern composite materials, particulate reinforced MMCs are finding increased applications due to their favourable mechanical properties. SiC reinforced aluminium is considered widely and other compositions for the ∗ Corresponding author, E-mail: [email protected].

Tribology of Composite Materials, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

158

K. Palanikumar, T. Rajasekaran and J. Paulo Davim

matrix are available commercially [1]. When the sole objective of an engineer in the design stage is to bring down all the transmission of systems with the lowest possible frictional and wear losses, in most of the situations friction and wear are considered to be an annoyance, except some exceptions namely brakes and clutches, etc., wherein friction and wear plays a indispensable role. But wear is given an important position in the aspects of tribology [2]. Though perfect accuracy could not be achieved due to unavoidable reasons, surfaces in engineering components seldom possess perfect geometry no matter how anxiously and expensively it is prepared because there will always be some sort of roughness over these surfaces which can be made apparent only when a higher magnification examination is exercised. In the case of these surfaces made to contact each other, it is the interest of tribologists to attempt in quantifying and predicting the nature of contact, its performance and its behavior for which it is designed. Most of the engineering products are found to be either deteriorate gradually or fails to function catastrophically because of the surface related problems such as wear, corrosion, etc. Metal matrix composite materials combine the properties of reinforcement possessing hard and brittle nature and the matrix possessing ductile and toughness nature so that to become a potential material for a wide range of applications. Due to their tribological as well as mechanical properties, aluminium matrix composites find exhaustive applications to name a few aerospace and automotive industries and also it is considered to be the most important material when reinforced with particulates [3]. Many parameters influence the wear characteristics of a metal matrix material but important parameters on which researchers concentrate are speed and load. In a technical report Anoop [3] has mentioned that the parameter that influences most is temperature and then the load. Also he pointed out that lower wear rate was recorded when applying lower loads considering 15% of SiCp as reinforcement and this may be due to the dominant oxidation wear mechanism than the adhesion mechanism. When Al is reinforced with SiC particle an improved wear resistance was evidenced on comparing an unreinforced one during dry sliding [4,5], but it is found that the rate of wear is decreasing when volume fraction of SiC is increased [6]. In this respect, Mehmet acilar [7] also agrees with Rao [6] and also adding that the wear rate of aluminium metal matrix composites reinforced with SiC particles produced by infiltration technique increases on increasing sliding speed as well as applied load. Aradhya [8] developed a finite element model for the purpose of predicting stresses causing localized yielding with respect to volume fraction of particulates and a model for change in fracture stress with the presence of SiC particles. When SiC reinforced 2124Al was studied for abrasive behavior of aged and non-aged specimens under the temperature range of 20 - 200۫ C the mass loss was found to be less at ambient temperature when compared with other temperatures. The wear mechanism found to be involved here was loosening particles causing micro cracking then leading to break off of particles. Further the author observed that there was no significant effect on wear rate due to the change of temperature [9]. When 4147Al alloy was taken with the comparative reinforcement of SiC and B4C particles for testing adhesion wear, the alloy with SiC particles have shown the better result on wear resistance in comparison with the effect which was shown by B4C particles [10]. Basavarajappa [11] in a short communication investigated to find out the significance of the design parameters during dry sliding wear quotes that the mass loss increases when the load is increased due to the fact that the SiC is quite strong during compression than that in the tension causing the transfer of material from the pin when load is increased. Shaoyang

Tribology of Composite Materials, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Modeling and Analysis on Wear Behaviour of Metal Matrix Composites

159

Zhang [12] conducted dry sliding test to investigate the friction and wear behavior of same brake material against two aluminium matrix materials with different sizes of reinforcement namely 3.5 and 34µm and observed better performance with the latter one than the former one. When Modi [13] attempted to investigate two squeeze cast Al alloys one with SiC particles reinforced and the other with SiC fiber reinforced for their resistance to dry wear under different pressures he observed lower wear rate for the alloy with SiC particle reinforced than the fiber reinforced one. Aluminium metal matrix composites presented still better wear resistance when it is reinforced by duplex reinforcement with SiC and carbon fiber than the one reinforced with SiC particle alone [14]. Chen Zhenhua [15] conducted experiment on spray deposited AlSi/SiC composites to sort out the effect of the silicon content and thermo-mechanical treatment on the dry sliding wear behavior and agreed that the rate of wear decreases with the increase of silicon content and adds that an appreciable change in wear resistance can be achieved through thermo-mechanical treatment. Sun Zhiqiang [16] evaluated the dry sliding wear behavior of silicon particles reinforced aluminium matrix composites illustrates the reason for change in wear trend with varying load is that inadequate bonding at interface of aluminium matrix composite and silicon carbide particles. It is interesting to note that Kwok [17] noticed distinctly three segments of friction and wear behavior in which initially the rate of wear is found to be less, then instant catastrophic failure taking place when critical speed is experienced and finally occurrence of melting of material respectively. The use of harder inter-metallic materials as reinforcements also found to be useful in improving the performance of metal matrix materials in terms of wear resistance [18]. Ranjit Bauri [19] recorded mild wear during the application of lower loads and severe wear at higher loads when he studied the sliding behavior of Al-Li-SiCp composites. When I˙zciler [20] carried out experiment to study the wear behavior of SiC reinforced Al alloy composite with two abrasives SiC and Al2O3 concluded that the former offers higher rate of wear when compared to the latter one. This may be due to the higher hardness it possesses. From the above studies, it is noticed that studies on wear behaviour of metal matrix composites is an important area of research. In the present chapter, an attempt has been made to study the wear behaviour of LM25 reinforced with different volume fraction of SiC particles at different loads and speeds for various volume fraction of SiC. The experiments are conducted in a comprehensive way using Taguchi’s design of experiments. Response surface analysis is carried out to study the wear loss of Al/SiC composites. The analysis of results is carried out using analysis of variance and presented in this study.

2. EXPERIMENTAL INVESTIGATION The experiments are planned using Taguchi’s orthogonal array in the design of experiments (DoE), which helps in reducing the number of experiments. The experiments are conducted according to a 3-level L27 orthogonal array. The parameters considered for the present investigation are: (1) % volume fraction of SiC, (2) Speed, (3) load, out of which % volume fraction of SiC is specially applied to MMC materials. The interactions may also play some role in deciding the wear of materials. After careful analysis and screening, only the square effects and two factor interactions are considered. Once examined factors and

Tribology of Composite Materials, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

160

K. Palanikumar, T. Rajasekaran and J. Paulo Davim

interactions are determined, the next step is to determine how many values have to be examined for each factor. Since all three factors are multi level variables and their outcome effects are not linearly related, it is decided to use three level tests for each factor [21]. Possible limits are chosen in order to observe the wear loss on the work piece material. The parameters used and their values are given in Table 1. Table 1. Control parameters and their levels used for experiments Notation

Parameters

V

% Volume fraction of SiC

Levels 1 25

S

Speed, rpm.

300

400

500

L

Load, N

29.43

39.24

49.05

2 15

3 10

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

By using Taguchi’s orthogonal array [21,22], the most suitable array is L27, which needs 27 runs and has 26 degrees of freedoms (DOF). It can conduct three values of parameters. To check the DOF in the experimental design, for the three values test, the three main factors take 6 DOFs (3x2) and the remaining DOFs are taken by interactions. The 3 level L27 orthogonal array is shown in Table 2, where the numbers 1, 2 and 3 stand for the values of the factors. This array specifies 27 experimental runs and has 13 columns. To avoid aliasing and overlap of the interactions with the main factors, the factors were assigned to the L27 columns as in Table 2 according to L27’s linear graph shown in Figure 1. The columns chosen for the main factors are 1, 2, and 5.

Figure 1. L27 Linear graph. Tribology of Composite Materials, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

Modeling and Analysis on Wear Behaviour of Metal Matrix Composites

161

Table 2. L27 Orthogonal array used for experiments Column numbers

Trial No. 1 2 3 4 5 6 7 8 9 10 11 12 13

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

14 15 16 17 18 19 20 21 22 23 24 25 26 27

V

S

VxS

VxS

L

VxL

VxL

SxL

-

-

SxL

-

-

1

2

3

4

5

6

7

8

9

10

11

12

13

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

2

2

2

2

2

2

2

2

2

1

1

1

1

3

3

3

3

3

3

3

3

3

1

2

2

2

1

1

1

2

2

2

3

2

3

1

2

2

2

2

2

2

3

3

3

1

3

1

1

2

2

2

3

3

3

1

1

1

2

1

2

1

3

3

3

1

1

1

3

3

3

2

3

2

1

3

3

3

2

2

2

1

1

1

3

1

3

1

3

3

3

3

3

3

2

2

2

1

2

1

2

1

2

3

1

2

3

1

2

3

1

3

1

2

1

2

3

2

3

1

2

3

1

2

1

2

2

1

2

3

3

1

2

3

1

2

3

2

3

2

2

3

1

1

2

3

2

3

1

3

1

3

2

2

3

1

2

3

1

3

1

2

1

2

1

2

2

3

1

3

1

2

1

2

3

2

3

2

2

3

1

2

1

2

3

3

1

2

2

2

2

2

3

1

2

2

3

1

1

2

3

3

3

3

2

3

1

2

3

1

2

2

3

1

1

1

1

3

1

3

2

1

3

2

1

3

2

1

2

1

3

1

3

2

2

1

3

2

1

3

2

3

2

3

1

3

2

3

2

1

3

2

1

3

1

3

3

2

1

3

1

3

2

2

1

3

3

3

3

3

2

1

3

2

1

3

3

2

1

1

1

1

3

2

1

3

3

2

1

1

3

2

2

2

2

3

3

2

1

1

3

2

3

2

1

2

1

2

3

3

2

1

2

1

3

1

3

2

3

2

3

3

3

2

1

3

2

1

2

1

3

1

3

1

Tribology of Composite Materials, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

162

K. Palanikumar, T. Rajasekaran and J. Paulo Davim Table 3. Experimental Results

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Trial No.

Wear loss x 10-3, g/min*.

% Volume fraction of SiC (V)

Speed, rpm. (S)

Load (L), N

1.

25

300

29.43

1.33

2.

25

300

39.24

2.36

3.

25

300

49.05

2.82

4.

25

400

29.43

1.81

5.

25

400

39.24

3.76

6.

25

400

49.05

5.71

7.

25

500

29.43

2.08

8.

25

500

39.24

5.93

9.

25

500

49.05

7.69

10.

15

300

29.43

0.97

11.

15

300

39.24

1.83

12.

15

300

49.05

3.15

13.

15

400

29.43

1.46

14.

15

400

39.24

3.44

15.

15

400

49.05

3.31

16.

15

500

29.43

1.85

17.

15

500

39.24

3.11

18.

15

500

49.05

4.85

19.

10

300

29.43

0.72

20.

10

300

39.24

2.31

21.

10

300

49.05

3.08

22.

10

400

29.43

2.36

23.

10

400

39.24

3.28

24.

10

400

49.05

4.95

25.

10

500

29.43

1.52

26.

10

500

39.24

2.86

27.

10

500

49.05

5.05

*

Average of 3 results.

LM 25 aluminium alloy confirms to BS 1490:1988 LM 25 (7Si 0.33Mg 0.3Mn 0.5Fe 0.1Cu 0.1Ni 0.2Ti) reinforced with green bonded silicon carbide particles having average dimension of 25 μm with different volume fractions manufactured through stir casting route is used for experimentation. The microstructure of the specimens used in this work is

Tribology of Composite Materials, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

Modeling and Analysis on Wear Behaviour of Metal Matrix Composites

163

presented in Figure 2. The experiments are conducted in a pin-on-roller wear tester. The wear behavior of the composite against the hardened steel roller has been evaluated. The diameter and width of the hardened steel roller used in the wear tester is 60 mm and 12 mm, respectively. The tests are carried out in dry conditions, for five minutes for each specimen. The wear mass loss of the samples is determined by an electronic balance with an accuracy of 0.001 mg. The wear rate of the specimen has been obtained by measuring the weight loss of the specimen. Figure 3 shows the experimental setup used for experimentation

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Figure 2. Microstructure of the specimens.

Figure 3. Friction and wear testing machine used for the experimentation.

