Electroslag Remelting Towards Clean Steel 9819932564, 9789819932566

Table of contents :
Preface
Acknowledgments
Contents
1 Introduction to Electroslag Remelting
References
2 Clean Steel Production by Electroslag Remelting
2.1 Cleanliness Target
2.2 Oxygen and Oxide Inclusions
2.3 Sulfur and Desulfurization
2.4 Hydrogen
2.5 Nitrogen and Phosphorus
2.6 Summary
References
3 Deoxidation of ESR and Its Correlation with Oxide Inclusions
3.1 Background
3.2 Oxygen Transfer During ESR Process
3.3 Thermodynamic Considerations on Deoxidation of ESR
3.4 Deoxidation Kinetics of ESR
3.4.1 Development of the Kinetic Model
3.4.2 Kinetic Model Parameters
3.4.3 Application of the Developed Kinetic Model to ESR Practice
3.5 Evaluation of the Dependence of Oxygen on the Processing Parameters of ESR
3.5.1 Initial Oxygen Content of Steel Electrode
3.5.2 Oxide Inclusions in Steel Electrode
3.5.3 Remelting Atmosphere
3.5.4 Deoxidation Schemes of ESR
3.5.5 Role of Slag Compositions
3.5.6 Reoxidation of Liquid Steel
3.5.7 Melting Rate and Filling Ratio of ESR
3.6 Summary
References
4 Reoxidation of Liquid Steel During ESR and Its Effect on Oxide Inclusions
4.1 Background
4.2 Experimental Work
4.3 Oxygen Content of the Steel
4.4 Inclusions in the Consumable Steel Electrode
4.5 Inclusions in the Remelted Ingots
4.6 Transient Inclusions in the Liquid Metal Pool During ESR Refining
4.7 Evolution Mechanism of Inclusions During ESR Refining
4.8 Other Cases of Inclusion Evolution During ESR
4.9 Summary
References
5 Desulfurization in Electroslag Remelting
5.1 Background
5.2 Desulfurization Basis of ESR
5.3 Dependence of Desulfurization on the Processing Parameters of ESR
5.3.1 Initial Sulfur Content of Consumable Electrode
5.3.2 Remelting Atmosphere
5.3.3 Slag Composition
5.3.4 Deoxidation Schemes of ESR
5.3.5 Melting Rate of ESR
5.3.6 Electrical Parameters of ESR
5.4 Desulfurization Associated with Sulfide Inclusion Evolution During ESR
5.5 ESR for Sulfur-Bearing Steel Production
5.6 Summary
References
6 Sulfide and Nitride Inclusion Evolution During ESR
6.1 Sulfide Inclusions
6.2 Sulfide in Oxide–Sulfide Complex Inclusions
6.2.1 From Original Attached State to Patch-Type and Shell-Type Sulfide
6.2.2 From Original Patch-Type to Shell-Type Sulfide
6.3 Nitride Inclusions
6.4 Summary
References
7 Evolution of Original Oxide Inclusions During ESR
7.1 Sites of Oxide Inclusion Removal During ESR
7.2 Al2O3 and MgO·Al2O3 Inclusions
7.3 Calcium Aluminate Inclusions
7.4 Manganese Silicate Inclusions
7.5 Role of Processing Parameters of ESR on Inclusions
7.5.1 Deoxidation Schemes of ESR
7.5.2 Slag Composition
7.5.3 Melting Rate of ESR
7.5.4 Electrical Parameters of ESR
7.6 Newly-Formed Inclusions in Remelted Ingot
7.7 Illustration of Inclusion Removal and Fresh Inclusion Generation During ESR
7.8 Summary
References
8 Evolution of Oxide Inclusions in Si–Mn-Killed Steel During ESR
8.1 Background
8.2 P-ESR Trials of Si–Mn Deoxidized Steel
8.3 Inclusions in Consumable Steel Electrode
8.4 Inclusions in the Liquid Metal Pool and Remelted Ingots
8.5 Evolution Mechanism of Oxide Inclusions During Protective Atmosphere Electroslag Remelting
8.6 Summary
References
9 Modification of Alumina and MgO·Al2O3 Inclusions by Calcium Treatment During ESR
9.1 Background
9.2 Modification of MgO·Al2O3 Spinel Inclusions in Steel During P-ESR
9.2.1 P-ESR Procedure and Inclusion Characterization
9.2.2 Characteristics of Non-metallic Inclusions
9.2.3 Evolution and Modification of MgO·Al2O3 Spinel Inclusions in the P-ESR Process
9.2.4 MgO·Al2O3 Spinel Inclusion Modification in Industrial P-ESR Practice
9.3 Simultaneous Modification of Alumina and MgO·Al2O3 Inclusions During P-ESR
9.3.1 Inclusions in the Consumable Steel Electrode
9.3.2 Inclusions and Primary Carbides in Remelted Ingots
9.3.3 Composition Distribution of Oxide Inclusions in Remelted Ingots
9.3.4 Transient Inclusions in Liquid Metal Pool During P-ESR Refining
9.3.5 Proposed Mechanism for Modification of Inclusions by Calcium During P-ESR Process
9.4 Evaluation of Al2O3 and MgO·Al2O3 Inclusion Modification Prior to ESR
9.5 Summary
References
10 Role of Calcium Modification of Oxide Inclusions During ESR on Primary Carbides
10.1 Background
10.2 Effect of Oxide Inclusions Modification on Primary Carbides in Tool Steel
10.2.1 Precipitate Formation in the Steel
10.2.2 Microstructure of As-Cast Remelted Ingot
10.2.3 Primary Carbides in As-Cast ESR Ingots
10.2.4 Microstructure and Toughness of Annealed Steel
10.3 MgO·Al2O3 Spinel Inclusions Modification on Primary Carbonitrides in Nickel-Base Superalloy
10.3.1 Two-Dimensional Observation of Inclusions and Primary Carbonitrides
10.3.2 Three-Dimensional Observation of Inclusions and Primary Carbonitrides
10.3.3 Oxide Inclusions Evolution
10.3.4 Effect of Oxide Inclusions Modification on Primary Carbonitrides
10.4 Summary
References

Citation preview

Chengbin Shi Jing Li Shufeng Yang

Electroslag Remelting Towards Clean Steel

Electroslag Remelting Towards Clean Steel

Chengbin Shi · Jing Li · Shufeng Yang

Electroslag Remelting Towards Clean Steel

Chengbin Shi University of Science and Technology Beijing Beijing, China

Jing Li University of Science and Technology Beijing Beijing, China

Shufeng Yang University of Science and Technology Beijing Beijing, China

ISBN 978-981-99-3256-6 ISBN 978-981-99-3257-3 (eBook) https://doi.org/10.1007/978-981-99-3257-3 Jointly published with Metallurgical Industry Press The print edition is not for sale in China (Mainland). Customers from China (Mainland) please order the print book from: Metallurgical Industry Press. ISBN of the Co-Publisher’s edition: 978-7-5024-9118-5 © Metallurgical Industry Press 2023 This work is subject to copyright. All rights are solely and exclusively licensed by the Publisher, whether the whole or part of the material is concerned, specifically the rights of reprinting, reuse of illustrations, recitation, broadcasting, reproduction on microfilms or in any other physical way, and transmission or information storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar methodology now known or hereafter developed. The use of general descriptive names, registered names, trademarks, service marks, etc. in this publication does not imply, even in the absence of a specific statement, that such names are exempt from the relevant protective laws and regulations and therefore free for general use. The publishers, the authors, and the editors are safe to assume that the advice and information in this book are believed to be true and accurate at the date of publication. Neither the publishers nor the authors or the editors give a warranty, expressed or implied, with respect to the material contained herein or for any errors or omissions that may have been made. The publishers remain neutral with regard to jurisdictional claims in published maps and institutional affiliations. This Springer imprint is published by the registered company Springer Nature Singapore Pte Ltd. The registered company address is: 152 Beach Road, #21-01/04 Gateway East, Singapore 189721, Singapore

Preface

Steel industry is constantly striving to produce advanced steel with the properties meeting different application requirements. Electroslag remelting (ESR) has been developed into an integral part in the production of some varieties of special steels and alloys, mainly because of its ability to provide high cleanliness level and excellent homogeneity of solidified ingot structure (reducing segregation, shrink holes, etc.) simultaneously. Electroslag remelting is one of the most important technologies for liquid steel secondary refining. Although the steel cleanliness generally is significantly improved through the ESR, with the increasing demands for more excellent comprehensive performance of steel, the cleanliness requirement of the steel produced by ESR is facing new challenges. The issue of steel cleanliness improvement by ESR is one of the most difficult aspects in the high-quality steel processing. The cleanliness improvement of the steel during ESR process is an eternal topic. The requirement of the steel cleanliness level is dependent on the application of a specific steel. With increasing the demands for the comprehensive performance of steel, constantly improving the steel cleanliness has always been a burning issue for the steel manufacturers. Great efforts have been devoted to optimize the processing parameters, auxiliary materials quality, slag chemistry, deoxidizing agent, and so on, in the steelmaking process to improve the steel cleanliness. The non-metallic inclusions and impurity element contents in steel are two important aspects in the production of clean steel. The presence of non-metallic inclusions in the steel is inevitable. Decreasing the oxygen content in steel could not only lower the amount of oxide inclusions, but also suppress the oxidation loss of alloying elements in the steel. This monograph constitutes a summary of the work of the author in the field of electroslag remelting for advanced clean steel production. This monograph constitutes nine chapters. Chapter 1 describes the fundamentals, advantages, exclusive characteristics of ESR in comparison with other technologies for liquid steel refining, and the absolute advantages of ESR in some varieties of advanced materials production. Meanwhile, the existing problems and future development directions of ESR technologies are put forward. Chapter 2 presents the outline of the general indexes of

v

vi

Preface

steel cleanliness and the aspects involved in the cleanliness control during the ESR process. Chapter 3 summarizes the oxygen transfer behavior and deoxidation during the ESR process. The is described first. The thermodynamic and kinetics considerations on the deoxidation of ESR are discussed. In view of the indivisible correlation of the deoxidation with oxide inclusions, the role of deoxidation operations of ESR on the oxide inclusions is also evaluated. A general concluding remark and perspective for future work are present. Chapter 4 is the description of the liquid steel reoxidation during protective atmosphere ESR of the steel. The role of liquid steel reoxidation during P-ESR on the oxide inclusion is presented. Further, thermodynamic modeling is established to study the interactions of gas-slag-metal-inclusion phases. The experimental determination and thermodynamic considerations are employed to elucidate the evolution mechanism of oxide inclusions during the P-ESR process. Chapter 5 presents the advances in the desulfurization practices and production of sulfur-bearing steel at a high sulfur level during ESR, as well as their relation with the ESR processing parameters. The interrelation between the desulfurization and the sulfide inclusion evolution during ESR is discussed. The remaining challenges for future work are proposed and assessed. Chapter 6 includes the evolution and removal of (Mn, Cr)S inclusions, MnS inclusions, sulfide in oxide-sulfide complex inclusions, and AlN and TiN inclusions during the ESR process based on the ESR trials and thermodynamic analysis. Chapter 7 presents the earlier work on the sites of oxide inclusion removal during ESR and the evolution of Al2 O3 , MgO·Al2 O3 , and calcium aluminate inclusions during the ESR. The generation of manganese silicate inclusions in the steel is described. The evolution of manganese silicate inclusions in the steel is presented in Chap. 8. Chapter 8 contains the evolution of oxide inclusions in Si-Mn deoxidized steel during protective atmosphere ESR and the effect of slag compositions on the oxide inclusion evolution. The evolution mechanism of oxide inclusions is ascertained through monitoring the change in the transient inclusion in combination with thermodynamic calculation. Chapter 9 presents the transformation trajectory of self-developed online calcium modification of MgO·Al2 O3 spinel, Al2 O3 , and low-MgO MgO·Al2 O3 inclusions during the P-ESR. The possible extent and mechanism of calcium modification of these oxide inclusions during the P-ESR process were ascertained. Chapter 10 describes the role of Al2 O3 and MgO·Al2 O3 inclusions modification during the P-ESR process on nitrides and primary carbides in steel, as well as the microstructure and toughness properties of the annealed steel. Furthermore, the modification of oxide inclusions is extended to refine the carbonitrides in Inconel 718 superalloy produced by P-ESR. Beijing, China

Chengbin Shi Jing Li Shufeng Yang

Acknowledgments

This monograph summarizes the selected achievements of the author’s academic research in the field of electroslag remelting for advanced steel production since 2009. It is such a great pleasure to have the opportunity to express my gratitude to all those who are involved directly or indirectly in making this work a success for their invaluable guidance and help during each stage of this work. The author gratefully appreciates Prof. Hanjie Guo and Prof. Jie Fu for providing the motivational push and discussions on the research work. I am grateful to Dr. Xichun Chen, Prof. Jung-wook Cho, and Prof. Joo Hyun Park for the discussion on my research activities. Gratitude is also extended to my graduate students Mr. Shijun Wang, Mr. Yi Huang, Mr. Xin Zhu, and Mr. Xin Zheng for their assistance in the preparation of the monograph for publication. Most of the work presented in this document would not have been possible without the help, suggestions, and advice from the specialists at the special steel plants. Herein, I would like to express my deep gratitude to all of them, in particular, Dr. Fang Jiang, Mr. Shujie Liu, Mr. Baoshan Guo, and Mr. Fuli Zhang. The financial support for the research achievements presented in this monograph by the National Natural Science Foundation of China (Grant Nos. 51444004, 51504019, 51874026, 52074027, 52274314), China Postdoctoral Science Foundation (Grant Nos. 2014M560047 and 2016T90035), and the Fundamental Research Funds for the Central Universities (Grant Nos. FRF-TP-14-009A1, FRF-TP-15010A2, and FRF-TP-18-004A3) is greatly acknowledged. I am also thankful to the financial support from the State Key Laboratory of Advanced Metallurgy, USTB (Grant Nos. 41603017, 41621024). I would like to record my deep gratitude to all members of my family for their faithful support and encouragement through the years. Chengbin Shi

vii

Contents

1

Introduction to Electroslag Remelting . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

1 9

2

Clean Steel Production by Electroslag Remelting . . . . . . . . . . . . . . . . . 2.1 Cleanliness Target . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.2 Oxygen and Oxide Inclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.3 Sulfur and Desulfurization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.4 Hydrogen . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.5 Nitrogen and Phosphorus . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.6 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

11 11 12 16 16 18 18 19

3

Deoxidation of ESR and Its Correlation with Oxide Inclusions . . . . . 3.1 Background . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.2 Oxygen Transfer During ESR Process . . . . . . . . . . . . . . . . . . . . . . . 3.3 Thermodynamic Considerations on Deoxidation of ESR . . . . . . . 3.4 Deoxidation Kinetics of ESR . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.4.1 Development of the Kinetic Model . . . . . . . . . . . . . . . . . . 3.4.2 Kinetic Model Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . 3.4.3 Application of the Developed Kinetic Model to ESR Practice . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.5 Evaluation of the Dependence of Oxygen on the Processing Parameters of ESR . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.5.1 Initial Oxygen Content of Steel Electrode . . . . . . . . . . . . 3.5.2 Oxide Inclusions in Steel Electrode . . . . . . . . . . . . . . . . . . 3.5.3 Remelting Atmosphere . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.5.4 Deoxidation Schemes of ESR . . . . . . . . . . . . . . . . . . . . . . 3.5.5 Role of Slag Compositions . . . . . . . . . . . . . . . . . . . . . . . . .

21 21 22 24 27 28 33 34 37 37 38 40 42 46

ix

x

Contents

3.5.6 Reoxidation of Liquid Steel . . . . . . . . . . . . . . . . . . . . . . . . 3.5.7 Melting Rate and Filling Ratio of ESR . . . . . . . . . . . . . . . 3.6 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4

5

6

Reoxidation of Liquid Steel During ESR and Its Effect on Oxide Inclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.1 Background . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.2 Experimental Work . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.3 Oxygen Content of the Steel . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.4 Inclusions in the Consumable Steel Electrode . . . . . . . . . . . . . . . . 4.5 Inclusions in the Remelted Ingots . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.6 Transient Inclusions in the Liquid Metal Pool During ESR Refining . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.7 Evolution Mechanism of Inclusions During ESR Refining . . . . . . 4.8 Other Cases of Inclusion Evolution During ESR . . . . . . . . . . . . . . 4.9 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Desulfurization in Electroslag Remelting . . . . . . . . . . . . . . . . . . . . . . . . 5.1 Background . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.2 Desulfurization Basis of ESR . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.3 Dependence of Desulfurization on the Processing Parameters of ESR . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.3.1 Initial Sulfur Content of Consumable Electrode . . . . . . . 5.3.2 Remelting Atmosphere . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.3.3 Slag Composition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.3.4 Deoxidation Schemes of ESR . . . . . . . . . . . . . . . . . . . . . . 5.3.5 Melting Rate of ESR . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.3.6 Electrical Parameters of ESR . . . . . . . . . . . . . . . . . . . . . . . 5.4 Desulfurization Associated with Sulfide Inclusion Evolution During ESR . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.5 ESR for Sulfur-Bearing Steel Production . . . . . . . . . . . . . . . . . . . . 5.6 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Sulfide and Nitride Inclusion Evolution During ESR . . . . . . . . . . . . . . 6.1 Sulfide Inclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.2 Sulfide in Oxide–Sulfide Complex Inclusions . . . . . . . . . . . . . . . . . 6.2.1 From Original Attached State to Patch-Type and Shell-Type Sulfide . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.2.2 From Original Patch-Type to Shell-Type Sulfide . . . . . . . 6.3 Nitride Inclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.4 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

63 64 66 66 73 73 75 76 76 78 82 82 84 89 90 93 93 94 97 97 99 105 108 108 109 111 111 112 113 117 117 123 123 131 132 136 137

Contents

7

8

9

Evolution of Original Oxide Inclusions During ESR . . . . . . . . . . . . . . 7.1 Sites of Oxide Inclusion Removal During ESR . . . . . . . . . . . . . . . 7.2 Al2 O3 and MgO·Al2 O3 Inclusions . . . . . . . . . . . . . . . . . . . . . . . . . . 7.3 Calcium Aluminate Inclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7.4 Manganese Silicate Inclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7.5 Role of Processing Parameters of ESR on Inclusions . . . . . . . . . . 7.5.1 Deoxidation Schemes of ESR . . . . . . . . . . . . . . . . . . . . . . 7.5.2 Slag Composition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7.5.3 Melting Rate of ESR . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7.5.4 Electrical Parameters of ESR . . . . . . . . . . . . . . . . . . . . . . . 7.6 Newly-Formed Inclusions in Remelted Ingot . . . . . . . . . . . . . . . . . 7.7 Illustration of Inclusion Removal and Fresh Inclusion Generation During ESR . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7.8 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Evolution of Oxide Inclusions in Si–Mn-Killed Steel During ESR . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8.1 Background . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8.2 P-ESR Trials of Si–Mn Deoxidized Steel . . . . . . . . . . . . . . . . . . . . 8.3 Inclusions in Consumable Steel Electrode . . . . . . . . . . . . . . . . . . . . 8.4 Inclusions in the Liquid Metal Pool and Remelted Ingots . . . . . . . 8.5 Evolution Mechanism of Oxide Inclusions During Protective Atmosphere Electroslag Remelting . . . . . . . . . . . . . . . . 8.6 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Modification of Alumina and MgO·Al2 O3 Inclusions by Calcium Treatment During ESR . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9.1 Background . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9.2 Modification of MgO·Al2 O3 Spinel Inclusions in Steel During P-ESR . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9.2.1 P-ESR Procedure and Inclusion Characterization . . . . . . 9.2.2 Characteristics of Non-metallic Inclusions . . . . . . . . . . . . 9.2.3 Evolution and Modification of MgO·Al2 O3 Spinel Inclusions in the P-ESR Process . . . . . . . . . . . . . . . . . . . . 9.2.4 MgO·Al2 O3 Spinel Inclusion Modification in Industrial P-ESR Practice . . . . . . . . . . . . . . . . . . . . . . . . 9.3 Simultaneous Modification of Alumina and MgO·Al2 O3 Inclusions During P-ESR . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9.3.1 Inclusions in the Consumable Steel Electrode . . . . . . . . . 9.3.2 Inclusions and Primary Carbides in Remelted Ingots . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9.3.3 Composition Distribution of Oxide Inclusions in Remelted Ingots . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

xi

141 142 143 150 152 152 152 154 157 158 160 163 165 165 169 169 170 171 172 175 180 180 183 183 186 186 186 189 197 199 201 201 207

xii

Contents

9.3.4

Transient Inclusions in Liquid Metal Pool During P-ESR Refining . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9.3.5 Proposed Mechanism for Modification of Inclusions by Calcium During P-ESR Process . . . . . . 9.4 Evaluation of Al2 O3 and MgO·Al2 O3 Inclusion Modification Prior to ESR . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9.5 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10 Role of Calcium Modification of Oxide Inclusions During ESR on Primary Carbides . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10.1 Background . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10.2 Effect of Oxide Inclusions Modification on Primary Carbides in Tool Steel . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10.2.1 Precipitate Formation in the Steel . . . . . . . . . . . . . . . . . . . 10.2.2 Microstructure of As-Cast Remelted Ingot . . . . . . . . . . . . 10.2.3 Primary Carbides in As-Cast ESR Ingots . . . . . . . . . . . . . 10.2.4 Microstructure and Toughness of Annealed Steel . . . . . . 10.3 MgO·Al2 O3 Spinel Inclusions Modification on Primary Carbonitrides in Nickel-Base Superalloy . . . . . . . . . . . . . . . . . . . . . 10.3.1 Two-Dimensional Observation of Inclusions and Primary Carbonitrides . . . . . . . . . . . . . . . . . . . . . . . . . 10.3.2 Three-Dimensional Observation of Inclusions and Primary Carbonitrides . . . . . . . . . . . . . . . . . . . . . . . . . 10.3.3 Oxide Inclusions Evolution . . . . . . . . . . . . . . . . . . . . . . . . . 10.3.4 Effect of Oxide Inclusions Modification on Primary Carbonitrides . . . . . . . . . . . . . . . . . . . . . . . . . . 10.4 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

208 210 213 214 214 219 219 220 221 223 227 227 233 234 236 240 242 244 244

Chapter 1

Introduction to Electroslag Remelting

Abstract Electroslag remelting (ESR) combines liquid metal refining and solidification structure control in one set. As the last processing step of liquid metal refining and the original stage of various phases formation, ESR is regarded as the key stage in refining metal, macro- and microstructure of as-cast ingot. This chapter provides an introductory overview of the ESR technologies. The history of ESR technologies is traced. The technical principle of ESR is described. The advantages and limitations of ESR technologies are critically assessed. Perspective and remaining issues are noted. The reader is introduced with the high level overview provided by this chapter, prepared to build an understanding of ESR technologies.

Electroslag remelting (ESR) is increasingly used to produce some varieties of special steel and alloy which are usually utilized for critical components, such as in the fields of thermal and nuclear power generation, aerospace, oil and gas extraction, tool and crankshaft manufacturing. ESR has always been an important processing technology that contributes the largest yield in the family of special smelting technologies. Based on the original electroslag remelting technology, electroslag remelting has developed various new branches successively, mainly including protective atmosphere ESR, drawing-ingot-type ESR, electroslag casting with liquid metal, single power two circuits electroslag remelting with a current-carrying mold, electroslag cladding, pressurized electroslag remelting, vacuum electroslag remelting. Among ESR family, conventional ESR and protective atmosphere electroslag remelting have always been in vigorous development, and other branches are still in their infancy. Electroslag remelting was first put into industrial operation at Deneprospetsstal electrometallurgical plant (Zaporozh’e, Ukraine) and the NKMZ heavymanufacturing plant (Kramatorsk, Ukraine) in 1958 around the world [1]. Independent work of the development of the first electroslag remelting furnace was finished by Prof. Jue Zhu from Beijing University of Iron and Steel Technology (now known as University of Science and Technology Beijing) and co-workers in 1959 in China, successfully producing ball-bearing steel for aviation industry [2]. The industrial production of ESR was firstly finished in 1960 in China [2, 3]. It took approximately 10 years before ESR became an acknowledged process for mass production of high-quality steel and alloy worldwide. © Metallurgical Industry Press 2023 C. Shi et al., Electroslag Remelting Towards Clean Steel, https://doi.org/10.1007/978-981-99-3257-3_1

1

2

1 Introduction to Electroslag Remelting

The development of ESR experienced a recession period in 1980s and 1990s, partially because of the booming development of other liquid steel refining technologies (such as ladle furnace (LF) refining, Ruhrstahl-Heraeus (RH) refining). In recent ten years, ESR is always in full swing because the increasing requirements for the mechanical properties of some special steel and alloy, as well as the irreplaceable role of ESR in the production of some varieties of special steel and alloy (Other steelmaking and casting technologies cannot produce or the their product quality falls far short of the performance requirements.). Electroslag remelting has some unique superiorities, in comparison with other technologies for liquid steel secondary refining, such as LF, RH and vacuum oxygen decarburization (VOD). First, ESR is a unit in which the melting of consumable electrode, reaction reactions of liquid steel refining and the solidification of liquid steel take place simultaneously. The quality of steel product is significantly improved by ESR through refining liquid steel and solidified structure of as-cast ingot. Second, the refining process of liquid steel consists of the following steps: the formation of liquid metal films and their collection into metal droplets at the electrode tip, traversing of the metal droplets through the molten slag pool and the converging of metal droplets in the liquid metal pool. In addition, ESR generally is the last processing procedure for refining liquid metal and controlling solidification structure of as-cast ingot in the steel and alloy manufacturing process. ESR can optimize the cleanliness and the solidification structure of the ingot simultaneously, which is superior to other refining processing technologies. With remarkable refining capacity, ESR can minimize the secondary pollution of liquid steel caused by the refractory materials, and the solidification condition of liquid steel is superior. The directional solidification of liquid steel in the mold during ESR provides favorable conditions for eliminating the defects such as shrinkage, porosity and non-metallic inclusion which usually occur in ingot casting and continuous casting slab. Figure 1.1 presents the schematic diagram of a typical ESR apparatus. The consumable electrode is produced by cast (or forged) after Basic Oxygen Furnace (BOF), Electric Arc Furnace (EAF) steelmaking or Vacuum Induction Melting (VIM) route (combining with secondary refining in many cases). The consumable electrode is placed in a water-cooled copper mold that contains slag. ESR slag generally is CaF2 -CaO-Al2 O3 -based system with minor additions of MgO, TiO2 , and/or SiO2 to tailor the slag for the specific remelting requirements. In conventional ESR, the electrical current passes from power → consumable electrode → molten slag pool → liquid metal pool → solidified ingot → power as a circuit. The current superheats the electrically resistive slag. The steel electrode tip melts in the molten slag as the temperature increases resulting from Joule heating up to the melting point of the steel, and liquid metal films form continuously and subsequently collect into liquid metal droplets. These liquid metal droplets thereafter detach from the electrode tip, and then pass through the molten slag and collect in the liquid metal pool below the molten slag. Eventually, the liquid metal pool cools down and solidifies from its bottom to produce ingot. ESR enables the production of ingots with the desired shape, i.e., round, square, rectangular.

1 Introduction to Electroslag Remelting

3

Fig. 1.1 Schematic diagram of protective atmosphere electroslag remelting apparatus

The shortcoming of conventional ESR is that the entire refining process is carried out under atmospheric condition. In this case, it is unfavorable to reduce the loss of alloying elements and increase the cleanliness of the ingot. In order to supplement the weakness of conventional ESR, protective argon gas atmosphere ESR was developed. The first protective inert gas atmosphere ESR with an airtight protective gas hood, a retractable base plate and electrode change system was put in operation at the Accaierie Valbruna plant in Vicenza in 1997 [4, 5], and the first protective inert gas atmosphere ESR with single electrode and static mold was installed at Uddeholm Tooling in 1997 [5]. In protective argon gas atmosphere ESR, high purity argon gas is introduced into the gas protective cap of protective argon gas atmosphere furnace, so the whole remelting process is performed under the condition of air isolation. In addition, pressurized nitrogen electroslag remelting at different nitrogen partial pressures is recognized as a promising technology for producing high nitrogen steel. Vacuum electroslag remelting has been developed to further improve the cleanliness of the steel in terms of the oxygen and inclusion contents. The development of vacuum electroslag remelting is still in its infancy, and more studies are still needed for industrialization. Electroslag remelting belongs to secondary refining technology. Consumable electrode is the raw material of ESR. Consumable electrode can be produced through

4

1 Introduction to Electroslag Remelting

melting by other smelting technologies, such as EAF → ladle furnace (LF) refining → vacuum degassing (VD), BOF → LF, induction furnace → LF → VD. The main aims of ESR are to improve the cleanliness and solidification structure of ascast ingot, eventually achieve high quality of steel and alloy products. The main advantages of ESR are featured in the following aspects: (1) Lowering impurity element content, such as oxygen, sulfur and hydrogen. (2) Lower the amount of non-metallic inclusions. (3) Improve the solidification structure of as-cast ingot. As the electrode melts during ESR, liquid metal pool and molten slag pool move upward along the copper mold, which is favorable to the directional solidification of liquid steel. (4) Low segregation degree of alloying elements in as-cast ingot. The solidification structure is dense and the composition is uniform because fresh liquid steel is constantly replenished in liquid meta pool during the solidification process. The excellent performance of electroslag remelting products is largely dependent on the inherent quality (such as cleanliness, dense solidification structure and uniform composition), which is attributed to the superior refining efficiency of electroslag remelting refining, including effective removal of non-metallic inclusions, chemical reactions of slag-steel refining and sound solidification structure [6, 7]. For example, the strength of ultrahigh strength steel suitable for aerospace application after electroslag remelting are greatly improved due to the reduction of the amount and size of inclusions as well as the improvement of inclusion distribution uniformity [8]. Electroslag remelting has been an important technology to produce some varieties of special steel and alloy for many years. Control of solidification process in electroslag remelting is an important way to improve the properties of the products. The high cleanliness of steel and alloy can be obtained by electroslag remelting refining, which is attributed to the superior conditions of chemical reactions during the ESR process. Electroslag remelting is not only the casting process, but also the metal refining process. Due to the effective adsorption of inclusions by molten slag and chemical reactions at the slag-steel interface in the electroslag remelting, the contents of non-metallic inclusions and sulfur is greatly reduced, and consequently the remaining inclusions in the ingot are generally fine and evenly distributed [9]. The thermodynamic condition determines the chemical reaction in the electroslag remelting process, and the kinetic condition determines the progress of these chemical reactions. Electroslag remelting has the following exclusive characteristics in comparison with other technologies of secondary refining of liquid steel. (1) The melting, casting and solidification of liquid metal occur in a clean environment, which reduces the contamination to liquid steel. The liquid steel processing is under the molten slag pool, which minimizes the increase in the hydrogen and nitrogen contents and the reoxidation of liquid steel caused by air atmosphere. Because the melting and solidification process are completed in a water-cooled copper mold, there is no contamination of molten steel by refractory materials, which is a common issue in other liquid metal refining and casting processing.

1 Introduction to Electroslag Remelting

5

Table 1.1 Comparison of the contact conditions of slag-metal during ESR [10] Sites

Contact area of slag-metal (mm2 )

Weight of metal droplet, metal pool (g)

Specific area (mm2 /g)

Action time (s)

Specific area and time (mm2 s/g)

Electrode tip

14,820

4.9

3218

0.125

410.2

Metal droplet

230.1

4.9

47.9

0.088

4.26

Liquid metal pool

4520

28,200

0.16

991

159

Experimental conditions

Electrode diameter 100 mm, mold diameter 280 mm, current 4500 A, voltage 55 V, slag NAF-6, GCr15 bearing steel

(2) Superior conditions for the thermodynamics and kinetics of chemical reactions. ➀ Sufficient contact between liquid steel and molten slag. The contact between liquid steel and molten slag during ESR takes place at three stages: (I) formation of liquid metal films and their collection into metal droplets at the electrode tip, (II) the process by which metal droplets pass through molten slag pool, (III) interface between molten slag pool and liquid metal pool. Table 1.1 shows the comparison of slag-metal contact conditions at different sites during ESR of GCr15 bearing steel. It can be seen that the contact area of slag-metal at the electrode tip is 3218 mm2 /g at this experimental condition. The contact area of slag-metal 47.9 mm2 /g when metal droplets pass through the molten slag pool, which is much higher than that in other steelmaking processes. For example, the contact area of slag-metal is 0.3–0.7 mm2 /g in 10–30 tonne-scale EAF steelmaking. ➁ High temperatures for chemical reactions. The temperature of molten slag pool in electroslag remelting process has a decisive influence on the refining efficiency, electrode melting rate, liquid metal pool shape, forming and solidification. The surface temperature of the molten slag pool during ESR is in the range of 1620–1850 °C, and the temperature of molten slag pool generally is about 1750 °C [10]. ➂ Strong stirring in molten slag pool. The kinetics conditions of strong stirring in the molten slag pool enable the full contact of slag and steel in the electroslag remelting process. Meanwhile, due to the stirring effect of electromagnetic force, the contact of the slag and steel is constantly updated, which enhances the chemical reactions and promotes the removal of impurities and non-metallic inclusions. The factors that cause the strong stirring of the molten slag pool include electrodynamic force, electromagnetic shrinkage effect

6

1 Introduction to Electroslag Remelting

Fig. 1.2 Schematic illustration of dendrite formation in the mushy zone during the ESR process [11]

force, mass action, convection of molten slag, the force caused by gas overflow and expansion [10]. (3) Directional solidification from bottom to top ensures the uniform and dense solidification structure of as-cast ingot. The melting of electrode and the solidification of molten metal take place simultaneously in the electroslag remelting process. There is always liquid metal molten pool and heated molten slag pool on the solidified metal, which not only keeps heat but also has enough liquid metal filling shrinkage holes, which consequently can effectively eliminate the loose and shrinkage holes. The gas and inclusions in the liquid metal can float up. Therefore, the structure of as-cast ingot generally is dense and uniform. Depending on the ingot diameter, a directional solidification showing a dendritic structure is typically achieved in as-cast ingot (see Fig. 1.2). For heavy forging ingots with diameters up to 3600 mm or at high melting rates of ESR, equiaxed dendrites is observed at the center region of the ingot [11, 12]. (4) Ingot surface quality is sound due to the formation of a thin slag film between solidified steel and mold wall during the remelting operation. In the future, electroslag remelting will show irreplaceable advantages in the following aspects: (1) Electroslag remelting will have a monopoly position in the production of medium and heavy ingot. (2) Electroslag remelting have absolute advantages in the production of high quality tool steel, dual phase heat resistant stainless steel, ultra-high strength steel containing nitrogen, tube billet and cold roll steel.