The wear loss is calculated by using the following relation:

Weightloss =

Weight Before − Weight After Time

, g / min

(1)

The experimental parameters used and the corresponding responses are given in Table 3. Tribology of Composite Materials, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

164

K. Palanikumar, T. Rajasekaran and J. Paulo Davim

3. MODELING OF TRIBOLOGICAL PARAMETERS USING RESPONSE SURFACE APPROACH In engineering applications modeling of process parameters is an important criterion. Mathematical models are used to find the relation between the input parameters and output response. For modeling and analysis, different techniques are used. Response surface method (RSM) is one of the important reliable modeling techniques based on statistics used in many engineering applications which is used for finding the relation between the input process parameters and output response wear loss for this present work. In many engineering fields, there is a relationship between a output variable of interest ‘y’ and a set of controllable variables {x1, x2, ……… xn}. In some systems, the nature of the relationship between y and x values might be known. Then, a model can be written in the form

y = f ( x1, x 2 ,... x n ) + ε

(2)

where ‘ ε ’ represents the noise or error observed in the response y. If we denote the expected response be E(y) = f ( x1, x 2 ,... x n ) = η

(3)

then the surface represented by

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

η = f ( x1, x 2 ,... x n )

(4)

is called response surface. In most RSM problems, the form of relationship between the response and the independent variable is unknown. Thus the first step in RSM is to find a suitable approximation for the true functional relationship between y and the set of independent variables is employed. Usually a second order model is utilized in response surface methodology [23]. k

k

i =1

i =1

y = β 0+ ∑ β i xi + ∑ β ii xi2 + ∑∑ β ij xi x j + ε i

(5)

j

The β coefficients, which should be determined in the second order model, are obtained by the least square method. The response surface methodology can be used to find the values of the controllable parameters that results in optimization of response or discover what values for the x values will result in a product (process) satisfying several requirements or specifications. [24]. The second order response surface representing the wear loss (W, g/min) can be expressed as a function of tribological parameters such as % volume fraction of SiC (V), speed (S), and load (N). The relationship between the wear loss and related parameters are expressed as follows:

Tribology of Composite Materials, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

Modeling and Analysis on Wear Behaviour of Metal Matrix Composites W=

165

β 0 + β1 (V) + β 2 (S) + β 3 (L) + β 4 (V2)+ β 5 (S2)+ β 6 (L2) + β 7 (VS)

+ β 8 (VL) + β 9 (SL)

(6)

The analysis of variance (ANOVA) method is used to evaluate the confidence interval and adequacy of the model. Analysis of variance essentially consists of partitioning the total variation in an experiment into components ascribable to the controlled factors and error. Table 4 shows the results of ANOVA. In ANOVA table, the sum of squares is used to estimate the square of deviation from the grand mean. Mean squares are estimated by dividing the sum of squares by degrees of freedom. F-ratio is an index used to check the adequacy of the model in which calculated value of F should be greater than the F-table value. The model is adequate at 95% confidence level since the F calculated value is greater than the F-table value. Finally, the values of R2 represent the coefficient of correlation (R2 value for the present investigation is 92.59%) The larger value of R2 is always desirable [22]. In view of the analysis results shown in, it can be concluded that the confidence of the regression model is satisfactory with a value larger than 95%. Based on the above analysis, the final model obtained is as follows: Wear loss (W)* 10-3 = +2.07825 -0.71292 * V + 6.78650E-003* S +0.024560 * L +0.010262 * V2 -3.67939E-005* S2 -1.75945E-003 * L 2 +6.86930E-004* V* S +3.40510E -003* V * L +5.17233E-004* S * L

(7)

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

where W= wear loss(g/min); V= % volume fraction of SiC, S = speed, rpm, and L = load (N). The diagnostic checking of developed model can be checked by residual analysis. Residual is the difference between the observed values and predicted or fitted values. Table 4. Adequacy checking of response surface model Source Regression Linear Square Interaction Residual Error Total

DF 9 3 3 3 17 26

Seq SS 67.646 57.9465 2.5114 7.1881 5.4163 73.0623

Adj SS 67.64598 3.90941 2.51141 7.18807 5.41627

Adj MS 7.51622 1.30314 0.83714 2.39602 0.3186

F 23.59 4.09 2.63 7.52

P 0 0.023 0.084 0.002

The normal probabilities of residuals are shown in Figure 4. The normal probability plot is used to verify the normality assumption. As shown in figure (Figure 4), the data are spread roughly along the straight line. Hence, it can be concluded that the data are normally distributed [25]. Figure 5 shows predicted results against the actual results. Figure 5 is used to show the correlation between the results. From Figure 5, it is asserted that the predicted results are very close to the experimental results and hence the response surface models are suitable for predicting wear loss of Al/SiC-MMC composites.

Tribology of Composite Materials, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

166

K. Palanikumar, T. Rajasekaran and J. Paulo Davim

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Figure 4. Normal Probability plot of residuals.

Figure 5. Correlation graph.

4. RESULTS AND DISCUSSION Metal matrix composites composed of two different matrix such as soft aluminium matrix and hard abrasive SiC particles. MMCs have been able to replace conventional monolithic alloys in applications were energy saving and light weight are important concern. The presence of hard abrasive SiC particles improves the tribological properties of MMC materials. Further MMCs have good specific strength and specific modulus, this make them good candidate materials for many engineering applications where sliding contact is expected. All mechanical components that undergo sliding or rolling contact, such as bearings, gears, seals, guides, piston rings, splines, brakes and clutches, are subject to some degree of wear. Wear is a surface phenomenon that occurs by the displacement and detachment of material, because it usually implies a progressive loss of weight and alteration of dimensions over a period of time. [26]. Wear surface analysis of MMC materials is carried out by optical microscope. Figure 6 shows the microstructure of the specimens taken at different load for 10% volume fraction

Tribology of Composite Materials, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

Modeling and Analysis on Wear Behaviour of Metal Matrix Composites

167

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

SiC composites. When the applied load on the specimen is low as shown in Figure 6(a), the wear surfaces are reduced. At low load, the plastic deformation is comparatively low and it produces smaller wear marks. The wear surface indicates the abrasive wear probably due to the abrasive action of the SiC material. Figure 6(b) shows the microstructure of the specimen taken at maximum applied load. More local eroded area are observed in the figure. The wear grooves size is more and is different for high load. Figure 7 shows the microstructure of the specimen with 15% volume fraction of SiC and Figure 8 shows the microstructure of 25% volume fraction of SiC. The figures shows the same trend as that of the 10% SiC. The increase of load increases the wear on the specimens. Also the figures 6 – 8 indicates that the amount of wear loss is more for 10% volume fraction SiC whereas 25% volume fraction of SiC shows less wear loss. From the figures it has been asserted that the increase of load increases the wear loss and increase of SiC percentage reduce the wear loss in using MMC composites.

Figure 6. Wear surface at minimum and maximum load for 10% volume SiC composites.

Figure 7. Wear surface at minimum and maximum load for 15% volume SiC composites. Tribology of Composite Materials, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

168

K. Palanikumar, T. Rajasekaran and J. Paulo Davim

Figure 8. Wear surface at minimum and maximum load for 25% volume SiC composites.

The influence of different parameters on wear loss is analysed by using analysis of variance. Table 5 illustrates the analysis of results to be used for finding the significance of the three factors, its square effects and their interactions affecting the wear loss of the Al/SiC composites. Table 5 is obtained by fitting a second order quadratic response surface regression model. In order to determine the significance of the individual parameters, F-ratio is used. The larger the absolute value of the F-ratio the more significant the factor will be.

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Table 5. Analysis of variance for factors Source

Sum of Squares

DF

Mean Square

F-Value

Prob > F

Model

67.65

9

7.52

23.62

< 0.0001

V

3.02

1

3.02

9.5

0.0068

S

16.53

1

16.53

51.95

< 0.0001

L

39.82

1

39.82

125.11

< 0.0001

2

1.52

1

1.52

4.79

0.0429

2

0.81

1

0.81

2.55

0.1286

2

0.17

1

0.17

0.54

0.4723

VS

3.3

1

3.3

10.38

0.005

VL

0.78

1

0.78

2.45

0.1356

SL

3.09

1

3.09

9.71

0.0063

Residual

5.41

17

0.32

Cor Total

73.06

26

V S

L

Tribology of Composite Materials, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

Significant

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Modeling and Analysis on Wear Behaviour of Metal Matrix Composites

Figure 9. (Continued)

Tribology of Composite Materials, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

169

170

K. Palanikumar, T. Rajasekaran and J. Paulo Davim

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Figure 9. Contour graph of wear loss at different volume fraction of SiC.

Figure 10. 3-D response graph of wear loss for varying parameters speed and load.

Tribology of Composite Materials, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

Modeling and Analysis on Wear Behaviour of Metal Matrix Composites

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Figure 11. 3-D response graph of wear loss for varying parameters Vol% SiC and load.

Figure 12. 3-D response graph of wear loss for varying parameters Vol% SiC and speed.

Tribology of Composite Materials, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

171

172

K. Palanikumar, T. Rajasekaran and J. Paulo Davim

The probability represents the probability of the coefficient, the smaller the value the more significant it represents [27]. From the table, it can be asserted that load is the main factor which influences the wear loss of MMC materials followed by speed and % volume fraction. Among the interactions considered the interaction between the parameters VS and SL are influencing the wear loss of Al/SiC MMC composites. Eq. (7) is plotted in Figure 9 as contours for each of the response surfaces by keeping the % volume fraction of SiC at different level. These response contours can help in the prediction of the wear loss at any zone of the experimental domain. It is clear from these figures that the wear loss reduces with the increase of % volume fraction of SiC. However, it increases with the increase of speed and load. Figures 10-12 show the effect for two varying parameters by keeping the third variable at middle level. Figure 10 shows the effects of speed at different load on the wear loss. With a fixed value of speed the wear loss increases with the increase of load. During sliding, at higher speed, more material flow along with SiC particles has been noticed which in turn produced high wear loss. From the figure, it can be concluded that low speed and load are preferred. Figure 11 shows the effects of % vol SiC at different load on wear loss. With a fixed value of % SiC the wear loss increases with the increase in load. Figure 12 shows the effects of vol % SiC against load on wear loss. The observed wear loss is better only at high vol % SiC. The high volume % SiC improve the resistance against wear and hence the wear loss is minimal at high volume fraction of SiC. From the analysis, it is found that the load is more significant factor than other parameters. Further more, the wear loss reduces with the increase of %vol SiC and it increases as the speed increases and load increases. The wear loss produced on the Al/SiC work piece is mainly due to the load.

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

CONCLUSION The wear loss of Al/SiC metal matrix composite material has been investigated according to the orthogonal array in experimental design. Based on the experimental and analytical results, the following conclusions are drawn: 1. The effect of tribological parameters on the wear loss is evaluated with the help of pin-on roller wear tester. Taguchi method of experimental design is used. 2. The analysis of wear loss on % volume SiC is carried out using contour graphs. 3. The analysis of tribological parameters have been carried out using 3-dimensional surface plots. 4. The results indicated that load and speed are the dominant factors which influence the wear loss of Al/SiC metal matrix composites. 5. A second order response surface model for wear loss is established from the observed data. The predicted values and measured values are fairly close to each other, which indicate that the developed model can be effectively used to predict the wear loss of Al/SiC composites with 95% confidence intervals. Using such model a remarkable saving in time and cost has been obtained.

Tribology of Composite Materials, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

Modeling and Analysis on Wear Behaviour of Metal Matrix Composites

173

6. The results revealed that the interaction between the parameters % volume of SiC and speed (VS); and the interaction between the parameters speed and load (SL) also influence the tribological behavior of Al/SiC metal matrix composites. 7. The increase in % volume fraction of SiC reduces the wear loss of Al/SiC composite materials.

REFERENCES [1]

[2] [3]

[4] [5] [6]

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

[7]

[8]

[9]

[10]

[11]

[12]

[13]

C. Antonio, C.A. and Davim, P.J. (2002). Optimal Cutting Conditions in Turning of ParticulateMetal Matrix Composites Based on Experiment and a Generic Search Model. Composites Part A,33:213–219. D. Dowson, Wear oh where, Wear, 103 (1985) 189 - 203. S. Anoop, S. Natarajan, S.P. Kumaresh Babu, Technical Report, Analysis of factors influencing dry sliding wear behaviour of Al/SiCp–brake pad tribosystem, Materials and Design 30 (2009) 3831–3838. B. N. Pramila Bai, B. S. Ramasesh and M. K. Surappa, Dry sliding wear of A356-AlSiCp composites, Wear, I57 (1992) 295-304. J.-Q. Jiang, R.-S. Tan, Dry sliding wear of an alumina short fiber reinforced Al-Si alloy against steel, Wear 195 (1996) 106-111. Rao, R.N., Das, S., Effect of matrix alloy and influence of SiC particle on the sliding wear characteristics of aluminium alloy composites, Materials and Design (2009), doi: 10.1016/j.matdes.2009.09.032. Mehmet Acilar, Ferhat Gul, Effect of the applied load, sliding distance and oxidation on the dry sliding wear behavior of Al-10Si/SiCp composites produced by vacuum infiltration technique, Materials and Design 25 (2004) 209–217. K. S. S. Aradhya and M. K. Surappa, Estimation of mechanical properties of 6061 AlSiCp composites using finite element method, Scripta Metallurgica Vol. 2S, pp. 817822, 1991. M. Murato˘glu, M. Aksoy, Abrasive wear of 2124Al–SiC composites in the temperature range 20–200 ◦C, Journal of Materials Processing Technology 174 (2006) 272–276. R. Ipek, Adhesive wear behaviour of B4C and SiC reinforced 4147 Al matrix composites (Al/B4C–Al/SiC), Journal of Materials Processing Technology 162–163 (2005) 71–75. S. Basavarajappa, G. Chandramohan, J. Paulo Davim, Application of Taguchi techniques to study dry sliding wear behavior of metal matrix composites, Materials and Design 28 (2007) 1393–1398. Shaoyang Zhang, Fuping Wang, Comparison of friction and wear performances of brake material dry sliding against two aluminum matrix composites reinforced with different SiC particles, Journal of Materials Processing Technology 182 (2007) 122– 127. O. P. Modi, B. K. Prasad, A. H. Yegneswaran, M. L. Vaidya, Dry sliding wear behaviour of squeeze cast aluminium alloy-silicon carbide composites Materials Science and Engineering, A 151 (1992) 235-245 .

Tribology of Composite Materials, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

174

K. Palanikumar, T. Rajasekaran and J. Paulo Davim

[14] Ure˜na, J. Rams, M. Campo, M. Sánchez, Effect of reinforcement coatings on the dry sliding wear behavior of aluminium/SiC particles/carbon fibers hybrid composites, Wear 266 (2009) 1128–1136. [15] Chen Zhenhua, Teng Jie, Chen Gang, Fu Dingfa, Yan Hongge, Effect of the silicon content and thermomechanical treatment on the dry sliding wear behavior of spraydeposited Al-Si/SiCp composites, Wear 262 (2007) 362-368. [16] Sun Zhiqiang, Zhang Di, Li Guobin, Evaluation of dry sliding wear behavior of silicon particles reinforced aluminum matrix composites b, Materials and Design 26 (2005) 454–458. [17] J.K.M. Kwok, S.C. Lim, High-speed tribological properties of some Al/SiCp composites: I. Frictional and wear-rate characteristics, Composites Science and Technology 59 (1999) 55-63. [18] J.C. Walker, W.M. Rainforth, H. Jones, Lubricated sliding wear behaviour of aluminium alloy composites, Wear 259 (2005) 577–589. [19] Ranjit Bauri, M.K. Surappa, Sliding wear behavior of Al–Li–SiCp composites, Wear 265 (2008) 1756–1766. [20] M. I˙zciler, M. Muratoglu, Wear behaviour of SiC reinforced 2124 Al alloy composite in RWAT system, Journal of Materials Processing Technology 132 (2003) 67–72. [21] Ross T.J, Taguchi techniques for quality engineering, McGraw-Hill, Newyork, 1989. [22] D.C.Montgomery, Design and analysis of experiments, John Wiley and Sons, NewYork; 1991. [23] Jae-seob Kwak, Application of Taguchi and response surface methodologies for geometric error in surface grinding process, International Journal of Machine Tools and Manufacture (2005), 45:327-334. [24] K. Palanikumar; R. Karthikeyan, Optimal Machining Conditions For Turning Of Particulate Metal Matrix Composites Using Taguchi And Response Surface Methodologies, Machining Science and Technology, (2006),10:4, 417 – 433 [25] Meet MINITAB (2003). Release 14 for Windows, September, MINITAB, Inc., State College, PA, USA. [26] S. Basavarajappa, G. Chandramohan, R. Subramanian, A. Chandrasekar, Dry sliding wear behaviour of Al 2219/SiC metal matrix composites, Materials Science-Poland, Vol. 24, No. 2/1, 2006. [27] Jack G.Zhou, Daniel Herscovici and Calvin C. Chen (2000), Parametric process optimization to improve the accuracy of rapid prototyped stereolithography parts, International Journal of Machine Tools and Manufacture, 40, 363-379.

Tribology of Composite Materials, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

In: Tribology of Composite Materials Editor: J. Paulo Davim

ISBN 978-1-62100-999-3 © 2012 Nova Science Publishers, Inc.

Chapter 8

TRIBOLOGY OF GLASS-CERAMIC BONDED COMPOSITE MATERIALS M. J. Jackson* Center for Advanced Manufacturing, MET, College of Technology, Purdue University, West Lafayette, Indiana, US

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

ABSTRACT This review describes the tribological issues surrounding use of glass-ceramic bonded composites materials, such as vitrified grinding wheels, and is discussed in three ways. The first explains the relationship between grinding wheel wear and performance and discusses various physical and chemical interactions between abrasive grains and workpiece materials and how it relates to grinding wheel performance; the second investigates the issues surrounding the composition of the vitrified bonding systems and how it affects wheel wear, whilst the third explores the effect of reactions in vitrified bonds during thermal treatment of the grinding wheel that affect the performance of vitrified grinding wheels during subsequent grinding of engineering materials. The paper provides a timely review of the microstructural aspects of vitrified grinding wheels.

Keywords: Tribology, Composite Materials, Grinding Wheels

1. INTRODUCTION The grinding process is accompanied by wear of the glass-ceramic abrasive wheel, and the rate of this wear plays an important role in determining the efficiency of the grinding process and the quality of the workpiece [1-27]. The structure of a vitrified grinding wheel is composed of abrasive grains, a bonding system, and a large number of pores. Figure 1 shows a typical porous composite grinding wheel structure. Krabacher [28] stated that wear *

Contact details: E-mail: [email protected]; tel: 001 765 494 0365; fax: 001 765 494 6219.

Tribology of Composite Materials, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

176

M. J. Jackson

mechanisms in grinding wheels appear to be similar to that of single-point metal cutting tools, the only difference being in the size of swarf. The general form of the wheel-wear curve with volume of workpiece material removed is similar to that of single-point cutting tools. [16, 2830]. The wear behaviour observed is similar to that observed in other wear processes – high initial wear is followed by steady-state wear. A third accelerating wear regime usually indicates ‘catastrophic’ wear where the wheel requires re-dressing. Accelerating wear is usually accompanied by workpiece burn. The performance index usually used to characterize wheel-wear resistance is the ‘grinding ratio’, or G-ratio, and is the ratio of the volume of workpiece removed to the volume of grinding wheel removed, thus, G = Vw Vs

(1)

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

G-ratios cover a wide range of values ranging from less than 1 for vanadium-rich highsteels [24] to over 60,000 when internally grinding bearing races using cBN wheels [31]. Attempts have been made on vitrified wheels to address the problems related to the wear of abrasive grits in terms of the theory of brittle fracture [10, 32]. The conclusions of various researchers lead us to believe that the variety of different and interacting wear mechanisms involved, namely plastic flow of abrasive, crumbling, chemical wear, etc., makes grinding wheel wear too complex to be explained using a single theoretical model.

  Figure 1. Microstructure of a composite glass-ceramic grinding wheel. A - denotes abrasive grain, B denotes vitrified bonding phase, and C represents distributed porosity.

Tribology of Composite Materials, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

Tribology of Glass-Ceramic Bonded Composite Materials

177

Figure 2. Grinding wheel wear mechanisms: (1) abrasive wear – A denotes a wear flat generated by abrasion; (2) bond bridge fracture – A denotes the abrasive grain, B denotes the interfacial bond layer, and C denotes a crack passing through the bond bridge; (3) abrasive grain fracture – A denotes crystallographic grain fracture; and (4) interface fracture between abrasive grain and bond bridge.

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

2. WEAR OF ABRASIVE COMPOSITES The four different wheel wear mechanisms to which overall wheel wear can be attributed to the following are shown in Figure 2 [10, 28, 33-39], viz: (i) Attritious wear (grit dulling); (ii) Fracture of bond bridges; (iii) Mechanical failure of grits and grit flaking; and (iv) Fracture at the interface between grit and bond.

2.1 Attritious Wear Grit dulling is the gradual deterioration of abrasive cutting edges leading to loss of sharpness. The sources of minute scale wear are: (i) Attritious wear due to mechanical friction [10, 40-41]; (ii) Plastic flow experienced by the abrasive at high temperatures and pressures [16, 32, 36-37]; (iii) Crumbling due to thermal or mechanical shock [16, 32, 36-37]; and (iv) Chemical reaction between abrasive and workpiece material at elevated temperatures and in the presence of grinding fluids [28, 32, 39].

Tribology of Composite Materials, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

178

M. J. Jackson

The last mechanism can induce lowering the resistance of the grit to other wear mechanisms. Grit dulling leads to the growth of wear flats on active grits, which increases the area of contact and the amount of rubbing between grit and workpiece. At the point of grit dulling very high temperatures existing in the area of contact greatly enhances adhesion and chemical reaction between the two surfaces. If grit or bond post fracture does not occur the plateau area on the grit widens, and hence the wear rate increases. If fracture is further delayed, as with hard grade wheels, the wheel becomes glazed and the workpiece tends to burn. It has been shown experimentally [39] that chemical affinity between the abrasive and workpiece material can be used as a guide for the selection of grinding wheels. Their observations of solid diffusion of silicon carbide into ferrous materials explains the catastrophic wear rates exhibited by these ‘workpiece-wheel’ combinations. The most common method for measuring wear flat area is by measuring wear flats at the grinding wheel surface using optical or electron microscopic techniques [10, 40]. Hahn [16] observed and analysed theoretically the effect of the wear flat area through attritious wear during plunge grinding of various workpiece materials. Hahn concluded that grinding forces will gradually increase during wear-flat formation up to a point where the wheel will restore its sharpness due to grit fractures.

2.2 Fracture Wear

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Grit and bond fracture are usually considered simultaneously for the following reasons. (i) They are of the same nature, i.e. fracture of brittle materials and hence the theory of brittle fracture is applicable to both bond and grit [10, 32, 42]. The applied thermal and mechanical loads usually under cyclic conditions cause initiation and further development of cracks which leads to fracture and the formation of new irregular surfaces; (ii) They are related to the dressing methods used and occur simultaneously. The initial and final stages of wheel life between dressing exhibit exclusively fracture wear which is a combination of grit and bond fracture; and (iii) The relative amounts of bond and grit fracture cannot always be found. An investigation into precision grinding [10], where light grinding conditions were involved, employed a soft wheel which gave a high percentage of bond fracture whereas a hard wheel gave mainly partial grit fracture – attritious wear occurring in both cases. However, the combination of grinding parameters such as equivalent chip thickness and workpiece material determines the effective wheel hardness, and so no single feature of the grinding process can be used to predict the fracture pattern of the wheel in advance. The main difficulty in relating wheel wear due to fracture to the particular grinding condition arises from the lack of knowledge about the loads applied to both the grit and bond and their response to these applied loads.