1 Introduction to Electroslag Remelting

7

(3) In the production of super alloy (superalloy, corrosion-resistant alloy, precision alloy, electrothermal alloy), electroslag remelting and vacuum arc remelting are in a competitive situation. Electroslag remelting will take the absolute advantage in advanced material production. (4) Electroslag remelting is in the developing stage in nonferrous metal production. (5) Electroslag remelting has a unique position in the production of hollow ingot and special-shaped casting ingot, such as alloy tube used in petrochemical industry and 3-d special-shaped casting-crankshaft. (6) The development prospect of electroslag metallurgy is that the electroslag technology goes out of the remelting with single mold and combines with the steelmaking process into an online process, becoming a step in the link of melting, refining and continuous casting. However, in order to meet the requirements of the development of electroslag metallurgy and the future needs of special steel and alloy, the development of electroslag remelting technologies in advancing situations still needs to be considered from the following aspects [2]: (1) Larger scale of ESR ingot. In the development of energy engineering, petrochemical engineering, machinery, especially nuclear power industry, a large number of high-end forgings are needed. In order to further develop the electroslag metallurgy technologies in China, it is necessary to attach importance to the second generation of electroslag remelting to realize the optimal design of equipment, controllable atmosphere, controllable remelting process and controllable solidification process. Meanwhile, it is required to strengthen the level of computer-control and fundamental research. (2) Developing electroslag rapid remelting technology for high-speed steel production and protective argon gas atmosphere electroslag remelting for die steel. (3) Developing protective inert gas atmosphere electroslag continuous casting technology for high-alloying steel, and reducing the great change in the slag compositions and the deterioration of slag properties in the long-term ESR process. (4) Generalizing computer-control electroslag casting technology for heavy roll and crankshaft steel. (5) Conducting the studies on the effect of protective argon gas atmosphere electroslag remelting on the microstructure and properties of superalloy and nonferrous metal. (6) Carrying out fundamental research on the second-generation electroslag remelting, such as the characteristics of gas atmosphere, slag and liquid metal three-phase reactions, control of the characteristics of rapid melting and liquid metal solidification, control of trace elements in non-ferrous metal. (7) Carrying out studies on the solidification characteristics of molten steel in ESR process, and in the solidification of heavy ingots and composite roll, as well as the flow and solidification characteristics of molten steel during electroslag centrifugal casting, and developing computer-control electroslag centrifugal casting technology.

8

1 Introduction to Electroslag Remelting

(8) Performing studies on CaF2 -containing slag systems, and developing fluoridefree and low-fluoride slag for ESR. Conventional commercial ESR slag contains a large amount of CaF2 (typically 40–70 mass%), aiming to reduce melting temperature and viscosity of the slag. Although CaF2 plays an important role in ESR slag, the evaporation of fluoride from slag melts during ESR process has always been an extremely serious issue because it poses serious contamination of environment [13–15], health hazard to plant operators [15, 16], corrosion of plant equipment [17], as well as the change in slag chemistry. The chemistry variation generally lowers the reliability of operating practice and degrades the quality of remelted ingot, especially for large-scale ESR [18]. Unlike fluoride-free mold fluxes, only a few works have been reported regarding the development of fluoride-free or low-fluoride slag for ESR. Narita et al. [19] verified the availability of CaO–Al2 O3 –(MgO) slags in ESR in terms of operation stability, elements loss and power consumption, and found that the surface quality of ingot was strongly improved by 20 mass% CaF2 addition in slag. Mao et al. [20] showed the applicability of CaO–Al2 O3 –(MgO) slags in ensuring as-cast ingot quality, but the difficulties in liquid slag starting and the stability of ESR refining remained to be solved. The problems caused by the application of these fluoride-free slags in ESR production have been reviewed by Li [10], including the difficulties in liquid slag starting, and control of the as-cast surface quality, inclusions removal, desulfurization, etc. These issues are closely related to the thermo-physical properties of the slag, i.e., viscosity, crystallization characteristics, and electrical conductivity. To date, fluoride-free slag has not been applied in the ESR practical production yet. Although electroslag remelting technology has many advantages, it also has some limitations, such as low melting and solidification rates leading to a low productivity, oxidation of consumable electrode, the loss of some alloying elements in the steel and alloy, and environmental pollution. In large-scale ingot production, some new problems are facing, such as porosity of as-cast structure, solute segregation of solidified ingot and unsatisfied cleanliness of ingot. Further exerting the advantages of ESR, and improving and removing the limitation of ESR is very important in the development of ESR technology at present and in future. In parallel with the continuous development of new ESR technologies, it is also quite necessary to optimize the current ESR processing parameters. It is of great scientific and leading significance to exert great efforts to increase the cleanliness level of the steel to further improve its mechanical properties. Although many investigations on the ESR in various aspects have been carried out, it still needs more work to improve the cleanliness of liquid metal during the ESR for meeting the increasing requirements for advanced steel production.

References

9

References 1. Paton B E, Medovar L B. Improving the electroslag remelting of steel and alloys[J]. Steel in Trans., 2008, 38(12): 1028–1032. 2. Fe J. Development of first and secondary generation of electro-slag metallurgical technology[J]. Special Steel, 2010, 31(1): 18–23. 3. Li Z B. A new advance electroslag metallurgy in the 21th century[J]. Special Steel, 2004, 25(5): 1–5. 4. Holzgruber W. Overview of 50 years of development in electroslag remelting in Austria[J]. BHM Berg-und Hüttenmänn. Monatsh., 2016, 161(Suppl 1): 2–11. 5. Holzgruber W, Holzgruber H. Development trends in electroslag remelting[C]. Medovar Memorial Symposium, 2001: 71–77. 6. Weber V, Jardy A, Dussoubs B, et al. A comprehensive model of the electroslag remelting process: description and validation[J]. Metall. Mater. Trans. B, 2009, 40(1): 271–280. 7. Vaish A K, Iyer G V R, De P K, et al. Electroslag remelting – its status, mechanism and refining aspects in the production of quality steels[J]. J. Metall. Mater. Sci., 2000, 42(1):11–29. 8. Maity S K, Ballal N B, Goldhahn G, et al. Development of ultrahigh strength low alloy steel through electroslag refining process[J], ISIJ Int., 2009, 49(6): 902–910. 9. Hernandez-Morales B, Mitchell A. Review of mathematical models of fluid flow, heat transfer, and mass transfer in electroslag remelting process[J]. Ironmak. Steelmak., 1999, 26(6): 423-438. 10. Li Z B. Electroslag Metallurgy Theory and Practice[M]. Metallurgical Industry Press, Beijing, China, 2010. (in Chinese) 11. Radwitz S, Scholz H, Friedrich B, et al. Influencing the electroslag remelting process by varying fluorine content of the utilized slag[C]. Proc. European Metallurgical Conf. 2015, Vol. 2, GDMB Society of Metallurgists and Miners, Düsseldorf, 2015: 887–896. 12. Shi C B, Zheng X, Yang Z B, et al. Effect of melting rate of electroslag rapid remelting (ESRR) on the microstructure and carbides in a hot work tool steel[J]. Metals Mater. Int., 2021, 27(9): 3603–3616. 13. Nakada H, Nagata K. Crystallization of CaO–SiO2 –TiO2 slag as a candidate for fluorine free mold flux[J]. ISIJ Int., 2006, 46(3): 441–449. 14. Takahira N, Hanao M, Tsukaguchi Y. Viscosity and solidification temperature of SiO2 –CaO– Na2 O melts for fluorine free mould flux[J]. ISIJ Int., 2013, 53(5): 818–822. 15. Klug J L, Hagemann R, Heck N C, et al. Fluorine-free mould powders for slab casting: crystallization control in the CaO–SiO2 –TiO2 –Na2 O–Al2 O3 system[J]. Steel Res. Int., 2012, 83(12): 1186–1193. 16. Persson M, Seetharaman S, Seetharaman S. Kinetic studies of fluoride evaporation from slags[J]. ISIJ Int., 2007, 47(12): 1711–1717. 17. Omoto T, Iwamoto Y, Yamaji H. Development of environment friendly mold powder[J]. Shinagawa Tech. Rep., 2002, 45: 85–92. 18. Xiang D L. Some problems meriting attention in large-scale ESR[J]. Heavy Casting Forging, 2011, 1: 26–35. (in Chinese) 19. Narita K, Onoye T, Ishii T, et al. A metallurgical study of oxide-base slag used for electroslag remelting[J]. Tetsu-to-Hagané, 1978, 64(10): 1568–1577. 20. Mao H X, Li Z B. A metallurgical study on low fluorine and fluorine - free slag for electroslag remelting[J]. Central Iron Steel Res. Inst. Tech. Bull., 1983, 3(4): 597–611.

Chapter 2

Clean Steel Production by Electroslag Remelting

Abstract This chapter presents the general indexes of steel cleanliness. The detriments of oxide inclusions, sulfur, hydrogen, nitrogen and phosphorus to the steel are described. The outline of the aspects involved in the oxygen, oxide inclusions, sulfur, hydrogen, nitrogen and phosphorus control during the ESR process are discussed. The limitation in lowering the nitrogen and phosphorus contents of liquid steel and alloy during the ESR is presented.

2.1 Cleanliness Target The term “clean steel” is commonly used to describe the steel that contains (1) low levels of oxygen, sulfur, phosphorus, nitrogen and hydrogen, (2) low non-metallic inclusion amount and controlled size distribution, morphology and composition of the inclusions, (3) minimized metallic impurity elements such as As, Sn, Sb, Se, Cu, Zn, Pb, Cd, Te and Bi. Steel cleanliness is a relative term. It varies with steel grades and the end use of the steel, as summarized in Ref. [1] for some varieties of steel products. Many grades of the steel produced by ESR also have strict requirements on the cleanliness. Steel cleanliness is an index that can be monitored to realize improvements for process optimization, which is included in a wide range of important operating conditions throughout the steelmaking and casting processes. Targeting high cleanliness of the steel has always been the key task of metallurgical workers. The key technical issue in clean steel manufacture is the control of inclusion amount, chemistry, morphology and size distribution. One of the aims of steelmaking and casting in steel plants has always been to minimize the inclusion amount in steel and to ensure that the chemistry and size distribution of remaining inclusions are strictly controlled.

© Metallurgical Industry Press 2023 C. Shi et al., Electroslag Remelting Towards Clean Steel, https://doi.org/10.1007/978-981-99-3257-3_2

11

12

2 Clean Steel Production by Electroslag Remelting

2.2 Oxygen and Oxide Inclusions The increasing demands for more excellent properties of tool steel urge steelmakers to constantly improve the steel cleanliness. The detriments of non-metallic inclusions to steel properties have been widely recognized, such as in the review articles by Zhang [2] for tire cord steels and by Park and Kang [3] for stainless steels. The sources of oxide inclusions in steel include: [1] deoxidation products as the reaction between the dissolved oxygen and the added deoxidizing agent, reoxidation products, slag entrapment, exogenous inclusions from other sources (including loose dirt, broken refractory brickwork and ceramic lining particles), chemical reactions (for example produce oxides from inclusion modification when calcium treatment is improperly performed.). Non-metallic inclusions lower the steel cleanliness and cause defects leading to worsening of desired mechanical properties of steel products. In addition, the presence of inclusions causes poor steel castability often resulting in slab downgrades and rejections, increased costs associated with recycling of liquid steel and refractories, and even shut down of the caster [4]. The detriments of non-metallic inclusions to steel differ according to the application of the specific steel grade. For example, Al2 O3 inclusions generally reduce the fatigue life of bearing steel and toughness of extrusion tool steel. In addition, MgO · Al2 O3 spinel inclusions cause sliver-like defect appeared on the cold sheet of stainless steel, sulfide inclusions lower the corrosion resistance of stainless steels. Anyway, large inclusions of any types are undesirable, while very fine dispersion can be either helpful or harmful. Total inclusion content (as measured by total oxygen content) has traditionally correlated with bearing life. Decreasing the total oxygen content generally gives significant increases in the bearing life. But this is not always the case. In addition to total oxygen content, the size, distribution and morphology of inclusions correlates well with bearing life. Figure 2.1 shows the correlation between bearing life and cleanliness expressed in terms of total oxygen level. Reducing total oxygen content of the steel from 30 to 5 ppm, which has been accomplished by many steel plants in practice improvement, has increased life by a factor of 100. High cleanliness of the steel is obtained by careful steelmaking operating practices, incorporating ladle slag control, state of the art in protection against oxygen pickup, effective inclusion removal in the ladle, etc. It can be seen from Fig. 2.1 that although the total oxygen content of the bearing steel produced by ESR is about twice as many as that produced by EAF + EBT or VAR, the bearing life produced by ESR, EAF + EBT or VAR is comparable. This is because total oxygen content as an index cannot represent the chemical compositions, types, size and distribution of oxide inclusions in the steel. The size and distribution of oxide inclusions in the ESR ingots are superior to that in the steel produced by other steelmaking technologies, therefore the life of the bearing made from ESR ingot is higher [6]. The study by Zhou et al. [7] shows that although the oxygen content of ESR bearing steel is higher than that of the bearing steel produced by continuous casting, the fatigue life of ESR bearing steel is higher than the latter because of its

2.2 Oxygen and Oxide Inclusions

13

Fig. 2.1 Evolution of bearing life as function of the total oxygen content in the steel produced by different steelmaking technologies [5]

less large inclusions and well dispersed inclusions. The reduction of the total oxygen content of the ESR bearing steel is vitally important for improving the bearing life. Reducing the total oxygen content of ESR bearing steel from about 0.0040 mass% to 0.0005 mass%–0.0015 mass% increases the bearing life by 0.3–0.5 times [8]. As a typical grade of special steels, tool steel has a strict requirement for its cleanliness. Tool steel, used as material to manufacture dies and tools, is special steel featuring high shock resistance, wear resistance, fatigue strength and hightemperature strength. The mechanical properties of tool steel are highly dependent on its manufacture processing to a great extent. In order to improve the cleanliness of tool steel, secondary refining processing is widely used, such as vacuum refining, ESR, vacuum arc remelting (VAR), to improve the cleanliness of the steel and reduce the contents of impurity elements and non-metallic inclusions. As the content of nonmetallic inclusions decreases, especially the brittle inclusions, the surface polishing properties of plastic die steel can be improved effectively. Targeting ultralow oxygen and sulfur contents (less than 0.0010 mass% and 0.0015 mass%, respectively) in the steel is an important aspect of high-quality tool steel production in steel plants around the world. Among the manufacturing processes currently employed for enhancing the steel cleanliness, ESR is one of the most widely used technology around the world since it permits removal of large detrimental defects. Taking the tool steel used to make dies as an example, according to the comparison of the product service performance, the advantage of the tool steel produced by outstanding enterprises is mainly embodied in its cleanliness, uniformity, microstructure, and dimension of steel product module. High cleanliness is represented by the low contents of sulfur, phosphorus, and oxide inclusions. High uniformity refers to the high ratio of transverse/longitudinal impact toughness, and the small hardness difference (less than 1 HRC) in the internal and external region of the die steel. Superior microstructure indicates fine carbides and grains. Precise dimension means high accurate size of die steel product module supplied to users. Increasing the cleanliness of steel can enhance the thermal cracking resistance of hot work tool steel, consequently improving the service life of the die steel [9]. Tool

14

2 Clean Steel Production by Electroslag Remelting

steel can be produced by different production routes: EAF → secondary refining, EAF → secondary refining → ESR, BOF → secondary refining, vacuum-induction melting → secondary refining → ESR. Among these routes, ESR is widely used in the production of high-quality tool steel in order to increase the cleanliness, compactness and uniformity of the steel, so as to greatly improve the isotropy and enhance the transverse toughness, plasticity, thermal fatigue resistance and fracture toughness of the die steel. The improvement of the properties of the die steel is related to the increase in the cleanliness and the reduction of non-metallic inclusions content as well as the significant reduction of solute segregation and primary carbides. ESR processing generally improves the impact property, elongation and section shrinkage of die steel, as well as the durability, heat resistance and the polishing property of die steel. For instance, AISI H13 die steel produced by ESR has been praised by many users because of its high cleanliness, uniform microstructure, excellent strength and superior toughness. The transverse toughness and plasticity of H13 die steel produced by EAF process is only equivalent to about 50% of the longitudinal performance, whereas this ratio of the die steel produced by ESR can reach higher than 80% [10]. The assessment of Very High Cycle Fatigue (VHCF) response of the conventional casting H13 steel and of a H13 steel subjected to ESR shows that all fatigue failures originated from the non-metallic inclusions [11]. Since the probability of finding critical defects in a loaded volume increases with the loaded volume, the loaded volume also significantly affects the VHCF response. This is generally referred to as the “size-effect” in VHCF. The enhancement of the steel cleanliness through the ESR process, moreover, limits the size-effect, mainly due to a significantly reduced inclusion-size range after the refining process [11]. ESR is widely employed in the production of high-quality die steel, such as the grades H13, S136 and NAK80. Increasing the cleanliness of die steel has become an important approach to improve the performance properties and service life of die steel. The cleanliness control during ESR of steel should be focused on the oxygen content and non-metallic inclusions for high-quality steel production. Non-metallic inclusions are generated in the steelmaking and casting process mainly during the deoxidation steps. Great efforts have been put forward to minimize the population of oxide inclusions by decreasing the oxygen content in the steel during steelmaking process. Total oxygen content is widely used to characterize steel cleanliness in terms of oxide inclusions and assess process improvement and quality control. Total oxygen is the sum of soluble oxygen in liquid steel and that present as oxide inclusions. The major advantage of this technique is the ease and speed of obtaining values that are directly related to cleanliness and easier to correlate with steelmaking processing parameters, such as processing time and the chemical compositions of the steel itself and slag chemistry. Electroslag remelting refining generally lower the oxygen content and remove non-metallic inclusions in steel considerably. During the ESR refining, the removal efficiency of oxide inclusions and generation of fresh oxide inclusions determine the oxygen level in the remelted ingot. The variation of the oxygen content in steel during ESR is closely related to the oxide inclusion evolution in this refining process.

2.2 Oxygen and Oxide Inclusions

15

The inclusion characteristics in the electroslag remelted steel and inclusion evolution during ESR refining have been extensively investigated in the last several decades. However, these previous studies were focused on the oxide inclusion characteristics under the condition of decreasing oxygen content (or without a mention of the oxygen level) in steel, and did not refer to the role of oxygen pickup during ESR on inclusion characteristics. The reoxidation of liquid steel do take place during ESR of low oxygen steel, and consequently led to the generation of fresh oxide inclusion [12]. The correlation of the oxygen and the oxide inclusion during ESR of the steel with a low oxygen content (< 15 ppm) is highly needed to be ascertained. Moreover, the information regarding the evolution of calcium aluminate inclusions during ESR is very scarce in the literatures. A thorough understanding of the evolution of calcium aluminate inclusions during ESR is quite necessary in order to take actions to control the inclusions not only in the ESR process, but also during the secondary refining of liquid steel for steel electrode production. As ESR is the last processing procedure for refining liquid steel in the steel product manufacturing process, the deoxidation and oxide inclusion evolution in the ESR process has always been a main focus. Deoxidation of liquid steel and alloy during the ESR is always an ongoing concern for producing advanced clean steel and alloy. Deoxidation of liquid steel during ESR is basically different from that in other steelmaking process operations. Deoxidation of ESR is dependent on one or more aspects including the initial oxygen content and oxide inclusions in the electrode, remelting atmosphere, deoxidation schemes, slag compositions, reoxidation degree, melting rate and filling ratio. To achieve a precise control of the inclusion amount, composition and its size distribution in the steel products, it is important to understand the inclusion characteristics at different steps of the steelmaking process. Towards this direction, steel samples are taken from the liquid steel during the steelmaking process and/or from the final steel products for inclusion characterization. Some common methods to evaluate the steel cleanliness are indirect methods that include total oxygen analysis, changes in slag composition or elemental fades/gains, clogging in submerged entry nozzles, and use of slag tracers. Direct methods to evaluate steel cleanliness are performed through inclusion characterization in steel samples. The characterization of inclusions consists of evaluating their size, distribution, morphology and chemical composition. There are a variety of techniques that can be used industrially to characterize the inclusion in steel, such as metallographic examination by scanning electron microscope (manual or automated SEM-EDS), laser techniques, remelt button analysis technique, optical emission spectrometry with pulse discrimination analysis (OES-PDA), ultrasonic testing, liquid metal cleanliness analyser, electrolytic extraction, confocal laser scanning microscope, cathode luminescence microscope (CLM), etc. Detailed reviews of inclusion characterization techniques were reported elsewhere [4, 13].

16

2 Clean Steel Production by Electroslag Remelting

2.3 Sulfur and Desulfurization The detriments of sulfur to the properties of steel and alloy have been widely recognized, for example, an increase in the sulfur content deteriorates the fracture toughness and hot ductility of steel [14, 15], and lowers the endurance strength of superalloy [16], as well as leads to sulfide-induced stress corrosion crack initiation in steel [17]. The desulfurization capacity of ESR is affected by various factors, such as the oxygen level of liquid metal [18], slag chemistry [19], remelting atmosphere [20] and sulfide inclusions [21]. ESR has been distinguished among other refining processes by its ability to remove sulfur typically down to a few hundredths or even thousandths of percent regardless of the electrode composition, in which the sulfur content is reduced by 50–80% from the electrode to the remelted ingot generally [22–24]. Two reactions, i.e., slag/ metal reaction (2.1) and gas/slag reaction (2.2), govern the sulfur removal in the ESR process. [S] + (O2− ) = (S2− ) + [O]

(2.1)

(S2− ) + 3/2{O2 } = {SO2 } + (O2− )

(2.2)

where [ ], ( ) and { } refer to a species in liquid metal, slag and gas phases, respectively. For some kinds of steel grades, a high level of sulfur content is expected, such as free-cutting steel, some of crankshaft steels and die steels. Electroslag remelting has an outstanding desulfurization capability. It is particularly difficult to achieve the uniformity of the sulfur at a high level in these sulfur-bearing steels in ESR industrial production.

2.4 Hydrogen Hydrogen in steel can cause the hydrogen embrittlement depending on its level, which degrades the mechanical properties and premature failure of the steel [25, 26]. Hydrogen embrittlement has been a subject of extensive studies over the past several decades. The features and proposed mechanisms of hydrogen embrittlement have been documented in review articles [25, 27]. The hydrogen content of ESR ingot is dependent on the partial pressure of water vapor in the atmosphere, gas/slag interface area, slag composition, moisture content of the slag, hydrogen content of the electrode, remelting rate, and slag quantity. At the early stage of ESR process, hydrogen pickup of steel ingot caused by the moisture introduced by the slag soon decreases. Then the hydrogen content of the ingot enters a stable state, and the hydrogen content is largely dependent on the water vapor pressure of the remelting atmosphere. Masui et al. [28] demonstrated that the

2.4 Hydrogen

17

Fig. 2.2 Effect of water vapor pressure on hydrogen content of steel in steady state during the ESR at the melting rate of about 100 kg/ h (other conditions are constant) [28]

hydrogen content of remelted ingot is √ proportional to the square root of water vapor pressure of the remelting atmosphere p(H2 O), as shown in Fig. 2.2. Minimizing the hydrogen content in ESR ingot is usually realized through lowering the hydrogen content of the consumable electrode, the moisture introduced by the slag components, and the moisture in the remelting atmosphere. To reduce the hydrogen content of the consumable electrode, vacuum treatment is employed during secondary refining of liquid steel (vacuum degassing (VD) or Ruhrstahl-Heraeus (RH) refining) for producing consumable steel electrode. In addition, the rust on the electrode steel surface has to be removed mechanically prior to ESR processing in order to prevent the decomposition of hydroxide at elevated temperatures during the ESR process. Slag roast prior to ESR is indispensable to lower the moisture introduced by the slag components during the ESR process. The study by Rohde and Lohr [29] shows that the hydrogen content of remelted ingot is almost equal to that in the electrode (0.0001–0.0002 mass%) when strictly roasted slag is used in protective inert atmosphere ESR. Pre-melted slag has been widely used for ESR. This kind of slag is also needed to be roasted prior to its use in the ESR. At present, pre-melted slag is roasted at around 800 °C for several hours at different ESR plants. There is still a lack of scientific slag roasting schemes for different slag compositions. For the production of the steel with a high requirement for the hydrogen content, it is better to adopt liquid slag starting for the ESR, in which the hydrogen content of the steel will be lowered to a low level effectively. In order to reduce the moisture in the remelting atmosphere of ESR, argon gas after drying is generally adopted in ESR for keeping a protective atmosphere. Inert gas drying can be done by passing the gas through a drying adsorbent, and then the dried inert gas is introduced into an airtight protective gas hood of ESR furnace. Moisture-free argon atmosphere is very favorable to minimize the hydrogen content in steel and alloy ingot in the normal remelting period. Compared with other aspects

18

2 Clean Steel Production by Electroslag Remelting

of cleanliness control during electroslag remelting, lowering hydrogen content has not been a tough issue. In generally, the hydrogen content in the steel and alloy after electroslag remelting could be kept less than 0.0002 mass% without particular difficulties.

2.5 Nitrogen and Phosphorus The removal of nitrogen during the ESR is a tough task. When the steel contains nitride-forming elements such as Ti and Nb, these elements can be combined with nitrogen to form titanium nitride or niobium nitride. Therefore, nitrogen in the liquid steel is no longer possible to be removed from the liquid steel in the form of bubbles. Nitride inclusions have high melting-point and low surface energy. It is diffusely distributed in liquid steel and difficult to be removed by floating up. For some kinds of steel grades, a high level of nitrogen content is expected in these steels, nitrogen is an inexpensive alloying element for increasing the corrosion resistance, tensile strength, heat resistance, corrosion resistance, and oxidation resistance [30, 31]. Pressurized electroslag remelting has been recognized as a promising technology for producing high nitrogen steel [32]. The transfer behavior of nitrogen at different nitrogen partial pressures and electrode immersion depths during pressurized electroslag remelting has been ascertained by Yu et al. [31]. Phosphorus is an undesirable element in most of the steels as it greatly decreases the mechanical properties of the steel, such as ductility, embrittlement, and weldability induce and induction of intergranular fracture [33, 34]. It is widely accepted that the dephosphorization of liquid metal is reached in an oxidation process. The dephosphorization and oxygen removal are not compatible in the ESR. The slag used in the ESR shows low dephosphorization ability due to the low iron oxide and manganese oxide contents and the absence of oxidizers [20], and very high temperature of slag pool (varying typically around from 1600 to 1800 °C) [35, 36]. Hence dephosphorization conditions are not valid in the conventional ESR. Therefore, in practical production, the dephosphorization of liquid metal is always achieved by other secondary refining processing technologies, rather than ESR.

2.6 Summary In this chapter the general indexes of steel cleanliness, i.e., oxygen, oxide inclusions, sulfur, hydrogen, nitrogen and phosphorus are presented as well as the processing parameters of ESR associated with these indexes. The detriments of non-metallic inclusions to steel properties have been widely recognized. Targeting high cleanliness of the steel has always been the key task of metallurgical workers. Deoxidation of liquid steel during ESR is basically different from that in other steelmaking process operations. The removal efficiency of oxide inclusions and generation of

References

19

fresh oxide inclusions during the ESR process determine the oxygen level in the ascast ingot. For metallographic characterization of non-metallic inclusions, various mature technologies have been developed, some of which supplement each other for better characterizations. Slag/metal reaction and gas/slag reaction govern the sulfur removal in the ESR process. The desulfurization ratio of ESR generally is 50–80%. It is particularly difficult to achieve the uniformity of the sulfur at a high level in sulfur-bearing steel in ESR industrial production. Minimizing the hydrogen content of the ingot is realized by lowering the hydrogen content of consumable electrode, the moisture of the slag and the moisture in the remelting atmosphere through vacuum treatment of liquid steel for producing the steel electrode, slag roast and the application of dried argon gas atmosphere. Nitrogen could hardly be removed from liquid metal during the ESR. Dephosphorization is invalid in the conventional ESR. The dephosphorization of liquid metal is always achieved by other secondary refining processing technologies prior to ESR.

References 1. Zhang L, Thomas B G. State of the art in evaluation and control of steel cleanliness[J]. ISIJ Int., 2003, 43(3): 271–291. 2. Zhang L. State of the art in the control of inclusions in tire cord steels – a review[J]. Steel Res. Int., 2006, 77(3): 158–169. 3. Park J H, Kang Y. Inclusions in stainless steels - a review[J]. Steel Res. Int., 2017, 88(12): 1700130. 4. Kaushik P, Pielet H, Yin H. Inclusion characterisation – tool for measurement of steel cleanliness and process control: Part 1[J]. Ironmak. Steelmak., 2009, 36(8): 561–571. 5. Birat J P. Impact of steelmaking and casting technologies on processing and properties of steel[J]. Ironmak. Steelmak., 2001, 28(2): 152–158. 6. Zhou D G, Wang C S, Qian M, et al. Improving bearing steel inclusion through Ca-Si deoxidation and acid-slag ESR[J]. Iron Steel, 1994, 29(7): 25–28. (in Chinese) 7. Zhou D G, Chen X C, Fu J, et al. Inclusions in electroslag remelting and continuous casting bearing steel[J]. J. Univ. Sci. Technol. Beijing, 2000, 22(1): 26–30. (in Chinese) 8. Wang C S, Liu S G, Xu M D, et al. Reducing oxygen content in electro-slag remelted bearing steel GCr15[J]. Special Steel, 1997, 18(3): 31–35. (in Chinese) 9. Ma D S, Zhou J, Chen Z Z, et al. Influence of thermal homogenization treatment on structure and impact toughness of H13 ESR steel[J]. J. Iron Steel Res., Int., 2009, 16(5): 56–60. 10. Jiang X J. Production practice of hot die steel H13 by BOF[J]. Baosteel Technology, 2007, (3): 29–32. (in Chinese) 11. Tridello A. VHCF response of two AISI H13 steels: effect of manufacturing process and size-effect[J]. Metals, 2019, 9(2): 133. 12. Shi C B, Wang H, Li J. Effects of reoxidation of liquid steel and slag composition on the chemistry evolution of inclusions during electroslag remelting[J]. Metall. Mater. Trans. B, 2018, 49(4): 1675–1689. 13. Bartosiaki B G, Pereira J A M, Bielefeldt W V, et al. Assessment of inclusion analysis via manual and automated SEM and total oxygen content of steel[J]. J. Mater. Res. Technol., 2015, 4(3): 235–240. 14. Morinaga M, Murata Y, Hashizume R, et al. Remarkable improvement in steam oxidation resistance due to the presence of sulfur in high Cr ferrtic steels[J]. ISIJ Int., 2001, 41(3): 314–316.

20

2 Clean Steel Production by Electroslag Remelting

15. Kizu T, Urabe T. Hot ductility of sulfur-containing low manganese mild steels at high strain rate[J]. ISIJ Int., 2009, 49(9): 1424–1431. 16. Li Q, Zhang H, Gao M, et al. Mechanisms of reactive element Y on the purification of K4169 superalloy during vacuum induction melting[J]. Int. J. Miner. Metall. Mater., 2018, 25(6): 696–703. 17. Bai X F, Sun Y H, Chen R M, et al. Formation and thermodynamics of CaS-bearing inclusions during Ca treatment in oil casting steels[J]. Int. J. Miner. Metall. Mater., 2019, 26(5): 573–587. 18. Shi C B. Deoxidation of electroslag remelting (ESR) – a review[J]. ISIJ Int., 2020, 60(6): 1083–1096. 19. Mehrabi K, Rahimipour M R, Shokuhfar A. The effect of slag types and melting rate on electro-slag remelting (ESR) processing[J]. Int. J. Iron Steel Soc. Iran, 2005, 2(1): 37-42. 20. Mattar T, EI-Fawakhry K, Haifa H, et al. Effect of nitrogen alloying on sulphur behaviour during ESR of AISI M41 steel[J]. Steel Res. Int., 2008, 79(9): 691–697. 21. Shi C B, Zheng D L, Guo B S, et al. Evolution of oxide-sulfide complex inclusions and its correlation with steel cleanliness during electroslag rapid remelting (ESRR) of tool steel[J]. Metall. Mater. Trans. B, 2018, 49(6): 3390–3402. 22. Narita K, Onoye T, Ishii T, et al. A metallurgical study of oxide-base slag used for electroslag remelting[J]. Tetsu-to-Hagané, 1978, 64(10): 1568–1577. (in Japanese) 23. Shi C B, Chen X C, Guo H J, et al. Assessment of oxygen control and its effect on inclusion characteristics during electroslag remelting of die steel[J]. Steel Res. Int., 2012, 83(5): 472–486. 24. Detrois M, Jablonski P D, Hawk J A. Evolution of tantalum content during vacuum induction melting and electroslag remelting of a novel martensitic steel[J]. Metall. Mater. Trans. B, 2019, 50(4): 1686–1695. 25. Nagumo M. Function of hydrogen in embrittlement of high-strength steels[J]. ISIJ Int., 2001, 41(6): 590–598. 26. Shibayama Y, Hojo T, Akiyama E. An evaluation method for hydrogen embrittlement of high strength steel sheets using U-bend specimens[J]. ISIJ Int., 2021, 61(4): 1104–1111. 27. Barth C F, Steigerwald E A. Evaluation of hydrogen embrittlement mechanisms[J]. Metall. Trans., 1970, 1(12): 3451–3455. 28. Masui A, Sasajima Y, Sakata N, et al. Some important factors affecting hydrogen pick-up and oxidation during ESR treatment[J]. Tetsu-to-Hagané, 1977, 63(13): 2181–2190. 29. Rohde L, Lohr D. Operational results of the 50 ton ESR plant of Thyssen Heinrichs[C]. Proceedings of the 5th International Conference on Vacuum Metallurgy and Electroslag Remelting Processes. Munich, 1976, 177. 30. Takahashi F, Momoi Y, Kajikawa K, et al. Effect of nitrogen content on cold working properties of high strength Mn-Cr-N steel made by pressurized ESR[J]. Tetsu-to-Hagané, 2014, 100(8): 943–950. 31. Yu J, Liu F B, Li H B, et al. Numerical simulation and experimental investigation of nitrogen transfer mechanism from gas to liquid steel during pressurized electroslag remelting process[J]. Metall. Mater. Trans. B, 2019, 50(6): 3112–3124. 32. Holzgruber W. New ESR technology for new improved products[C]. Proc. 7th Int. Conf. on Vacuum Metallurgy, Iron Steel Institute of Japan, Tokyo, 1982, 1452–1458. 33. Komazazki S, Watanabe S, Misawa T. Influence of phosphorus and boron on hydrogen embrittlement susceptibility of high strength low alloy steel[J]. ISIJ Int., 2003, 43(11): 1851–1857. 34. Yoshida N, Umezawa O, Nagai K. Influence of phosphorus on solidification structure in continuously cast 0.1 mass% carbon steel[J]. ISIJ Int., 2003, 43(3): 348–357. 35. Choudhary M, Szekely J. The modeling of pool profiles, temperature profiles and velocity fields in ESR systems[J]. Metall. Trans. B, 1980, 11(3): 439–453. 36. Jardy A, Ablitzer D, Wadier J F. Magnetohydronamic and thermal behavior of electroslag remelting slags[J]. Metall. Trans. B, 1991, 22(1): 111–120.