Tribology of Composite Materials, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Tribology of Glass-Ceramic Bonded Composite Materials

179

Tarasov [33] suggests that grit fracture occurs as a result of mechanical forces due to chip formation or thermal shock induced by instantaneous high temperatures. Hahn [16] proposed a thermal stress hypothesis to explain the fracture of abrasive grits. Plunge grinding tests were conducted under fixed normal force conditions. Hahn asserted that as wear progresses, measurements of torque indicated that the tangential force actually decreases, therefore grit fracture due to mechanical loading will not occur. Mechanical stresses were also considered as an explanation for wear rates of wheels tested. Bhattacharyya et al. [43] observed grit loss due to fracture using an electron microscope. They concluded that they could not differentiate between Peklenik’s ‘crystal splintering’, i.e., grit flaking due to thermal stress, and grit fragmentation. However, they did explain their results in terms of Hahn’s thermal shock hypothesis. Hahn’s experimental conditions suggested that attritious wear was expected to have a major contribution to the thermal shock hypothesis as Mohun [44] observed with abrasive discs. The wear measurements of Hahn [16] were based on the reduction in wheel diameter which Malkin and Cook [10] attributed to attritious wear. Wear rates recorded were of the order of 50μ inch/sec. on wheel diameter. For purely attritious wear, wheel wear rates of the order of 5μ inch/sec. are normally observed. This indicated that the wear mechanism was not solely due to attritious wear. The amount of fracture wear present may consist of fragments of uniform average size particles due to partial mechanical grit fracture, or thermal flaking. Malkin and Cook [10] collected wheel wear particles for each grade of wheel tested when grinding with a fixed set of operating conditions then analysed their size distribution statistically. They found that with a soft grade wheel (G-grade), approximately 85% of the total wheel wear was due to bond fracture whilst with a harder K-grade wheel, this value reduces to around 55%. Attritious wear particles accounted for only 4% of the total wear in both cases. The strongest evidence in support of the idea of fracture due to mechanical loading is that fracture occurs at some distance away from the cutting tips [34, 44]. Yoshikawa [32] concluded that the heat generated by cutting has no effect on grit fracture since the peak temperature of the grip occurs at the surface of the grit in contact with the workpiece where fracture would be initiated upon cooling according to the thermal stress hypothesis. The hypothesis does not take into account any difference in coefficient of thermal expansion between grit and bond materials, and also of the effect of thermal shocks due to the quenching action of grinding fluids on the grit leaving the cutting zone. Saito and Kagiwada [45] analysed the latter case and reported that the thermal stress in a grit due to a pulsating heat source showed that the magnitude of the maximum tensile stress is not large enough to cause fracture of the grit. Eiss [42] and Malkin and Cook [12] both adopted the mechanical loading approach. Eiss applied a theoretical model of an idealised grit and compared it with grinding data. Malkin and Cook [12] derived an expression, from first principles, for the probability of bond fracture against the bond stress factor, (Ft – 20Fn)/VB. Yoshikawa and Sata [34] and Yoshikawa [32] developed expressions for the probability of grit and bond post fracture as functions of grit stress, σgrit, and the product (1/VB . fgrit), where fgrit is the grit grinding force, VB is the percentage of bond in the wheel by weight, and 1/VB is the bond weakness factor. Although bond and grit fracture are similar mechanisms they have a different effect on the economics of the grinding process. The first mechanism results in a rapid loss of wheel, and the second mechanism, on a comparable scale with the uncut-chip thickness, generates

Tribology of Composite Materials, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

180

M. J. Jackson

sharp cutting edges and is known as the self-dressing action. Both mechanical and thermal stresses seem to be responsible for fracture wear. The effect of heat at the grit interface is responsible for locally changing the mechanical properties of the abrasive material. However, fragments of larger sizes are likely to occur through mechanical loading which governs both grit pull-out wear and the self-sharpening action.

2.3. Wheel Wear Mechanisms

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

In view of the large numbers of independent variables involved in grinding it is selfevident that the more dominant wheel wear mechanisms depend greatly on the conditions applied to each grinding application. Grinding wheel wear consists of a large variety of physical and chemical mechanisms of very different nature. The wear processes involved in grinding are classified as follows: −

attritious wear, i.e. progressive wear leading to loss of grit form leading to a deterioration of cutting ability, and to excessive heat generation, and;

−

fracture wear (of grit and bond) which restores the cutting ability of the wheel and removes worn grits and allows grinding to progress efficiently.

A typical sequence of events for a grinding wheel according to Tarasov [33], involves dulling of the active grits by attritious wear and then to regain sharpness by fracture wear until bond post failure releases the grit. After the wheel has been dressed, stage I of the wheel wear diagram illustrates the removal of weakened grits by the fracture wear mechanism. Tsuwa and Yasui [46] reported the existence of a layer on the wheel surface after dressing which is progressively removed after grinding has started. High wear rates are also exhibited during the final stage of grinding where catastrophic breakdown of the wheel occurs due to mechanical overloading of the grinding grits. The area exposed on the surface of active grits directly affects the magnitude of the grinding energy required for metal removal and, hence, the amount of power required before the onset of workpiece burn. The steady-state wear regime (stage II) occurs due to the combined effect of attritious wear and fracture wear. However, the ability of the grinding grits and bond posts to fracture, when the load exceeds some limit, determines the duration of this stage and prevents forces from becoming too excessive. Fracture also limits the amount of heat generated at the cutting zone and gives better workpiece quality. Grisbrook [30] found that the greatest amount of wheel wear results from diamond dressing rather than from wheel wear, which emphasises the need for longer steady-state periods and fewer dressings. Stetiu and Lal [47] found that the mechanism of wheel wear can be changed from attritious wear to one of fracture wear by selecting the appropriate wheel hardness without changing the grinding conditions. The particle-size distributions gave evidence of a selfdressing action. In terms of volumetric wear Malkin and Cook [10] found that in plunge-feed grinding experiments only 4% of the total wear volume was due to attritious wear, the rest being due to grit and bond fracture according to wheel grade when the grinding parameters are fixed. Tsuwa [35-37] recorded the changes occurring in wheel cutting edges during

Tribology of Composite Materials, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

Tribology of Glass-Ceramic Bonded Composite Materials

181

grinding. He stated that the section of grit that forms a cutting edge is always the worn surface of the grit, and that edges which fracture by a considerable amount cease to become active. It was concluded that self-dressing takes place only if the attritious wear is very great, or the operation is continued up to a very high degree of dulling of the active grits so that fracture edges become active. It is certain that the rate of wear is dependent upon this fracturing tendency and that forces upon the grit will undoubtedly increase. There is a fundamental difference between the two forms of grinding wheel wear. Attritious wear is undesirable and in that case all practical measures aim to reduce it. This form of wear is very similar to the corresponding mechanism which causes deterioration of the cutting ability of turning tools were large scale failure is unlikely to occur [48]. Fracture wear in grinding is advantageous in that it can be a controlled form of wear. Although it is a form of volumetric wheel loss it opposes attritious wear in terms of grinding efficiency. Fracture wear gives the wheel its functional significance and considering very little is known about fracture behaviour in abrasive particles, it is clearly important to understand how grit and bond composition affects fracture behaviour of grinding wheels and their performance.

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

2.4 Wheel Wear and Grinding Forces Grinding forces and their effect in wheel wear of grinding tools have been subjected to detailed investigations. The earliest dynamometer used for the measurement of grinding forces was probably one used by Marshall and Shaw [49]. For simplicity, two force components operate, namely: the normal force, Fn; and the tangential force, Ft. A notable feature of grinding force measurements is the high Fn/Ff ratio ranging from 1.5 to 3. This value seems high compared with conventional machining operations where a typical value lies between 0.5 and 1. The difference is attributed to the different effective cutting geometries and the pattern of metal removal by the grits. Grisbrook et al. [30] and Grisbrook [50] recorded the magnitude of grinding force components during the whole period of wheel life between dressings, and for fixed downfeed conditions. The typical change in force pattern can be divided into four phases. (i) An unstable phase where forces rise abruptly up to a peak, then fall to a steady-state value as the initial high wear rate (stage I), due to the effects of dressing, slows down. Davis and Rubenstein [51] showed that the rate of change in grinding force increases slightly after a transition point is reached then settles down to steady-state conditions. Pattison and Chisholm [29] showed that grinding forces are affected by the dressing conditions, which has been confirmed by Rowe et al. [52-53]; (ii) A phase where forces are constant and heat flow into the grit and workpiece is in equilibrium. The region coincides with the self-dressing action of the wheel. Bond bridge fracture does not normally occur in this region although bond strength does play an important part here as it determines the magnitude of the applied load the grit can sustain without fracture;

Tribology of Composite Materials, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

182

M. J. Jackson (iii) A third phase where there is a progressive build-up of power and grinding forces. The mechanisms that proceeds from the second phase assumes that grits adopt stable geometries and shapes which are harder to fracture, and from then on the number of fractures decreases and grits become dull. At this point, grits are prone to overheating and grinding becomes inefficient. As rubbing and ploughing increases, and since the metal removal rate is constant, the normal force component increases at a much higher rate than the tangential force component. However, higher forces at this stage are not accompanied by higher wear rates which indicates that the rate of wheel wear is a function of the absolute values of the forces and their relative magnitude, i.e., Ft/Fn, which is referred to as the grinding coefficient. Marshall and Shaw [49] suggested that a higher grinding coefficient produced a more efficient grinding process. This is explained by considering that the normal force component induces compressive stresses into the abrasive grit whilst the tangential force component causes tensile stresses to be exerted at the rake face of the grit. Therefore, a higher Ft/Fn ratio means a higher probability of grit fracture since the grit material has a lower tensile than compressive strength which means that cutting edges retain their sharp facets for a greater period of time; and

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

(iv) In this period, the rate of change of grinding forces becomes less and the effects of vibration become evident. This phase corresponds to stage III of the wheel-wear curve where wear rates are detrimental to economic grinding. The progress wear of grinding wheels operated under fixed normal force conditions has been studied by Hahn [16]. High force values lead to increases in depths of cut and rapid catastrophic wear due to fracture. For very low force intensities Hahn found that metal removal ceases quickly, and the wheel subsequently glazes and the wear rate becomes negligible. Lindsay [54] and Lindsay and Hahn [55] found that radial wheel wear is linearly proportional to the normal force existing between wheel and workpiece, irrespective of whether the force is applied under fixed force conditions, or under fixed feed conditions. Volumetric wear is reported in Lindsay and Hahn’s further work in precision grinding [56]. They related volumetric wear to be an approximate quadratic function of normal force intensity. Their conclusions showed that wheel wear could be related to any wheel-wear geometry, or conformity, in terms of interface force intensity and contact pressure.

2.5 Assessment of Grinding Forces and Wear The dominant wear mechanisms in grinding wheels are attritious and fracture wear. In the author’s opinion, fracture wear (bond and grit) should be considered the most important wear mechanisms for these reasons: (i) Fracture wear constitutes approximately 95% of the total volumetric wheel wear; (ii) Wear during the initial and final stages of the wheel-wear curve (I and III) is due to gross fracture;

Tribology of Composite Materials, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

Tribology of Glass-Ceramic Bonded Composite Materials

183

(iii) The dressing mechanism that affects wheel life and grinding forces is essentially fracture-sharpening of cutting edges; (iv) Fracture wear reduces grinding forces and heat generation which results in lower power consumption, and smaller workpiece distortion; and (v) Fracture wear, in the form of self-dressing, affects the duration of useful wheel life (region II of the wheel-wear curve). It should be noted that the mechanism of grinding wheel fracture is an extremely complex process caused by the action of thermal and mechanical stresses induced into the grit. Thermal stresses are thought to be responsible for grit flaking whilst larger fragments are associated with mechanical stresses that are directly related to forces acting upon the grinding grit. Fracture wear due to mechanical loading seems to be dependent on absolute force components and their relative magnitude. For fixed normal force operations the force is sufficiently large to allow fracture to occur at a controlled rate. Too high a force will cause catastrophic wear rates to dominate, whilst very small forces will impair metal cutting. For fixed-feed operations the grinding force components increase steadily during stage II of the wheel-wear curve. If the cutting conditions produce a friable wheel then metal cutting will take place efficiently. The rate of change of force depends upon the initial value of the force component ratio.