Chapter 3

Deoxidation of ESR and Its Correlation with Oxide Inclusions

Abstract This chapter presents the deoxidation of ESR and related affecting factors, which have been accomplished by the research groups around the world over the past decades. The oxygen transfer behavior during ESR process is described first. The thermodynamic and kinetics considerations on the deoxidation of ESR are described and the underlying mechanisms are discussed. Deoxidation during the ESR is therefore affected by multiple factors, such as initial oxygen content of steel electrode, metal compositions, deoxidizing agents, absorption ability of slag to oxide inclusions, remelting atmosphere, slag compositions, and oxide inclusion evolution. Next, the dependence of the oxygen on the processing parameters of ESR is reviewed and discussed. In view of the indivisible correlation of the deoxidation with oxide inclusions, the role of deoxidation operations of ESR on the oxide inclusions is also evaluated. Finally, a general concluding remark and perspective for future work are present.

3.1 Background Deoxidation of liquid steel and alloy is always an ongoing concern for eliminating the detriments of oxide inclusions to the processing and mechanical properties of the steel and alloy. Great efforts have been put forward to minimize the amount of oxide inclusions by decreasing the oxygen content of the steel during ESR process [1–6]. Deoxidation of liquid steel during ESR is basically different from that in other steelmaking process operations. The iron oxide activity of the slag is a measure of its oxygen potential during ESR [1, 2, 6, 7]. The intention of deoxidizing agent addition in ESR process is to deoxidize the molten slag [1–4, 8, 9]. In the ESR process, the oxygen level of liquid steel is determined by the interactions of atmosphere-slag-metal-inclusion. Deoxidation during ESR is therefore affected by multiple factors, such as alloy compositions, deoxidizing agents, absorption ability of slag to oxide inclusions, remelting atmosphere, slag compositions, and oxide inclusion evolution. Some of these factors could only be varied in a limited range in the ESR practical production. As for a specific ESR practice, the contribution ratio of each of these factors is different, namely major, minor or negligible role. © Metallurgical Industry Press 2023 C. Shi et al., Electroslag Remelting Towards Clean Steel, https://doi.org/10.1007/978-981-99-3257-3_3

21

22

3 Deoxidation of ESR and Its Correlation with Oxide Inclusions

Lowering the oxygen potential of slag could not only reduce the oxygen introduction from the slag to liquid metal, but also suppress the loss of alloying elements, such as Ti, Al, Si and Mn, in liquid metal during the ESR [3, 8, 10]. Deoxidation of ESR comes down to the removal degree of oxide inclusions in liquid metal [11]. The effects of the abovementioned factors, which influence the deoxidation of ESR, on the oxide inclusions therefore should also be addressed because different types of inclusions basically experience various evolution trajectories during the ESR, as demonstrated in the previous studies regarding Al2 O3 , MgO · Al2 O3 , nitride, sulfide, and calcium aluminate inclusions [11–16]. However, only a few works regarding the correlation of deoxidation and oxide inclusions have been reported. In view of the significance of decreasing oxygen level, the complexity of multiple factors affecting the deoxidation, and infeasibility in on-line monitor during on-going ESR, it is quite important to summarize the existing works on the deoxidation and associated oxide inclusions during ESR for the purpose of assisting future work. Electroslag remelting is generally operated using high frequency alternating current in production practices around the world so far. Although low frequencies of alternating current (up to < 5 Hz) and direct current of ESR exert apparent influences on the contents of oxygen and non-metallic inclusions in steel according to laboratory-scale ESR trials [17], the effect of power supply modes on the deoxidation of ESR is not summarized in this article.

3.2 Oxygen Transfer During ESR Process The oxygen in the ESR process arises from the following sources: (1) The original oxygen in the consumable electrode. (2) The oxide scale on the electrode steel surface. (3) The oxygen in the air atmosphere, which affects the oxygen content of liquid steel in two different ways: (i) atmospheric oxygen permeates directly through molten slag into liquid metal pool by diffusion of physically dissolved oxygen. It should be mentioned on the basis of experimentally measured permeability that the permeation of oxygen as O2 , which physically dissolves in molten slag, plays a minor role in oxygen transfer [18–20], (ii) atmospheric oxygen reacts with the steel electrode at the electrode-atmosphere surface to form iron oxide as the electrode is heated at elevated temperatures. (4) Reducible oxides in the slag. (5) The moisture in the atmosphere and in the slag reacts with O2− in the slag to form (OH− ), which thereafter introduces oxygen into liquid steel [21]. (6) The dissolved oxygen as the products of desulfurization reactions in the ESR process. A schematic illustration of the oxygen transfer in the ESR process is shown in Fig. 3.1. The relative contribution of each of these sources is dependent on the specific ESR conditions. In the case of ESR for producing heavy ingot, the oxide scale that originates from the reaction between atmospheric oxygen and the steel electrode at the electrode surface during ESR is the major source of oxygen [9]. As for the ESR operation that is performed in open air atmosphere, the oxygen in air atmosphere is

3.2 Oxygen Transfer During ESR Process

23

Fig. 3.1 Schematic illustration of oxygen transfer in the ESR process. Reproduced from Ref. [5]

responsible for the increase in the oxygen content of liquid metal in the abovementioned third way. The ferrous oxide formed on the electrode surface and the original oxide scale on the electrode surface enter into the slag pool during ESR, resulting in an increase in the oxygen potential of the molten slag. The oxygen potential of the ESR-type slag is typically determined by the FeO activity in the slag (in some cases, MnO is present in the slag, but in a quite small fraction). FeO in molten slag pool introduces oxygen into the liquid metal pool as expressed in Eq. 3.1. (FeO) = [O] + [Fe] ΔG  = 139,000 − 57.1T [22] (J/mol)

(3.1)

In this article, the square bracket [ ] indicates the component in liquid metal, and the bracket ( ) indicates the component in molten slag unless specially stated. In addition, FeO could also introduce oxygen into the liquid metal film during the liquid metal film formation and collection into metal droplets at the electrode tip. Ferrous oxide in slag thus plays an important role in introducing oxygen to the metal phase. In the case where inert gas atmosphere is employed in the ESR process, it is helpful to (i) prevent (at least substantially reduce) the formation of FeO, resulting from the chemical reaction between atmospheric oxygen and the steel electrode taking place on the electrode surface, which can indirectly lead to oxygen pickup in liquid steel, (ii) meanwhile, prevent the transfer of oxygen from the air atmosphere into liquid steel through the molten slag. Furthermore, the soluble oxygen [O] as the product of desulfurization reactions in the ESR process could also result in oxygen pickup in liquid metal as expressed in Eq. 3.2. ) ( ) ( [S] + O2− = S2− + [O]

(3.2)

24

3 Deoxidation of ESR and Its Correlation with Oxide Inclusions

3.3 Thermodynamic Considerations on Deoxidation of ESR In conventional steelmaking, the deoxidation of liquid steel is generally realized through adding deoxidizing agents into liquid steel to combine the dissolved oxygen as oxide inclusions first, and then removing oxide inclusions for obtaining lower oxygen content of liquid steel. Different kinds of deoxidizing agents, such as Al, Si– Mn, Si–Fe, Al–Ti and SiAlBa, are widely used for liquid steel deoxidation according to the requirements of individual steel grade. As for the deoxidation of ESR, Al and Ca-Si are widely used for lowering the oxygen potential of molten slag and the oxygen content of liquid metal. In the ESR refining practice, the presence of FeO in molten slag could hardly be prevented. The measured FeO content in the slag collected both during ESR and after ESR is lower than 1 mass% in the production practice [5, 23–25]. As a reducible oxide in molten slag, FeO in molten slag could introduce oxygen into liquid steel according to the reaction (FeO) = [O] + [Fe] at a low oxygen level of liquid steel. On the contrary, the oxygen in liquid steel is transported into the molten slag according to the reaction [O] + [Fe] = (FeO) in the case of a low oxygen potential of molten slag and high oxygen level of liquid steel. Lowering the oxygen potential of molten slag could suppress the oxygen supply into liquid steel by the molten slag during ESR. In the case of electroslag remelting under protective argon gas atmosphere, the dissolved oxygen level (oxygen pickup) of liquid steel is determined by the oxygen potential of molten slag (reducible oxide FeO content) [5]. The measured FeO content in the slag collected both during ESR and after ESR is smaller than 1 mass% in the practice. The oxygen content of liquid steel in equilibrium with FeO in the slag could be estimated from reaction (3.1) and its equilibrium constant as follows: K =

aFeO aFeO = aFe · aO aFe · f O [%O]

(3.3)

The oxygen content equilibrated with dissolved aluminum in liquid steel and Al2 O3 in the slag was estimated from the reaction (3.4) 2[Al] + 3[O] = (Al2 O3 ) ΔG  = −1,205,115 + 386.714T [26] (J/mol) (3.4) K =

aAl2 O3 aAl2 O3 = 2 3 aAl · aO ( f Al [%Al])2 · ( f O [%O])3

(3.5)

where aFeO and aAl2 O3 are the activities of FeO and Al2 O3 in slag. aFe is the activity of iron in liquid steel, which is unit relative to 1% solution standard state. f Al and f O are the activity coefficients of dissolved aluminum and oxygen in liquid steel, respectively, and can be expressed by the following formulas [27] lg f i =

) ( j j ei [% j] + ri [% j]2

(3.6)

3.3 Thermodynamic Considerations on Deoxidation of ESR j

j

25

where ei and ri are the first-order and second-order interaction parameters, respectively. The first-order interaction parameters used in the current calculation are summarized in Ref. [15]. The available second-order interaction parameters are as C Si Al follows: rAl = −0.004 [28], rAl = −0.0006 [28], rAl = −0.0011 + 0.17/T [28], Ni Cr Al rO = 0.00011 [29], rO = 0.00058 [30], rO = 0.0033 − 25/T [31]. Taking electroslag remelting of H13 tool steel as an example for the thermodynamic evaluation, in the thermodynamic calculation, the activities of FeO relative to pure liquid standard state and Al2 O3 relative to pure solid standard state in the CaF2 -containing slag at different temperatures were estimated with FactSage 6.4 using CON2 database (ThermFact/CRCT, Montréal, Canada). The pre-melted slag with the composition of 30.4 mass% CaF2 , 28.7 mass% CaO, 30.7 mass% Al2 O3 , 2.5 mass% MgO, 6.7 mass% SiO2 , and others (FeO < 0.4 mass%, S < 0.02 mass%, TiO2 < 0.03 mass%) is used for the thermodynamic analysis. The chemical composition of H13 tool steel is listed as: 0.41C–1.06Si–0.36Mn–5.17Cr– 0.96V–1.27Mo–0.0017Ca–0.0004Mg–0.012Al–0.0008O–0.0026S (in mass%). The liquidus temperature of the steel is calculated with Thermo-Calc software (TCFE7 database) to be 1748 K. The temperature of the liquid metal films in the ESR is taken as the liquidus temperature of the steel because their superheat could hardly exceed 20–30 °C [32–34]. The oxygen content in liquid steel estimated from the reaction [Fe] + [O] = (FeO) is plotted against the FeO content in slag, as shown in Fig. 3.2. The estimated oxygen content in liquid steel according to the reaction 3[O] + 2[Al] = (Al2 O3 ) is also present in Fig. 3.2 for comparison. It can be learned from Fig. 3.2 that the measured total oxygen, which includes both the dissolved oxygen (free oxygen) and the oxygen bonded as oxide inclusions, in the remelted ingot is much lower than the oxygen level determined by (FeO)–[O] equilibrium. The dissolved oxygen content determined by [Al]–[O] equilibrium is far lower than that by (FeO)–[O] equilibrium, as shown in Fig. 3.2. The dissolved oxygen in liquid steel is few parts per million, which is much lower than the total oxygen level (8 ppm in the steel electrode and 14 ppm in the remelted ingots) [15]. The dissolved oxygen level of liquid steel is determined by [Al]–[O] equilibrium. Although the reaction equilibrium between liquid steel and slag with respect to oxygen could hardly be reached in the ESR refining practice, the FeO content in the molten slag is indeed far higher than the equilibrium value. FeO in the slag consequently introduces the oxygen into liquid steel during the ESR process, resulting in the dissolved oxygen pickup in liquid steel. A similar finding is obtained for electroslag remelting of S136 steel (0.39%C–0.26%Si–0.43%Mn–13.37%Cr– 0.34%V–0.21%Mo–0.079%Al–0.018%S) using 60 mass% CaF2 –20 mass% CaO– 20 mass% Al2 O3 slag, as shown in Fig. 3.3. A detailed description of the calculation has been presented in Ref. [35]. Even a small amount of FeO can lead to high oxygen potential of the slag for introducing oxygen into liquid steel. Thoroughly removing the oxide scale on the steel electrode surface prior to ESR and lowering the oxygen potential of molten slag during protective inert atmosphere ESR are suggested to prevent (at least suppress) the oxygen pickup in liquid steel. It is a common operation for deoxidation of ESR through decreasing the oxygen potential of molten slag by adding deoxidizing agents to molten slag pool. It has

3 Deoxidation of ESR and Its Correlation with Oxide Inclusions

Fig. 3.2 Relationship between the dissolved oxygen concentration of liquid steel equilibrated with FeO in the slag and the FeO content as well as the oxygen level determined by [Al]–[O] equilibrium and the measured total oxygen content (T.O) in the remelted ingot [15]

0.012

0.009 [mass% O]

26

0.006

0.003

0.000 0.0

0.2

0.4 0.6 0.8 (mass% FeO)

1.0

1.2

0.20

0.24

0.0010 0.0008

[mass% O]

Fig. 3.3 Comparison of the oxygen content estimated from the reactions (FeO)–[O] equilibrium and [Al]–[O] equilibrium between S136 steel and slag at the liquidus temperature of the steel 1474 °C. Reprinted with permission from Ref. [35]

(FeO)-[O] equilibrium at 1873 K (FeO)-[O] equilibrium at 1748 K [Al]-[O] equilibrium at 1873 K [Al]-[O] equilibrium at 1748 K Measured T.O

(FeO)-[O] [Al]-[O]

0.0006 0.0004 0.0002 0.0000 0.00

0.04

0.08 0.12 0.16 (mass% FeO)

been verified from the plant trials when using the purified pre-melted slag (finally the total oxygen content < 10 ppm in the remelted ingots), in which the FeO content is less than 0.1 mass%, combining with deoxidizing agent aluminum particles addition to slag pool during the ongoing protective argon gas atmosphere ESR process [15]. In the case of deoxidizing agent addition during ESR process, due to the large temperature difference and heat transfer between molten slag and deoxidizing agents, solid deoxidizing agents melt immediately after their addition. Taking Al-containing deoxidizing agent for an example, aluminum dissolved in molten slag pool will lower the oxygen potential of molten slag to restrain the transport of oxygen by FeO in molten slag into liquid steel as expressed in Eq. 3.8. 2[Al]in slag + 3(FeO) = (Al2 O3 ) + 3[Fe]

(3.7)

3.4 Deoxidation Kinetics of ESR

27

Fig. 3.4 Schematic illustration of oxygen transfer and deoxidation in the ESR process. The solid black point represents the dissolved oxygen in the liquid metal film. The open arrow indicates the oxygen transfer path, except for the representation of the path from [Al]in slag to [Al] in liquid metal pool. The solid arrow indicates the direction of deoxidation [36]

According to reaction (3.7), the concentration of FeO in molten slag is reduced substantially, which destroys the oxygen transport by FeO into liquid steel. Meanwhile, the aluminum in molten slag brought by deoxidizing agent addition can also directly react with the oxygen in liquid metal film and liquid metal pool, as reaction 2[Al] + 3[O] = (Al2 O3 ), to reduce the oxygen content of liquid steel. In the case of a very low dissolved oxygen content (see Fig. 3.2 for example, the dissolved oxygen content in liquid steel is calculated to be only a few ppm based on (Al2 O3 )-[Al]–[O] equilibrium), deoxidizing agent Al could hardly lower the dissolved oxygen in liquid steel directly. A schematic illustration of the deoxidation mechanism by deoxidizing agent Al in the ESR process is present in Fig. 3.4. The removal of oxide inclusions, which are the products of liquid steel deoxidation of ESR, is affected by various factors in the ESR process. If these oxide inclusions are not removed during the ESR, the intention of deoxidizing agent addition could not be realized. The aim of deceasing the oxygen content of the steel is to reduce the oxide inclusion amount. Therefore, the deoxidation operations of ESR should also consider their effects on the oxide inclusions for low oxygen steel production, such as inclusion chemistry and amount.

3.4 Deoxidation Kinetics of ESR The kinetics of chemical reactions taking place in the ESR process has been studied in literatures. However, the works on this topic are still very limited. The kinetics studies that have been conducted up to date include the thermodynamic equilibrium model proposed by Hawkins [37], the single-stage reactor model developed by Etienne and Mitchell [38] for predicting the changes in the concentrations of alloying elements

28

3 Deoxidation of ESR and Its Correlation with Oxide Inclusions

experiencing metal-slag reactions, the mass transfer model of slag-metal reactions developed by Fraser and Mitchell [33, 39], the model for chemical reactions and mass transfer between slag and liquid steel developed by Wei and Mitchell [23, 24, 32, 40], the mass transfer model of desulfurization [41], and the deoxidation model developed by the present author [42]. The prediction of the mass transfer model developed by Fraser and Mitchell [33, 39] for calculating the oxidation loss of Mn in liquid steel by FeO in the slag agrees well with the experimental results in both the steady and the unsteady states. The model developed by Wei and Mitchell [32] has been successfully applied to predict the changes in the concentrations of Mn, Si and Al in liquid steel, as well as the concentrations of MnO, SiO2 , Al2 O3 and FeO in slag with the remelting time during ESR of low-alloy steel SAE 1020 [32, 43], and the concentrations of these components during ESR of Cr–Mo–V turbine rotor steel on an industrial unit [23]. The changes in the concentrations of Mn, Si, Al, Ti and Cr, as well as the concentrations of MnO, SiO2 , Al2 O3 , Cr2 O3 and FeO in slag during ESR of stainless steel 1Cr18Ni9 are also precisely predicted with this model [24, 40]. In this chapter, the deoxidation kinetics of ESR, which is taken from the results of the present author’s research group, is presented. According to the penetration theory, a kinetic model for oxygen transfer between molten slag and liquid steel during the ESR process is established in this chapter. The developed kinetic model is applied to predict the oxygen transfer rate between liquid steel and molten slag, and the role of processing parameters of ESR on the diffusion flux of oxygen transfer between molten slag and liquid steel.

3.4.1 Development of the Kinetic Model As consumable electrode melts during the ESR process, liquid metal films form on the downside of the consumable electrode and subsequently collect into liquid metal droplets. These liquid metal droplets periodically grow up to the critical size at the electrode tip during the ESR process, and then detach from the electrode tip and pass through the slag bath due to the difference in density. The chemical reactions between liquid steel and molten slag take place during the generation of liquid metal droplets through liquid metal films. A kinetic model for oxygen transfer between molten slag and liquid steel during the ESR process is developed based on the penetration theory. The main points of the penetration theory are as follows [44]: (1) The mass transfer among different phases is realized by fluid microelement (see Fig. 3.5). (2) Fluid is composed of numerous fluid microelements. Assuming the concentration of a component in each fluid microelement in fluid II is Cb . A fluid microelement in fluid II is taken to the interface between fluid II and other fluid (fluid I) by natural convection or turbulent flow. When the concentration of a

3.4 Deoxidation Kinetics of ESR

29

Fig. 3.5 Schematic illustration of solute penetration theory

component in fluid I is higher than the equilibrium concentration of the component in fluid II, this component in fluid I will transfer into the fluid microelement in fluid II. Consequently, at the interface, the concentration of the component in fluid II Cs is higher than its concentration in fluid II bulk Cb . (3) The residence time of fluid microelements at the interface is very short (0.01– 0.1 s). The residence time is termed as the lifetime of the fluid microelement te . Fluid microelements flows into fluid II again after residence time te , and the concentration of the component in fluid microelement reach Cb + ΔC. (4) This mass transfer process can be regarded as unsteady one-dimensional semiinfinite diffusion because the lifetime of the fluid microelement is very short at the interface and the penetration depth of the component into the fluid microelement is smaller than the thickness of the fluid microelement. The chemical reactions during the ESR are involved among molten slag, liquid steel and gas phases. The chemical reactions could take place at four reaction sites during the ESR process: ➀ electrode tip/slag interface, ➁ metal droplet/slag interface, ➂ liquid metal pool/slag interface, ➃ atmosphere/slag interface. Among these reaction sites, reaction site ➀ is the phase boundary with the largest refining potential for the chemical reactions, at which the reaction takes place predominantly during the liquid metal film formation at the electrode tip [45–47]. It is attributed to more preferable kinetics conditions at this site (namely, long chemical reaction time, and high ratio of surface to volume between the liquid metal film at the electrode tip and slag [47, 48]). Reaction site ➁ is of a little importance acting as chemical reaction site due to the very short residence time of metal droplet passing through the slag pool (virtually 0.01–0.1 s [47, 49, 50]). If protective Ar gas atmosphere is employed in the ESR process, the direct transfer of oxygen from the atmosphere into liquid steel

30

3 Deoxidation of ESR and Its Correlation with Oxide Inclusions

through the slag will be prevented. The oxygen transfer takes place only between molten slag and liquid metal in protective atmosphere ESR process. The oxygen transfer between molten slag and liquid metal at the electrode tip could be regarded as an independent reaction system during the ESR process. The following assumptions and simplifications have been incorporated in the current model: (1) The oxygen transfer between molten slag and liquid steel is accomplished during the formation of liquid metal films at the electrode tip. The slag-metal interface as the metal droplets pass through the molten slag pool and the liquid metal pool and molten slag pool can be ignored as the sites of oxygen transfer. (2) The mass transfer of the component during the formation of liquid metal films at the consumable electrode tip meets the solute penetration theory. (3) Chemical reactions occur at the slag-metal interface, and the diffusion of the components takes place in the slag bulk and metal bulk. The influence of fluid flow on mass transfer process is not considered separately, which will be involved in mass transfer coefficient. (4) Thermodynamic equilibrium of the chemical reactions at the interface is reached in the transient state. The initial and boundary conditions of semi-infinite diffusion are shown as follows: t = 0, x ≥ 0, c = cb 0 < t ≤ te , x = 0, c = cs , x = ∞, c = cb For unsteady one-dimensional semi-infinite diffusion, the solution to Fick’s second law is expressed as follows: ) ( c − cb x = 1 − erf √ cs − cb 2 Dt ) ( x c = cs − (cs − cb )erf √ 2 Dt

(3.8) (3.9)

where D is the diffusion coefficient. When x = 0 (at the interface), the diffusion flux of the component can be expressed by the following formulas: ( J = −D

∂c ∂x

[

)

)] ( ∂ x erf √ ∂x 2 Dt x=0 / D 1 = D(cs − cb ) · √ = (cs − cb ) π t π Dt

= D(cs − cb ) x=0

(3.10)

The average diffusion flux in the lifetime of the microelement te is represented as:

3.4 Deoxidation Kinetics of ESR

1 J= te

{te / 0

31

/ D D (cs − cb ) (cs − cb )dt = 2 πt π te

(3.11)

According to equation J = kd (cs − cb ) and Eq. 3.11, the mass transfer coefficient kd can be calculated using the following equation: / kd = 2

D π te

(3.12)

The oxygen transfer between molten slag and liquid steel can be expressed as reaction (3.1). The standard Gibbs free energy change of the reaction (3.1) is represented as follows: ΔG  = −RT ln

a[Fe] · a[O] aFeO

(3.13)

The activity of iron in liquid steel aFe relative to 1% solution standard state is taken as unit. Equation 3.13 can be rewritten as ΔG  = −RT ln

a[O] a[Fe] · a[O] = −RT ln aFeO aFeO

(3.14)

The introduction of oxygen by FeO from molten slag into liquid steel can be divided into the following steps: (1) Transport of FeO from the slag bulk to the slag-steel interface as (FeO)*. (FeO) = (FeO)∗

(3.15)

(2) Chemical reaction at the interface to form oxygen [O]*. (FeO)∗ = [O]∗ + [Fe]∗

(3.16)

(3) Transport of [O]* from the interface to the liquid metal bulk. [O]∗ = [O]

(3.17)

(4) Transport of [Fe]* from the interface to the liquid metal bulk. [Fe]∗ = [Fe]

(3.18)

When the oxygen potential of slag is higher than that of liquid steel, FeO in molten slag introduces oxygen to liquid steel. On the contrary, the oxygen in liquid steel is transferred to the molten slag. According to the penetration theory, the oxygen potential of fluid microelement in liquid metal film at the electrode tip varies as the

32

3 Deoxidation of ESR and Its Correlation with Oxide Inclusions

electrode

Fig. 3.6 Schematic illustration of microelement flowing of fluid at electrode tip

cb

cb

t=0 t=te

slag pool

fluid microelement flows down along the electrode tip. A schematic illustration of microelement flowing of fluid at electrode tip is shown in Fig. 3.6. The interval time between starting of the fluid microelement formation and detachment of the fluid microelement from electrode tip is defined as residence time te . Considering that the rate of chemical reactions at the slag–metal interface is sufficiently large at elevated temperatures, mass transfer is the rate-controlling step of oxygen transfer in ESR process. The diffusion flux of oxygen transfer from molten slag to liquid steel introduced by FeO is expressed as follows 1 J= te

{te / 0

/ D (cs − cb )dt = 2 πt

) DFeO ( cs,FeO − cb,FeO πte

(3.19)

where J is diffusion flux (mol/(m2 s)), cs, FeO is the concentration of FeO at slag-liquid steel interface (mol/m3 ), cb, FeO is the concentration of FeO in slag bulk (mol/m3 ), DFeO is diffusion coefficient of FeO (m2 /s), te is residence time (s). The residence time te on the electrode film can be calculated using the following equation [33] ( )1/3 ( ) ) ( 2π cos θ 2/3 μm Re 5/3 te = 3.35 3Wm gρm sinθ cos θ

(3.20)

where Wm is the melting rate of the electrode in volume (m3 /s), θ is the cone angle on the electrode tip (o ), ρm is the density of liquid steel (kg/m3 ), g is the acceleration of gravity (m/s2 ), Re is the radius of the electrode (m), μm is the viscosity of liquid steel (Pa s).

3.4 Deoxidation Kinetics of ESR

33

3.4.2 Kinetic Model Parameters The developed kinetic model is applied to electroslag remelting of S136 tool steel. The remelting trials T1 and T2 were conducted in argon gas atmosphere throughout the ESR process, whereas the trials T3 and T4 were performed in open air atmosphere. For three heats (Exp. T1, T3, and T4), the deoxidizing agent was added continually into the slag pool during the ESR process for slag deoxidation. The addition rate of deoxidizing agent is about 20 kg/t in the trials T1 and T4, as well as about 40 kg/t in the trial T3. A description of the experimental details of ESR has been presented Ref. [5]. The diffusion coefficient of FeO in 60 mass% CaF2 –20 mass% CaO–20 mass% Al2 O3 molten slag DFeO is taken as 1.5 × 10–8 m2 /s [32]. The model parameters are listed as follows [42]: melting rate of protective atmosphere ESR is 1.59 × 10–6 m3 / s, the measured cone angle θ on the electrode tip is 40°, the viscosity of liquid steel μm is 0.005 Pa·s, the density of liquid steel is 7.0 × 103 kg/m3 , the density of the molten slag ρs is 2.81 × 103 kg/m3 , the radius of the steel electrode Re is 0.04 m, the residence time of fluid microelement at the electrode tip te is calculated according to Eq. 3.20 to be 1.2 s. The oxygen is introduced into liquid steel by FeO in molten slag according to the reaction (3.1). Equation 3.14 is rewritten as follows: aFeO =

exp

a[O] ( −ΔG  ) = RT

f O [% O] ( ) exp −ΔG RT

(3.21)

where f O is the activity coefficient of soluble oxygen in liquid steel, and can be calculated with Eq. 3.6, together with the first-order interaction parameters summarized in Ref. [14]. The temperature of the liquid metal films at the steel electrode tip during ESR is close to the liquidus temperature of the steel [32]. The temperature of the liquid metal films at the electrode tip in the ESR was taken as the liquidus temperature of the steel (estimated as 1474 °C with FactSage software) because their superheat could hardly exceed 20–30 °C [32–34]. Combining Eq. 3.1 with Eq. 3.21, the temperature of the liquid metal films and the soluble oxygen content of liquid metal, the activity of FeO in molten slag can be calculated. It was reported by different researchers that the temperatures of molten slag pool in the ESR process are in the range of 1700–1850 °C [51–53]. Medina et al. [51] reported that the temperature of molten slag pool is dependent on the melting rate of the ESR. According to the comparison of the current experimental conditions with that in the experimental work by Medina et al. [51], the temperature of molten slag pool in the current study could be about 1800 °C. According to Eq. 3.21 and the calculated activity coefficient of FeO in molten slag γFeO , the concentration of FeO at slag-metal interface can be calculated as follows: xFeO =

aFeO γFeO

(3.22)

34

3 Deoxidation of ESR and Its Correlation with Oxide Inclusions

The activity coefficient of FeO in 60 mass% CaF2 –20 mass% CaO–20 mass% Al2 O3 molten slag with a small amount of FeO can be calculated by Eq. 3.23 [32]. ) ( ) 11,800 ( xCaO + xMgO · xSiO2 + 0.25xAlO1.5 T 4916 3562 123 xMnO xSiO2 + xSiO2 xAlO1.5 + xMnO xAlO1.5 + T T T

lg γFeO =

(3.23)

The volume fraction of FeO in molten slag can be calculated by means of the relationship between volume fraction concentration and mole fraction concentration, as expressed in Eq. 3.24. cFeO =

xFeO



(3.24)

x i Mi ρs

3.4.3 Application of the Developed Kinetic Model to ESR Practice The oxygen contents of the remelted ingots ESR-S1, ESR-S2, ESR-S3 and ESR-S4 produced in the ESR trials ST1, ST2, ST3 and ST4 are 0.0012 mass%, 0.0019 mass%, 0.0018 mass% and 0.0033 mass%, respectively. In combination with the experimental results of electroslag remelting of S136 steel presented in Ref. [5], the activities of FeO at the slag-metal interface at the electrode tip in these four ESR trials are calculated by Eq. 3.21 to be 0.0032, 0.0055, 0.0052 and 0.0102. The mole fraction concentrations of FeO at the slag-metal interface are calculated to be 0.0028, 0.0048, 0.0046 and 0.0077, and then the volume fraction concentrations of FeO cs, FeO are calculated to be 103 mol/m3 , 176 mol/m3 , 168 mol/m3 and 279 mol/m3 , respectively. According to the chemical compositions of the slag in the ESR trials (as shown in Table 3.1), the concentrations of FeO in molten slag bulk cb, FeO were calculated, and the results are summarized in Table 3.1. Figure 3.7 shows the relationship between the diffusion flux of oxygen transfer between molten slag and liquid steel phases, and the FeO concentration in the molten Table 3.1 Chemical composition of the slag and the concentration of FeO in slag bulk cb, FeO Trial No.

cb, FeO (mol/m3 )

Chemical composition of the slag (mass%) CaF2

CaO

Al2 O3

FeO

T1

56.79

19.40

23.05

0.04

16

T2

57.38

18.87

23.11

0.14

55

T3

60.04

15.94

21.62

0.11

44

T4

40.05

22.35

34.66

0.22

88

3.4 Deoxidation Kinetics of ESR 0.05 Trial Trial Trial Trial

T1 T2 T3 T4

0.45

-0.05

0.70

0.26

0.00

0.42

J (mol/(m2·s))

Fig. 3.7 Relationship between the diffusion flux of oxygen transfer between molten slag and liquid steel phases and the FeO concentration in the slag [35]

35

-0.10 0.0

0.5

1.0 1.5 (mass% FeO)

2.0

slag. It is learned from Fig. 3.7 that there is a critical value of FeO content in the slag. In the case where the FeO content of the slag is lower than this critical value, the oxygen transfer is from liquid steel to molten slag. Otherwise, the oxygen transfer is from molten slag to liquid steel. The critical value of the FeO concentration in the slag calculated using the developed kinetic model is 0.26 mass%, 0.45 mass%, 0.42 mass% and 0.70 mass%, respectively, as shown in Fig. 3.7. The diffusion flux of oxygen transfer from liquid steel to molten slag decreases with increasing the concentration of FeO in the slag up to the critical value as shown in Fig. 3.7, whereas the diffusion flux of oxygen transfer from molten slag to liquid steel increases with increasing the FeO concentration in the slag. It indicates that ferrous oxide plays an important role in transporting oxygen between molten slag and liquid metal phases. It is crucial for reducing the oxygen content of liquid steel to an extra-low level through lowering the FeO concentration in the slag before and during ESR process. In the case of a constant FeO concentration in the molten slag, the relationship between the diffusion flux of oxygen transfer between slag and liquid steel, and the original dissolved oxygen content in liquid steel at the electrode tip is present in Fig. 3.8. The diffusion flux of the oxygen from liquid steel to molten slag constantly increases with the increase in the dissolved oxygen content in liquid steel. Figure 3.9 shows the relationship between the diffusion flux of oxygen transfer between molten slag and liquid steel, and the radius of steel electrode. The FeO concentration in molten slag is constant for each trial. The melting rate of each PESR trial is 1.59 × 10−6 m3 /s. It can be seen that the diffusion flux of the oxygen from liquid steel to molten slag constantly decreases with the increase in the radius of steel electrode. It is attributed to the increase in the residence time of liquid metal film at the electrode tip with the increase in the radius of steel electrode. The melting rate of ESR is an important parameter. It determines not only the production efficiency of ESR, but also the as-cast ingot quality. The relationship

0.12 Trial Trial Trial Trial

0.09

2

Fig. 3.8 Relationship between the diffusion flux of oxygen transfer between molten slag and liquid steel, and the dissolved original oxygen content in liquid steel [35]

3 Deoxidation of ESR and Its Correlation with Oxide Inclusions

J (mol/(m ·s))

36

T1 T2 T3 T4

0.06 0.03

0.00 -0.03

0.002

0.004 0.006 [mass% O]

0.008

0.010

0.032 Trial Trial Trial Trial

0.024

T1 T2 T3 T4

2

J (mol/(m ·s))

Fig. 3.9 Relationship between the diffusion flux of oxygen transfer between molten slag and liquid steel, and the radius of steel electrode [35]

0.000

0.016

0.008

0.000 0.00

0.05

0.10

0.15 R (m)

0.20

0.25

0.30

between the diffusion flux of oxygen transfer between molten slag and liquid steel, and the melting rate of ESR is shown in Fig. 3.10. The same composition of the slag is used in the calculation for each ESR trial, and the radius of the steel electrode is 0.04 m. As shown in Fig. 3.10, the diffusion flux of the oxygen transfer from liquid steel to molten slag constantly increases with the increase in the melting rate of ESR. It is attributed to the decrease in the residence time of liquid metal film at the electrode tip and the increase in the mass transfer coefficient of FeO with increasing the melting rate. The diffusion flux of the oxygen from liquid steel to molten slag thus increases constantly.