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

2.6 Effect of Workpiece Material on Wheel Wear The suitability of abrasive materials to grind workpiece materials efficiently depends on their attritious wear resistance. The abrasive should be harder than the workpiece material being ground, however, hardness is not the dominant factor. In fact one would not use the two hardest known of natural abrasives, diamond and silicon carbide, to grind ferrous alloys. Attritious wear of grits is both mechanical and chemical [39, 57-61]. Chemical effects are significant when the abrasive is appreciably harder than the workpiece material and its associated metallurgical phases. At higher temperatures during cutting, chemical reactions may occur between workpiece, grinding fluid, the surrounding atmosphere and the abrasive and bond. Diamond is not suitable for grinding ferrous metals despite its hardness. This is attributed to attritious wear caused by reversion from diamond to graphite [62]. Degradation of diamond appears to be aggravated in the presence of iron low in carbon. Loladze and Bockuchava [63] listed five types of diamond wheel wear based on adhesion, abrasion and diffusion wear. Cubic boron nitride is more stable than diamond in the presence of ferrous metals. However, the success of C.B.N. on various steels is dependent on the complex carbide phases within the steel workpiece. The hardness of the carbides are quoted as ‘abrasive numbers’ which are essentially weighted averages of their Vickers’ hardness values [64]. When grinding ferrous metals with aluminium oxide abrasives, the most important chemical reaction usually involves the oxidation of iron and the reaction of the oxide with the abrasive to form spinel, FeAl2O4 [57], thus: 2 Fe+O2+2Al2O3 → 2 FeAl2O4 Tribology of Composite Materials, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

(2)

184

M. J. Jackson

Spinel is an intermediate compound between the oxidized workpiece material and aluminium oxide. A high attritious wear rate on steel in humid air rather than dry air was found to be due to the catalytic effect of water on the oxidation of iron [57]. However, despite the role of oxygen and water to promote adhesion and attrition during grinding, their elimination by grinding in a vacuum has a dramatic effect on the process [6567]. Chemical reaction between workpiece and abrasive was reported to be reduced. However, loading of the wheel surface increased. Surface oxidation and corrosion in normal grinding environments tends to reduce adhesion between metal particles and the workpiece. This same effect might explain the difficulties encountered when grinding high-temperature oxidation-resistant metals, including stainless steels, nickel-based alloys, and titanium [68]. Silicon carbide abrasives are harder than aluminium oxide abrasives but are inferior when grinding ferrous materials. The main chemical reaction tends to involve the dissociation of silicon carbide [59, 69], which promotes attritious wear when grinding titanium and other non-ferrous metals. In addition to chemical affinity, mechanical factors contribute significantly to attritious wear. When grinding carbon and alloy tool steels, the G-ratio is greatly reduced when grinding the material in its fully hardened state [70], which suggests that a mechanical effect is taking place. However, hardness is not indicative of grindability especially when the material’s hardest phases are softer than the abrasive. High-speed tool steels contain complex carbides that tend to reduce grindability of alumina abrasive wheels. The hardest carbides in high-speed steels are carbides of tungsten, molybdenum and vanadium. The volume fraction, C*, of these carbides in tool steels relative to that of tungsten carbide by itself can be approximated in terms of the weight percentages of tungsten (W), molybdenum (Mo), and vanadium (V) as [70];

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

C* = W + 1.9Mo + 6.3V

(3)

The relative grinding ratio G* is the G-ratio for different tool steels expressed as a percentage relative to that of an M2 tool steel. The results tend to indicate that higher carbide content reduces grindability of alumina grinding wheels. This can be compared favourably with a similar correlation between G-ratio and the vanadium content for grinding high-speed tool steels with aluminium oxide wheels [24]. Higher G-ratios have been obtained for materials produced by powder metallurgical methods that results in a fine dispersion of small hard carbides which tends to be less abrasive than large hard carbides [71].

2.7 Effect of Abrasive and Bond Composition on Wheel Performance Attempts were made to describe the process of wheel wear in terms of mechanical stresses applied to abrasive grits during stage II of the wheel-wear curve. Graham and Voutsadopoulos [72] presented data which supported the argument that fracture-type wear is the most important wear mechanism related to the loss of abrasive material from the grinding operation as a consequence of high stresses induced in the abrasive grit by grinding forces. Grit fracture was assumed to be caused by induced tensile stresses of relatively small magnitude. These workers used existing experimental data to correlate G-ratio to a number of process variables using a finite-element model of an idealised wedge. The wedge was assumed to be rigidly held in an infinitely strong bond. Graham and Voutsadopoulos [72]

Tribology of Composite Materials, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

Tribology of Glass-Ceramic Bonded Composite Materials

185

applied a tangential force of 5.4lbs force and a normal force of 10.41lbs force to produce a stress pattern showing lines of constant maximum shear stress (isochromatics). They applied Griffith’s brittle fracture criterion to the post-processed finite element results in order to locate areas of compressive, tensile and neutral stresses within the model grit. The area of brittle failure was also located on this diagram in order to illustrate points at which the abrasive material is likely to fail, i.e. tensile fracture of the grit. A good correlation was found between the maximum tensile stress in the grit and the G-ratio using experimental data contained in the literature. However, Graham and Voutsadopoulos assumed that grit fracture was the pre-dominant wear mechanism assuming that grit to be rigidly held. Wear during this period of grinding (stage II) is mixed, i.e., bond and grit fracture which explains why the correlation is close but not exact. The vitrified bond is not infinitely stronger than the grit, and its magnitude governs the duration of stage II wear. These workers postulated that wear during stages I, II and III was probably due to tensile stresses induced in the bond and the grit material, i.e., grit fracture and pull-out. The main criticism of the work is due mainly to the assumption of a rigidly held wedge finite element model, and the correlation of results of this model to wear data relating to stages I, II and III of the wheel-wear curve.

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

3. VITRIFIED BONDING MATERIALS The tensile strength of a ceramic material is determined by the most serious flaw in it. Such flaws are known as Griffith flaws and can appear as cracks, pores, or irregular-shaped grains that have a sharp notch acting as a stress raiser around the notch to a high level. Any inclusion which has a different elastic modulus from the matrix will produce a small stress concentration in its vicinity, e.g., a perfectly spherical pore will increase the average tensile stress by a factor of three at the pore surface on a plane perpendicular to the tensile stress direction. The low strengths observed in many ceramics must be caused by the presence of sharp notches. In ceramics of moderate strength (> 70MN/m2), flaws are typically 100μ and will be frequently found to be pores. In ceramics showing high strength (350-700MN/m2), flaws are considerably smaller and are of grain size dimensions such as grain boundary cracks of fractured grains – grain sizes typically a few microns. The most serious flaw on a body is one situated at the surface orientated so that its maximum dimension is perpendicular to the applied tensile stress. Surface flaws are the most serious because the effective flaw length is the complete flaw length, whereas inside the volume of the stressed body, the effective flaw size is less than the flaw length. In addition, if the body is subject to bending, for example a bond post connected to two adhesive grains, it will experience the highest stress at its surface. Factors other than flaws that affect the strength of the body are related to the nature of the body, i.e., composition, grain size, and general porosity. These factors are important since they control the energy required to extend the flaw. The energy required for fracture initiation is higher than the energy required to form a new surface, i.e., the surface energy. This is partly because it includes the energy absorbing process of plastic deformation that occurs in the highly stressed region of the crack tip. Generally for ceramics the fracture initiation energy γ1, is usually a few times 10J/m2. This value varies with fracture surface roughness so that smooth surfaces, such as glassy materials, have low values of, γ1. In the

Tribology of Composite Materials, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

186

M. J. Jackson

ideal case of a dense, homogeneous material containing a single flaw, the stress multiplication factor of a flaw can be evaluated for some simple flaw geometries. The expression relating the stress at failure, σ, to the size of the flaw causing failure, and the failure energy for fracture initiation for an elliptical crack is,

σ failure =

2.E.γ 1 π .C

(4)

Where E is Young’s modulus, C is the crack length if it is a surface flaw or half a crack length of it is an internal flaw, and σfailure is taken as the average stress calculated from the specimen’s geometry and dimensions and the applied load. This is Griffith’s equation but for any flaw geometry it is expected that,

σ failureα

E.γ 1 πC

(5)

This equation should be valid for bodies containing several well-separated serious flaws. The stress, σfailure, includes any residual stresses in the body that are usually of unknown magnitude. Many experimental studies on ceramics have shown that strength depends on the total porosity, i.e., small and large pores have an effect. An empirical relationship between the failure stress and the porosity, p, of a body has been found to be:

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

σfailure = σ0 e-bp

(6)

Where, σ0, is the strength found by extrapolating the date to zero porosity, and, b, is a constant determined from the experimental data plotted as in, σfailure, versus porosity, p. The value of, b, has been found to vary considerably for the same ceramic material depending on the shape, size and distribution of the porosity [73]. The relationship is known as the Ryshkewitch-Duckworth [74, 75] equation and was shown by Knudsen [76] to be based on the increased average stress caused by the reduction in load-bearing area resulting from the porosity. However, Knudsen did not consider factors that control strength at zero porosity. Carniglia [73] considered that the flaw should be enclosed within a volume λ, i.e., the Saint Venant volume, such that at the periphery of this volume the stress re-distributing effect of the flaw was negligible. This volume was then considered in relation to the spacing, L, between general porosity. Carniglia showed that the Ryshkewitch-Duckworth equation was applicable only when λ >> L. This is, when the Griffith flaw that initiates fracture is larger than the pores which form the general porosity, and when the spacing between pores is small compared to the size of the Griffith flaw. Under these conditions, the average flaw stress acting on the Griffith flaw has increased by the reduction in load-bearing area and can be considered as uniform. When λ >> L, a local stress model, such as the Griffith model for an elliptical crack, can be used. That is the stress magnification produced by a single flaw, with no interference from other flaws that cause

Tribology of Composite Materials, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

Tribology of Glass-Ceramic Bonded Composite Materials

187

failure. The case when λ < L cannot be treated theoretically in a general manner, because each Griffith flaw is close to another stress re-distributing flaw and therefore the average stress around each flaw is highly variable. In practice, λ >> L would occur quite frequently for ceramics. Most ceramics contain a number of smaller pores and frequently more serious Griffith flaws. If the samples tested have constant Griffith fracture initiating flaws and the general porosity is variable, then the strength data should fit the Ryshkewitch-Duckworth equation. Carniglia [73] developed a more complex equation that showed the RyshkewitchDuckworth equation to be an approximation. If the general porosity remains constant but the size of the Griffith flaws is variable, then the fracture strength values, σfailure, should yield a

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

straight line when plotted against

(γ 1 / C ) . This approach should work because, γ1, and, E,

are determined for the material containing the same general porosity. Flaws can be classified as gross macroscopic, microscopic and sub-microscopic [77]. This means this classification encompasses size only. A flaw is considered to be gross if it is readily visible to the unaided eye, so that the origin of failure can be viewed, i.e., large surface cracks, inclusions at the surface, etc. A microscopic flaw is not easily identifiable and is usually a small crack, void or a small inclusion. A sub-microscopic flaw is identified using a scanning electron microscope. Flaws in ceramics result from several causes, i.e. pores might be caused by differential firing shrinkage on a small scale of size, burn-out of organic matter (dextrin and fillers), gaseous evolution caused by a reaction on firing, diffusion of gases or some other mechanism, etc. Differential shrinkage on a small scale of size, is caused by the non-uniformity of the characteristic properties on a related scale of size. The characteristic properties are porosity, particle size, composition, or particle alignment. Large-scale non-uniformities of any of these characteristics can cause the constituent materials to re-distribute during firing and may possibly lead to splitting of the abrasive wheel. The tensile stress may be relieved by the formation of one large fissure or possibly by the formation of numerous small fissures. A number of factors involved in pressing these wheels that could be responsible of the nonuniformity of porosity on a large scale include, segregation of fines, friction at the die wall, non-uniform powder packing on deposition, and variations in compaction ratio. In addition to these, lamination problems might occur, depending on the state of the powder and the pressing technique used. It is caused by the elastic recovery of the compact that occurs when the compaction pressure is removed, as a result of the entrapment of air, particularly in fine powders. Spontaneous micro-cracking on cooling from the firing temperature is a common source of cracks around large inclusions of a different phase. This micro-cracking occurs either because of stresses arising from either a mismatch in thermal expansion of the inclusion and the matrix, or is caused by a phase transformation of the inclusions. In general, pores in ceramic bodies may be described by pores, cracks, fissures, inclusions, large grains, and surface defects. These flaws might be combined and produce complicated fracture-initiating combinations.

3.1 Effect of Particle Size of Constituent Materials Parmalee and Morgan [78] found that decreasing the particle size of quartz altered the vitrification behaviour and strength of a ceramic body. The bodies examined were classed as coarse, commercial, and fine, with average diameters of quartz of 68, 45, and 11μm,

Tribology of Composite Materials, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

188

M. J. Jackson

respectively. The finer the quartz, the greater the reduction in porosity and the higher the strength. Koenig [79] studied the effects of feldspar, quartz, and kaolin particle size in a vitreous china body. It was found that firing shrinkage and flexural strength were greater for the finer-ground feldspar body, and that there was less water absorption. The bodies containing finely ground quartz, as well as feldspar, had considerably greater strength than the regular china clay body, and they showed greater strength in which fine quartz and regular feldspar, or regular quartz, and fine feldspar were used. Koenig concluded that more finely divided quartz affects vitrification quite markedly and that finely ground feldspar increases vitrification behaviour. Sane and Cook [80] discovered that ball milling for 100 hours reduced the final porosity of a clay-feldspar-quartz composition from 17.1% to 0.3% using the same firing conditions. The change is caused by intimate mixing of the constituents and the reduced distances fluxing ions are required to diffuse during firing to enhance chemical reactions. Increased densification of the fired body, due to smaller particle sizes, tended to produce a stronger body. During vitrification of the body, a large mass of viscous liquid is formed. The liquid wets solid particles that are pulled together under the action of surface tension when the liquid flows into the pores. When the firing temperature is increased, more melt is formed which is less viscous. However, the dissolution of quartz opposes this reduction in viscosity which helps grinding wheel manufacturers, as well as manufacturers of clay-based materials, to fire products over a wide range of soaking temperatures. Fine grinding of quartz produces more surface area of quartz per volume, which promotes its dissolution in the liquid phase, and consequently aids vitrification, which increases strength. Finer quartz particles have also been reported to inhibit inversion cracking at 573°C [81].

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

3.2 Effect of Mullite and Glass Content It is generally accepted that the development of interlocking fine mullite needles in the body increases the strength of the clay-based material. However, this hypothesis is not free from controversy. According to Zoellner [82] and Budnikov [83], mullite provides porcelain its strength. Zoellner dissolved pieces of porcelain in a cold 25-33% solution of hydrofluoric acid for several days. The vitreous matrix and quartz were dissolved and the crystalline phase remained. Zoellner assumed that these crystals were sillimanite, since mullite at that time was unknown. He suggested that increasing the firing time and temperature to increase the formation of these crystals would increase the strength of porcelain. Budnikov [83] published data concerning the strength of electrical porcelain and mullite, from which the strength of mullite is greater than porcelain. Geller [84] showed that the effect of firing increased the size of mullite crystals to such an extent that the strength of porcelain decreased. This was reported also by Krause and Keetman [84] and Eitel [86]. Grofcsik [87] reported that the total Al2O3 content of kaolinite transforms by exothermic reaction into mullite at about 960°C. Therefore, repeated or prolonged firing will not change the amount of mullite but may change its size. According to Krause and Keetman [85], the size of mullite crystals will increase with the logarithm of firing time at a suitable soaking temperature. They found that by maintaining the samples at 1400°C for 6,60,600, and 6000 minutes, the average length of acicular crystals of mullite increased in size to 5, 7.2, 11, and 14.2 μm, respectively.

Tribology of Composite Materials, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Tribology of Glass-Ceramic Bonded Composite Materials

189

Sane and Cook [80] milled a clay-based body (20% feldspar, 30% quartz, 42.5% kaolin and 7.5% ball clay) for different periods of time which were fired at three different temperatures. They discovered that increasing milling time would increase the wt.% glass in the body and reduced the amount of quartz. However, it was discovered that increasing the mullite content of the body, increased strength of the body. The glassy plane was considered the major component of porcelain, which was considered the weakest part. This aspect of the body has attracted much attention to the strength of clay-based materials. MattyasovskyZsolnay [88] assumed that the tensile strength of porcelain is influenced by stresses set-up in the glassy phase rather than by the amount and size of mullite crystals. He suggested that if mullite controls the strength of porcelain then an increase in the Al2O3 content (by increasing kaolinite fraction) to form more mullite would increase the strength of porcelain. Mattyasovsky-Zsolnay demonstrated that mechanical strength increased when quartz content was increased and reduced when kaolinite content was increased. This statement is in contradiction to the experimental data published by Weidman [89]. Weidman increased the kaolin content and the result was an increase in strength. Experimental samples containing 50, 60, 65 and 70% weight kaolin (corresponding to 30, 16, 10, 1% weight quartz and 20, 24, 25 and 29% weight feldspar) showed bending strengths of 71, 91, 94 and 130 M Pa, respectively. Kalnin et al. [90] published results on the strength and elasticity of quartz-free clay-based bodies using compositions of kaolin and nepheline syenite, focusing on mullite content and porosity. They concluded that elastic moduli and flexural strength of the bodies increased with the proportion of mullite present over the range 11-36% weight. In kaolin-rich preparations, i.e., 2:1 and 1:1 kaolin: nepheline syenite ratios, the amount of mullite obtained by powder x-ray methods agrees with the calculated values assuming complete decomposition of kaolin into mullite and silica. However, in the nepheline syenite system (1:2 ratio), the amount of mullite present is much less than expected, and it appears that in this case silica and alumina dissolved in vitrified nepheline syenite. Koch [91] noted that porcelain bodies should be fired such that the microstructure contains plate-like primary mullite, and a high proportion of fine acicular secondary mullite concentrated in the glassy phase and also at grain boundaries to give a felted structure. Lack [92] explained that the presence of a viscous, glassy phase aids the diffusion of cations so that mullite forms more regular-shaped crystals. Therefore, in sintering wheel bonds, high temperatures are needed to obtain well-developed mullite crystals. It has been shown that using mineralizers such TiO2 [93] and MgO [94], the mullite content can be increased. Primary mullite is formed by the decomposition of clays and secondary mullite is formed by re-crystallization. It is possible that both forms of mullite affect mechanical strength in many ways [95].

3.3 Effect of Quartz Content The study of the effect of quartz on the strength of clay-based materials has concentrated on finding an optimum size of quartz particle that inhibits ‘de-bonding’ from the matrix. The ‘pre-stress theory’ espoused by many researchers, was proposed to account for observed increases in strength that occurred with an increase in fine quartz content. It was assumed that fine quartz generates a compressive stress in the matrix and that any applied tensile force would be effectively reduced by the ‘pre-stress’. Therefore, higher tensile stresses would be

Tribology of Composite Materials, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

190

M. J. Jackson

required to cause failure [88, 96-98]. However, cracks were found around small particles after loading that were present before loading. This indicated that small quartz particles should have a weakening effect on the matrix rather than a strengthening effect [99]. The weakening effect of quartz was demonstrated by comparing the strength of the matrix material, composed of equal weight fractions of kaolin and nepheline syenite, fired to 1265°C and subsequently ground; with and without quartz. The ground matrix was isostatically pressed with an equal amount of quartz of different sizes, then fired to a pre-determined bulk density. The maximum strength was found to occur at a quartz particle size of 25μm having a sharp fall above and a slight decrease below this size. But the matrix material without the quartz was much stronger, which contradicts the ‘pre-stress theory’. Weyland [100] showed that the strength of clay-feldspar-alumina porcelain bodies greatly decreased when a small percentage of quartz was added. Another problem is that when Al2O3 replaced SiO2, in the composition of clay-based materials, the alumina-containing body showed higher strength [101-103]. German [104] suggested that the development of high strength with alumina was probably due to a reduction in the number of Griffith cracks that can be caused by silica inversion. Smothers [105] opposed the ‘pre-stress theory’. He performed hot-strength tests on both quartz and alumina-containing bodies. Above the transition temperature of quartz the strength of the quartz body was as high as that for the alumina body. Weyl [107] and Dunsmore et al. [108] have also shown that the strength of a quartz-containing body is higher above the displacive polymorphic change of quartz. On cooling a clay-based material from the soaking temperature, the quartz transforms at 573°C. This transformation, which involves a contraction of the quartz particles, produces stresses both in the quartz particles and in the matrix, and can cause circumferential cracks surrounding the quartz grains in a quasi-pore within the matrix. However, it has been reported in studies both porcelain bodies [106] and on model systems [109] that it is only above a critical size that circumferential cracking tends to occur during cooling. Investigations have led to relationships being proposed which relate the effect of quartz particle size on strength and the existence of an optimum grain size. However, the reported optimum value of the quartz size varies significantly between researchers. Krause [110-112] carried out experiments with finely milled and graded quartz and established that the maximum bend strength is obtained with quartz of particle sizes in the range 15-20μm. The data showed that up to 45μm particle size, the body with the highest % weight gave the strongest matrix. Ludas [113] reported that the highest mechanical strength of porcelain occurred when quartz was sieved to the size of 30-35μm. Beech and Norris [114] investigated a porcelain made using 40% weight quartz of a particle size 10-30μm which produced the highest strength. Beech and Norris also varied the firing conditions. Their results indicated that the maximum strength occurred at a temperature below the required for maximum bulk density. Grofcsik [87] used the disc compression test on circular discs to investigate the effect of quartz particle size on mechanical properties. Experimenting with a series of bodies of different composition, particle size and firing schedule, Grofcsik concluded that for fine quartz, increasing the quartz content increases strength. He also pointed out that having optimum grains of 20-60μm may be causing the grains to dissolve in the glassy phase during firing, whilst coarser grains allow harmful stresses to accumulate. Many researchers have published data on optimum particle sizes for quartz. However, it must be borne in mind that the strength of clay-based materials depends not only on quartz particle size, but also on other characteristics of the fired body.

Tribology of Composite Materials, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

Tribology of Glass-Ceramic Bonded Composite Materials

191

Therefore, it is difficult to relate the strength of clay-based materials on quartz particle size alone. Dinsdale and Wilkinson [115] related the properties of whiteware bodies to the size of constituent particles. By assuming, ε, to represent the mean size of the crystalline particles present in the fired system, they suggested the modulus of rupture, S, to be, S = K. ε -a

(7)

Where, K, is a constant and the index, a, would be expected from Griffith’s crack theory to be about 0.5. Work on single-phase crystalline materials has shown that there is an exponential relationship between the strength, S, and the true porosity, P, of the form. S = Soe-bp

(8)

Where, So, is the strength at zero porosity. Knudsen [76] presented the above equations as,

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

S = K. ε -a . e-bp

(9)

Where K, a, and, b, are empirical constants. Dinsdale and Wilkinson [115] agreed that the grain size distribution of the starting materials will strongly influence both the porosity and the grain size in the fired body. If the size of the filler material such as quartz is increased, packing may be increased, and unfired porosity reduced. However, fired strength is adversely affected. Improvements in fired strength can be attained by reducing filler size that is obtained at the expense of firing contraction. In a paper by Evans and Linzner [116], an acoustic emission study was carried out on a clay-based material as it was loaded to failure. This showed that the acoustic emission rate increased rapidly as the stress in the sample approached the failure value. The sound pulses were expected to arise from cracks that debond quartz particles, cracking of quartz particles, themselves and cracks linking up or starting to run after being arrested. The results of this study indicate the quartz particles that remain attached to the matrix after cooling, and are therefore residually stressed, can be detached when the applied stress reaches a sufficiently high value. As the highest stresses in the material, resulting from the applied stress, occur adjacent to the tips of many cracks in the sample, the formation of a process zone around a crack which starts to grow can be expected if bonded quartz particles are in its vicinity. In another acoustic emission study of porcelain [117], it was found that a maximum acoustic emission rate occurred at a temperature below the β - to α – quartz transition point. This suggests that the residual stress, resulting from the transition, increased as a result of the thermal expansion mismatch between the quartz particles and the matrix on cooling below 573°C. It was found that the temperature at which the maximum emission rate occurred was reduced as the quartz particles were made smaller. This is consistent with the assumption that smaller inclusions require higher stresses to become de-bonded. A further interesting finding was that a second maximum emission rate occurred at 200°C, the phase transition from β - to – α – cristobalite. The presence of cristobalite arises from the conversion of quartz during densification heat treatment. Results published by Oral et al., [118] showed that for a variety

Tribology of Composite Materials, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

192

M. J. Jackson

of clay-based material compositions, the strength of ring specimens showed considerable scatter in test pieces that had optimum heat treatments. Failure was reported to occur by cracks around quartz particles or by quasi-spherical pores.