3.5 Evaluation of the Dependence of Oxygen on the Processing Parameters … 0.036 0.030

2

J (mol/(m ·s))

Fig. 3.10 Relationship between the diffusion flux of oxygen transfer between molten slag and liquid steel, and the melting rate of ESR [35]

37

Trial T1 Trial T2 Trial T3 Trial T4

0.024 0.018

0.012 0.006 0.0

1.0x10

-6

2.0x10

-6

3.0x10

-6

4.0x10

-6

5.0x10

-6

3

Wm (m /s)

3.5 Evaluation of the Dependence of Oxygen on the Processing Parameters of ESR 3.5.1 Initial Oxygen Content of Steel Electrode It has been verified by many laboratory-scale experiments and plant trials that the oxygen content of the steel could be greatly reduced after ESR refining. For electroslag remelting of the steel with high oxygen contents, ESR lowers the oxygen content invariably even if the protective atmosphere and/or slag deoxidation are not employed in the ESR operation. However, there are controversial findings on whether the oxygen content in the remelted ingot is dependent on the original oxygen content of the steel electrode or not [11, 13, 15, 17, 48, 54–60]. Plöckinger [48] claimed that the oxygen content level was dependent only on the reactions between slag and liquid steel, and not concerned with the initial oxygen content of steel electrode. In recent decades, the oxygen content in the steel used as the electrode for ESR could be reduced to a low level because of the improvement of steelmaking technologies. Several previous studies demonstrated that the oxygen content of the steel increased after ESR of the steel electrode with a low oxygen content, such as the work by Paar et al. [17], Medina and Cores [54], Wang et al. [55], Chang et al. [56], and Li et al. [57]. Moreover, Zhou et al. [58] reported that the oxygen content of the bearing steel was in the range of 15–30 ppm after ESR in open air atmosphere irrespective of the oxygen level (ranging from 5 to 40 ppm) of the consumable steel electrode. More and more efforts have been devoted to improve the cleanliness of the steel electrode. It is expected that a low initial oxygen content of the steel electrode gives a lower oxygen level in the remelted ingot. In the face of such situation, oxygen pickup is the key focus, and should also be suppressed during the ESR. The degree of oxygen pickup during ESR is dependent on the operation parameters of ESR, such as slag chemistry, steel composition and remelting atmosphere, especially the oxygen

38

3 Deoxidation of ESR and Its Correlation with Oxide Inclusions

potential of the slag. The plant trials show that the oxygen content of martensitic stainless steel 8Cr17MoV is decreased from 58 ppm in the consumable steel electrode to 40 ppm after protective atmosphere ESR [13], and the oxygen content of Si-Mnkilled steel is decreased from 74 ppm in the steel electrode to 34–38 ppm after protective atmosphere ESR [61]. On the contrary, the oxygen content of H13 tool steel increases from 18 ppm in the consumable steel electrode to 21–34 ppm in remelted ingots [11], and from 8 ppm in the consumable steel electrode to 14–17 ppm after protective atmosphere ESR [15]. However, this is not always the case. A low initial oxygen content of the steel electrode (13 ppm) was reduced to 8–12 ppm after open air atmosphere ESR in the case where the FeO content in the slag was at a low level (0.1–0.2 mass%) [59]. It is really a tough case to lower the oxygen content to a lower level after ESR of low oxygen steel. In order to further lower the oxygen during ESR of the steel with low oxygen content, the strategic point is to prevent the oxygen pickup during the ESR process, in which the oxygen potential of slag is the key source.

3.5.2 Oxide Inclusions in Steel Electrode Deoxidation of liquid steel is directly related to the removal of oxide inclusions during ESR. In general, a majority of the original oxide inclusions in the steel electrode are removed during the ESR process, especially a preferential elimination of large inclusions [3, 59, 60]. Apart from some types of original oxide inclusions in the steel electrode (for example, MnO–SiO2 –Al2 O3 inclusions [61]), other types of the original oxide inclusions could partially survive from the steel electrode to ESR ingot [62, 63]. The factors, which affect the oxide inclusion evolution during ESR, greatly influence the oxygen content of ESR ingot. Different deoxidation schemes (deoxidizing agent type, addition amount, and addition time, etc.) are generally adopted for liquid steel deoxidation according to the requirements of individual steel grade when producing the steel electrode for ESR, which consequently generate different types of oxide inclusions in the steel electrode, such as Al2 O3 , MgO · Al2 O3 and various calcium aluminate inclusions. The removal degree of these oxide inclusions during ESR largely determines the total oxygen content of ESR ingots. Different types of original oxide inclusions basically experience various evolution trajectories during ESR, as demonstrated in the previous studies regarding Al2 O3 , MgO · Al2 O3 , MnO–SiO2 –Al2 O3 and calcium aluminate inclusions [5, 11–13, 15, 16, 61], consequently giving different removal degrees of oxide inclusions during the ESR of the steel. The types, compositions and size of the original oxide inclusions in the steel electrode exert a significant influence on the refining consequence of ESR in terms of the oxygen level of the remelted ingots [64]. The cleanliness and oxide inclusions in the bearing steel ZGCr15 after electroslag remelting of the steel electrode deoxidized by different amounts of Al, Si–Ca, Si–Fe, Si–Mn–Ca or Al–Mn–Si in induction furnace melting which caused the generation

3.5 Evaluation of the Dependence of Oxygen on the Processing Parameters …

39

of different types of original oxide inclusions were compared [64]. It shows that the deoxidization of the steel electrode using Si–Ca gives the most effective refining in terms of the oxygen level and oxide inclusions in remelted ingots, whereas the deoxidization of the steel electrode using Al leads to the least effective refining, which is attributed to easier removal of low-melting-point silicate inclusions during the ESR process. The deoxidation of 316LC stainless steel during open air atmosphere ESR was investigated by Ahmadi et al. [65], through deoxidizing the steel electrodes by aluminum or by aluminum and calcium-silicon wires. The results show that it is more effective for removal of calcium aluminate inclusions than alumina inclusions during the ESR, which is because the removal of alumina in the slag pool by chemical reactions is almost impossible, whereas calcium aluminate inclusions can float up easily in the liquid metal pool, consequently leading to a lower oxygen content of remelted ingot. Unfortunately, the comparison by Ahmadi et al. [65], was not conducted at the same oxygen level of the steel electrodes. Meanwhile, the differences in the elimination degrees of calcium aluminate inclusions and alumina inclusions during ESR were attributed to floatation and adherence by Ahmadi et al. [65]. However, the removal of inclusions through their floatation in liquid metal pool is possible, but it contributes in a very small manner with regard to the removal of inclusions in the ESR process [66]. It was reported in some studies that removal of inclusions by floatation in liquid metal pool was unlikely because of their fine size [67]. For Si–Mn deoxidized steel, the oxide inclusions are manganese silicates which are liquid at steelmaking temperatures in many cases [68–71]. The present author investigated the evolution of oxide inclusions in Si–Mn-killed steel during protective atmosphere ESR [61]. The results show that the oxide inclusions in the steel electrode are ternary (7.1–26.4 mass%) MnO-(53.3–82.8 mass%) SiO2 -(8.3–23.1 mass%) Al2 O3 without exception, which are fully removed during the protective argon gas atmosphere ESR in two ways: a portion of these inclusions are dissociated in its individual chemical species into liquid steel as the liquid metal films form on the downside of the electrode tip and collect into liquid metal droplets subsequently during ESR, which consequently causes soluble oxygen pickup in liquid steel, whereas the others are removed by absorbing them into molten slag before liquid metal droplets collect in liquid metal pool during ESR. MgAl2 O4 and Al2 O3 inclusions readily form in the liquid metal pool as a result of the reaction between alloying elements and the dissolved oxygen in liquid steel that dissociates from MnO–SiO2 –Al2 O3 inclusions [61]. Consequently, in this case, the oxygen content of the remelted ingot is dependent on the ratio of MnO–SiO2 –Al2 O3 inclusions dissociation to absorption of MnO–SiO2 –Al2 O3 inclusions into molten slag during the ESR process. It is different from the removal of Al2 O3 and MgAl2 O4 inclusions, which are removed through absorbing them into molten slag [67, 72]. The deoxidation degree of liquid steel virtually is different in these cases. The change in the tantalum content during vacuum induction melting (VIM) and subsequent ESR of martensitic steel CPJ7 was investigated by Detrois et al. [73]. Their results demonstrate that the tantalum content of CPJ7 steel decreases by 25% during ESR, which is attributed to the formation of Ta2 O5 inclusions during VIM and

40

3 Deoxidation of ESR and Its Correlation with Oxide Inclusions

subsequent 95% reduction in the number density of Ta2 O5 inclusions (be transferred to the slag) during ESR. A significant decrease in the oxygen content is observed for all ESR trials with an oxygen concentration of around 28 ppm in the ESR ingots. In their work, ESR of the CPJ7 steel electrodes results in, on average, 49% decrease in the oxygen content, which correlates to the removal of Ta2 O5 inclusions in the steel electrode during ESR.

3.5.3 Remelting Atmosphere Electroslag remelting is accompanied with chemical reactions among gas-slag-metalinclusion phases. The chemical reactions at slag-metal interface are closely related to the oxygen in air atmosphere. The oxygen transfer from air atmosphere to slag-metal phases is schematically described in Fig. 3.1. In the case of the remelting leaving the slag surface and steel electrode open to the air atmosphere, atmospheric oxygen reacts with steel electrode at the electrode-atmosphere surface to form iron oxide as the electrode tip is heated at elevated temperatures during ESR. The formed iron oxide introduces oxygen into the liquid metal, as illustrated in Sect. 3.2. Apart from this trajectory of oxygen transfer, oxygen gas can be physically dissolved in molten slag as diatomic molecules [18, 74–76], oxygen molecules thereafter dissolve into liquid steel below the slag pool after permeating through the molten slag [20]. The physical dissolution of oxygen gas and its permeability through the molten slag have been measured by several researchers [18–20, 76, 77]. The oxygen transfer rate through CaO–SiO2 –Al2 O3 melts was measured to be in the range of 3 × 10–19 to 6 × 10–18 mol/(cm s), and 5 × 10–12 to 5 × 10–8 mol/(cm s) through CaF2 – CaO–SiO2 –Al2 O3 melts [19, 77]. The physical solubility of oxygen in molten slag is largely dependent on the slag composition but not on the temperature [18]. The physical permeability of oxygen through CaF2 –Al2 O3 and CaF2 –Al2 O3 –CaO molten slag for ESR was measured using an oxygen concentration cell with ZrO2 solid electrolyte by Wei and Liu [20]. The measured permeability of oxygen is 1 × 10–20 – 6 × 10–19 mol/(cm s) and 1 × 10–21 –5 × 10–18 mol/(cm s) in the temperature range of 1673–1873 K (1400–1600 °C) under pure oxygen gas atmosphere at 0.1 MPa. These results show that the oxygen in air atmosphere could penetrate through molten slag as oxygen molecules into liquid steel, but the oxygen transfer in this trajectory makes a very small contribution. The oxygen pickup in this way is virtually prevented when protective inert gas atmosphere is employed in the ESR process. Protective inert gas atmosphere is increasingly used in ESR practice around the world, in order to prevent the oxygen introduction into liquid metal and the loss of alloying elements in steel and alloy. The relationship between the oxygen content of electroslag remelted 304 stainless steel and the partial pressure of oxygen in gas atmosphere during ESR is shown in Fig. 3.11. The oxygen contents of the remelted ingots increase with increasing the partial pressure of oxygen in gas atmosphere irrespective of the slag systems. The oxygen content of the steel is reduced to a lower level by using the gas atmosphere

3.5 Evaluation of the Dependence of Oxygen on the Processing Parameters …

41

with sufficiently low oxygen content in the ESR process. The oxygen contents of the ingots produced by using different slag compositions are apparently different because of different oxygen gas permeability of these slag systems at the same oxygen partial pressure. The ESR practice conducted in open air atmosphere normally leads to higher FeO content in the slag than the remelting in protective atmosphere [59], which results in a higher oxygen content of liquid steel. For protective atmosphere ESR, inert gas is introduced into the gas protective cap of protective atmosphere ESR equipment to protect the remelting atmosphere against the surrounding air. Protective inert gas atmosphere could suppress the oxygen pickup during the ESR process. In comparison with the remelting in protective argon gas atmosphere, it was reported by Wang et al. [55] that the oxygen content of bearing steel GCr15 increased by 6–12 ppm in the case of the remelting in open air atmosphere. Chang et al. [79] compared the remelting atmosphere of ESR, and found that the oxygen content increased from 20 ppm in the steel electrode to 36 ppm accompanied by a significant loss of Al and Si in the steel after ESR in open air atmosphere, whereas the oxygen content increased from 20 to 24 ppm with an accompanying slight loss of Al and Si in steel in the case of the remelting in protective argon gas atmosphere. Unfortunately, the chemical compositions and FeO content of the slag are not present and evaluated in the studies by Wang et al. [55] and Chang et al. [79], which could be the main source of oxygen pickup in these results. Fig. 3.11 Relationship between the oxygen content of remelted AISI304 stainless steel ingots and the partial pressure of oxygen in the gas atmosphere [78]

42

3 Deoxidation of ESR and Its Correlation with Oxide Inclusions

Targeting a low FeO content in the slag throughout the ESR process is indispensable for producing low oxygen steel and alloy. Although the slag with a low FeO content (FeO + TiO2 ≤ 0.2 mass%) is used in the ESR practices, the oxygen content of the steel still increases by more than three times after ESR, accompanying with an increase in the amount of oxide inclusions [17]. The oxygen pickup in these trials is attributed to the absence of the protective atmosphere for ESR operation, leaving the molten slag and the hot part of the electrode in permanent contact with air [17]. Even though protective argon gas atmosphere is employed in many ESR experiments and production practices, the oxygen content in the gas atmosphere is seldom monitored and controlled. The ESR production practices demonstrate that the oxygen content is decreased by ESR in protective argon gas atmosphere (the oxygen content is lower than 100 ppm in the remelting atmosphere) from 17 ppm in the steel electrode to 8 ppm in the ingot of 600 mm in diameter [80]. The application of protective argon gas atmosphere, in which the oxygen content could be monitored and lowered to a low level, plays an important role in producing ultralow oxygen steel. Vacuum electroslag remelting has been developed to further improve the cleanliness of the steel in terms of the oxygen and inclusion contents [81]. The work by Huang et al. [82] shows that the oxygen content increases from 13 ppm in the steel electrode to 24 ppm and 18 ppm after ESR process in argon atmosphere (0.1 MPa) and vacuum atmosphere (0.01 MPa), respectively. The reduction in the oxygen increase when using vacuum atmosphere is attributed to carbon deoxidization, which only occurs in vacuum atmosphere. The development of vacuum electroslag remelting is still in its infancy, and more studies on the mechanism of deoxidation are still needed. Pressurized electroslag remelting is recognized as a promising technology for producing high nitrogen steel [83]. In the research fields of pressurized electroslag remelting, the precious studies are mainly focused on the microstructure and mechanical properties of high nitrogen steel produced by pressurized electroslag remelting [83, 84]. There is a lack of studies on the physicochemical phenomena of pressurized electroslag remelting, such as thermodynamic conditions and kinetics of chemical reactions. The transfer behavior of nitrogen at different nitrogen partial pressures and electrode immersion depths during pressurized electroslag remelting has been first ascertained by Yu et al. [85]. The role of pressurized atmosphere of ESR on the deoxidation of ESR and oxide inclusions has not been studied in published literatures yet. Thus future work is quite needed on this topic.

3.5.4 Deoxidation Schemes of ESR Iron oxide in molten slag is the source of oxygen potential for oxygen pickup and the main driving force for the oxidation reactions in the ESR process, as illustrated in Sect. 3.2, which has also been manifested by other researchers according to their ESR trials [1, 2, 86]. It is demonstrated that the chemical reactions between FeO in the slag and deoxidizing agent predominate at high oxygen potential (FeO > 0.7 mass% in slag), whereas exchange reactions between deoxidizing agents and

3.5 Evaluation of the Dependence of Oxygen on the Processing Parameters …

43

the slag components alter the inclusion compositions at low oxygen potential (FeO < 0.2 mass% in slag) [2]. However, the presence of iron oxide in the molten slag during the ESR refining of steel is unavoidable. The sources of iron oxide include the component in the initial slag and the oxide scale formed on the electrode surface before and during ESR as a result of the reaction between the steel electrode and the oxygen in air atmosphere. Even a small amount of FeO in slag can result in a high level of oxygen in liquid steel, as illustrated in Sect. 3.3. The laboratory-scale ESR trials with a frequency of 4.5 Hz show that even though the ESR trials of the low oxygen steel electrode are performed in open air atmosphere, the oxygen content is reduced, instead increased, from 13 ppm in the steel electrode to 8–12 ppm in the ingots of 165 mm in diameter because of a low FeO content in the slag (0.1–0.2 mass%) [59]. Continuous or periodic addition of deoxidizing agents into the slag pool is widely used for deoxidation of ESR [1, 2, 4, 5, 9, 51, 54]. The main function of deoxidizing agents addition during the ESR process is to reduce the concentration of FeO in the slag [1, 2, 4, 9, 51, 54]. The added deoxidizing agents, more or less, could hardly be avoided to enter into the liquid steel in the ESR process [5, 9, 87]. Deoxidizing agents are normally in the form of small pellets or grains for slag deoxidation in the ESR process. The high efficiency of this technique has been verified by many ESR experiments, plant trials and production practices. In many ESR practices, inert atmosphere and deoxidizing agent addition are employed simultaneously for reducing the oxygen content of the metal to an extremely low level. Deoxidation schemes of ESR play an important role not only in the contents of oxygen and sulfur, the composition of liquid steel, but also in the oxide inclusions. In addition, the slag composition is also sensitive to deoxidation because of the accumulation of deoxidation products in slag, which causes the change in the chemistry of the slag [2]. This variation of slag compositions could cause the change in liquid steel composition through slag-metal reactions. Aluminum is the most commonly used deoxidizing agent not only in conventional steelmaking process, but also in the ESR refining. Yoshinao et al. [88] lowered the oxygen content of high nitrogen stainless steel to 20–30 ppm after ESR when using aluminum deoxidation for CaF2 –CaO–Al2 O3 slag, and a higher deoxidation efficiency was obtained for the case of the remelting when using the slag with a lower Al2 O3 content and higher CaO content. Shao et al. [89] successfully kept the homogeneity of aluminum and silicon contents in the remelted ingots because of the decreased FeO and SiO2 in the slag through adding aluminum powders into the slag pool during ESR of high speed steel M2Al. The deoxidation ability of Al, Ca–Si and Al–Si during laboratory-scale ESR of SAE4340 steel in protective argon gas atmosphere using 50 mass% CaF2 –20 mass% CaO–30 mass% Al2 O3 slag was compared by Mitchell et al. [4]. The results show that FeO in slag plays a key role in the chemical reactions in the ESR process. In the case of a low oxygen potential (FeO < 0.2 mass%), the aluminum content of the steel increases because of the reaction between calcium and Al2 O3 in slag when calcium-silicon is used for slag deoxidation. At the medium oxygen potential of the slag (0.4–0.6 mass% FeO), the deoxidation effectiveness of aluminum and calcium is comparable.

44

3 Deoxidation of ESR and Its Correlation with Oxide Inclusions

For designing proper schemes of deoxidizing agent addition, different addition rates of Al-based deoxidizing agents were attempted for the deoxidation of protective argon gas atmosphere ESR by the present author [5, 12, 90]. The oxygen content of high-Al steel was lowered from 34 to 10 ppm after protective argon gas atmosphere ESR with Al-based deoxidizing agent addition (30 mass% Al + 15 mass% Al2 O3 + 15 mass% CaF2 + 40 mass% iron powders) of 1.2 kg per ton steel [14, 87]. The mechanical mixtures of 20 mass% Al, 20 mass% Al2 O3 , 20 mass% CaF2 and 40 mass% iron powders were added continually into the slag pool for slag deoxidation of protective argon gas atmosphere ESR of S136 tool steel. The oxygen content was decreased from 89 ppm in the steel electrode to 12 ppm in the ESR ingot in the case of a deoxidizing agent addition rate of about 20 kg/t [5]. The oxygen content was decreased from 29 ppm in the steel electrode to 8 ppm in the 5 tonne-scale ingot produced by protective argon gas atmosphere ESR with Al-based deoxidizing agent addition for slag deoxidation [90]. The deoxidizing agents not only lower the oxygen potential of the slag, but also react with the soluble oxygen in liquid steel [5, 87]. It is schematically described in Fig. 3.4. Deoxidizing agent Ca–Si is generally used for the deoxidation of ESR, in which the contents of aluminum, calcium, titanium and/or Al2 O3 -containing inclusions in the steel should be strictly controlled [51]. Plöckinger [48] claimed that Si deoxidation equilibrium was depended on the activity of SiO2 in the slag during ESR, and the oxygen content of the steel could be reduced to a low level only at a low SiO2 content in the slag. Medina and Cores [54] added Ca–Si (70 mass%) into the slag pool to reduce the FeO content in the slag during ESR for producing the steel with a good homogeneity of alloying elements. Their results show that the oxygen content of the steel increases after ESR in all cases (from initial 24–42 ppm to 42–83 ppm), the oxygen contents of the ingots microalloyed with Ti are determined by A1 and Ti contents. However, the reason for the oxygen pickup is not included in the study by Medina and Cores [54]. The present author claim that the chemical reactions between microalloyed elements Ti, Si, Al and the slag should be responsible for the oxygen increase in their study. Medina et al. [51] lowered the FeO content of the slag by using Ca–Si during electroslag remelting of microalloyed steel. It is found that the oxygen content of the ingots is determined by the contents of calcium and aluminum in the steel. A thermodynamic equilibrium is established between the molten slag and liquid steel, and not between liquid steel and inclusions in the case of deoxidation with Ca–Si, which contributes to the absence of the loss of alloying elements. The present author compared the effect of different calcium addition rates on the oxygen content of steel refined by protective argon gas atmosphere ESR. The results show that calcium addition makes no contribution to further lowering the oxygen content of the steel in comparison with the absence of calcium addition in the protective argon gas atmosphere ESR [12, 13]. For electroslag remelting of the steel with a low oxygen content, slag deoxidization could suppress the oxygen pickup in the steel. The electroslag remelting of GCr15 bearing steel with the oxygen content lower than 6 ppm in open air atmosphere was conducted by periodic addition of Ce–La for slag deoxidization. The degree of the

3.5 Evaluation of the Dependence of Oxygen on the Processing Parameters …

45

oxygen pickup in the steel after ESR was reduced by nearly a half in comparison with the ESR trials without Ce-La addition [55]. The addition rate of deoxidizing agents (termed deoxidation rate in some previous articles) is an important index in affecting the deoxidation of ESR. The addition rate is dependent on one or more of the FeO content in slag, slag composition, steel composition, oxygen content of the electrode, and yield of deoxidizing agents. Improper types and addition rates of deoxidizing agents will cause oxygen pickup in liquid steel as the deoxidation products and/or chemical reaction products between deoxidizing agent and slag components. In the case where the activity of FeO in slag is reduced to a low level after slag deoxidation, the oxide components of the slag could be reduced by the alloying elements or by excessive deoxidizing agents, resulting in an increase in the contents of some alloying elements in the ingot [54, 91]. The present author carried out a serious of ESR deoxidation trials, and found that aluminum pickup took place in the steel during ESR when using Al-based deoxidizing agent for slag deoxidation [5, 12, 14]. In some cases, aluminum contents even roughly doubled in the steel after protective argon gas atmosphere ESR with Al-based deoxidizing agent addition [12]. The experimental work by Reyes-Carmona and Mitchell [1] shows that the aluminum content of the steel increases with increasing the addition rate of deoxidizing agent aluminum, which also results in the increase in the Al2 O3 content in the complex oxide inclusions. For deoxidation of ESR when using different deoxidizing agents, the experimental work by Wang et al. [92] demonstrated that the deoxidation of ESR using aluminum powders, Ca–Si powders or RE–Mg–Si caused the pickup of aluminum in CrNiMoV steel. Ca–Si is the relatively proper deoxidizing agent, and its proper addition rate is 1.5 kg per ton steel in order to control the oxygen potential of the molten slag and prevent aluminum pickup in ESR of CrNiMoV steel. For electroslag remelting of T8MnA steel, the fluctuation of Si and Mn contents in the remelted ingots is originated from the reactions between SiO2 and MnO in the slag and deoxidizing agent Al for slag deoxidation during the ESR process [93]. Slag deoxidation is also widely employed to prevent the loss of concerned alloying elements in steel and alloy in the ESR process. This operation has been validated for successfully keeping constant Ti content in the stainless steel [94, 95]. Decreasing the oxygen potential of the slag by periodic addition of aluminum shots (amounting to 0.1% of the ingot weight) into the slag pool during ESR in open air atmosphere was conducted by Chatterjee et al. [3]. This operation eliminated the loss of silicon, manganese and chromium in 15CDV6 steel, otherwise 5–8% losses took place in the case without slag deoxidation during ESR. The chemical composition was uniform over the whole ingot in the longitudinal and transverse directions. The amount of inclusions was lowered from the index 0.48 to 0.12–0.17. The inclusions larger than 6 μm decreased from 40% of the total amount to 5%. The loss of Si and Mn was prevented and low Al content (< 0.01 mass%) in steel was kept during ESR for 180 tonne-scale 26Cr2Ni4MoV steel production when using Ca–Al deoxidation for the earlier stage and Al deoxidation for the later stage [8]. Wang et al. [96] confirmed that the loss of Si and Mg in CrNiMoV steel was prevented in the case of ESR using

46

3 Deoxidation of ESR and Its Correlation with Oxide Inclusions

Al for slag deoxidation, meanwhile homogeneous distribution of Al content in the ESR ingot was obtained. Many efforts have been made to investigate deoxidation of ESR, but only a few studies have carried out to ascertain the effect of deoxidation operations of ESR on the characteristics of inclusions. There is a critical deoxidation rate for a specific ESR operation, which could result in a change in the chemical reactions between deoxidizing agent and FeO in the slag as well as other slag components. It was revealed in the previous study that excessive Al-based deoxidizing agent addition resulted in Al2 O3 inclusions formation during protective atmosphere ESR of H13 tool steel, which survived in the remelted ingots ultimately [90]. It was also noted in the work by Mehrabi et al. [97] that excessive addition rate of deoxidizing agent for deoxidation of ESR led to the generation of fresh alumina inclusions. ReyesCarmona and Mitchell [1] recognized that below the critical addition rate of Ca–Si (10 kg per ton steel), the ESR refining system was ruled by the deoxidation reactions which were caused by a high activity of FeO in the slag. Above the critical addition rate, there is a low FeO activity of the slag, and deoxidizing agents react with the slag components CaO, SiO2 and Al2 O3 , which lead to higher Al contents in remelted ingots and formation of aluminate inclusions. Higher deoxidation rates using Al during ESR were observed to cause higher Al2 O3 content in aluminate inclusions [1]. The present author claims that the influence of deoxidizing agents on inclusions is not only dependent on the composition and addition rate of the deoxidizing agent, but also on the compositions of the original oxide inclusions in electrode. Deoxidation of ESR for producing heavy ingot is still a challenging issue. For heavy ingot production, the types and addition rate of deoxidizing agents for slag deoxidation are mainly dependent on the variation of slag compositions during each ESR heat in dozens of hours. The complex deoxidation of ESR with aluminum at an addition rate of 0.35–0.25 kg every five minutes and Ca–Si with an addition rate of 0.10–0.05 kg every five minutes has been confirmed to be a proper deoxidization technology for producing 2.25Cr1Mo steel of 56–79 tonne-scale of each ingot, in which the oxygen of the steel is decreased from 89 to 26 ppm [9]. In the future work, the types and the addition rates of deoxidizing agents should be designed for the deoxidation of more steel and alloy grades in the different scale ESR production. For doing this, the online monitor of the change in the slag chemistry during ESR is indispensable. The process modelling of deoxidation for heavy remelted ingot will be a promising work.

3.5.5 Role of Slag Compositions The slag for ESR is usually CaF2 –CaO–Al2 O3 -based system with minor additions of MgO, TiO2 and/or SiO2 to tailor the slag for the specific remelting requirements. The functions of the slag in the ESR process have been summarized elsewhere [98]. Deoxidation of ESR is strongly dependent not only on both the types and addition rates of deoxidizing agents, but also on the slag compositions. Different

3.5 Evaluation of the Dependence of Oxygen on the Processing Parameters …

47

compositions of the ESR-type slags basically have different capacities for adsorption and dissolution of oxide inclusions [99], which cause different deoxidation degrees of liquid steel [11, 99]. The increasing demand for more excellent performance of tool steel has urged metallurgists to further improve the steel cleanliness. Slag compositions exert a great influence on oxide inclusions characteristics (chemistry, number density, and removal, etc.) during conventional secondary refining of liquid steel in industrial practice [100–102]. There is no exception for ESR refining technology. ESR is greatly different from conventional refining technologies in terms of their practical operations and refining principle. It is quite necessary to pay attention to the role of slag composition on the steel cleanliness and inclusions during ESR.

FeO Content of the Slag FeO in slag determines the oxygen content of the remelted ingots in many cases [58, 103]. The deoxidation operations in the ESR process are mainly to lower the FeO content in the slag through adding deoxidizing agents to the slag pool, as summarized above. Machining the steel electrode to a metallic bright surface could prevent the introduction of FeO into the slag to a higher level. It has been verified by Wang et al. [55] and Liu et al. [104] that this operation decreases the oxygen content of liquid steel effectively to a lower level for ESR of 1.2 tonne-scale bearing steel in comparison with the ESR without the operation for removing oxide scale on the steel electrode surface. The essence of removing oxide scale on steel electrode surface is to lower the increase of the FeO content in the slag so as to prevent the oxygen pickup in liquid steel.

SiO2 Content of the Slag The conventional slag used for electroslag remelting of superalloy is required to avoid the presence of SiO2 in order to prevent the chemical reactions between SiO2 and strong oxidizing alloying elements in the alloy such as Al and Ti. It is very difficult and costly to keep extremely low SiO2 content in ESR slags during practical slag manufacturing process due to the impurity in raw materials. For electroslag remelting of most steels, however, SiO2 is a permissible constituent in the ESR slags [105]. In fact, it is suggested that a certain amount of SiO2 addition in the CaF2 –CaO–Al2 O3 slags can meet several requirements of drawing-ingot-type electroslag remelting of steel [98]. In the case of same types of deoxidizing agents (Al or Ca–Si) and addition rate, it has been verified by Reyes-Carmona and Mitchell [1] that the non-metallic inclusions change from only calcium aluminate inclusions to calcium aluminate and Ca–Al silicate inclusions when using 55 mass% CaF2 –15 mass% CaO–15 mass% Al2 O3 –15 mass% SiO2 slag for the ESR instead of 50 mass% CaF2 –20 mass% CaO–30 mass% Al2 O3 slag.