3.4 Ceramic Bonding Materials and Bond Strength During grinding, the action of tangential and normal grinding forces create stresses within the abrasive grit and adjacent bond posts which inevitably causes bond-post failure. When the bond strength is optimized, this process takes place gradually, i.e., when abrasive grits have lost their ability to cut the workpiece. The most frequently used bonding materials for vitreous-bonded grinding wheels containing clay minerals, quartz, and feldspar, which contain crystalline phases that reduce their melting points. As previously mentioned, bonding materials used for alumina wheels resemble high-strength tough enamels (vitreous), and those for silicon carbide wheels are referred to as stoneware or soft porcelain bonds. In alumina grinding wheels, the bond not only dissolves other crystalline phases in the bond, but also dissolves the surface of the alumina grain. The bond must not produce any ‘rounding’ of the grains, therefore, the bond must produce the same ‘hardness’ at the interface of the bond/grit couple. This explains why it is possible to fire grinding wheels at temperatures 200 or 300°C higher than the melting point of the bonds without discharging the bonding material or deforming the wheel itself. Guilleaume [120] characterised vitreous bonds using the Seger formulae. Guilleaume tested 60 types of clay bond which were described by the formula:

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

RO.(1.25-3.0Al2O3).(4.5-10SiO2)

(10)

Where RO is the sum of alkali oxides contained in the bond, i.e., CaO, MgO, MnO etc. These bond compositions were composed of a ‘clay substance’, feldspar, and quartz with powdered marble and magnesite as mineralizers. Guilleaume [121] first examined fired bonds without the admixture of alumina grains, then determined the strength of bar specimens containing the alumina grains. Specimens were fired at different temperatures and specimens with variations in bond composition fired at constant temperature. The effects on strength and hardness were examined. Guilleaume’s results gave only a limited understanding since no information was provided on mineral composition and fired bond composition. Furthermore, firing conditions characterised by Seger cones (Sg) do not represent well-defined conditions. Firing conditions characterised by the same Sg temperature can produce grinding wheels with very different properties. Franz [122] reported that a deeper analysis on the properties of bonding materials on the performance of grinding wheels is required. He examined the bonding materials used by Guilleaume and reported that on firing grinding wheels containing bonds of the same composition in a small temperature range (between Sg8 and Sg12 – a temperature difference of 100°C), deviations in bond hardness were as high as 7 degrees [121]. Rieke and Haeberle [123] noted that the strength of specimens prepared with bond materials that have low melting points are higher than those wheels prepared with classic bond materials.

Tribology of Composite Materials, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Tribology of Glass-Ceramic Bonded Composite Materials

193

Filonenko and Lavrov [119] pointed out that during the cooling period of firing, when grinding wheels are quickly cooled from peak temperature to 800°C, de-vitrification is inhibited. Thereafter, slow cooling is required in order to relieve internal stresses. The mechanical strength of the bond is dependent on the amount of its vitreous phase. According to Filonenko and Lavrov [119] of all the crystalline compounds, only spinel (MgO.Al2O3) is capable of raising the strength of vitreous bonds. This compound surrounds alumina grains, as a quasi-embedding material, with octahedral spinel crystals smaller than 8μm diameter formed in the alumina-rich melt zone developed at the interface between bond and grit. The high strength bonds used were located in the proximity of the SiO2 peak in the Na2O-SiO2Al2O3 ternary diagram. As a consequence of the alumina-rich melting zone, the following compounds may be formed: anorthite, cordierite, mullite, spinel, plagioclases, anastase, rutile, heamatite and magnetite [119]. The composition of ceramic bonds is important for maintaining strength. In order to achieve complete penetration of the abrasive grain surface, the viscosity of the melt plays an extremely important part in complete adhesion between grit and bond. The flow characteristics of ceramic bonds have been investigated by a number of researchers. Moser [124] used a heating microscope to observe, qualitatively, the changes in contact angle of various bonding materials up to 1460°C. Bond materials and bond containing alumina grains were examined using three crude bond compositions, viz, (i) illite bond, (ii) modified illite bond, and (iii) fritted borosilicate bond. It was found that the fritted borosilicate bond melted at a lower temperature and produced better wetting characteristics. Moser reasoned that increased wetting of bond to grit would increase wheel strength. Hartline [125] conducted work on the strength and fracture of alumina abrasive wheels and concluded that the fracture process involved in grinding wheels proceeds through the bond, and fractography conducted on bond posts showed a signs of de-vitrification and porosity. Hartline stated that increases in strength were achieved by developing uniform bond post strengths and/or higher strength wheel structure. Experimental grinding wheel bonds were examined by Barry, Lay, and Morrell [126] using novel glass-ceramics and conventional feldspar bonds. It was concluded that the strongest bonds were those based on traditional clay-feldspar-quartz composition, and that strength was dependent on wetting and flow properties of the bond on the surface of the grit. Ogawa and Okamoto [127] found that increasing the feldspar content of bonds increased the adhesion of bond to grit in alumina wheels by producing a less viscous mixture at the soaking temperature. The development of superabrasive grinding wheels has resulted in a number of papers published on ceramic bonding of diamond and cBN [128-130]. Yang et al. [129] have published results concerning the strength of vitreous-bonded cBN wheels. They concluded that the strength of their specimens was dependent on glass composition and, to a much lesser extent, on porosity. The bonds examined were full of pores with an optimum composition of 51% SiO2, 15% Al2O3, 26% B2O3, 3% Na2O and 5% CaO. Jackson, Barlow and Mills [131] concluded that the strength of vitreous-bonded alumina wheels depended on the K2O-CaO ratio. They reasoned that the increased mass of glassnetwork modifiers would release liberated gases within the bonds by reducing the bond’s viscosity. It appears that a reduction in bond viscosity not only releases gases liberated during the breakdown of clays and fluxes, but also improves interface strength by improving wetting and flow characteristics of the glassy bond.

Tribology of Composite Materials, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

194

M. J. Jackson

3.5 Effect of Interfacial Cohesion on Bond Strength and Wheel Wear

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Interfacial cohesion in vitreous-bonded grinding wheels is as important, if not more important, as the effect of bond composition on their strength and performance. In grinding wheels, interfacial strength dictates whether a grinding grain cuts efficiently or not at all. According to Hondros [132], a knowledge of interfacial properties holds the key to the design of bulk properties. Moseley, Briggs and Lewis [133] compared the cohesion of various borosilicate bonding materials in contact with white-fused alumina (containing 99% wt Al2O3) and impurity phases 3% wt TiO2, 1% wt SiO2 and MgO and CaO, Fe2O3 and ZrO2 in smaller quantities. They compared the properties of each grit/bond composition by measuring their fracture toughness and related these values qualitatively to observations of fracture surfaces and interfaces. Fired test specimens were inspected which showed preferential etching along crystallographically controlled directions in white alumina grit. This was observed to be dissolution of planar blocks of sodium aluminate or β-alumina (Na2O.11Al2O3) present in α –alumina (essentially pure Al2O3), established by x-ray diffraction of samples. β-alumina is thought to be detrimental to grit strength which, when in small well-dispersed amounts, can control the self-sharpening effect during grinding. However, in large amounts leads to loss of strength of the grit. Bragg, Gottfried and West were the first to determine the crystal structure of β-alumina in 1931. Beevers and Ross [134] came to the same conclusion as Bragg et al. when they discovered that the crystal structure and chemical composition do not readily agree with each other. Deviations in the chemical composition, noted by various workers, report: Na2O.9Al2O3; Na1.5Al10.83O17; and Na2O6Al2O3. According to Harata [135] the non-stoichiometric composition alters the measured lattice parameters. Based on his own measurements, β-alumina is represented by the formula, (1.16+ x) Na2O.11Al2O3

(11)

Where x denotes the molar fraction of Na2O, which varies from 0.19 to 0.59. Below this limit, α– Al2O3 separates from β-alumina, above it NaAlO2 dissociates. Moser [136] performed electron microscopic studies on white-fused alumina grits. He observed that βalumina was present as spots and bands on the surface of the grits, and noted that heating uncoated grits up to a temperature of 1000°C led to a lower amount of β-alumina due to the evaporation of Na2O above 900°C, thus leaving α –alumina. He further observed that grits coated with borosilicate bonding material contained a higher level of Na2O at the interface. The most likely cause of increasing the strength of white alumina grinding wheels is the dissolution of β-alumina that would locally enrich the melted bond with Na2O which promotes fluidity in alumino-borosilicate and alumino-alkalisilicate glasses. The glassy phase would then fill dissolution bands created in the surface of the grit thus promoting a better mechanical bond between grit and glass phase, i.e. enhanced shear resistance when subjected to grinding forces. In their samples, Moseley et al. [133] also identified small (50μm dia x 10μm thick) Ca-rich platelets, of which two types were identified; alite (Ca3SiO5) and an unnamed oxide, NaCaAlO3, known to have several polymorphic forms. S.E.M. analysis performed by Moseley et al. on brown-fused alumina grit, showed that the grit contained 9698 wt. %, Al2O3, 1-2% wt. TiO2 and up to 0.5 wt. SiO2. The amount of TiO2 in solution is

Tribology of Composite Materials, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Tribology of Glass-Ceramic Bonded Composite Materials

195

inconsistent with earlier work which had determined that the maximum solubility of TiO2 in Al2O3 is less than 0.3 mol.%. at 1300°C [137]. However, it was stated that MgO may increase solubility of TiO2 in alumina. Examination of this section of brown-fused grit showed the appearance of blade-like inclusions at the grit-bond interface. This morphology is consistent with rutile needles observed in some synthetic sapphires [138]. These inclusions accounted for the variability in TiO2 content and hence its solubility in alumina. Further examination of the fired specimens in both white and brown-grit samples, showed evidence of crystal formation in the glassy bond. These crystals were lath-shaped with square sections distributed evenly throughout the glass. X-ray diffraction indicated an alumino-borate with high alumina content. The best match was with Al18 B4 O33. Titanium was considered deleterious to fracture toughness due to the presence of rutile needles on the surface of the grit and titania in the bond itself. Moseley et al. suggested using oxides to allow the formation of titanates instead of rutile needles. Unfortunately, Mn-doped glasses did not improve strength appreciably. A second attempt was made to form a protective coat around the grit. Oxides such as MgO, ZnO, and CoO were used to form spinel layers between grit and bond. Unfortunately, the test specimens were lower in strength. They concluded that white-fused alumina grits were stronger due to planar dissolution of β-alumina, whilst brown-fused alumina grits were considered weaker due to rutile needle formation a the interface between bond and grit. Similar interfacial studies were conducted on alumina abrasives [139] and silicon carbide vitreous-bonded wheels [140]. Comparison of wheel performance using wheels with brownfused and white-fused alumina grits was examined by Reichenbach [26]. Interface studies of metals in contact with superabrasive grits have recently been conducted by Scott et al. [141]. These workers provided substantial quantitative information on the wetting behaviour of copper-based alloys on diamond. This work was further extended by Evens et al., [142] whom correlated G-ratio of abrasive buttons with interfacial bond strength. Evens et al. concluded that increases in titanium and tin concentrations enhance wetting to the (III) plane of diamond, and that titanium segregated to the interface where alloy compositions promoted wetting on the diamond surface and to the free surface where alloy compositions did not promote wetting. The interfacial layer was characterized by electron probe analysis to be composed of bronze (α + δ) eutectoid and CuSnTi2. Good interfacial bonding was found not to be associated with good wetting. The interfacial bonding layer with promoted wetting was identified as a 100nm reaction product (TiC), which was associated with low temperatures, i.e., 900-950°C. The best composition was found to be a mixture of 20wt.% copper –10wt.% tin-titanium alloy which wets and bonds well to diamond at 900-950°C. Evens et al., [142] stated that better bonding and wetting was achieved with nickel-coated diamond grinding buttons. Shilo et al. [143] conducted a comprehensive study on the wettability of glass on substrates of cBN They found that wettability was dependent on the reaction product at the grit/bond interface, which was identified as B2O3. In the case of sodium-borosilicate glasses, with Na2O/B2O3 ratios less than or equal to 0.5, boron atoms tend to co-ordinate from ternary to tetrahedral form owing to the introduction of oxygen atoms by alkali oxides. The best wetting conditions were achieved at 900°C using glasses of the system Na2O-B2O3-PbOSiO2. The addition of 4% wt.-7 % wt. Li2O was responsible for the best wetting condition that was due to the network modifying effect of Li2O.