48

3 Deoxidation of ESR and Its Correlation with Oxide Inclusions

The slag with different SiO2 contents makes a difference to the steel cleanliness by affecting the liquid steel compositions, and indirectly changes the oxide inclusions. Mitchell et al. [2] conducted ESR trials using the slag with different SiO2 contents (from 5 to 23 mass%), and found that the slag with low SiO2 activity gave rise to an inclusion population which is predominantly alumina or low-calcium aluminates in low-alloy steel, and the SiO2 content of the inclusions increased as the SiO2 activity of the slag increased, leading to the generation of aluminosilicate inclusions. The selected results from the author’s previous study on the influence of varying SiO2 contents in the slag on the refining efficiency of P-ESR are presented in this chapter. The chemical compositions of the pre-melted slag used in ESR experiments are presented in Table 3.2. The mass ratio of CaO/Al2 O3 in different slags was kept at 0.88. The SiO2 content in the slags ranged from 1.9 to 12.0 mass%. The slag with 1.9 mass% SiO2 was prepared without any intentionally added SiO2 . Pre-melted slags with four different SiO2 contents designated as F1, F2, F3 and F4 were used in ESR trials T1, T2, T3 and T4, respectively. Pre-melted slag was roasted at 973 K (700 °C) in an electrical resistance furnace for 8 h to remove the moisture in the slag before ESR experiments. The chemical composition of the consumable steel electrode is listed in Table 3.3. The oxide scale on electrode steel surface was basically removed mechanically prior to ESR experiments. The whole ESR process was conducted in protective argon atmosphere. During protective atmosphere ESR refining, a steel sample was taken from the liquid metal pool in mold using a vacuum sampling tube that was made of quartz (6 mm in inner diameter), followed by quenched in water. The sampling from liquid metal pool was performed at the time when about two thirds of the whole refining process was finished. The temperature in liquid metal pool is in gradient distribution (typically varying around 1873 K (1600 °C)) not only along radial direction but also along axial direction, which strongly affects the shape and size of liquid metal pool and consequently influences the solidification quality of ascast ingot [106–109]. The as-cast ingots produced in the ESR trials using these four slag systems were designated as ESR-1, ESR-2, ESR-3, and ESR-4, respectively. Table 3.2 Chemical compositions of the slag used in ESR experiments (mass%) Slag No.

CaF2

CaO

Al2 O3

MgO

SiO2

F1

29.3

30.5

34.5

3.8

1.9

F2

28.1

29.2

33.1

3.6

6.0

F3

27.2

28.3

32.0

3.5

9.0

F4

26.3

27.4

30.9

3.4

12.0

Table 3.3 Chemical composition of the consumable steel electrode (mass%) C

Si

Mn

Cr

Mo V

Ti

Ca

Mg

Al

O

S

N

0.40 1.040 0.35 5.10 1.2 0.96 0.0078 0.0019 0.0002 0.0099 0.0018 0.0032 0.0078

3.5 Evaluation of the Dependence of Oxygen on the Processing Parameters …

49

The steel samples that were cut from remelted ingots were prepared for chemical analysis. The contents of soluble Al, Ca, Mg and Si in remelted ingots were measured by the inductively coupled plasma atomic emission spectroscopy (ICP-AES). The total oxygen and sulfur contents in the steel were measured by the inert gas fusioninfrared absorptiometry. The nitrogen content was determined by inert gas fusionthermal conductivity method. The microscopic observations of the metallographic samples that taken from the consumable electrode and as-cast ESR ingots have been described in Sect. 4.2. The contents of alloying elements in the remelted ingots produced in the protective atmosphere ESR trials using above four slag systems are listed in Table 3.4. The total oxygen content increased from 18 ppm in the steel electrode to 21–34 ppm in the remelted ingot. For the ESR refining process in which the reoxidation of liquid steel takes place, the oxygen content of the steel increases appreciably, but the total oxygen contents of the remelted ingots are independent on the varying SiO2 contents in the slag. The aluminum content is observed to increase apparently accompanying with a slight loss of silicon in steel after ESR, compared with their concentrations in consumable electrode. It is noted that aluminum pickup tends to decrease with the increase in SiO2 content in slag, whereas the loss of silicon in steel becomes smaller. Silicon loss and aluminum pickup during ESR are virtually prevented through intentionally adding 9 mass % SiO2 or more to the slag. The following reaction should be responsible for the change in alloying element content because the initially low aluminum content in steel and low activity of SiO2 in slag provide a driving force for aluminum pickup. 3[Si] + 2(Al2 O3 )slag = 4[Al] + 3(SiO2 )slag ΔG  1 = 658,300 − 107.2T [28] (J/mol)

(3.25) K =

4 3 aAl · aSiO 2 3 2 aSi · aAl 2 O3

=

3 ( f Al [%Al])4 · aSiO 2

(3.26)

2 ( f Si [%Si])3 · aAl 2 O3

where aSiO2 and aAl2 O3 are the activities of SiO2 and Al2 O3 in slag, respectively. f Al and f Si are the activity coefficients of dissolved aluminum and silicon in liquid steel, respectively, and can be calculated by Eq. 3.6. The activities of oxide components relative to pure solid standard states in the slag melts at 1873 K (1600 °C) were estimated with FactSage 7.1 (ThermFact/ Table 3.4 Chemical compositions of remelted ingots (mass%) Ingot No.

Si

Ca

Al

Mg

O

S

N

ESR-1

0.990

0.0005

0.057

< 0.0004

0.0034

0.0013

0.0079

ESR-2

1.019

0.0005

0.022

< 0.0004

0.0021

0.0014

0.0080

ESR-3

1.033

0.0005

0.018

< 0.0004

0.0026

0.0015

0.0075

ESR-4

1.034

0.0004

0.013

< 0.0004

0.0024

0.0017

0.0080

50

3 Deoxidation of ESR and Its Correlation with Oxide Inclusions

CRCT, Montréal, Canada). The calculated values are shown in Fig. 3.12. Both SiO2 and Al2 O3 activities were observed to linearly increase roughly with increasing SiO2 content in the slag. The activity of SiO2 is much lower than that of Al2 O3 in slag. The Gibbs free energy change for reaction (3.25) with regard to four ESR trials was calculated, in combination with the reported interaction parameters [13, 22, 28, 30, 110, 111] included in Eq. 3.6, to be − 172.1 kJ/mol, − 121.7 kJ/mol, − 100.5 kJ/ mol and − 80.3 kJ/mol, respectively. It indicated that aluminum pickup and silicon loss in steel did originate from the reduction of Al2 O3 from slag by dissolved silicon in liquid steel according to reaction (3.25). A small amount of SiO2 addition in slag was confirmed to be effective in preventing silicon loss and aluminum pickup during ESR. (1) Inclusions in Consumable Steel Electrode In view of the significant importance of the chemistry of inclusions in steel, a specific attention was focused on the chemistry evolution of inclusions in the steel before, during and after ESR. Figure 3.13 presents the SEM-EDS element mappings of typical inclusions observed in consumable steel electrode. All oxide inclusions in the consumable steel electrode were identified as CaO–Al2 O3 –SiO2 with a small amount of MgO. Furthermore, EPMA analysis was performed to quantitatively determine the compositions of oxide inclusions. The MgO content in CaO–Al2 O3 –SiO2 –MgO inclusions was identified to be extremely low (approximately 1 mass%). The inclusions in consumable electrode (termed original inclusions) are mostly 3–8 μm in size. A detailed description of the compositions of oxide inclusions is presented in Sect. 4.4. (2) Inclusions in the Remelted Ingots Three types of oxide inclusions were found in the remelted ingots produced in the protective atmosphere ESR trials using the slag with 1.9 mass%, 6.0 mass% and 0.25 Activity of component in slag

Fig. 3.12 Activities of oxide components relative to pure solid standard states in the slag with various SiO2 contents at 1873 K (1600 °C) estimated with FactSage 7.1

Al2O3 CaO

SiO2 MgO

0.20 0.15 0.10 0.05 0.00 F1

F2 F3 Slag Number

F4

3.5 Evaluation of the Dependence of Oxygen on the Processing Parameters …

51

Fig. 3.13 SEM-EDS element mappings of typical inclusions in consumable steel electrode: a (Ca, Mn)S adhering to CaO–Al2 O3 –SiO2 –MgO, b patch-type (Ca, Mn)S associated with CaO–Al2 O3 – SiO2 –MgO inclusion

9.0 mass% SiO2 , i.e., CaO–Al2 O3 –SiO2 –MgO, CaO–Al2 O3 –MgO and MgAl2 O4 . Al2 O3 inclusions are occasionally observed in the remelted ingots. The SEM micrograph, element mappings and EDS spectrums of the typical inclusions observed in the remelted ingot when using the slag with the 1.9 mass% SiO2 are present in Sect. 4.5. There is no difference in the types of the inclusions observed in the remelted ingots for these three cases of protective atmosphere ESR. In some cases, SiO2 was detected by EDS in CaO–Al2 O3 –MgO inclusions, but its content was far lower than 1 mass%. These inclusions were categorized as CaO–Al2 O3 –MgO type. No CaO– Al2 O3 –SiO2 –MgO inclusions with the compositions similar to those in consumable electrode were observed in ESR ingots. It was identified that MgO content was approximately 1mass% in these calcium aluminate inclusions. Most of the observed CaO–Al2 O3 –SiO2 –MgO and CaO–Al2 O3 –MgO inclusions are 2–6 μm in size. The calcium aluminate inclusions with the size ranging from 6 to 12 μm take up a very small proportion. The size of MgAl2 O4 inclusions is about 1.5 μm. Both EDS and EPMA analysis show that SiO2 content in CaO–Al2 O3 –SiO2 – MgO inclusions greatly decreases with an accompanying considerable increase in

52

3 Deoxidation of ESR and Its Correlation with Oxide Inclusions

Al2 O3 content after ESR refining, in comparison with the compositions of the oxide inclusions in consumable electrode. In the case of low SiO2 level (1.9 mass%) in slag, only a trace amount of SiO2 (slightly smaller than 1 mass%) was detected in CaO–Al2 O3 –SiO2 –MgO inclusions in remelted ingot. As increasing SiO2 content in the slag, SiO2 content in inclusions increased in the remelted ingots. It should be pointed out that these quaternary CaO–Al2 O3 –SiO2 –MgO inclusions (defined as type I) contain roughly same amounts of CaO and MgO as that in original oxide inclusions. The CaO content in CaO–Al2 O3 –MgO inclusions is far lower than that in CaO–Al2 O3 –SiO2 –MgO inclusions (type I) according to EDS and EPMA analysis, whereas Al2 O3 content shows an opposite feature. CaO–Al2 O3 –MgO inclusions exhibit much lower CaO/Al2 O3 mass ratio than CaO–Al2 O3 –SiO2 –MgO inclusions. In addition to CaO–Al2 O3 –SiO2 –MgO inclusions (type I), the other type of quaternary CaO–Al2 O3 –SiO2 –MgO inclusions (type II) was found in the remelted ingot ESR-4 (see Fig. 3.14c for an example), in which CaO and MgO contents were nearly same as that in CaO–Al2 O3 –MgO inclusions. CaO–Al2 O3 –SiO2 –MgO inclusions (type II) contain approximately 5 mass% SiO2 , and significantly lower CaO content and higher Al2 O3 content in comparison with that in CaO–Al2 O3 –SiO2 –MgO inclusions (type I). Unlike the inclusions observed in other three steel ingots, no CaO– Al2 O3 –MgO inclusions are present in the remelted ingots produced in the protective atmosphere ESR trials using the slag with 12.0 mass% SiO2 . The relative proportion (in number) of each type of oxide inclusions is presented in Fig. 3.15. CaO–Al2 O3 –SiO2 –MgO inclusions account for nearly 30% of the total inclusions in each remelted ingot. CaO–Al2 O3 –MgO inclusions are the predominant type in ESR-1. The relative proportion of CaO–Al2 O3 –MgO inclusions decreases with increasing SiO2 content in the slag (but still greater than 1/3 of the total inclusions in ESR-4). MgAl2 O4 inclusions take up 17% in ESR-1, and its relative fraction increases with increasing SiO2 content in the slag (about 37% of the total inclusions in ESR-3 and ESR-4). Todoroki and Mizuno [112] reported a similar trending showing that SiO2 in the slag enhanced the formation of MgAl2 O4 spinel inclusions in 304 stainless steel through laboratory slag-metal reaction experiments using MgO crucible. Al2 O3 inclusions were occasionally observed in remelted ingots. These Al2 O3 inclusions were classified as MgAl2 O4 inclusions when counting inclusion relative proportion. The oxide inclusions were observed to exhibit a narrow variation in their compositions in individual remelted ingot. The average compositions of CaO–Al2 O3 –SiO2 – MgO inclusions in each ESR ingot were displayed on CaO–Al2 O3 –SiO2 ternary phase diagram to reveal the chemistry evolution trajectory of oxide inclusions before and after ESR refining, as shown in Fig. 3.16. MgO was not taken into account due to its negligibly small content. The average composition of CaO–Al2 O3 –SiO2 –MgO inclusions in consumable electrode is also presented in Fig. 3.16 for comparison. The compositions of CaO–Al2 O3 –SiO2 –MgO inclusions in remelted ingots locate in high-melting-temperature region [> 1873 K (1600 °C)], indicating the transformation of original CaO–Al2 O3 –SiO2 –MgO from liquid state to solid CaO–Al2 O3 – SiO2 –MgO after ESR. This is due to a considerable decrease in SiO2 content in parallel with a considerable increase in Al2 O3 content in oxide inclusions after ESR,

3.5 Evaluation of the Dependence of Oxygen on the Processing Parameters …

53

Fig. 3.14 SEM-EDS element mappings of typical inclusions observed in ingot ESR-4: a CaO– Al2 O3 –SiO2 –MgO (type I), b dual-phased CaO–Al2 O3 –SiO2 –MgO + CaS, c CaO–Al2 O3 –SiO2 – MgO (type II). (EDS spectrums in Figs. (d) and (e) correspond to the centre point of the inclusions shown in Figs. (b) and (c), respectively)

54

3 Deoxidation of ESR and Its Correlation with Oxide Inclusions cps/eV

Al

20

Acquisition

(d)

18 16 14 12

Ca

10 8 6

2

Mg

Cr O

4 S Ca

Fe Fe

Si

S

Cr

0 1

2

3

4

5

6

7

8

9

10

keV 22

cps/eV

Acquisition

20

Al

0.30atom% Si

18 16 14

(e)

Mo

12 Fe

10 8 6 4 2

Cr V O Ca

Ca

Mg

Cr

Fe Si

Mo

V

0 1

2

3

4

5

6

7

8

9

10

keV

Fig. 3.14 (continued)

as indicated by the symbols of data points shown Fig. 3.16. The liquidus temperatures of the inclusions seem to decrease with increasing SiO2 content in slag. As for CaO–Al2 O3 –MgO inclusions, the concentrations of CaO and Al2 O3 are in range from 12 mass% to 19 mass% and 77 mass% to 86 mass%, respectively. According to CaO–Al2 O3 –MgO phase diagram [12], the compositions of these inclusions lie in high-melting-temperature region [> 1973 K (1700 °C)]. The examples of MgAl2 O4 and Al2 O3 inclusions (BSE images and EDS spectrums) observed in the remelted ingots are shown in Fig. 3.17. EDS spectrums showed that Mg content in these inclusions was extremely low. Approximately 2 mass% MgO was detected in MgAl2 O4 inclusions by EPMA analysis. This observation is quite similar with the findings reported elsewhere [13, 113]. According to MgO–Al2 O3 binary phase diagram [114], these observed MgAl2 O4 inclusions are in two-phase region of spinel + alumina, and could be considered as a mixture of Al2 O3 and MgO · Al2 O3 spinel (with only a trace amount). Some of the MgAl2 O4 inclusions

3.5 Evaluation of the Dependence of Oxygen on the Processing Parameters …

100 Frequency of Inclusion (%)

Fig. 3.15 Relative proportion (in number) of the oxide inclusions in remelted ingots. CASM and CAM represent CaO–Al2 O3 –SiO2 –MgO (type I) and CaO–Al2 O3 –MgO inclusions, respectively. (In ESR-4, CaO–Al2 O3 –MgO inclusions without exception contain about 5 mass% SiO2 )

55

CASM MgAl2O4 CAM

80 60 40 20 0

ESR-1

ESR-2 ESR-3 Remelted Ingots

ESR-4

Fig. 3.16 Distribution of average compositions of CaO–Al2 O3 –SiO2 –MgO (CASM) inclusions in consumable electrode and remelted ingots on ternary CaO–Al2 O3 –SiO2 phase diagram, (temperature in degree Celsius, and MgO was deleted from analyses). Solid marks represent the average composition of inclusions in steel samples

56

3 Deoxidation of ESR and Its Correlation with Oxide Inclusions

were observed to serve as nucleation sites for (Ti, V)(C, N), promoting nucleation and growth of carbonitride (consequently several times larger than oxide sites). Fujimura et al. [115], Park et al. [116], and Shi et al. [117] reported that MgAl2 O4 spinel facilitated the formation of TiN and (Ti, V)N by acting as heterogeneous nucleation agent because of low lattice disregistry between the substrate and the nucleated phase. (3) Transient Inclusions in the Liquid Metal Pool During ESR Refining Figure 3.18 shows the BSE images and EPMA analyzed compositions of the inclusions observed in the sample collected from the liquid metal pool during ESR. An example of CaO–Al2 O3 –MgO inclusion in liquid metal pool is presented in Fig. 3.19. The composition of this inclusion is quite uniform (meanwhile without sulfide phase in it). No appreciable differences were observed in the compositions, size and morphology between the oxide inclusions in liquid metal pool and those in remelted ingot for each ESR trial. A detailed description of inclusion characteristics is presented in Sect. 4.6.

cps/eV Al

Acquisition

16

Fe

(k)

14 12 10 8 6 4 2 0

Ti V Cr O

Mg Fe

Cr

F

Ti

1

2

3

4

V

5

6

7

8

9

10

keV

Fig. 3.17 Examples of MgAl2 O4 and Al2 O3 inclusions observed in remelted ingots: a ESR-1, c–d ESR-2, e–g ESR-3, h–j ESR-4. (EDS spectra in Fig. (k) corresponds to the oxide inclusions shown in Fig. (e))

3.5 Evaluation of the Dependence of Oxygen on the Processing Parameters …

57

Fig. 3.18 BSE images and EPMA analyzed compositions of inclusions observed in liquid metal pool during ESR. a CaO–Al2 O3 –MgO, b CaO–Al2 O3 –SiO2 –MgO (type II), c CaO–Al2 O3 –SiO2 – MgO (type I), d MgAl2 O4 . (EPMA analysis is expressed in mass percent)

cps/eV

Al

30

Acquisition

(b)

0.16 atom% Si

25 20 15 10

Ca Cr O

5

Fe

Mg Fe

Ca

Cr

Si

0 1

2

3

4

5

6

7

8

9

10

keV

Fig. 3.19 Example of CaO–Al2 O3 –MgO inclusion (SEM-EDS element mappings) observed in the steel sampled from the liquid metal pool during ESR. (EDS spectrum in Fig. (b) corresponds to the inclusion shown in Fig. (a))

(4) Evolution Mechanism of Inclusions During ESR Refining The microscopic observations show that the content of SiO2 in CaO–Al2 O3 –SiO2 – MgO inclusions dramatically decreases in remelted ingot accompanying with a considerable increase in Al2 O3 content, in comparison with the oxide inclusions

58

3 Deoxidation of ESR and Its Correlation with Oxide Inclusions

in the steel before ESR. The reduction of SiO2 in the original CaO–Al2 O3 –SiO2 – MgO inclusions by dissolved aluminum in liquid steel was assessed according to reaction (2.4). 4[Al] + 3(SiO2 )inclusion = 3[Si] + 2(Al2 O3 )inclusion ΔG  4 = −658,300 + 107.2T [28] (J/mol) K =

2 3 aAl · aSi 2 O3 3 4 aSiO · aAl 2

=

2 aAl · ( f Si [%Si])3 2 O3 3 aSiO · ( f Al [%Al])4 2

(3.27)

(3.28)

where aAl2 O3 and aSiO2 are the activities of Al2 O3 and SiO2 in oxide inclusion, respectively. The activity coefficients of dissolved silicon and aluminum in liquid steel f Si and f Al can be calculated using Eq. 3.6. The activities of Al2 O3 and SiO2 relative to pure solid standard states in oxide inclusion were calculated with FactSage 7.1 (FToxid database), based on the average compositions of CaO–Al2 O3 –SiO2 –MgO inclusions in consumable steel electrode. The interaction parameters used in the calculation of activity coefficients were described in Sect. 3.1. The Gibbs free energy change for reaction (3.27) was calculated to be − 132.9 kJ/mol, − 74.5 kJ/mol, − 59.2 kJ/mol and − 38.9 kJ/mol for these four trials, which was an indication of the reduction of SiO2 from original oxide inclusions by dissolved aluminum in liquid steel. This was supported by the experimental observations. The increasing values of Gibbs free energy change for reaction (3.27) was an indication of decreasing tendency of the reduction reaction between SiO2 in oxide inclusions and dissolved Al in liquid steel, as evidenced by the determined SiO2 contents (became less marked with increasing SiO2 content in the slag) in oxide inclusions, which originated from a considerable decreasing Al pickup in liquid steel caused by increasing SiO2 content in the slag. The driving force for reaction (3.27) is the difference in oxide activities between the slag and oxide inclusions. It was the slag-steel reaction, through affecting the aluminum content in liquid steel, that contributed to the compositional evolution of CaO–Al2 O3 –SiO2 –MgO inclusions during ESR. SiO2 in oxide inclusions could not be completely reduced by dissolved Al in liquid steel thermodynamically, resulting in the transformation of inclusions from CaO–Al2 O3 –MgO–SiO2 to CaO–Al2 O3 –MgO. It was revealed from compositional analysis that the reduction in SiO2 content was approximately equal to the increment of Al2 O3 expected from the reaction stoichiometry of 4[Al] + 3(SiO2 ) = 3[Si] + 2(Al2 O3 ) in CaO–Al2 O3 –SiO2 –MgO inclusions (type I), in comparison with the compositions of original CaO–Al2 O3 –SiO2 –MgO inclusions in the steel before ESR refining. These determined compositional changes of CaO–Al2 O3 –SiO2 –MgO inclusions before and after ESR did support the proposed mechanism that the reduction of SiO2 from original CaO–Al2 O3 –SiO2 –MgO inclusions by dissolved aluminum in liquid steel.

3.5 Evaluation of the Dependence of Oxygen on the Processing Parameters …

59

The Gibbs free energy change for reaction (3.29) was calculated to evaluate the driving force for the reaction between dissolved aluminum in liquid steel and CaO in original CaO–Al2 O3 –SiO2 –MgO inclusion by combining the standard Gibbs free energy change for reaction (3.29), first-order and second-order interaction parameters summarized elsewhere [13, 22, 28]. 2[Al] + 3(CaO)inclusion = 3[Ca] + (Al2 O3 )inclusion ΔG  6 = 733,500 − 59.7T [118] (J/mol)

(3.29)

The concentration of dissolved oxygen concentration estimated from [Si]–[O] equilibrium and [Al]–[O] equilibrium was used, instead of the measured total oxygen concentration in steel, to calculate the activity coefficients of dissolved calcium based O on the first-order and second-order interaction parameters [28] eO Ca = −9000, r Ca = Ca,O 6 6 3.6 × 10 , and rCa = 2.9 × 10 . In the calculation of the Gibbs free energy change for reaction (3.29), the concentration of dissolved calcium in liquid steel should be used. The measured concentration of total calcium in steel includes the dissolved calcium (free calcium) and insoluble calcium combined as inclusions. The measured total calcium content in remelted ingots is 5 ppm. The dissolved calcium in the steel is very difficult to be determined. The dissolved calcium content in steel is far lower than total calcium content [119, 120]. For convenience, 1 ppm was used in the current calculation. The calculated values of the Gibbs free energy change for reaction (3.29) are extremely positive (275.0 kJ/mol, 162.0 kJ/mol, 107.8 kJ/mol, and 20.5 kJ/mol for trials T1, T2, T3, and T4). In the case of higher dissolved calcium content in steel, the calculated Gibbs free energy change is larger. The thermodynamical calculation for reaction (3.29) suggested that CaO in original CaO–Al2 O3 –SiO2 –MgO inclusions in steel electrode could not be reduced by dissolved aluminum in liquid steel during ESR. The CaO content in CaO–Al2 O3 –MgO and CaO–Al2 O3 –SiO2 –MgO inclusions (type II) was identified to be far lower than that in original CaO–Al2 O3 –SiO2 –MgO inclusions. It therefore was concluded that the oxide inclusions containing lower CaO content were not originated from the transformation of original CaO–Al2 O3 – SiO2 –MgO inclusions through reduction reaction by dissolved aluminum in liquid steel, but were newly-formed oxide inclusions. Most of the original inclusions in the consumable electrode were removed during the ESR process (as indicated by the significant decrease in calcium content in the steel before and after ESR), and the others were transformed to CaO–Al2 O3 –SiO2 – MgO inclusions (type I) which remained until in the ESR ingot. The reoxidation of liquid steel during ESR leads to considerable pickup of dissolved oxygen, which provides a driving force for the generation of fresh oxide inclusions, thereby resulting in the generation of CaO–Al2 O3 –MgO, Al2 O3 , MgAl2 O4 and CaO–Al2 O3 –SiO2 – MgO inclusions (type II). The preceding experimental determination did confirm that CaO–Al2 O3 –MgO, Al2 O3 , MgAl2 O4 and CaO–Al2 O3 –SiO2 –MgO (type II) inclusions were newly-formed inclusions.

60

3 Deoxidation of ESR and Its Correlation with Oxide Inclusions

It should be stressed that CaO–Al2 O3 –MgO inclusions could be modified MgAl2 O4 inclusions, but the contribution in this trajectory to the formation of CaO– Al2 O3 –MgO inclusions during ESR process was expected to be quite small because of a significant difference in the size between CaO–Al2 O3 –MgO inclusions and MgAl2 O4 inclusions as well as their size distribution (as described above in the second section of Sect. 3.5.5.2), considering that calcium modification of MgAl2 O4 inclusion could hardly increase the size of the original inclusion appreciably [13]. Calcium modification of MgAl2 O4 spinel and low-MgO-containing MgAl2 O4 inclusion to CaO–Al2 O3 –MgO inclusions during ESR process has been presented in detail in the previous publication [13]. It is demonstrated that the aluminum pickup in liquid steel is drastically decreased in accompanying with the decrease in silicon loss with increasing SiO2 content in the slag for ESR, which originates from a decreasing tendency of the reduction reaction of Al2 O3 from the slag by silicon in liquid steel. Silicon loss and aluminum pickup during ESR are virtually prevented by adding 9 mass % and more SiO2 to the slag. The oxygen content in the steel appreciably increases during ESR, but keep roughly constant in remelted ingots even though the slag with different SiO2 contents is used in ESR. The oxide inclusions in the steel before ESR are liquid CaO–Al2 O3 –SiO2 –MgO, which are mostly 3–8 μm in size. Three types of oxide inclusions are present in both liquid metal pool and remelted ingots, i.e., CaO–Al2 O3 –MgO, CaO–Al2 O3 –SiO2 – MgO, and MgAl2 O4 (about 1.5 μm in size). In the case of remelting using the slag with 12 mass% SiO2 , CaO–Al2 O3 –MgO inclusions invariably contain approximately 5 mass% SiO2 . Most of these calcium aluminate inclusions are 2–6 μm in size. CaO– Al2 O3 –SiO2 –MgO inclusions (type I) originate from the reduction of SiO2 from the original oxide inclusions in consumable electrode by dissolved Al in liquid steel during ESR. CaO–Al2 O3 –MgO, MgAl2 O4 , and CaO–Al2 O3 –MgO–SiO2 (type II) inclusions are generated by the reactions taking place inside liquid steel in liquid metal pool as reoxidation products. The relative proportion (in number) of MgAl2 O4 inclusions in steel increases with increasing SiO2 content in the slag for ESR refining, whereas CaO–Al2 O3 –MgO inclusions show a decreasing trend in relative proportion. The relative proportion of CaO–Al2 O3 –SiO2 –MgO inclusions keeps nearly a constant (a little less than 30% of the total inclusions), even though varying the SiO2 contents in ESR slag. The SiO2 content in original CaO–Al2 O3 –SiO2 –MgO inclusions is considerably reduced, in parallel with an increase in Al2 O3 content, after ESR of the steel. The oxide inclusions change from liquid to solid state during ESR. Increasing SiO2 addition in slag preferentially results in a decreasing pickup of Al in liquid steel, consequently contributes to a decreasing reduction of SiO2 from oxide inclusions during ESR.

Al2 O3 Content of the Slag Al2 O3 is an indispensable component in almost all commercial ESR-type slag (mostly 20–30 mass%). It greatly affects the viscosity and electrical conductivity of the slag

3.5 Evaluation of the Dependence of Oxygen on the Processing Parameters …

61

[121]. The effect of varying Al2 O3 content in slag on the steel cleanliness has been scarcely studied. For ESR of the steel with the oxygen content of 13 ppm, Schneider et al. [59] recognized that increasing the activity of Al2 O3 in the slag caused the increase in the oxygen content of the steel, in which the oxygen content almost doubled to 24 ppm in some cases, even though the FeO content in the slag was kept at a very low level (0.1 mass%). Chang et al. [122] proposed that Al2 O3 in the slag was an important source of oxygen pickup during ESR of the steel with low oxygen (18 ppm), in which the oxygen pickup was originated from the decomposition of Al2 O3 from the slag. In their study, FeO and SiO2 (4–6 mass%) in the slag was not considered for evaluating the factors affecting the oxygen content of liquid steel.

CaF2 Content of the Slag Commercial ESR-type slag contains a large amount of CaF2 (typically 40– 70 mass%), aiming to reduce the melting temperature and viscosity of the slag [123]. Although CaF2 plays an important role in the ESR-type slag, the evaporation of fluoride from the slag melts during ESR process has always been an extremely serious issue because it poses serious contamination of environment [124–126], health hazard to plant operators [126, 127], corrosion of plant equipment [128], as well as the change in the slag chemistry. The variation of slag chemistry generally causes the fluctuation of the viscosity and other thermo-physical properties of the slag, which thereby could change the contents of some elements in liquid steel and degrade the reliability of ESR operating practice and the quality of remelted product, especially for large-scale ESR [129]. However, the development of fluoride-free or low-fluoride slag for ESR is still in its infancy. Radwitz et al. [130] compared the oxygen contents of the steel remelted using CaF2 –CaO–Al2 O3 slag with varying CaF2 contents (mass%CaO/mass%Al2 O3 = 1). The results show that a lower oxygen content is obtained by protective atmosphere ESR of the steel with 25 ppm oxygen in the case of a coupled increase in the CaO and Al2 O3 contents and decrease in the CaF2 content in the slag. In the case of protective atmosphere ESR using high CaF2 slag, the oxygen content of the ingot is slightly higher than that of the steel electrode. The total amount of inclusions in the steel decreases after protective atmosphere ESR, but increases with increasing the CaF2 content in the slag and decreasing the contents of CaO and Al2 O3 accordingly.

MgO Content of the Slag The commercial ESR-type slag generally contains about 3 mass% MgO. To evaluate the refining characteristics of ESR when using the slag with four different contents of MgO, a slight overpressure (0.12 MPa) argon gas atmosphere ESR of 21CrMoV5-7 steel was performed [6]. The chemical compositions of the slags are shown in Table 3.5.

62

3 Deoxidation of ESR and Its Correlation with Oxide Inclusions

Table 3.5 Chemical compositions of the slag used in the ESR (mass%) Slag No.

CaF2

CaO

Al2 O3

MgO

SiO2

FeO

S1

59.34

18.98

18.64

2.03

0.22

0.046

S2

59.34

17.48

17.13

5.07

0.01

0.04

S3

59.34

15.05

14.69

9.98

0.03

0.04

S4

59.34

12.62

12.25

14.89

0.04

0.03

Figure 3.20 shows the total oxygen contents of remelted ingots over the ingot height when using different slag compositions. In Fig. 3.20, the average amount of total oxygen in the electrodes is highlighted in black, whereas the range of measured oxygen values in the electrodes is illustrated in grey. There are no changes in the total oxygen (T.O) contents over the ingot length, despite the variation of aluminum and silicon contents in the steel and different deoxidizing abilities of aluminum and silicon. It is further recognizable that the oxygen levels in the remelted ingots are in the range of that in the electrode or even slightly higher and could not be significantly reduced [6]. The comparison of the average total oxygen contents depending on the content of MgO shows that the oxygen level is not strongly influenced by the variation in the slag composition (see Fig. 3.21), only slightly higher contents are detected at 10 mass% MgO [6]. With regard to the inclusions in the ingots, the amount of the inclusions smaller than 4 μm increases with increasing the MgO contents in the slag and even exceeds the amount of the inclusions in the steel electrode [6]. With the increase in the MgO contents in the slag, the total amount of inclusions increases, which increases up to similar values as that in the steel electrode due to the presence of a large amount of

Fig. 3.20 Total oxygen content after remelting using various slags over the ingot height [6]

3.5 Evaluation of the Dependence of Oxygen on the Processing Parameters …

63

Fig. 3.21 Comparison of average total oxygen content of remelted ingots depending on the MgO contents of the slag for ESR [6]

small inclusions in the case of higher MgO content in the slag. More studies are still needed to further clarify the mechanism as well as the effect of MgO on the inclusion chemistry.

3.5.6 Reoxidation of Liquid Steel The difference in the oxygen potential between liquid steel and molten slag, as well as gas phase could contribute to the reoxidation of liquid steel during protective atmosphere ESR [11]. The oxygen content of steel nearly doubles (21–34 ppm) after protective atmosphere ESR of the steel with low oxygen content (18 ppm), indicating the occurrence of the reoxidation of liquid steel during the protective atmosphere ESR [11]. As part of the original oxide inclusions is removed during the ESR refining, the oxygen level contributed by the reoxidation of liquid steel virtually is higher than the difference in the measured oxygen contents between consumable electrodes and remelted ingots. The introduction of the oxygen from atmosphere and FeO (the newly-formed during on-going ESR, and oxide scale that not being removed from steel electrode surface) could hardly be prevented during ESR, and these aspects are the sources of oxygen pickup in the steel. The reoxidation of liquid steel during protective atmosphere ESR leads to considerable pickup of soluble oxygen in liquid steel, which provides a driving force for the generation of fresh oxide inclusions. In the author’s previous study, the formation of fresh CaO–Al2 O3 –MgO, Al2 O3 , MgAl2 O4 and CaO–Al2 O3 –SiO2 –MgO inclusions during protective atmosphere ESR of tool steel takes place, as a result of the chemical reactions occurring inside liquid steel in the liquid metal pool caused by reoxidation of liquid steel [11]. Electroslag remelting of 2.4 tonne-scale bearing steel G20CrNi2Mo in protective argon gas atmosphere show that the FeO content in the slag increases from

64

3 Deoxidation of ESR and Its Correlation with Oxide Inclusions

0.20 mass% at the beginning of ESR to 0.45 mass% at the end of ESR, and the oxygen content increases from 12 ppm in the electrode to 16–21 ppm in the ingots, whereas aluminum decreases from 0.040 mass% to 0.031 mass%–0.019 mass% [131]. A kinetic model for predicting the variation of oxygen and aluminum contents in bearing steel G20CrNi2Mo was developed based on the penetration and film theories by Li et al. [131]. The model reveals that the increase in the soluble oxygen in liquid steel mainly occurs during the metal droplets formation and falling. The rate-determining step of the reoxidation of liquid steel lies in the mass transfer of FeO at the slag side of the slag-steel interface. With the increase in the FeO content from 0.20 to 0.45 mass%, the mass transfer resistance of FeO decreases obviously, thus resulting in an increase in the oxygen content and aluminum oxidation. The slag deoxidation with Al–Mg alloy during protective atmosphere ESR of the steel with an oxygen content of 15 ppm has been validated to successfully suppress the reoxidation of liquid steel, but this operation fails to reduce the oxygen content of the steel [132]. Although these authors [132] emphasized the role of FeO in the slag on the oxygen pickup in liquid steel, they did not provide insight about the FeO content of the slag and its correlation with slag deoxidation operation.