Tribology of Composite Materials, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

196

M. J. Jackson

4. REACTIONS IN CERAMIC BONDS Reactions in bond compositions used in this study are similar to those that occur in claybased materials used in whiteware bodies. The reactions outlined in this chapter form the basis for studies on the effect of bond composition on wheel performance. Clays provide the wheel bond with plasticity for forming and strength in the green state. The two principal clays used for wheel bonds are china clay and ball clay. Both clays are primary clays formed from decomposed feldspar. The main mineral constituents in both china and ball clay is kaolinite (Al2O3. 2SiO2.2H2O). However, clays differ in purity and plasticity. Kaolinite has a layered structure and the particles reflect this by having a plate-like morphology. Fluxes are used in bonds in order to lower the firing temperature by reacting primarily with the clay to form a viscous liquid phase that promotes densification. The flux is usually a feldspathic mineral such as feldspar, nepheline syenite, Cornish stone, talc and sometimes lithia. The complete range of fluxes has been reviewed by Royle [144]. The fluxing effect of feldspar in clay-based materials was studied by Schramm and Hall [145]. Orthoclase (K2O.Al2O3.6SiO2) and albite (Na2O.Al2O3.6SiO2) and a combination of these fluxes in the formation of clay-based materials. In high strength bonds, used for conventional and superabrasives vitreous products, the fluxing effect is provided by powdered glass frits and borax in addition to feldspar. These additions provide the bond with fluxes that melt over a wider temperature range than those using traditional bond materials. This reduces the firing temperature that tends to promote increased adhesion between grit and bond.

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

4.1 Densification and Phase Analysis When a grinding wheel has been formed, it is fired at an appropriate soaking temperature to be densified in order to mature to the optimum state. During firing the bond materials, abrasive, and fillers initially behave independently of one another. Water occurring in the pores and in the clay is driven off below 250°C, whilst at 500°C, the organic matter has been burnt off.

4.1.1

Theoretical Phase Analysis – Use of Equilibrium Diagrams

The overall final composition of clay-based grinding wheel bonds is composed of unreacted quartz, mullite, glass and sometimes cristobalite depending on the reaction conditions. The formation and growth of mullite crystals is thought to occur in the following manner: (i) formation in the kaolinite platelets; (ii) formation in the feldspar platelets; (iii) formation in the mica platelets; (iv) growth by recrystallisation; and (v) formation during cooling.

Tribology of Composite Materials, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Tribology of Glass-Ceramic Bonded Composite Materials

197

Reaction rates and mechanisms of mullite formation may be different when the mullite is formed in the mica, kaolinite and feldspar platelets. The formation rate in one region is influenced by formations in other regions unless concentration gradients of atoms are distributed in such a way as to promote equal reaction rates throughout the total reaction volume. It is clear that the formation of mullite in clay-based bonds is very complicated. The formation of mullite can be explained qualitatively using equilibrium diagrams, and quantitatively by comparing equilibrium phases with experimental results. The application of equilibrium diagrams to ceramic manufacturing processes has been discussed by many authors. However, when applied to clay-based materials, Bowen [146], Hall and Insley [147] and Foster [148] express reservations in using them due to the coarseness of the clay particles, the high viscosity of molten feldspar [149, 150], and the slow diffusion of alkali ions which tend to prevent equilibrium and homogeneity of the mass. Shelton [151] used equilibrium diagrams with success to correlate the properties of fired whiteware bodies to the amount of eutectic melt formed. Dietzel and Padurow [152] used phase diagrams in their discussion of quartz dissolution in porcelain. Although equilibrium is not obtained for clay-based bonds, equilibrium diagrams are useful for describing reaction rates towards equilibrium. The equilibrium diagram used in this work is the K2O-Al2O3-SiO2 ternary system derived by Schairer and Bowen [153]. When heating a body consisting of quartz, kaolinite and potash feldspar, there are no appreciable reactions that take place between quartz, feldspar and kaolinite relicts until the temperature approaches the eutectic temperature of the total system, 985°C. In each isolated sub-system below 985°C, quartz has transformed to its high temperature form β-quartz and the feldspar grains have transformed to sanidine. Kaolinite has lost its chemically bonded water and the decomposition products have been transformed to γ-Al2O3 and an amorphous phase with a high SiO2 content. This is the primary composition of the reaction system. Considering complete equilibrium conditions within the total system immediately under the eutectic temperature there are three solid phases, feldspar, tridymite, and mullite. When the system is heated to the eutectic temperature a fourth phase is formed, a melt phase with a composition of the eutectic. With continued heating at constant temperature the amount of the melt phase is increased until one of the phases is consumed. If feldspar is consumed the composition point of the melt moves with rising temperature along the boundary line between the primary phase regions of tridymite and mullite until one of these two phases is consumed. The composition point then enters the remaining solid phase in the direction of the composition of the total system. If tridymite is first consumed at the phase reaction in the eutectic, the composition point of the melt moves along the boundary line between the feldspar and mullite regions. If neither of these two phases is consumed before 1140°C, the tenary invariant point is reached where feldspar is consumed and leucite occurs as a new phase. After this phase reaction is completed the composition follows the boundary line until one of the solid phases has been consumed. The consumption of phases is dependent on the rate of reaction of the body until equilibrium conditions have been reached. These conditions are applicable under conditions of equilibrium. For clay-based bonds equilibrium is rarely achieved and as a consequence, the grain size of the raw material is the most important reaction rate variable. Quartz shows very low reactivity up to the eutectic temperature. At that temperature, the quartz phase present is not affected by the other phases.

Tribology of Composite Materials, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

198

M. J. Jackson

4.1.2. Formation of Mullite in Kaolinite Clays

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

The effect of heat on ceramic raw materials has been studied by many workers as far back as 1887 when le Chatelier [183] charted discontinuities in the thermal analysis of kaolinite. Ford [184] and Todor [185] presented differential thermal analysis curves for china clay and ball clay. The major phase transformations for potash feldspar, as measured using a DuPont 1600 D.T.A. cell, of kaolinite → metakaolinite → spinel-type phase → mullite occurs as heating continues. There appears to be a controversy regarding the products formed in heat treatment. One theory suggests that, on dehydroxylation, kaolinite forms a mixture of alumina and amorphous silica. The other theory considers the dehydrated product as an aluminosilicate. To explain the exothermic reaction it is believed that γ-Al2O3 is formed and in the other case, an Al-Si spinel is formed. Comeforo et al., [154] show the appearance of hexagonal particles far above the temperature for dehydroxylation, illustrating a residual structure still present in the non-crystalline compound of metakaolin. Many researchers have investigated reactions that occur in clay-based materials containing these compounds. Weiss et al., [155] announced that the cubic phase that appeared at 900°C from kaolinite was an Al-Si spinel. These workers isolated the spinel phase by leaching amorphous SiO2 from the fired kaolinite and found that its chemical formula agreed with the theoretical formula. The Al-Si spinel analyzed by Weiss et al. is different from the γAl2O3 spinel because its lattice constant is 0.002nm lower owing to the replacement of Alions by smaller Si-ions present in the cubic γ-Al2O3 spinel structure [156]. However, other researchers have supported the crystallization of γ-Al2O3 as the cause of the exothermic peak near 100°C [157-161]. There are two other exothermic peaks due to the crystallization of primary mullite and formation of cristobalite from amorphous silica that occur at 1150°C and 1250°C respectively. The transformation of pure kaolinite to mullite has been reviewed by Chaudhuri [162].

4.1.3. Effect of Heat on Feldspar and Quartz α - quartz is known to invert to β-quartz at 573.3°C [163]. The transition is accomplished by a 0.8% volume expansion and is reversible and rapid. Any cristobalite present in the matrix is transformed (α → β structure) between 200-270°C [81]. The breakdown and melting of potash feldspar has been studied by Morey and Bowen [164] who reported incongruent melting at 1170°C. However, the thermal behaviour of feldspar has had little attention paid to it. The solubility of quartz in feldspar [165] and the solubility of quartz in clay and feldspar [166] has been investigated extensively. These workers report that mullite (3Al2O3.2SiO2) crystallizes in molten feldspar.

4.1.4. Effect of Heat on Clay-based Materials Clay-based materials principally contain the oxides SiO2, Al2O3, K2O and Na2O. Equilibrium diagrams were used by Lundin [167] to study the formation of mullite in detail. For mullite content, the experimental values were 3.4% lower than those calculated from the phase diagram. For glass content, the difference was 5% lower. Lundin concluded that the quartz had only partly reacted. The difficulty in attaining equilibrium in clay-based materials

Tribology of Composite Materials, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

Tribology of Glass-Ceramic Bonded Composite Materials

199

is thought to be due to the high viscosity of molten feldspar, retarded dissolution of quartz, and the slow diffusion of partially mixed and partially melted constituents. When the temperature increases mullite dissolves partially in the melt if equilibrium is achieved. At around 1000°C the surface tension of the liquid draws unreacted/partially reacted particles together that reduces porosity and increases bulk density. A loss in surface area causes shrinkage in the body. The porosity is initially interconnected and is referred to as open porosity. However, the reduction in the volume of the body produces closed porosity. When the open porosity is removed, the body is vitrified. The densification of the body is slow because of the high viscosity of the liquid phase. The decrease in the viscosity is expected from the increase in temperature that is partially offset by the enrichment of the liquid with silica by partial dissolution of quartz. In practice, the slow dissolution of quartz results in bonds having wide range of firing. The initial densification of clay-based bonds can be modelled simply using two contacting spherical particles. The rate of initial neck growth is,

x = r

3.γ t .t 2.η .ρ

(12)

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

The increase in contact diameter is proportional to t, and the increase in area between particles should be directly proportional to time. The factors affecting the rate of densification are surface tension, γt, viscosity, η, and particle size, r. The volume shrinkage, or linear shrinkage which takes place is determined by the approach between particle centers is given by the model as,

ΔV ΔL 9.γ t .t =3 = Vo Lo 4.η .r

(13)

This equation shows that the initial rate of shrinkage is directly proportional to the viscosity and particle size. As firing develops, more liquid is formed and mullite crystals appear. Primary mullite is formed from clay relicts whilst secondary mullite is formed from the melt. At 1200°C, a considerable amount of mullite is formed in quartz-rich bonds. Shelton and Meyer [168] reported that increased rates of heating cause less liquid and mullite formation, less quartz corrosion, and more pores. An optimum heating rate of 50°C-90°C per hour was preferred. The crystallization of glasses in the K2O-SiO2-Al2O3 system was studied extensively by Hermansson and Carlesson [169]. They concluded that crystallization from this high viscosity ternary system is possible. Tuttle and Cook [170] claimed to have found quartz, mullite, cristobalite and wollastonite as the crystalline constituents of kaolin-flintfeldspar blends. However, in many cases only quartz and mullite, and occasionally cristobalite are the phases present in clay-based kaolin-quartz-feldspar composites.

4.1.5. Effects of Cooling When densification occurs, the cooling rate is reduced in order to prevent thermal stress cracking of the body. It is better to reduce the cooling rate when crystalline inversions occur

Tribology of Composite Materials, Nova Science Publishers, Incorporated, 2009. ProQuest Ebook Central,

200

M. J. Jackson

that involve volume changes. The inversion ranges for quartz and cristobalite are 550-580°C and 200-300°C, respectively. When quartz-containing bonds begin to cool from the soaking temperature, it is considered that the liquid phase relieves stresses resulting from thermal expansion mismatch between itself and the phases β-quartz, β-cristobalite and mullite to at least 800°C. At 800°C, stresses will develop in quartz particles and the matrix which causes micro-cracking. The shrinkage behaviour of quartz and the glass phase has been described by Storch et al. [171]. Between the temperature range 573