3.5.7 Melting Rate and Filling Ratio of ESR Liquid metal films form at the electrode tip during the ESR process, and thereafter collect as liquid metal droplets. This stage of ESR plays a predominant role in refining of liquid metal during the ESR process. The melting rate of ESR largely determines the thickness of the metal film at the electrode tip, and the residence time of the liquid metal films and metal droplets at the electrode tip [33]. In addition, the local solidification time and the advance rate of the solidifying front are strongly affected by the melting rate of ESR [133, 134], which therefore influence the formation and removal of fresh inclusions in the liquid metal pool during the ESR process [135]. Electroslag remelting of 316LC stainless steel with an oxygen content of 223 ppm shows that increasing the melting rates of ESR lowers the cleanliness of the steel in terms of oxide inclusions amount and the oxygen content of the steel [65]. Ahmadi et al. [65] deduced that the increase in the inclusion amount and oxygen content was originated from a faster pass of the liquid metal droplets through the slag pool with increasing the melting rates of ESR, which therefore led to a decrease in the elimination of oxide inclusions. This issue is open to question. The inclusion removal during ESR takes place predominantly at the stage of liquid metal films formation and their collection into droplets at the electrode tip, whereas the stage when the liquid metal droplets pass through the slag pool and the process in the liquid metal pool contributes in a small manner (does not play an important role) [64, 66, 135– 138]. The present author insists that the increase in the inclusions amount and oxygen content of the steel is mainly originated from the decrease in the residence time of the liquid metal films and metal droplets at the electrode tip with the increase in the

3.5 Evaluation of the Dependence of Oxygen on the Processing Parameters …

65

melting rates of ESR, resulting in the decrease in the effective refining of liquid steel for inclusion removal. In the case of protective atmosphere ESR of the steel with ultra low oxygen content (8 ppm), the present author’s previous study [15] shows that the oxygen content of the steel nearly doubles (up to 14–17 ppm) after protective atmosphere ESR, indicating the occurrence of the reoxidation of liquid steel during the ESR process. In these cases, the melting rates (350, 400, 450, and 500 kg/h) of ESR make a negligible difference in the steel cleanliness in terms of the oxygen, sulfur, and nitrogen contents in the steel. The filling ratio (ratio of the cross-sectional area of the electrode to the crosssectional area of the mold) of ESR could be varied in a limited range according to the specific ESR requirements. The dependence of the oxygen content of the remelted ingot on the filling ratios of ESR has been studied scarcely. The electroslag remelting of bearing steel GCr15 with an oxygen content of 10 ppm was performed by Liu et al. [104] to compare the effect of two filling ratios on the oxygen content of the steel. The results reveal that the increase in the filling ratio leads to a lower oxygen content of the remelted ingot, as shown in Fig. 3.22. It is attributed to the decrease in the flank area of the steel electrode with increasing the filling ratio of ESR. The present author claims that the smaller flank area of the steel electrode lowers the amount of the oxide scale generated on the steel electrode surface during the ESR process, which lowers the oxygen pickup of liquid steel resulting from FeO in the slag. Unfortunately, the FeO content of the slag is not provided in the article of Liu et al. [104]. Fig. 3.22 Relationship between the filling ratios of ESR and the oxygen content of the remelted ingots when using CaF2 –Al2 O3 slag in the ESR process [104]

66

3 Deoxidation of ESR and Its Correlation with Oxide Inclusions

3.6 Summary Although there are controversial findings on whether the oxygen content of remelted ingot is dependent on the oxygen content of the steel electrode or not, keeping a low oxygen content of the steel electrode is quite necessary for ultralow oxygen steel production. As for protective atmosphere of ESR, the oxygen concentration in the gas atmosphere should be detected online to keep an effective inert atmosphere. The oxygen potential of the slag has to be minimized to prevent the reoxidation of liquid steel even if protective inert atmosphere is employed for ESR. Thoroughly removing the oxide scale on the electrode surface and adding deoxidizing agents for slag deoxidation are indispensable operations in ESR practice for producing low and ultralow oxygen steel. It is suggested that the addition rate of deoxidizing agents for successful deoxidation of a particular ESR practice should be based on the change in the oxygen potential of the slag and soluble oxygen content of liquid steel. The melting rates exert a negligible effect on the oxygen content in the case of protective atmosphere ESR of the ultra low oxygen steel, unlike the remelting of high oxygen steel. Filling ratio of ESR has a minor influence on the oxygen content of remelted ingot. Deoxidation of ESR for heavy ingot production is still a challenging issue. Monitoring the change in the slag chemistry during ESR is indispensable for designing proper types and addition rates of deoxidizing agents. Oxide inclusion removal during the ESR process largely determines the oxygen content of remelted steel. The studies on the evolution trajectories of different types of oxide inclusions and its correlation with soluble oxygen and total oxygen of the steel during ESR process are still in its infancy. For example, the dissolution of oxide inclusions or adsorption of oxide inclusions by molten slag at different stages of ESR process, and the respective contribution to oxide inclusions removal in these ways have to be ascertained. Quantitative analysis of oxide inclusion removal in different trajectories will remain a challenge. It will provide guidance for designing the deoxidation schemes for refining liquid steel in the production of the consumable electrode for ESR.

References 1. Reyes-Carmona F, Mitchell A. Deoxidation of ESR slags[J]. ISIJ Int., 1992, 32(4): 529–537. 2. Mitchell A, Reyes-Carmona F, Samuelsson E. The deoxidation of low-alloy steel ingots during ESR[J]. Trans. ISIJ, 1984, 24(7): 547–556. 3. Chatterjee M, Balasubramanian M S N, Gupt K M, et al. Inoculation during electroslag remelting of 15CDV6 steel[J]. Ironmak. Steelmak., 1990, 17(1): 38–42. 4. Mitchell A, Carmona F R, Wei C. Deoxidation in the electroslag process[J]. Iron Steelmaker, 1982, 9(3): 37–41. 5. Shi C B, Chen X C, Guo H J, et al. Assessment of oxygen control and its effect on inclusion characteristics during electroslag remelting of die steel[J]. Steel Res. Int., 2012, 83(5): 472– 486.

References

67

6. Radwitz S, Scholz H, Friedrich B, et al. Proc. 2015 Int. Symp. Liq. Met. Process. Cast., Proc. 2015 Int. Symp. on Liquid Metal Processing and Casting, The Minerals, Metals and Materials Society, Warrendale, PA, 2015, pp. 153–162. 7. Kay D A R, Mitchell A, Ram M. Thermodynamic properties of iron oxides in CaF2 -CaO-Fet O electroslag fluxes[J]. J. Iron Steel Inst., 1970, 208(2): 141–146. 8. Xiang D, Zhu X, Wang K. Influence of deoxidation on controlling low Al during ESR process[J]. J. Iron Steel Res., 1989, 1(4): 9–14. (in Chinese). 9. Wen P, Xiang D. Research on deoxidization technology of electroslag remelting process for heavy ingot with high quality[J]. Heavy Casting and Forging, 2016, (1): 23–25. (in Chinese). 10. Hou D, Jiang Z, Dong Y, et al. Effect of slag composition on the oxidation kinetics of alloying elements during electroslag remelting of stainless steel: part-1 mass-transfer model[J]. ISIJ Int., 2017, 57(8): 1400–1409. 11. Shi C B, Wang H, Li J. Effects of reoxidation of liquid steel and slag composition on the chemistry evolution of inclusions during electroslag remelting[J]. Metall. Mater. Trans. B, 2018, 49(4): 1675–1689. 12. Shi C B, Chen X C, Guo H J, et al. Control of MgO·Al2 O3 spinel inclusions during protective gas electroslag remelting of die steel[J]. Metall. Mater. Trans. B, 2013, 44(2): 378–389. 13. Shi C B, Yu W T, Wang H, et al. Simultaneous modification of alumina and MgO·Al2 O3 inclusions by calcium treatment during electroslag remelting of stainless tool steel[J]. Metall. Mater. Trans. B, 2017, 48(1): 146–161. 14. Shi C B, Chen X C, Guo H J. Characteristics of inclusions in high-Al steel during electroslag remelting process[J]. Int. J. Miner. Metall. Mater., 2012, 19(4): 295–302. 15. Shi C B, Zheng D L, Guo B S, et al. Evolution of oxide–sulfide complex inclusions and its correlation with steel cleanliness during electroslag rapid remelting (ESRR) of tool steel[J]. Metall. Mater. Trans. B, 2018, 49(6): 3390–3402. 16. Li S, Cheng G, Miao Z, et al. Evolution of oxide inclusions in G20CrNi2Mo carburized bearing steel during industrial electroslag remelting[J]. ISIJ Int., 2018, 58(10): 1781–1790. 17. Paar A, Schneider R, Zeller P, et al. Effect of electrical parameters on type and content of non-metallic inclusions after electro-slag-remelting[J]. Steel Res. Int., 2014, 85(4): 570–578. 18. Sasabe M, Goto K S. Permeability, diffusivity, and solubility of oxygen gas in liquid slag[J]. Metall. Trans., 1974, 5(10): 2225–2233. 19. Sasabe M, Kinoshita Y. Permeabilities of oxygen through molten CaO–SiO2 –Al2 O3 system with Fe2 O3 or CaF2 [J]. Tetsu–to–Hagané, 1979, 65(12): 1727–1736. (in Japanese). 20. Wei J, Liu Z. Study on oxygen transfer through molten slags of CaF2 +Al2 O3 and CaF2 +Al2 O3 +CaO systems for ESR[J]. Acta Metall. Sin., 1994, 30(8): 350–360. (in Chinese). 21. Masui A, Sasajima Y, Sakata N, et al. Some important factors affecting hydrogen pick-up and oxidation during ESR treatment[J]. Tetsu-to-Hagané, 1977, 63(13): 2181–2190. (in Japanese). 22. The Japan Society for the Promotion of Science, The 19th Committee on Steelmaking. Steelmaking Data Sourcebook[M], Gordon and Breach Science Publishers, New York, NY, 1988. 23. Wei J. Mass-transfer model analysis of ESR process on industrial unit[J]. Acta Metall. Sin., 1987, 23(3): 126–134. (in Chinese). 24. Wei J. Oxidation of alloying elements during ESR of stainless steel[J]. China J. Met. Sci. Technol., 1989, 5(4): 235–246. 25. Li S, Cheng G, Miao Z, et al. Effect of slag on oxide inclusions in carburized bearing steel during industrial electroslag remelting[J]. Int. J. Miner. Metall. Mater., 2019, 26(3): 291–300. 26. Andersson M A T, Jönsson P G, Nzotta M M. Application of the sulphide capacity concept on high-basicity ladle slags used in bearing-steel production[J]. ISIJ Int., 1999, 39(11): 1140– 1149. 27. Wagner C. Thermodynamics of Alloys[M]. Cambridge: Addison-Wesley Press, 1952, 51. 28. Ohta H, Suito H. Activities in CaO-SiO2 -Al2 O3 slags and deoxidation equilibria of Si and Al[J]. Metall. Mater. Trans. B, 1996, 27(6): 943–953. 29. Ohta H, Suito H. Calcium and magnesium deoxidation in Fe-Ni and Fe-Cr alloys equilibrated with CaO-Al2 O3 and CaO-Al2 O3 -MgO slags[J]. ISIJ Int., 2003, 43(9): 1293–1300.

68

3 Deoxidation of ESR and Its Correlation with Oxide Inclusions

30. Park J H, Todoroki H. Control of MgO·Al2 O3 spinel inclusions in stainless steels[J]. ISIJ Int., 2010, 50(10): 1333–1346. 31. Itoh H, Hino M, Ban-ya S. Assessment of Al deoxidation equilibrium in liquid iron[J]. Tetsu– to–Hagané, 1997, 83(12): 773–778. 32. Wei J H, Mitchell A. Changes in composition during A. C. ESR – I. theoretical development[J]. Acta Metall. Sin., 1984, 20(5): 261–279. 33. Fraser M E, Mitchell A. Mass transfer in the electroslag process: part 1 mass-transfer model[J]. Ironmak. Steelmak., 1976, 3(5): 279–287. 34. Mitchell A, Szekely J, Elliott J F. Application of mathematical modelling to the ESR process[C]. Electroslag refining: proceedings of a conference on electroslag refining[M]. London: The Iron and Steel Institute, 1973, 3–15. 35. Shi C B. Behaviour and control technique of oxygen and inclusions during protective gas electroslag remelting process[D]. University of Science and Technology Beijing, Beijing, 2012, 31–33. 36. Shi C B, Chen X C, Luo Y W, et al. Materials Processing Fundamentals[J]. The Minerals, Metals and Materials Society, Warrendale, PA, 2013, 31. 37. Hawkins R J, Swinden D J, Pocklington D N. Relevance of laboratory experiments to the control of composition in production-scale ESR[C]. Electroslag refining: proceedings of a conference on electroslag refining. London: The Iron and Steel Institute, 1973, 21–34. 38. Etienne M, Mitchell A. Slag-metal transfer in electroslag melting[C]. 28th Electric Furnace Conference Proceeding, Iron and Steel Soc of AIME, Warrendale, PA, 1970, 28–32. 39. Fraser M E, Mitchell A. Mass transfer in the electroslag process: part 2 mass-transfer cofficients[J]. Ironmak. Steelmak., 1976, vol. 3(5): 288–301. 40. Wei C. Application of the mass transfer model to ESR process of high alloy steel[J]. J. Xian Inst. Metall. Construction Eng., 1985, 44(4): 69–80. (in Chinese). 41. Hou D, Jiang Z H, Dong Y, et al. Mass transfer model of desulfurization in the electroslag remelting process[J]. Metall. Mater. Trans. B, 2017, 48(3): 1885–1897. 42. Shi C B , Guo H J, Chen X C, et al. Kinetic study on oxygen transfer in gas shielded electroslag remelting process[J]. Special Steel, 2013, 34(2): 11–15. (in Chinese). 43. Wei J H, Mitchell A. Changes in composition during A. C. ESR—II. laboratory results and analysis[J]. Acta Metall. Sin., 1984, 20(5): 280–287. (in Chinese). 44. Hua Y X. Introduction to kinetics of metallurgical processes[M]. Beijing: Metallurgical Industry Press, 2004. 131–132. 45. Cooper C K, Ghosh D, Kay D A R, et al. ESR reaction sites[C]. 28th Electric Furnace Conf. Proc., Iron and Steel Society of AIME, Warrendale, PA, 1970, 8. 46. Mellberg P and Sandberg H. On the mechanism of desulphurization during A.C. and D.C. electroslag remelting[J]. Scand. J. Metall., 1973, 2(3): 121–124. 47. Holzgruber W, Petersen K, Schneider P E. Comparison of products and economics from large VAR, ac-ESR and dc-ESR ingots[C]. Trans. Int. Vac. Metal. Conf., 1968, 499–523. 48. Plöckinger E. Electroslag remelting-a modern tool in metallurgy[J]. J. Iron Steel Inst., 1973, 211(8): 533–541. 49. Klyuev M M, Spitzberg V M. Removal and formation of non-metallic inclusions in metal produced by electro-slag remelting[J]. Stahl (in English), 1969, 2(2): 168–171. 50. Kato M, Hasegawa K, Nomura S, et al. Transfer of oxygen and sulfur during direct current electroslag remelting[J]. Trans. ISIJ, 1983, 23(7): 618–627. 51. Medina S F, López F, Coedo A G. Manufacture by electroslag remelting of microalloyed steels with two precipitating metal elements[J]. Ironmak. Steelmak., 1997, 24(4): 329–336. 52. Medina S F, Andres M P. Electrical field in the resistivity medium (slag) of the ESR process: influence on ingot production and quality[J]. Ironmak. Steelmak., 1987, 14(3): 110–121. 53. Kawakami M, Nagata K, Yamamura M, et al. Profiles of temperature, voltage and local heat generation in slag phase and metal pool of ESR unit under operation[J]. Tetsu-to-Hagané, 1977, 63(13): 2162–2171. 54. Medina S F, Cores A. Thermodynamic aspects in the manufacturing of microalloyed steels by the electroslag remelting process[J]. ISIJ Int., 1993, 33(12): 1244–1251.

References

69

55. Wang C S, Liu S G, Xu M D, et al. The oxygen content of electroslag remelted GCr15 steel was reduced[J]. Special Steel, 1997, 18(3): 31–35. (in Chinese). 56. Chang L Z, Yang H S, Li Z B. Study on oxygen behavior in electroslag remelting process[J]. Steelmak., 2010, 26(5): 46–50. (in Chinese). 57. Li J J, Zhang J W. Measures to reduce oxygen content of 30CrNiMo8 by electroslag remelting in unprotected atmosphere[J]. Shanxi Metall., 2019, (2): 52–61. (in Chinese). 58. Zhou D G, Xu W G, Wang P, et al. Study on oxygen control and action during electroslag remelting of bearing steel[J]. Iron Steel, 1998, 33(3): 13–17 (in Chinese). 59. Schneider R S E, Molnar M, Gelder S, et al. Effect of the slag composition and a protective atmosphere on chemical reactions and non-metallic inclusions during electroslag remelting of a hot-work tool steel[J]. Steel Res. Int., 2018, 89(10): 1800161. 60. Padki G M, Balasubramanian MSN, Gupt K M, et al. Effect of ESR and carbon content on mechanical properties of a high-strength steel (15CDV6) for aerospace applications[J]. Ironmak. Steelmak., 1983, 10(4): 180–184. 61. Shi C B, Park J H. Evolution of oxide inclusions in Si-Mn-killed steel during protective atmosphere electroslag remelting[J]. Metall. Mater. Trans. B, 2019, 50(3): 1139–1147. 62. Persson E S and Mitchell A. Differences in inclusion morphology between ESR remelted steel with and without tracer in the slag[C]. Proc. Liq. Met. Process. Cast. Conference 2017. The Minerals, Metals & Materials Society, Philadelphia, USA, 2017, 373–379. 63. Persson E S, Karasev A, Jönsson P. Studies of three dimension inclusions from ESR remelted and conventional cast steel[J]. Liquid Metal Processing and Casting Conf. 2017. The Minerals, Metals & Materials Society, Pittsburgh, PA, 2017, 353–360. 64. Li Z, Zhang J, Che X. Control of content and composition of nonmetallic inclusion in ESR steel[J]. J. Iron Steel Res., 1997, 9(2): 7–12. (in Chinese). 65. Ahmadi S, Arabi H, Shokuhfar A, et al. Evaluation of the electroslag remelting process in medical grade of 316L stainless steel[J]. J. Mater. Sci. Technol., 2009, 25(5): 592–596. 66. Mitchell A. Oxide inclusion behaviour during consumable electrode remelting[J]. Ironmak. Steelmak., 1974, 1(3): 172–179. 67. Baligidad R G, Prakash U, Rao V R, et al. Electroslag remelting of Fe-28 at.-%Al intermetallic alloy[J]. Ironmak. Steelmak., 1994, 21(4): 324–331. 68. Kang Y B, Lee H G. Inclusions chemistry for Mn/Si deoxidized steels: thermodynamic predictions and experimental confirmations[J]. ISIJ Int., 2004, 44(6): 1006–1015. 69. Wang K, Jiang M, Wang X, et al. Formation mechanism of SiO2 -type inclusions in Si-Mnkilled steel wires containing limited aluminum content[J]. Metall. Mater. Trans. B, 2015, 46(5): 2198–2207. 70. García-Carbajal A, Herrera-Trejo M, Castro-Cedeño E I, et al. Characterization of inclusion populations in Mn-Si deoxidized steel[J]. Metall. Mater. Trans. B, 2017, 48(6): 3364–3379. 71. Park J S, Park J H. Effect of slag composition on the concentration of Al2 O3 in the inclusions in Si-Mn-killed steel[J]. Metall. Mater. Trans. B, 2014, 45(3): 953–960. 72. Fu J. An investigation of mechanism on the removal of oxide inclusions during ESR process[J]. Acta Metall. Sin., 1979, 15(4): 526–539. (in Chinese). 73. Detrois M, Jablonski P D, Hawk J A. Evolution of tantalum content during vacuum induction melting and electroslag remelting of a novel martensitic steel[J]. Metall. Mater. Trans. B, 2019, 50(4): 1686–1695. 74. Greene C H, Gaffney R F. Apparatus for measuring the rate of absorption of a bubble in glass[J]. J. Amer. Ceram. Soc., 1959, 42(6): 271–275. 75. Sasabe M, Kinoshita Y. Permeability of oxygen through molten CaO-SiO2 -Al2 O3 , Fe2 O3 CaO-SiO2 -Al2 O3 and CaF2 -CaO- SiO2 -Al2 O3 Systems[J]. Trans. ISIJ, 1980, 20(12): 801– 809. 76. Sasabe M, Kitamura S. Transport phenomena of oxygen through molten CaO-SiO2 system containing zinc and/or nickel oxide[J]. ISIJ Int., 1993, 33(1): 133–139. 77. Sasabe M, Kinoshita Y. Permeability, diffusivity and solubility of oxygen gas in molten CaO-SiO2 -Al2 O3 system[J]. Tetsu-to-Hagané, 1978, 64(9): 1313–1322. (in Japanese).

70

3 Deoxidation of ESR and Its Correlation with Oxide Inclusions

78. Holzgruber W. Overview of 50 years of development in electroslag remelting in Austria[J]. BHM Berg-und Hüttenmänn. Monatsh, 2016, 161(Suppl 1): 2–11. 79. Chang L Z, Yang H S, Li Z B. Effect of argon protection on quality of electroslag remelting ingot of low alloy steel[J]. Special Steel, 2009, 30(4): 59–60. (in Chinese). 80. Zheng D L, Li J, Shi C B, et al. Evolution of oxide–sulfide complex inclusions in extralow-oxygen steel during electroslag remelting[C]: Proc. 2019 Annual Meeting for Secondary Refining, The Chinese Society for Metals, Beijing, 2019, pp. 528–536. (in Chinese). 81. Sekiya A, Nakayama S, Taketsuru T. Deoxidation of stainless steel on vacuum electroslag remelting process[J]. Denki-Seiko, 1995, 66(1): 47–53. (in Japanese). 82. Huang X, Li B, Liu Z, et al. Oxygen transport behavior and characteristics of nonmetallic inclusions during vacuum electroslag remelting[J]. Vacuum, 2019, 164(6): 114–120. 83. Stein G, Menzel J. Nitrogen alloyed steels – a new generation of materials with extraordinary properties[J]. Int. J. Mater. Prod. Technol., 1995, 10: 290–302. 84. Pant P, Dahlmann P, Schlump W, et al. A new nitrogen alloying technique - a way to distinctly improve the properties of austenitic steel[J]. Steel Res., 1987, 58(1): 18–25. 85. Yu J, Liu F B, Li H B, et al. Numerical simulation and experimental investigation of nitrogen transfer mechanism from gas to liquid steel during pressurized electroslag remelting process[J]. Metall. Mater. Trans. B, 2019, 50(6): 3112–3124. 86. Narita K, Onoye T, Ishii T, et al. A metallurgical study of oxide-base slag used for electroslag remelting[J]. Tetsu-to-Hagané, 1978, 64(10): 1568–1577. (in Japanese). 87. Shi C B, Chen X C, Guo H J. Oxygen control and its effect on steel cleanliness during electroslag remelting of NAK80 die steel[C]. AISTech 2012 Proc., Association for Iron & Steel Technology, Warrendale, PA, 2012, 947–957. 88. Yoshinao K and K. Yasuyuki: Creation of high nitrogen-bearing stainless steel by pressurized ESR method with aluminum deoxidation[J]. Proceedings of International Conference on High Nitrogen Steels 2006, The Chinese Society for Metals, Jiuzhaigou, Sichuan, 2006, pp. 360– 365. 89. Shao Q L. Electroslag remelting W6Mo5Cr4V2Al high speed steel[J]. Hebei Metall., 1999, 5: 48–49. (in Chinese). 90. Shi C B, Chen X C, Wang F, et al. Steel cleanliness control during protective gas electroslag remelting of H13 die steel[J]. Mater. Sci. & Technol. 2012 Conference and exhibition Proceedings. The Minerals, Metals and Materials Society, Warrendale, PA, 2012, 1261–1268. 91. Pateisky G, Biele H, Fleischer H J. The reactions of titanium and silicon with Al2 O3 -CaOCaF2 slags in the ESR process[J]. J. Vac. Sci. Tech., 1972, 9(6): 1318–1322. 92. Wang A R, Yu Y G, Fu J. Control of magnesium content in CrNiMoV steel during the ESR process[J]. J. Univ. Sci. Technol. Beijing, 1997, 19(2): 134–137. (in Chinese). 93. Zhang L J, Wang C T, Jiang R L, et al. Control measures of composition change in electroslag remelting T8MnA process[J]. Jiangxi Metall., 1995, 15(2): 6–8. (in Chinese). 94. Hou D, Jiang Z, Dong Y, et al. Effect of slag composition on the oxidation kinetics of alloying elements during electroslag remelting of stainless steel: part-2 control of titanium and aluminum content[J]. ISIJ Int., 2017, 57(8): 1410–1419. 95. Hou D, Liu F, Qu T, et al. Behavior of alloying elements during drawing-ingot-type electroslag remelting of stainless steel containing titanium[J]. ISIJ Int., 2018, 58(5): 876–885. 96. Wang A R, Yu Y G and Fu J. Aluminum control in CrNiMoV steel during electroslag remelting process[J]. J. Univ. Sci. Technol. Beijing, 1998, 5(2): 87-–9. 97. Mehrabi K, Rahimipour M R, Shokuhfar A. The effect of slag types and melting rate on electro-slag remelting (ESR) processing[J]. Int. J. Iron Steel Soc. Iran, 2005, 2(1): 37–42. 98. Shi C B, Li J, Cho J W, et al. Effect of SiO2 on the crystallization behaviors and in-mold performance of CaF2 -CaO-Al2 O3 slags for drawing-ingot-type electroslag remelting[J]. Metall. Mater. Trans. B, 2015, 46(5): 2110–2120. 99. Dong Y W, Jiang Z, Cao Y, et al. Effect of slag on inclusions during electroslag remelting process of die steel[J]. Metall. Mater. Trans. B, 2014, 45(4): 1315–1324. 100. Song M, Nzotta M, Sichen D. Study of the formation of non-metallic inclusions by ladle glaze and the effect of slag on inclusion composition using tracer experiments[J]. Steel Res. Int., 2009, 80(10): 753–760.

References

71

101. Song M, Nzotta M, Sichen D. Effect of ladle slag of previous heat on number of non-metallic inclusions in ladle treatment of tool steel[J]. Ironmak. Steelmak., 2011, 38(8): 584–589. 102. Rocha V C, Pereira J A M, Yoshioka A, et al. Evaluation of secondary steelmaking slags and their relation with steel cleanliness[J]. Metall. Mater. Trans. B, 2017, 48(3): 1423–1432. 103. Huang X, Li B, Z. Liu. Three-dimensional mathematical model of oxygen transport behavior in electroslag remelting process[J]. Metall. Mater. Trans. B, 2018, 49(2): 709–722. 104. Liu S G, Xu M D, Liu F X, et al. Oxygen and its control in electroslag remelting of high carbon chromium bearing steel[J]. Special Steel, 1993, 14(3): 45–48. (in Chinese). 105. Allibert M, Wadier J F, Mitchell A. Use of SiO2 -containing slags in electroslag remelting[J]. Ironmak. Steelmak., 1978, 5(5): 211–216. 106. Oguti Y, Tanbe Y, Miyama S, et al. Temperature Distributions in the slag and metal pools in a laboratory-scale ESR furnace[J]. Tetsu-to-Hagané, 1977, 63(13): 2152–2161. 107. Fu J, Chen C, Chen E, et al. Temperature distribution in slag pool during electroslag remelting[J]. Acta Metall. Sin., 1979, 15(1): 44–50. 108. Choudhary M, Szekely J. The modeling of pool profiles, temperature profiles and velocity fields in ESR systems[J]. Metall. Trans. B, 1980, 11(3): 439–453. 109. Dong Y, Hou Z, Jiang Z, et al. Study of a single-power two-circuit ESR process with currentcarrying mold: mathematical simulation of the process and experimental verification[J]. Metall. Mater. Trans. B, 2018, 49(1): 349–360. 110. Ono-Nakazato H, Taguchi K, Maruo R, et al. Silicon deoxidation equilibrium of molten Fe–Mo alloy[J]. ISIJ Int., 2007, 47(3): 365–369. 111. Park J H, Lee S B, Kim D S, et al. Thermodynamics of titanium oxide in CaO–SiO2 –Al2 O3 – MgOsatd –CaF2 slag equilibrated with Fe–11mass%Cr melt[J]. ISIJ Int., 2009, 49(3): 337–342. 112. Todoroki H, Mizuno K. Effect of silica in slag on inclusion compositions in 304 stainless steel deoxidized with aluminum[J]. ISIJ Int., 2004, 44(8): 1350–1357. 113. Verma N, Pistorius P C, Fruehan R J, et al. Calcium modification of spinel inclusions in aluminum-killed steel: reaction steps[J]. Metall.Mater. Trans. B, 2012, 43(4): 830–840. 114. Eisenhüttenleute V D. Slag Atlas[M]. 2nd ed., Cambridge: Woodhead Publishing Limited, 1995, 44. 115. Fujimura H, Tsuge S, Komizo Y, et al. Effect of oxide composition on solidification structure of Ti added ferritic stainless steel[J]. Tetsu–to–Hagané, 2001, 87(11): 707–712. 116. Park J S, Lee C H, Park J H. Effect of complex inclusion particles on the solidification structure of Fe-Ni-Mn-Mo alloy[J]. Metall. Mater. Trans. B, 2012, 43(60): 1550–1564. 117. Shi C B, Zhu Q T, Yu W T, et al. Effect of oxide inclusions modification during electroslag remelting on primary carbides and toughness of a high-carbon 17 mass% Cr tool steel[J]. J. Mater. Eng. Perform., 2016, 25(11): 4785–4795. 118. Suito H, Inoue R. Thermodynamics on control of inclusions composition in ultraclean steels[J]. ISIJ Int., 1996, 36(5): 528–536. 119. Yoshioka T, Shimamura Y, Karasev A, et al. Mechanism of a CaS formation in an Al-killed high-S containing steel during a secondary refining process without a Ca-treatment[J]. Steel Res. Int., 2017, 88(10): 1700147. 120. Kang Y J, Li F, Morita K, et al. Mechanism study on the formation of liquid calcium aluminate inclusion from MgO-Al2 O3 spinel[J]. Steel Res. Int., 2006, 77(11): 785–792. 121. Hara S, Hashimoto H, Ogino K. Electrical conductivity of molten slags for electro-slag remelting[J]. Trans. ISIJ, 1983, 23(12): 1053–1058. 122. Chang L Z, Shi X F, Cong J Q. Study on mechanism of oxygen increase and countermeasure to control oxygen content during electroslag remelting process[J]. Ironmak. Steelmak., 2014, 41(3): 182–186. 123. Shi C B, Shin S H, Zheng D L, et al. Development of low-fluoride slag for electroslag remelting: role of Li2 O on the viscosity and structure of the Slag[J]. Metall. Mater. Trans. B, 2016, 47(6): 3343–3349. 124. Nakada H, Nagata K. Crystallization of CaO–SiO2 –TiO2 slag as a candidate for fluorine free mold flux[J]. ISIJ Int., 2006, 46(3): 441–449.

72

3 Deoxidation of ESR and Its Correlation with Oxide Inclusions

125. Takahira N, Hanao M, Tsukaguchi Y. Viscosity and Solidification Temperature of SiO2 –CaO– Na2 O Melts for Fluorine Free Mould Flux[J]. ISIJ Int., 2013, 53(5): 818–822. 126. Klug J L, Hagemann R, Heck N C, et al. Fluorine-free mould powders for slab casting: crystallization control in the CaO-SiO2 -TiO2 -Na2 O-Al2 O3 system[J]. Steel Res. Int., 2012, 83(12):1186–1193. 127. Persson M, Seetharaman S, Seetharaman S. Kinetic sudies of fuoride eaporation from Slags[J]. ISIJ Int., 2007, 47(12): 1711–1717. 128. Omoto T, Iwamoto Y, Yamaji H. Development of environment friendly mold powder[J]. Shinagawa Tech. Rep., 2002, 45: 85–92. 129. Xiang D L. Some Problems Meriting Attention in Large-scale ESR[J]. Heavy Casting Forging, 2011, (1): 26–35. (in Chinese). 130. Radwitz S, Scholz H, B. Friedrich, et al. Influencing the electroslag remelting process by varying fluorine content of the utilized slag[C]. Proc. Eur. Metall. Conference 2015, vol. 2, GDMB Society of Metallurgists and Miners, Düsseldorf, Germany, 2015, 887–896. 131. Li S J, Cheng G G, Miao Z Q, et al. Kinetic analysis of aluminum and oxygen variation of G20CrNi2Mo bearing steel during industrial electroslag remelting process[J]. ISIJ Int., 2017, 57(12): 2148-2156. 132. Wang H, Shi C M, Li J, et al. Evolution of CaO–MgO–Al2 O3 –CaS–(SiO2 ) inclusions in H13 die steel during electroslag remelting process[J]. Ironmak. Steelmak., 2018, 45(1): 6–16. 133. Hernandez-Morales B, Mitchell A. Review of mathematical models of fluid flow, heat transfer, and mass transfer in electroslag remelting process[J]. Ironmak. Steelmak., 1999, 26(6): 423– 438. 134. Rao L, Zhao J H, Zhao Z X, et al. Macro- and microstructure evolution of 5CrNiMo steel ingots during electroslag remelting process[J]. J. Iron Steel Res. Int., 2014, 21(7): 644–652. 135. Kay D A R, Pomfret R J. Removal of oxide inclusions during AC electroslag remelting[J]. J. Iron Steel Inst., 1971, 209(12): 962–965. 136. Fu J, Zhu J. Change of oxide inclusions in electroslag remelting process[J]. Acta Metall. Sin., 1964, 7(3): 250–262. (in Chinese). 137. Li Z B, Zhou W H, Li Y D. Mechanism of inclusion removal by electroslag remelting[J]. Iron Steel, 1980, 15(1): 20–26. (in Chinese). 138. Zhou D G, Chen X C, Fu J, et al. Inclusions in bearing steel produced by electroslag remelting and continuous casting[J]. J. Univ. Sci. Technol. Beijing, 2000, 22(1): 26–30. (in Chinese).

Chapter 4

Reoxidation of Liquid Steel During ESR and Its Effect on Oxide Inclusions

Abstract This chapter presents the findings of the reoxidation of liquid steel during protective atmosphere electroslag remelting of the steel with a low oxygen content. The formation and transformation of oxide inclusions are closely related to the soluble oxygen content of liquid steel. The effect of liquid steel reoxidation during ESR on the oxide inclusion chemistry is also presented. To reveal the inclusion evolution during ESR process, the inclusions in the samples taken from the steel electrode, liquid metal pool during ESR process, and remelted ingot are characterized. Further, thermodynamic modelling is established to study the interactions of gas-slag-metalinclusion phases. The experimental determination and thermodynamic considerations are employed to elucidate the evolution mechanism of non-metallic inclusions in steel during the ESR process.

4.1 Background In steelmaking processing, high requirements are imposed on the steel cleanliness. Various issues should be taken into account for clean steel production, such as deoxidation scheme optimization, reoxidation prevention, slag composition. The oxygen content of the steel and alloy is dramatically reduced by ESR generally. However, this is not always the case, the increase in the oxygen content of the steel after ESR has been validated by many studies [1–3]. The reoxidation of liquid steel is a common practice during electroslag remelting. With the advances in the steelmaking technologies in recent a few decades, the oxygen content of steel usually could be lowered to a low level (a few parts per million) during secondary refining of liquid steel for producing the electrode for ESR. For low (or ultralow) oxygen steel and alloy production, further reducing the oxygen content of the low oxygen electrode by ESR is a tight task, even if a protective atmosphere is employed throughout the ESR process. Unlike the reoxidation of the liquid steel in other steelmaking process operations, in which the reoxidation may occur by several reasons, such as air exposure from open eyes in the ladle and tundish [4, 5] and reaction with refractories [6], the reoxidation of liquid steel during ESR is dependent in a large part on the oxygen © Metallurgical Industry Press 2023 C. Shi et al., Electroslag Remelting Towards Clean Steel, https://doi.org/10.1007/978-981-99-3257-3_4

73

74

4 Reoxidation of Liquid Steel During ESR and Its Effect on Oxide Inclusions

potential of the slag [7], besides the dissociation of original oxide inclusions [8] and the decomposition of components from slag [3], as well as permeation of atmospheric oxygen directly through molten slag into liquid metal pool by diffusion of physically dissolved oxygen which has been confirmed to play a minor role in soluble oxygen pickup [9, 10]. The oxygen potential of the ESR-type slag is typically determined by the iron oxide activity of the slag (in some cases, MnO is present in the slag, but in a quite small fraction). The reoxidation of liquid steel always takes place during ESR of low-oxygen steel because of a big difference in the oxygen potential between liquid steel and the slag phase, as well as gas phase, even if protective inert atmosphere is employed throughout the ESR process [11, 12]. Reoxidation of liquid steel is a major problem in generating oxide inclusions [13–15]. Reoxidation also results in consumption of alloying elements such as Al, Mn and Si in liquid steel, leading to oxide inclusion formation [16]. The reoxidation of liquid steel during ESR leads to considerable soluble oxygen pickup, which provides a driving force for the generation of fresh oxide inclusions. Some of the reoxidation results did indeed show that the amount of oxide inclusions did not decrease after ESR, but increased [13]. The results from open air atmosphere ESR trials show that the oxygen content of hot work tool steel almost doubled to 0.0024 mass%, leading to the formation of a large amount of new Al2 O3 inclusions [13]. In addition to generation of fresh inclusions by reoxidation, the reoxidation of liquid steel during ESR could contribute to modification of the original oxide inclusion chemistry (resulting in their full liquefaction) and the increase in its size. The soluble oxygen supplied from the reoxidation of liquid steel and concerned elements in liquid steel react with the original semiliquid CaO–Al2 O3 –MgO inclusions that had not been removed in ESR process. With the progress of the transformation reaction, the MgO content of the oxide inclusion is diluted continuously, and consequently full liquid CaO–Al2 O3 –MgO–SiO2 inclusions with a larger size are generated [12]. The reoxidation of liquid steel during conventional ESR and protective inert atmosphere ESR has become a normal phenomenon for low and ultralow oxygen steel production. Reoxidation of liquid steel during protective argon gas atmosphere ESR that gives rise to generation of new oxide inclusions and modification of CaO–Al2 O3 –MgO inclusions has been confirmed in Ref. [11, 12]. More work is still needed to clarify the role of the reoxidation of liquid steel during ESR on the inclusions possessing different compositions and sizes together with its association with the compositions steel and slag, as well as reveal its mechanism according to thermodynamics and kinetics.

4.2 Experimental Work

75

4.2 Experimental Work The steel with a low oxygen content was used as the consumable electrode for protective atmosphere ESR. The consumable electrode was produced as follows: BOF → Ladle furnace (LF) refining → Ruhrstahl Heraeus (RH) refining → continuous casting. The billets produced by continuous casting were forged into steel rod used as consumable electrodes for ESR. The chemical composition of the consumable electrode is listed in Table 4.1. The oxide scale on electrode steel surface was basically removed mechanically prior to ESR experiments. The pre-melted slag (29.3 mass% CaF2 , 30.5 mass% CaO, 34.5 mass% Al2 O3 , 3.8 mass% MgO, 1.9 mass% SiO2 ) was roasted at 973 K (700 °C) in an electrical resistance furnace for 8 h to remove the moisture in the slag before ESR trials. The whole ESR process was conducted in protective argon atmosphere. During ESR refining, a steel sample was taken from the liquid metal pool in mold using a vacuum sampling tube that was made of quartz (6 mm in inner diameter), followed by quenched in water. The sampling from liquid metal pool was performed at the time when about two thirds of the whole refining process was finished. The temperature in liquid metal pool is in gradient distribution (typically varying around 1873 K (1600 °C)) not only along radial direction but also along axial direction, which strongly affects the shape and size of the liquid metal pool and consequently influences the solidification quality of as-cast ingot [17–20]. Metallographic samples were taken from the consumable electrode and midheight of each as-cast ESR ingot at the mid-radius position. These steel samples were mechanically ground by silicon carbide papers, and polished using diamond paste. Inclusions exposed on the cross section of the polished steel sample were analyzed in terms of their chemistry, size and morphology by scanning electron microscope (SEM, FEI Quanta-250; FEI Corporation, Hillsboro, OR) equipped with energy dispersive X-ray spectrometer (EDS, XFlash 5030; Bruker, Germany). To reveal the transient evolution of inclusions in ESR process, the steel samples collected from liquid metal pool by vacuum sampling tubes were mounted with epoxy resin, polished, and then examined by SEM–EDS. Furthermore, quantitative analysis of the chemistry of inclusions in consumable electrode, the steel samples taken from liquid metal pool and ESR ingots was performed using electron probe microanalyzer (EPMA, 1720, SHIMADZU, Japan). The instrument was operated at the acceleration voltage of 10 kV. For each steel sample, approximately 100 inclusions were randomly selected to characterize their chemistry, size, and morphology. Table 4.1 Chemical composition of the consumable steel electrode (mass%) C

Si

Mn

Cr

Mo V

Ti

Ca

Mg

Al

O

S

N

0.40 1.040 0.35 5.10 1.2 0.96 0.0078 0.0019 0.0002 0.0099 0.0018 0.0032 0.0078

76

4 Reoxidation of Liquid Steel During ESR and Its Effect on Oxide Inclusions

4.3 Oxygen Content of the Steel After protective atmosphere ESR, the total oxygen content increased from 18 ppm in the steel electrode to 34 ppm in the remelted ingot. The difference in the oxygen potential between liquid steel with ultralow oxygen content and slag phase, as well as gas phase contributed to reoxidation of liquid steel during ESR. In view of the fact that part of the original oxide inclusions were removed during ESR refining (the detailed discussion will be presented in Sect. 3.3), the oxygen level contributed by the reoxidation of liquid steel virtually is higher than the difference in the measured oxygen contents between consumable electrode and remelted ingot. The introduction of oxygen from atmosphere and FeO (the newly-formed during on-going ESR, and oxide scale that not being removed from electrode steel surface) could hardly be prevented during ESR, which were the sources of oxygen pickup in steel. The details of oxygen transfer in ESR process have been ascertained in Chap. 3.2.

4.4 Inclusions in the Consumable Steel Electrode The oxide inclusions in the consumable steel electrode are CaO–Al2 O3 –SiO2 with a small amount of MgO without exception. An example of the inclusions observed in the steel electrode is shown in Fig. 4.1. EPMA analysis was performed to quantitatively determine the compositions of oxide inclusions. The backscattered electron (BSE) images and EPMA analyzed results of typical inclusions are shown in Fig. 4.2. The MgO content in CaO–Al2 O3 – SiO2 –MgO inclusions was identified to be extremely low (approximately 1 mass%). It should be mentioned that all oxide inclusions are invariably associated with patchtype sulfide inclusions. The inclusions in consumable electrode (termed original

Fig. 4.1 SEM–EDS element mappings of a typical inclusion in consumable steel electrode: (Ca,Mn)S adhering to CaO–Al2 O3 –SiO2 –MgO

4.4 Inclusions in the Consumable Steel Electrode

77

Fig. 4.2 BSE images and EPMA analyzed compositions of inclusions observed in consumable electrode. Figs. (a), (b) and (c) shows the examples of inclusions (EPMA point analysis of inclusion composition, shown as the symbol “+”, is expressed in mass percent)

inclusions) are mostly 3–8 µm in size. In order to minimize the interference caused by an excessive excitation of surrounding steel matrix and sulfide phase, both EDS and EPMA point measurements were focused at the center of each oxide inclusion. The EDS analysis of micron-sized inclusions was inevitability distorted by the surrounding steel matrix and different parts of multi-phased inclusion. The details of steel matrix effect on EDS analysis of micron-sized inclusions have been ascertained by Pistorius et al. [21] According to attempts varying different acceleration voltages and measurement time, SEM–EDS analysis could not clarify whether any Mn, Ca and S did exist or not in different parts of these complex micro-inclusions. Therefore, EPMA line scanning of elements Ca, S, and Mn across an inclusion and EPMA element mappings of the outside layer of oxide-sulfide inclusion were carried out to reveal these elements in different parts of multi-phased inclusions. As examples, Figs. 4.3 and 4.4 show the EPMA line scanning and element mappings of oxide-sulfide inclusions, respectively. It was revealed from EPMA line scanning and element mappings that Mn and S did only appear on the edges of an oxide-sulfide complex inclusion, and Ca was present in both oxide and sulfide phases. According to CaS–MnS binary phase diagram, [22] the sulfide phase in these inclusions was considered to be (Ca,Mn)S solid solution. The compositions of oxide inclusions (oxide phase in dual-phased CaO–Al2 O3 – SiO2 –MgO+(Ca,Mn)S inclusions) in consumable electrode are depicted on CaO– Al2 O3 –SiO2 ternary phase diagram, as shown in Fig. 4.5. MgO in these oxide inclusions was deleted from analyses for simplification in view of its negligibly low content (approximately 1 mass%). CaO–Al2 O3 –SiO2 phase diagram was calculated with FactSage 7.1 (FToxid database). The region surrounded by bold red line in this phase diagram is the low-melting-temperature region [< 1873 K (1600 °C)]. The compositions of almost all oxide inclusions locate in the low-melting-temperature region, as presented in Fig. 4.5, indicating their liquid state in liquid steel at steelmaking temperatures.

78

4 Reoxidation of Liquid Steel During ESR and Its Effect on Oxide Inclusions 4000 Ca

Intensity (counts)

2000

0 800

2

4

6 S

400 0

2

4

6 Mn

60 30 0

0

1

2

3

4

5

6

7

8

Distance (um) 4000 Ca

Intensity (counts)

2000

0 600

2

4

6

8

10

12

14 S

400 200 0

2

4

6

8

10

12

14 Mn

60 30 0

0

2

4

6

8

10

12

14

16

Distance (um)

Fig. 4.3 EPMA element (Ca, S, and Mn) line scanning across an inclusion in the consumable electrode. The dashed lines shown in Figs. (a) and (b) indicate the trace of the line scan

Fig. 4.4 EPMA element mappings of the outside layer of duplex oxysulfide inclusion in the consumable electrode

4.5 Inclusions in the Remelted Ingots Figure 4.6 presents the SEM micrograph, element mappings and EDS spectrums of typical inclusions observed in the remelted ingot. Three types of oxide inclusions were found in the remelted ingot, i.e., CaO–Al2 O3 –SiO2 –MgO, CaO–Al2 O3 –MgO

4.5 Inclusions in the Remelted Ingots

79

Fig. 4.5 Distribution of oxide inclusion compositions, analyzed with EPMA, in the consumable electrode on the CaO–Al2 O3 –SiO2 ternary phase diagram (temperature in degree Celsius), (MgO was deleted from analyses). Pink solid circles represent the composition of inclusions, and some of these pink solid circles are overlapped with each other. The phase diagram was calculated with FactSage 7.1

and MgAl2 O4 . In some cases, SiO2 was detected by EDS in CaO–Al2 O3 –MgO inclusions, but its content was far lower than 1 mass%. These inclusions were categorized as CaO–Al2 O3 –MgO type. No CaO–Al2 O3 –SiO2 –MgO inclusions with the compositions similar to those in consumable electrode were observed in ESR ingots. It was identified that MgO content was approximately 1 mass% in these calcium aluminate inclusions. Most of the observed CaO–Al2 O3 –SiO2 –MgO and CaO–Al2 O3 – MgO inclusions are 2–6 µm in size. The calcium aluminate inclusions with the size ranging from 6 to 12 µm take up a very small proportion. The size of MgAl2 O4 inclusions is about 1.5 µm. CaO–Al2 O3 –MgO inclusions are the predominant type in the remelted ingot, and CaO–Al2 O3 –SiO2 –MgO inclusions account for 30% of the total inclusions. MgAl2 O4 inclusions take up 17% in the number proportion. Both EDS and EPMA analysis showed that SiO2 content in CaO–Al2 O3 –SiO2 – MgO inclusions greatly decreased with an accompanying considerable increase in Al2 O3 content after ESR refining, in comparison with the compositions of the oxide inclusions in consumable electrode. These quaternary CaO–Al2 O3 –SiO2 – MgO inclusions contain only a trace amount of SiO2 (slightly smaller than 1 mass%), and roughly same amounts of CaO and MgO as that in original oxide inclusions. The CaO content in CaO–Al2 O3 –MgO inclusions is far lower than that in CaO– Al2 O3 –SiO2 –MgO inclusions according to EDS and EPMA analysis, whereas Al2 O3

80

4 Reoxidation of Liquid Steel During ESR and Its Effect on Oxide Inclusions

content shows an opposite feature. CaO–Al2 O3 –MgO inclusions exhibit much lower CaO/Al2 O3 mass ratio than CaO–Al2 O3 –SiO2 –MgO inclusions. Nearly half proportion of CaO–Al2 O3 –SiO2 –MgO inclusions was invariably associated with CaS as a poor discontinuous shell (see Fig. 4.6 for example). Unlike the sulfides in consumable steel electrode, no patch-type sulfide inclusions associated with oxide inclusions were observed in remelted ingots. The oxide inclusions were observed to exhibit a narrow variation in their compositions in the remelted ingot. The average composition of CaO–Al2 O3 –SiO2 –MgO

Fig. 4.6 SEM–EDS element mappings of typical inclusions observed in the remelted ingot: a CaS adhering to CaO–Al2 O3 –SiO2 –MgO, b CaS adhering to CaO–Al2 O3 –MgO (EDS spectrums in Figs. (c) and (d) correspond to the inclusions shown in Figs. (a) and (b), respectively)

4.5 Inclusions in the Remelted Ingots

81

cps/eV 24

Acquisition

Al

22

(c)

20 18 16 14 Ca

12 10 8 Mg

6 4 2

Fe

Cr O

S

Fe

Ca

S

Si

Cr

0 1

2

3

4

5

6

7

8

9

10

keV 18

cps/eV

Acquisition

Al

16

(d)

14 12 10 8 6 4 2

S

Mg

Cr V O

Ca

Ca

Fe

Fe

S

V

Cr

0 1

2

3

4

5

6

7

8

9

10

keV

Fig. 4.6 (continued)

inclusions in the ingot is displayed on the CaO–Al2 O3 –SiO2 ternary phase diagram shown in Fig. 3.16 in Sect. 3.5.5.2 (see the result for ESR-1). In comparison with the compositions of oxide inclusion in the consumable electrode shown in the CaO– Al2 O3 –SiO2 ternary phase diagram (see Fig. 4.5), the compositions of CaO–Al2 O3 – SiO2 –MgO inclusions in the remelted ingot locate in the high-melting-temperature region [> 1873 K (1600 °C)], indicating the transformation of original CaO–Al2 O3 – SiO2 –MgO from liquid state to solid CaO–Al2 O3 –SiO2 –MgO after ESR. As for CaO–Al2 O3 –MgO inclusions in the remelted ingot, the concentrations of CaO and Al2 O3 are in range from 12 mass% to 19 mass% and 77 mass% to 86 mass%, respectively. According to the CaO–Al2 O3 –MgO ternary phase diagram, the compositions of these inclusions lie in high-melting-temperature region [> 1973 K (1700 °C)].

82

4 Reoxidation of Liquid Steel During ESR and Its Effect on Oxide Inclusions

4.6 Transient Inclusions in the Liquid Metal Pool During ESR Refining A particular attention was focused on the compositional evolution of oxide inclusions by analyzing the transient inclusions during ESR process. The inclusions in the steel samples collected from liquid metal pool during ESR were determined by SEM– EDS and EPMA. Figure 4.7 presents the examples of transient inclusions (SEM micrographs and EDS spectrums) in liquid metal pool. The oxide inclusions in liquid metal pool of different ESR heats were identified as CaO–Al2 O3 –SiO2 –MgO, CaO– Al2 O3 –MgO, and a small amount of MgAl2 O4 . No sulfide inclusions were present in the samples collected from liquid metal pool. Al2 O3 inclusions were occasionally observed in liquid metal pool, as that in remelted ingots. Most of the observed Al2 O3 appears spherical morphology (mostly smaller than 1 µm), which differs from the morphology of Al2 O3 inclusions in ESR ingots.

4.7 Evolution Mechanism of Inclusions During ESR Refining The preceding experimental results in Sect. 2.3.4 shows that the content of SiO2 in CaO–Al2 O3 –SiO2 –MgO inclusions dramatically decreases in remelted ingot accompanying with a considerable increase in Al2 O3 content, in comparison with the oxide inclusions in the steel before ESR. The compositional change of CaO–Al2 O3 –SiO2 – MgO inclusions is attributed to the reduction of SiO2 in the original inclusions by dissolved aluminum in liquid steel according to Reaction (2.4), resulting in a considerable increase in the Al2 O3 content. The thermodynamic assessment of Reaction (4.1) is presented in Sect. 3.5.5.2. [23] (J/mol) 4[Al] + 3(SiO2 )inclusion = 3[Si] + 2(Al2 O3 )inclusion G ⊕ 4 = −658, 300 + 107.2T

(4.1)

The Gibbs free energy change for Reaction (4.2) was calculated to evaluate the driving force for the reaction between dissolved aluminum in liquid steel and CaO in the original CaO–Al2 O3 –SiO2 –MgO inclusion by combining the standard Gibbs free energy change for Reaction (4.2), first-order and second-order interaction parameters summarized elsewhere [23, 24]. [25] 2[Al] + 3(CaO)inclusion = 3[Ca] + (Al2 O3 )inclusion G ⊕ (J/mol) 6 = 733, 500 − 59.7T

(4.2)

The concentration of dissolved oxygen concentration estimated from [Si]–[O] equilibrium and [Al]–[O] equilibrium was used, instead of the measured total oxygen concentration in steel, to calculate the activity coefficients of dissolved calcium based O O = −9000, rCa = on the first-order and second-order interaction parameters [23] eCa

4.7 Evolution Mechanism of Inclusions During ESR Refining

83

cps/eV Acquisition

Al

35

0.17atom% Si

30

(g)

25 20 Fe

15 10 5

Ca

Mg

Cr O Fe Ca

Cr

Si

0 1

2

3

4

5

6

7

8

9

10

keV

Fig. 4.7 Typical inclusions observed in the sample collected from the liquid metal pool during ESR. a CaO–Al2 O3 –MgO, b CaO–Al2 O3 –MgO, c CaO–Al2 O3 –SiO2 –MgO, d Al2 O3 , e Al2 O3 , f MgAl2 O4 (EDS spectrum in Fig. (g) corresponds to the oxide inclusion shown in Fig. (a)) Ca,O 3.6 × 106 , and rCa = 2.9 × 106 . In the calculation of the Gibbs free energy change for Reaction (4.2), the concentration of dissolved calcium in liquid steel should be used. The measured concentration of total calcium in steel includes the dissolved calcium (free calcium) and insoluble calcium combined as inclusions. The measured total calcium content in remelted ingots is 5 ppm. The dissolved calcium in the steel is very difficult to be determined. The dissolved calcium content in steel is far lower than total calcium content [26, 27]. For convenience, 1 ppm was used in the current calculation. The calculated values of the Gibbs free energy change for Reaction (4.2) are extremely positive (275.0 kJ/mol). In the case of higher dissolved calcium content in steel, the calculated Gibbs free energy change is larger. The thermodynamical

84

4 Reoxidation of Liquid Steel During ESR and Its Effect on Oxide Inclusions

calculation for Reaction (4.2) suggested that CaO in original CaO–Al2 O3 –SiO2 – MgO inclusions in steel electrode could not be reduced by dissolved aluminum in liquid steel during ESR. The CaO content in CaO–Al2 O3 –MgO inclusions was identified to be far lower than that in original CaO–Al2 O3 –SiO2 –MgO inclusions. It therefore was concluded that the oxide inclusions containing lower CaO content were not originated from the transformation of original CaO–Al2 O3 –SiO2 –MgO inclusions through reduction reaction by dissolved aluminum in liquid steel, but were newly-formed oxide inclusions. Deoxidation of liquid steel during ESR comes down to removing oxide inclusions in this refining process. Most of the original inclusions in consumable electrode were removed during ESR process (as indicated by the significant decrease in calcium content from 19 to 5 ppm in the steel before and after ESR), and the others were transformed to CaO–Al2 O3 –SiO2 –MgO inclusions which remained until in the ESR ingot. The reoxidation of liquid steel during ESR led to considerable pickup of dissolved oxygen (as ascertained in Sect. 4.3), which provided a driving force for the generation of fresh oxide inclusions, thereby resulting in the generation of CaO–Al2 O3 –MgO, Al2 O3 and MgAl2 O4 inclusions. The preceding experimental determination did confirm that CaO–Al2 O3 –MgO, Al2 O3 , and MgAl2 O4 inclusions were newly-formed inclusions. It should be stressed that CaO–Al2 O3 –MgO inclusions could be modified MgAl2 O4 inclusions, but the contribution in this trajectory to the formation of CaO– Al2 O3 –MgO inclusions during ESR process was expected to be quite small because of a significant difference in the size between CaO–Al2 O3 –MgO inclusions and MgAl2 O4 inclusions as well as their size distribution (as presented in Sect. 4.5), considering that calcium modification of MgAl2 O4 inclusion could hardly increase the size of the original inclusion appreciably.

4.8 Other Cases of Inclusion Evolution During ESR The oxygen content nearly doubled (14–17 ppm) after protective atmosphere ESR of the steel with ultra low oxygen content (8 ppm), indicating the occurrence of the reoxidation of liquid steel in the current ESRR trials. As shown in Sect. 3.3, the oxygen level in liquid steel is confirmed to be determined by [Al]–[O] equilibrium during protective atmosphere ESR of the steel with ultra low oxygen content, whereas FeO in the slag transferred oxygen into the liquid steel even though the FeO content was very low (0.4–0.6 mass%) in the slag, resulting in a significant oxygen pickup in the steel. The oxide inclusions were identified as CaO–Al2 O3 –MgO (consisting of 27.5– 40.3 mass% CaO, 36.4–52.0 mass% Al2 O3 , and 14.9–19.6 mass% MgO (a small amount (< 2 mass%) of SiO2 was detected in some cases, see Fig. 4.8 for an example)). Both the manual SEM–EDS and automated SEM–EDS analysis from the two-dimensional determination on the polished cross sections showed that the inclusions observed in the steel electrode are about 2 µm.

4.8 Other Cases of Inclusion Evolution During ESR

85

Fig. 4.8 Element mappings of a typical inclusion in the consumable steel electrode: CaO–Al2 O3 – MgO–SiO2 with associated CaS

The morphology of the oxide inclusions is angular. The analysis of oxide inclusions compositions cannot be taken to be exact because of the distorting effect of the steel matrix on the microanalysis of these oxides. Despite this inherent uncertainty, the change in the inclusions compositions was absolutely observed in the current study, as present later in this article. The compositions of these oxide inclusions were located in the region of 100–75% liquid at 1873 K (1600 °C) on the CaO– Al2 O3 –MgO ternary phase diagram (referring to the CaO–Al2 O3 –MgO ternary phase diagram reported in Reference [28]). The morphology and element distribution of the typical inclusions in the remelted billets were determined by SEM–EDS, as shown in Figs. 4.9 and 4.10. These inclusions were identified to be oxide-sulfide complex inclusions. The sulfide phase was identified as CaS. The oxide in the oxide–sulfide complex inclusion was identified as CaO–Al2 O3 – SiO2 –MgO. The elements are homogeneously distributed in these oxide inclusions. In addition, the oxide inclusions with heterogeneous compositions were occasionally found. Examples of EDS element mappings of such inclusions are present in Fig. 4.10. These inclusions were mixtures of CaS adhering to multi-phased inclusion of CaO–Al2 O3 –MgO+CaO–Al2 O3 –SiO2 –MgO. After ESRR, not only singlephased CaS, but also isolated oxide-only inclusions were not observed in the remelted billets irrespective of the melting rates of ESRR. It should be stressed that the original CaO–Al2 O3 –MgO inclusions were no longer observed in the steel after ESRR refining. After ESRR of the steel, the contents of CaO and SiO2 dramatically increased and Al2 O3 content increased mildly in the oxide inclusions whereas the MgO content greatly decreased accordingly. According to the EDS determination, it was found that both CaS patch-type and CaS shell-type oxide inclusions contained less than 3.5 mass% MgO. The compositions of the oxides in the complex inclusions were projected on the CaO–Al2 O3 –SiO2 ternary phase diagram, as shown in Fig. 4.11. MgO was not taken into account in analysis because of its very small content. The

86

4 Reoxidation of Liquid Steel During ESR and Its Effect on Oxide Inclusions

Fig. 4.9 Element mappings of oxide-sulfide complex inclusions observed in the remelted ingots (patch-type CaS associated with CaO–Al2 O3 –SiO2 –MgO inclusion)

region surrounded by the bold red line in the phase diagram is the low-meltingtemperature region (< 1773 K (1500 °C)) calculated with FactSage 7.1 using FToxid database. The compositions of CaS patch-type oxide inclusions in the remelted ingots are located in or nearby the low-melting-temperature region whereas the compositions of CaS shell-type oxide inclusions fall in the fully liquid region (< 1773 K (1500 °C)), suggesting their liquid state during ESRR process. The meltingtemperatures of the oxide inclusions with shell-type CaS are much lower than that of the oxide inclusions with patch-type CaS. Figure 4.12 shows the size distribution of the inclusion in the steel samples taken from the remelted ingots. It is clear that the size of the inclusions significantly increases after ESRR of the steel, in comparison with the original inclusions in the steel electrode. Although the inclusions contain more or less CaS, the determination of the size distribution of oxide inclusion could hardly be distorted by CaS phase due to its small fraction. As shown in Fig. 4.12, the inclusions are mainly in the size range of 2–4 µm in these four ingots (a little more than 50% of the total inclusions). The inclusions smaller than 2 µm take up 20–30% in the number proportion, followed

4.8 Other Cases of Inclusion Evolution During ESR

87

Fig. 4.10 Element mappings of the typical oxide inclusions with heterogeneous compositions. a and b show the examples of such inclusions: CaS adhering to a complex inclusion of CaO– Al2 O3 –MgO+CaO–Al2 O3 –SiO2 –MgO

by the inclusions with the size of 4–8 µm (about 17% in the number proportion). The proportion of the inclusions larger than 8 µm is very small (about 3%). The chemistry, morphologies and size of oxide-sulfide complex inclusions considerably changed during the ESRR process. The compositions of the original oxide inclusions in the steel electrode are located in the liquid + solid field at 1873 K (1600 °C) on the CaO–Al2 O3 –MgO ternary phase diagram. Apart from those had been removed during ESRR, the remaining semiliquid CaO–Al2 O3 –MgO inclusions were fully transformed to liquid CaO–Al2 O3 –SiO2 –MgO inclusions during ESRR (occasionally the oxide inclusions with non-uniformly distributed elements were also found as shown in Fig. 4.10), in accompanying with a considerable increase in their size. In comparison with the original oxide inclusions, the contents of CaO and SiO2 obviously increased and Al2 O3 content increased mildly in the oxide inclusions in the remelted ingots whereas the MgO content decreased sharply. This evolution was thought to originate from the slag-steel-inclusion reactions. The dissolved oxygen was constantly supplied into liquid steel during the reoxidation of liquid steel in the ESRR process. Meanwhile, the calcium that was

88

4 Reoxidation of Liquid Steel During ESR and Its Effect on Oxide Inclusions

Number proportion of inclusions (%)

Fig. 4.11 Composition distribution of oxide inclusions in the remelted ingots on the CaO–Al2 O3 – SiO2 phase diagram (MgO was subtracted from analyses). Pink-filled circle and open circle symbols represent the compositions of patch-type and shell-type oxide inclusions, respectively (some of the circles overlap each other in the phase diagram); (Color Fig.online)

70 ESR-1 ESR-2 ESR-3 ESR-4

60 50 40 30 20 10 0 1 to 2

2 to 4 4 to 8 Inclusion diameter (µm)

> 8

Fig. 4.12 Size distribution of the inclusions detected in the polished cross sections of the samples taken from the remelted ingots. The as-cast billets designated as ESR-1, ESR-2, ESR-3, and ESR-4 correspond to different melting rates of ESRR, respectively

4.9 Summary

89

returned to liquid steel by the dissociation of CaS in the original inclusions and original CaS-only inclusions also provided a driving force for the modification of the semiliquid CaO–Al2 O3 –MgO inclusions, apart from the original dissolved calcium in liquid steel. The total oxygen content in the steel electrode is rather low (8 ppm). The soluble oxygen content in the solidified steel electrode is virtually zero. The dissolved oxygen supplied from the liquid steel reoxidation during ESRR process, the dissolved calcium, aluminum and silicon diffused to the interface of the preexisting semiliquid CaO–Al2 O3 –MgO inclusion and liquid steel. These dissolved oxygen, calcium, silicon and aluminum reacted with the semiliquid CaO– Al2 O3 –MgO inclusions. With the successive diffusion of these elements from liquid steel through liquid product layer to the reaction site, the transformation reaction of the oxide inclusion progressed. As the progress of the transformation reaction, the size of the oxide inclusion increased and the MgO content of the oxide inclusion was diluted continuously, which consequently generated CaO–Al2 O3 –MgO–SiO2 inclusions with a larger size. The proposed mechanism of oxide inclusion transformation was supported by the experimental observations that multi-phased oxides coexisted in the partially modified inclusion (see Fig. 4.10 for examples). In view of the high levels of soluble aluminum and silicon in the steel, the extent of oxide inclusion transformation is expected to largely depend on the concentration of available dissolved oxygen in liquid steel (only a few parts per million of calcium would be sufficient to form calcium aluminate inclusions in liquid steel [27]). A contact angle of 90° between inclusion and liquid steel acts as the threshold whether the inclusion can be removed smoothly in liquid steel or not [29–32]. The contact angle of liquid oxide inclusions (liquid CaO–Al2 O3 –SiO2 –MgO in the current work) in liquid steel is much smaller than 90° [33]. Such low contact angle deteriorated the removal tendency of oxide inclusions from liquid steel in the current ESRR process. These modified inclusions remained in the remelted ingots.

4.9 Summary In this chapter the reoxidation of liquid steel during ESR and its effect on oxide inclusions evolution during the protective atmosphere ESR process are presented. The oxygen content of the steel increases appreciably after P-ESR because of the reoxidation of liquid steel during the protective atmosphere ESR. The introduction of oxygen from atmosphere and FeO in slag are the sources of oxygen pickup in steel. The oxygen level in liquid steel is determined by [Al]–[O] equilibrium during ESR of the steel with ultra low oxygen content, whereas FeO in the slag transfers oxygen into the liquid steel even though the FeO content is very low (0.4–0.6 mass%) in the slag, resulting in a significant oxygen pickup in the steel. Oxide inclusions in the steel before ESR are liquid CaO–Al2 O3 –SiO2 –MgO. Three types of oxide inclusions are present in both liquid metal pool and remelted ingots, i.e., CaO–Al2 O3 –MgO, CaO–Al2 O3 –SiO2 –MgO, and MgAl2 O4 (about

90

4 Reoxidation of Liquid Steel During ESR and Its Effect on Oxide Inclusions

1.5 µm in size). Most of these calcium aluminate inclusions are 2–6 µm. CaO– Al2 O3 –SiO2 –MgO inclusions (type I) originated from the reduction of SiO2 from the original oxide inclusions in consumable electrode by dissolved Al in liquid steel during ESR. CaO–Al2 O3 –MgO, MgAl2 O4 , and CaO–Al2 O3 –MgO–SiO2 (type II) inclusions are generated by the reactions taking place inside liquid steel in liquid metal pool as reoxidation products. The SiO2 content in original CaO–Al2 O3 –SiO2 – MgO inclusions is considerably reduced, in parallel with an increase in Al2 O3 content, after ESR of the steel. The oxide inclusions change from liquid to solid state during the P-ESR. It is the reoxidation of liquid steel during protective atmosphere ESR that contributes to the modification of the original oxide inclusion chemistry (resulting in their full liquefaction) and the increase in its size. Reoxidation of liquid steel should be strictly prevented through lowering the slag oxygen potential during protective atmosphere ESR. Low contact angle of liquid oxide inclusions in liquid steel deteriorates the removal tendency of these oxide inclusions from liquid steel during the ESR process. These inclusions remain in the steel after the protective atmosphere ESR.

References 1. Medina S F, Cores A. Thermodynamic aspects in the manufacturing of microalloyed steels by the electroslag remelting process[J]. ISIJ Int., 1993, 33(12): 1244–1251. 2. Wang C S, Liu S G, Xu M D, et al. Reducing oxygen content in electro-slag remelted bearing steel GCr15[J]. Special Steel, 1997, 18(3): 31–35. (In Chinese). 3. Chang L Z, Shi X F, Cong J Q. Study on mechanism of oxygen increase and countermeasure to control oxygen content during electroslag remelting process[J]. Ironmak. Steelmak., 2014, 41(3): 182–186. 4. Thunman M, Eckert S, Hennig O, et al. Study on the formation of open-eye and slag entrainment in gas stirred ladle[J]. Steel Res. Int., 2007, 78(12): 849–856. 5. Chatterjee S, Li D, Chattopadhyay K. Tundish open eye formation: a trivial event with dire consequences[J]. Steel Res. Int., 2017, 88(9): 1600436. 6. Zinngrebe E, Small J, Laan S V D, et al. Microstructures and formation of tundish clogging deposits in Ti-alloyed Al-killed steel[J]. Metall. Mater. Trans. B, 2020, 51(5): 2321–2338. 7. Shi C B. Deoxidation of electroslag remelting (ESR) – a review[J]. ISIJ Int., 2020, 60(6): 1083–1096. 8. Shi C B, Park J H. Evolution of oxide inclusions in Si-Mn-killed steel during protective atmosphere electroslag remelting[J]. Metall. Mater. Trans. B, 2019,50(3): 1139–1147. 9. Sasabe M, Goto K S. Permeability, diffusivity, and solubility of oxygen gas in liquid slag[J]. Metall. Trans., 1974, 5(8): 2225–2233. 10. Wei J, Liu Z. Study on oxygen transfer through molten slags of CaF2 +Al2 O3 and CaF2 +Al2 O3 +CaO systems for ESR[J]. Acta Metall. Sin., 1994, 30(8): 350–360. (in Chinese). 11. Shi C B, Wang H, Li J. Effects of reoxidation of liquid steel and slag composition on the chemistry evolution of inclusions during electroslag remelting[J]. Metall. Mater. Trans. B, 2018, 49(5): 1675–1689. 12. Shi C B, Zheng D L, Guo B S, et al. Evolution of oxide–sulfide complex inclusions and its correlation with steel cleanliness during electroslag rapid remelting (ESRR) of tool steel[J]. Metall. Mater. Trans. B, 2018, 49(6): 3390–3402.

References

91

13. Schneider R S E, Molnar M, Gelder S, et al. Effect of the slag composition and a protective atmosphere on chemical reactions and non-metallic inclusions during electro-slag remelting of a hot-work tool steel[J]. Steel Res. Int., 2018. 89(10): 1800161. 14. Tiekink W, Boom R, Overbosch A, et al. Some aspects of alumina created by deoxidation and reoxidation in steel[J]. Ironmak. Steelmak., 2010, 37(7): 488–495. 15. Kim T S, Chung Y, Holappa L, et al. Effect of rice husk ash insulation powder on the reoxidation behavior of molten steel in continuous casting tundish[J]. Metall. Mater. Trans. B, 2017, 48(3): 1736–1747. 16. B Coletti, Gommers B, Vercruyssen C, et al. Reoxidation during ladle treatment[J]. Ironmak. Steelmak., 2003, 30(2): 101–105. 17. Oguti Y, Tanbe Y, Miyama S, et al. Temperature distributions in the slag and metal pools in a laboratory-scale ESR furnace[J]. Tetsu-to-Hagane´, 1977, 63(13): 2152–2161. 18. Fu J, Chen C, Chen E, et al. Temperature distribution round the molten slag pool in the process of electroslag refining[J]. Acta Metall. Sin., 1979, 15(1): 44–50. 19. Choudhary M, Szekely J. The modeling of pool profiles, temperature profiles and velocity fields in ESR systems[J]. Metall. Mater. Trans. B, 1980, 11(3): 439–453. 20. Dong Y, Hou Z, Jiang Z, et al. Study of a single-power two-circuit ESR process with currentcarrying mold: mathematical simulation of the process and experimental verification[J]. Metall. Mater. Trans. B, 2018, 49(1): 349–360. 21. Pistorius P C, Verma N. Matrix effects in the energy dispersive x-ray analysis of CaO-Al2 O3 MgO inclusions in steel[J]. Microsc. Microanal., 2011, 17(6): 963–971. 22. Piao R, Lee H G, Kang Y B. Activity measurement of the CaS–MnS sulfide solid solution and thermodynamic modeling of the CaO–MnO–Al2 O3 –CaS–MnS–Al2 S3 system[J]. ISIJ Int., 2013, 53(12): 2132–2141. 23. Ohta H, Suito H. Activities in CaO-SiO2 -Al2 O3 slags and deoxidation equilibria of Si and Al[J]. Metall. Mater. Trans. B, 1996, 27(6): 943–953. 24. The Japan Society for the Promotion of Science. The 19th Committee on Steelmaking: Steelmaking Data Sourcebook[M]. Gordon and Breach Science Publishers, New York, 1988. 25. Suito H, Inoue R. Thermodynamics on control of inclusions composition in ultraclean steels[J]. ISIJ Int., 1996, 36(5): 528–536. 26. Yoshioka T, Shimamura Y, Karasev A, et al. Mechanism of a CaS formation in an Al-killed high-S containing steel during a secondary refining process without a Ca-treatment[J]. Steel Res. Int., 2017, 88(10): 1700147. 27. Kang Y J, Li F, Morita K, et al. Mechanism study on the formation of liquid calcium aluminate inclusion from MgO·Al2 O3 spinel[J]. Steel Res. Int., 2006, 77(11): 785–792. 28. Yang W, Guo C, Li C, et al. Transformation of inclusions in pipeline steels during solidification and cooling[J]. Metall. Mater. Trans. B, 2017, 48(5): 2267–2273. 29. Zheng X, Hayes P C, Lee H G. Particle removal from liquid phase using fine gas bubbles[J]. ISIJ Int., 1997, 37(11): 1091–1097. 30. Cho J S, Lee H G. Cold model study on inclusion removal from liquid steel using fine gas bubbles[J]. ISIJ Int., 2001, 41(2): 151–157. 31. Arai H, Matsumoto K, Shimasaki S, et al. Model experiment on inclusion removal by bubble flotation accompanied by particle coagulation in turbulent flow[J]. ISIJ Int., 2009, 49(7): 965– 974. 32. Yoshioka T, Ideguchi T, Karasev A, et al. The effect of a high al content on the variation of the total oxygen content in the steel melt during a secondary refining process[J]. Steel Res. Int., 2018, 89(2): 1700287. 33. Yoshioka T, Nakahata K, Kawamura T, et al. Factors to determine inclusion compositions in molten steel during the secondary refining process of case-hardening steel[J]. ISIJ Int., 2016, 56(11): 1973–1981.

Chapter 5

Desulfurization in Electroslag Remelting

Abstract Electroslag remelting (ESR) gives a combination of liquid metal refining and solidification structure control. One of the typical aspects of liquid metal refining during ESR for the advanced steel and alloy production is desulfurization. It involves two patterns, i.e., slag–metal reaction and gas–slag reaction (gasifying desulfurization). This chapter presents the advances in desulfurization practices of ESR are reviewed. The effects of processing parameters, including the initial sulfur level of consumable electrode, remelting atmosphere, deoxidation schemes of ESR, slag composition, melting rate, and electrical parameters on the desulfurization in ESR are assessed. The interrelation between desulfurization and sulfide inclusion evolution during ESR is discussed, and advancements in the production of sulfur-bearing steel at a high-sulfur level during ESR are described. The remaining challenges for future work are also proposed.

5.1 Background As the last processing stage of liquid metal refining, the role of electroslag remelting (ESR) processing parameters on the refining efficiency is always an ongoing concern. The detriments of sulfur to the properties of steel and alloy have been widely recognized. For example, an increase in sulfur content deteriorates the fracture toughness and hot ductility of steel [1, 2], lowers the endurance strength of superalloy [3], and leads to the initiation of sulfide-induced stress corrosion crack in steel [4]. The desulfurization capacity of ESR is affected by various factors, such as oxygen levels of liquid metal [5], slag chemistry [6, 7], remelting atmosphere [8, 9], and sulfide inclusions [10, 11]. ESR has a strong desulfurization ability through which the sulfur content is reduced by 50–80% from an electrode to a remelted ingot generally [12–15]. Although studies on desulfurization by ESR are not as many as those on deoxidation and inclusions in ESR, great efforts have been devoted to lowering the sulfur contents of steel and alloy during ESR. Increasing demands for more excellent properties of steel have urged manufacturers to further improve steel and alloy cleanliness. Therefore, the desulfurization fundamentals of ESR should be fully elucidated. © Metallurgical Industry Press 2023 C. Shi et al., Electroslag Remelting Towards Clean Steel, https://doi.org/10.1007/978-981-99-3257-3_5

93

94

5 Desulfurization in Electroslag Remelting

Targeting a low sulfur content is an important aspect of clean steel and alloy production. In addition, maintaining a high-sulfur content is required for some steel grades, such as free-cutting steel and some high-speed steels. For example, sulfur can act as a free-cutting element to improve the cutting performance of steel and increase the abrasive resistance of high-speed steel [16, 17]. However, there are technological difficulties for sulfur-bearing steel production by ESR. In this process, the sulfur contents always vary greatly [17]. In the present work, the desulfurization fundamentals and underlying mechanisms of ESR are assessed. The main factors that influence the desulfurization capacity of ESR are discussed. The interrelation between desulfurization and sulfide inclusion evolution during ESR is also evaluated. The key points of sulfur level maintenance for sulfur-bearing steel production by ESR are presented. A general concluding remark and perspectives for future work are proposed.

5.2 Desulfurization Basis of ESR ESR is generally operated using high frequencies (50 or 60 Hz) of an alternating current (AC) power supply in production practices worldwide. The ESR described in this article is in an AC mode unless otherwise specified. Steel and alloy produced by ESR usually require a low or ultralow-sulfur content. However, for some steel grades, such as crankshaft steel, a high-sulfur content is required. Desulfurization during ESR occurs in two ways. (1) Slag–metal reaction involves the removal of sulfur from liquid metal to molten slag. It is expressed as follows:     [S] + O2− = S2− + [O]

(5.1)

(2) Gas–slag reaction occurs via the oxidation of sulfur ions at the atmosphere/slag interface as a result of their diffusion from molten slag layer to the atmosphere/ slag interface and the exposition of the slag surface to atmospheric oxygen. This reaction is described as follows: (S2− ) + 3/2{O2 } = {SO2 } + (O2− )

(5.2)

where [], (), and {} refer to a species in liquid metal, slag, and gas phases, respectively. The equilibrium constant of Reaction (5.1) is expressed as K =

a(S2− ) · a[O]

a[S] · a(O2− )

(5.3)

5.2 Desulfurization Basis of ESR

95

where K is equilibrium constant, a(S2− ) and a(O2− ) are the activities of sulfur ion and oxygen ion in molten slag, respectively. a[O] and a[S] are the activities of soluble oxygen and sulfur in liquid metal, respectively. Equation (5.3) can be rewritten as a(S2− ) a[S]

=K·

a(O2− ) a[O]

(5.4)

This deduction suggests that the transfer of sulfur from metal to slag is promoted by a higher basicity of slag and a lower content of soluble oxygen in liquid metal. The equilibrium constant of Reaction (5.2) is expressed as: K =

p{SO2 } · a(O2− ) 3 2 p{O/ 2 } · a(S2− )

(5.5)

3/2

p{O2 } p{SO2 } =K· a(S2− ) a(O2− )

(5.6)

where p is partial pressure of gaseous component. According to Eq. (5.6), gasifying desulfurization is enhanced by a higher partial pressure of oxygen in the atmosphere and a lower basicity of slag. SO2 generated through gasifying desulfurization (the desulfurization by gas–slag reaction) is diluted in open air atmosphere of ESR. The dilution of SO2 consequently contributes a lower partial pressure of SO2 in the atmosphere, which provides a favorable condition for desulfurization during ESR. The schematic of an ESR apparatus is shown in Fig. 5.1. Desulfurization of liquid metal takes place at four reaction sites during ESR: ➀ electrode tip/slag interface, ➁ metal droplet/slag interface, ➂ liquid metal pool/slag interface, and ➃ atmosphere/ slag interface. The reaction sites of desulfurization during ESR is schematically presented in Fig. 5.2. The desulfurization of slag–metal reactions occurs at reaction sites ➀, ➁, and ➂, and the gas–slag reaction takes place at reaction site ➃. These reaction sites contribute, more or less, to the overall desulfurization during ESR. Among these reaction sites, reaction site ➀ is the phase boundary with the largest refining potential for desulfurization by ESR. In this site, desulfurization occurs predominantly during the formation of a liquid metal film at the electrode tip [18–20]. It is attributed to more preferable kinetic conditions, namely, long chemical reaction time and high surface-to-volume ratio between a liquid metal film at the electrode tip and slag [20–22]. Reaction site ➁ is of a little importance as a desulfurization site because of the very short residence time of metal droplets passing through the slag pool (virtually 0.01–0.1 s [20, 23]). Desulfurization at the liquid metal pool/ slag interface makes little contribution to the desulfurization during ESR because of reaction time limitation [24]. Desulfurization by ESR includes chemical reactions and mass transfer of components in liquid metal, molten slag, and gas phases. Using a kinetic model developed

96

5 Desulfurization in Electroslag Remelting

Fig. 5.1 Schematic of an ESR apparatus

Fig. 5.2 Desulfurization reaction sites of ESR. ➀ Electrode tip/slag, ➁ metal droplet/slag, ➂ metal pool/slag, and ➃ atmosphere/slag

5.3 Dependence of Desulfurization on the Processing Parameters of ESR

97

based on the penetration and film theories for the ESR of 1Cr21Ni5Ti stainless steel, Hou et al. [25] proposed that remelted steel has a minimum sulfur content because of kinetics limitation at a given sulfur content of steel electrode, such a low sulfur content cannot be further reduced even though increasing the sulfur distribution ratio between metal and slag phases or decreasing the sulfur content of the slag. Desulfurization of liquid steel during ESR is determined by one or several factors, such as slag compositions, remelting atmosphere, melting rate of ESR, and oxygen content of liquid steel. For a specific ESR process, some of these factors make a negligible contribution, but one of them controls desulfurization.

5.3 Dependence of Desulfurization on the Processing Parameters of ESR 5.3.1 Initial Sulfur Content of Consumable Electrode For low-sulfur steel production, a low-sulfur electrode should be prepared for ESR. This condition has been verified with the experimental data of open air atmosphere ESR, as shown in Figs. 5.3 and 5.4. The sulfur content of remelted steel is reduced to an average of 0.007 wt% after remelting of the steel electrode with 0.015 wt% sulfur (see Fig. 5.3), whereas increasing the sulfur content of the consumable electrode to 0.025 wt% results in a relatively higher sulfur content of remelted ingots (0.008– 0.012 wt%), in which the sulfur contents are dependent on the differences in the slag compositions [26]. It is learned from Fig. 5.4 that, for refining three different grades of tool steel (namely, D3, L6, and M2), the sulfur content of the ESR ingot is lower in the case of electroslag remelting of the electrode with a lower sulfur content [27]. It is more 0.015 Sulfur content of ESR ingot / wt%

Fig. 5.3 Relationship between the sulfur content of consumable electrodes and sulfur content of ESR ingots [26]. The chemical composition of slag is presented in mass fraction

70CaF2-15CaO-15Al2O3 80CaF2-20CaO 70CaF2-30Al2O3

0.012

0.009

0.006

0.003 0.010

0.015

0.020

0.025

0.030

Sulfur content of comsumable electrode / wt%

98

Sulfur content of ESR ingot / wt%

0.012

0.015wt% S in Elec. 0.018wt% S in Elec.

(a)

0.009

0.006

0.003

0.000

Slag 1

Sulfur content of ESR ingot / wt%

0.04

Slag 2 Slag number

Slag 3

0.025wt% S in Elec. 0.055wt% S in Elec.

(b)

0.03

0.02

0.01

0.00

Slag 1

0.012 Sulfur content of ESR ingot / wt%

Fig. 5.4 Relationship between the sulfur contents of steel electrodes and ESR ingots produced by ESR when using different slag systems: a tool steel D3; b tool steel L6; c high speed steel M2. Elec. represents steel electrode. Adapted from published data [27]

5 Desulfurization in Electroslag Remelting

Slag 2 Slag number

Slag 3

0.015wt% S in Elec. 0.018wt% S in Elec.

(c)

0.009

0.006

0.003

0.000

Slag 1

Slag 2 Slag number

Slag 3

5.3 Dependence of Desulfurization on the Processing Parameters of ESR Table 5.1 Chemical composition of the slags used in ESR trials

99

Slag

CaF2

CaO

Al2 O3

Slag 1

50 ± 2.5

20 ± 2.5

30 ± 2.5

Slag 2

55 ± 2.5

20 ± 2.5

25 ± 2.5

Slag 3

50 ± 2.5

5 ± 2.5

45 ± 2.5

Adapted from published data [27] wt%

pronounced in the ESR of the electrodes with high original sulfur contents (0.025 and 0.055 wt%) than in the ESR of electrodes with low original sulfur contents, as presented in Fig. 5.4b. The chemical composition of the slag used in ESR trials is shown in Table 5.1 [27]. For the ESR of the tool steel D3 and the high-speed steel M2 electrodes with identical initial sulfur contents, differences in the sulfur contents of ingots are expected to originate from variations in steel electrode compositions, especially oxygen contents, but the data of oxygen contents are not presented in the reference [27]. As a result, these differences lead to various activities of soluble oxygen and activity coefficients of sulfur in liquid steel. However, Padki et al. [28] showed that the sulfur contents of all remelted ingots range from 0.017 to 0.019 wt% irrespective of the initial sulfur contents of steel electrode (0.02–0.38 wt%) in which 70 wt%CaF2 –30 wt%Al2 O3 slag is used in the ESR trials. A low sulfur level of a consumable electrode generally gives a low sulfur content of remelted ingots, but this condition is not always the case. This low sulfur level of remelted ingots also depends on slag composition and oxygen content [29–31]. The ESR of a steel electrode with a high-sulfur level (hundreds of parts per million or more) has been described in References [26–28]. With advances in steelmaking technologies, the sulfur content of steel electrode can be lowered to a low level (a few parts per million) during the secondary refining of liquid steel for electrode production. For low- or ultralow-sulfur steel and alloy production, the sulfur content of low-sulfur electrodes should still be further reduced through ESR. Laboratoryscale open air atmosphere ESR trials with a mold diameter of 165 mm and a low frequency of 4.5 Hz of an AC power supply show that the sulfur content of steel is reduced to 0.0004–0.0008 wt% despite a very low initial sulfur level (0.0013 wt%) in electrodes [32], in which the variations in the sulfur content of ESR ingots are dependent on the slag composition as shown in Table 5.2. More work is needed to study the desulfurization of electrodes with low sulfur content in large-scale ESR trials and at a high frequency of the AC in ESR, which are the usual modes of ESR production practices.

5.3.2 Remelting Atmosphere The previous studies on the desulfurization in different remelting atmospheres of ESR are summarized in Table 5.3 [9, 12, 13, 23, 26, 32–39]. The oxygen partial pressure of remelting atmosphere influences the removal of sulfur from molten slag

100

5 Desulfurization in Electroslag Remelting

Table 5.2 Chemical composition of the slags used in ESR trials [32] wt% Slag

CaF2

CaO

Al2 O3

MgO

SiO2

3C3A

31.5

29.5

33.5

3

1.5

3C3A1S

29

27

30.5

3

10

4C4A

14.5

37.5

41.5

4

1.5

3A

68

—

30

—

2

to gas phase during ESR. Gasifying desulfurization also helps enhance the overall desulfurization capacity of ESR, as verified by open air atmosphere ESR trials [6, 9, 12, 26, 33] and the desulfurization in protective argon atmosphere ESR trials [12, 23]. The study by Schneider et al. [32] shows that the desulfurization capacity of ESR is less pronounced under a protective nitrogen atmosphere, in comparison with that in open air atmosphere ESR. In direct-current ESR, the desulfurization ratio is low under a low oxygen partial pressure in ESR, and a high desulfurization ratio is obtained in open air atmosphere ESR [23]. In ESR with an AC power supply of M41 steel, up to 64% of the initial sulfur content of the charge is removed by gas–slag reaction [9]. Hlineny and Buzek [8] indicated that most of the sulfur that removed from liquid metal enters the gas phase of open air atmosphere ESR. As oxidative desulfurization at the atmosphere/slag interface is restrained by the protective inert atmosphere of ESR, sulfur ions as the desulfurization products of slag–metal reactions constantly accumulate in molten slag with the progress of the desulfurization during ESR, which can cause a decrease in the desulfurization ability or an increase in the sulfur content of remelted ingots [12, 23, 40]. Kang et al. [39] reported that the desulfurization ratio (34%) contributed by ESR in open air atmosphere is higher than that (27%) by the remelting in a nitrogen atmosphere under a slight overpressure. This higher desulfurization degree (the ratio of the difference between the sulfur content of electrode and the sulfur content of ingot to the sulfur content of electrode) is attributed to gasifying desulfurization in ESR, which results in a lower degree of sulfur ion accumulation in molten slag, consequently increasing the desulfurization of the molten slag indirectly. The ESR of GH4169 superalloy shows that the desulfurization degree in argon atmosphere is lower than that in open air atmosphere. However, the desulfurization degree in ESR in an argon atmosphere combined with calcium addition is higher than that in the open air atmosphere, as shown in Fig. 5.5 [35]. The sulfur content of GH4169 superalloy decreases as the calcium content of the superalloy increases [35]. The change in the sulfur content of slag against the height of remelted ingots is shown in Fig. 5.6. Even though the sulfur content of the slag is apparently higher in protective argon atmosphere ESR combined with calcium addition than that in open air atmosphere ESR, the desulfurization degree in the former case, in which gasifying desulfurization is absent, is indeed higher than that in the latter case (see Figs. 5.5 and 5.6). The high desulfurization ratio in the case of calcium addition during ESR is attributed

—

60CaF2 –20CaO–20Al2 O3

Chen et al. [35]

0.0018

0.019

—

—

—

—

—

—

—

—

—

—

0.078–0.079 0.074–0.076*

—

—

—

—

—

—

—

—

—

—

—

0.0032–0.0068 —

—

0.0004–0.0006 0.0005–0.0009 —

0.004–0.005

0.005

33.3CaF2 –33.3CaO–33.3Al2 O3

0.007

0.009

55CaF2 –30CaO–15Al2 O3

0.012

0.008–0.010 0.012

0.015

0.009

0.025 0.025

0.008 0.006–0.007

0.025 0.015

0.005–0.006

75CaF2 –15CaO–10Al2 O3

65CaF2 –15CaO–20Al2 O3

Ahmadi et al. [34]

Mattar et al. [9]

70CaF2 –30Al2 O3

Eissa and 70CaF2 –15CaO–15Al2 O3 EI-Mohammadi [26] 80CaF2 –20CaO

0.033–0.062

0.015

0.095

69.0CaF2 –29.8Al2 O3 –0.06SiO2 –0.4CaCO3

Kato et al. [23]

0.0032

0.008–0.023 0.018

Narita et al. [12] 54.2CaF2 –19.5CaO–24.6Al2 O3 –0.8MgO–1.1SiO2 –0.11T.Fe

70CaF2 –30CaO

—

0.015–0.025

—

—

80CaF2 –20CaO

Same result as the remelting in air atmosphere

0.040–0.047

0.070–0.142 0.065–0.126

0.13

2012

2009

2008

1998

1983

1978

1970

Year

(continued)

Sulfur Sulfur content of ingot in different atmospheres/ content in wt% electrode/ Air Argon Nitrogen wt%

90CaF2 –10CaO

Chemical composition of slag (wt%)

Cooper and Kay 100CaF2 [33] 95CaF2 –5CaO

Author

Table 5.3 Previous studies on desulfurization in different remelting atmospheres of ESR

5.3 Dependence of Desulfurization on the Processing Parameters of ESR 101

Note T.Fe represents total Fe content in slag. *ESR in Ar–5vol%O2 atmosphere

3C3A (Table 2)

3C3A (Table 2)

Schneider et al. [32] 0.0013

0.005

40CaF2 –30CaO–30Al2 O3

Kang et al. [39]

—

— —

0.0004 —

2015

2015

2014

2012

Year

0.0009

—

2018

0.0029–0.0037 2018

0.0012–0.0016 —

0.0016–0.0022 —

0.0010–0.0016 —

0.0009–0.0018 —

0.0014–0.0035 —

0.0050–0.0079 —

0.0012

0.0029–0.0033 —

—

— 0.0070

See Table 5

Slag S6 (Table 4)

Radwitz et al. [38]

—

—

Slag S5 (Table 4)

0.0078

0.0012

—

Slag S1 (Table 4)

Radwitz et al. [37]

0.0018

0.0020–0.0052 0.0020–0.0047 —

—

(48–52)CaF2 –(18–22)Al2 O3 –(18–22)CaO–(4–6)MgO–(4–6)SiO2

Chang et al. [36]

0.0180

Slag S3 and S4 (Table 4)

60CaF2 –20CaO–20Al2 O3

Shi et al. [13]

Sulfur Sulfur content of ingot in different atmospheres/ content in wt% electrode/ Air Argon Nitrogen wt%

Slag S2 (Table 4)

Chemical composition of slag (wt%)

Author

Table 5.3 (continued)

102 5 Desulfurization in Electroslag Remelting

5.3 Dependence of Desulfurization on the Processing Parameters of ESR 0.0020 Open air atmosphere Argon atmosphere Argon atmosphere with Ca addition

Sulfur content / wt%

Fig. 5.5 Change in the sulfur content of remelted ingots against the height of ingots [35]

103

0.0016 Sulfur content of electrode

0.0012

0.0008

0.0004 0.0002

0

30

60

90

120

Height of the remelted ingot / mm

0.05

Sulfur content of slag / wt%

Fig. 5.6 Change in the sulfur content of slag corresponding to the height of ingots [35]

Open air atmosphere Argon atmosphere Argon atmosphere with Ca addition

0.04

0.03 0

30

60

90

120

Height of the remelted ingot / mm

to the chemical reaction between soluble calcium and sulfur in liquid metal [35], as expressed in the following reaction: [Ca] + [S] = (CaS)

(5.7)

This finding is a progress in the desulfurization by protective argon atmosphere ESR. However, the chemical analysis of the sulfur content in the ingots at such low sulfur contents is associated with some uncertainties. More work is needed to assess the thermodynamics and kinetics of the chemical reactions between soluble calcium and sulfur in ESR.

104

5 Desulfurization in Electroslag Remelting

The abovementioned studies [12, 35, 39] have shown that protective atmosphere of ESR lowers the desulfurization ability of ESR because gasifying desulfurization is prevented. This phenomenon is different from the desulfurization in open air atmosphere ESR. However, Shi et al. [13], Cooper and Kay [33], and Chang et al. [36] demonstrated that an argon atmosphere does not reduce the desulfurization degree of ESR compared with that in open air atmosphere. An experimental work has been conducted to compare the effects of the open air atmosphere and argon atmosphere of ESR on the sulfur content of tool steel (0.39C–0.26Si–0.43Mn– 13.37Cr–0.10Ni–0.21Mo–0.34V–0.079Al–0.0089O–0.020P–0.018S, wt%) [13]. Its results have demonstrated that sulfur content is substantially reduced from 0.0180 wt% in an electrode to 0.0020–0.0047 wt% in the ingots produced by protective argon atmosphere ESR, and 0.0020–0.0052 wt% in the ingots in the case of open air atmosphere ESR. These results are evidence of the comparable desulfurization efficiency in these two cases. For the ESR of the steel with low sulfur and oxygen contents (0.0018 wt% S and 0.0006 wt% O), the sulfur content of the remelted ingots produced by both protective argon atmosphere ESR and open air atmosphere ESR is 0.0012 wt% [36]. It indicates that remelting atmospheres exert no effect on the desulfurization degree of low-sulfur steel. More work is needed to establish the relationship between the remelting atmosphere and the desulfurization of low-sulfur steel by ESR when using other slag systems. With increasing demands for the cleanliness of steel and alloy, protective atmosphere remelting has become a standard for new ESR plant installations to lower the oxygen content of remelted ingots and suppress the loss of alloying elements during ESR. Protective atmosphere ESR is generally referred to the remelting in an argon atmosphere. Although gasifying desulfurization is prevented in an inert atmosphere, the desulfurization degree of ESR in a protective argon atmosphere is not reduced compared with that of the remelting in open air atmosphere provided that sulfides in the used molten slag during protective atmosphere ESR are still away from their saturation and the kinetics condition is comparable in these two cases. The decrease in the desulfurization degree of ESR, which is caused by a protective atmosphere, is highly dependent on the sulfide capacities of slag melts. In a sufficiently high sulfide capacity of slag, slag–metal reactions can already yield a high desulfurization capacity of ESR. On the contrary, gasifying desulfurization is needed to enhance desulfurization in ESR following the desulfurization by slag– metal reactions. In the latter case, the remelting atmosphere plays a significant role in desulfurization during ESR. More work is needed to quantitatively verify the role of protective argon atmosphere in the desulfurization during ESR by using the slag with different compositions. Pressurized ESR is a promising technology for producing high nitrogen steel [41–43]. In recent years, vacuum ESR has been developed to further improve the cleanliness of steel in terms of oxygen and inclusion contents [44, 45]. The nitrogen transfer behavior at different nitrogen partial pressures of pressurized ESR and the deoxidation of vacuum ESR have been studied [43, 45, 46]. Nevertheless, future work

5.3 Dependence of Desulfurization on the Processing Parameters of ESR

105

should be performed to reveal the roles of pressurized and vacuum atmospheres in the desulfurization in ESR.

5.3.3 Slag Composition The chemical composition of slag is the most important concern among the factors affecting the steel cleanliness in terms of the deoxidation, non-metallic inclusion removal, and desulfurization efficiency of ESR [5, 32, 47, 48]. ESR-type slag is generally CaF2 –CaO–Al2 O3 -based system with minor additions of MgO, TiO2 , and SiO2 to tailor the slag for specific remelting requirements. On the basis of the sulfide capacities of CaF2 –CaO–Al2 O3 melts, Mattar et al. [9] reported that the optimum desulfurization (highest sulfide capacity) for this slag system is attained with 20 wt% CaO and 80 wt% CaF2 . However, the experimental work supporting this assertion for desulfurization by ESR is lacking. Slag chemistries not only affect sulfur removal by slag–metal reaction but also influence the elimination of sulfur from molten slag to gas phase via gas–slag reaction. The thermodynamic analysis of slag–metal desulfurization and gas–slag desulfurization, as expressed in Eqs. (5.4) and (5.6), shows that high basicity (high a(O2− ) ) and low oxygen potential (low a[O] ) are favorable to the desulfurization by slag– metal reaction, whereas low basicity (low a(O2− ) ) and high oxygen partial pressure (high p{O2 } ) are beneficial to the desulfurization by gas–slag reaction. These two desulfurization reactions not only affect each other, but also have inherent contradictions. From the viewpoint of kinetics, the desulfurization rates of slag–metal and gas–slag reactions increase as slag viscosity decreases. The slag viscosity is mainly dependent on slag chemistries. A low slag viscosity causes a strong stirring action, which is mainly caused by electromagnetic forces, and a strong diffusion of sulfur from molten metals through the slag layer and to the atmosphere/slag interface [9]. Consequently, the desulfurization rate of gas–slag reaction increases. A sufficient amount of CaO should be added to the slag to achieve pronounced desulfurization during ESR. Sulfur removal via slag–metal reaction is enhanced by increasing the CaO/Al2 O3 mass ratio of the slag because of a high CaO activity in ESR [9]. The ESR of free-cutting steel (0.13 wt% S) when using the CaF2 –CaO slag with different CaO contents (0, 5, 10, 20, and 30 wt%) shows that the sulfur content of steel apparently decreases as CaO contents increase expectedly in open air atmosphere ESR [33]. The desulfurization degree decreases at the later stage of open air atmosphere ESR because of an increase in the oxygen content of liquid steel through increasing the FeO content of the slag with the progress of the ESR, especially for the case of low-CaO slag [33]. In the protective argon atmosphere ESR of the steel with 0.0029 wt% sulfur, the desulfurization degree of ESR decreases as the SiO2 content in the slag increases [30]. A similar finding is observed in the protective argon atmosphere ESR of the steel with 0.0032 wt% sulfur [31]. It is attributed to a decrease in the CaO activity as the SiO2 content of the slag increases [30, 31]. Meanwhile, the desulfurization

106

5 Desulfurization in Electroslag Remelting

kinetics of ESR deteriorates because of the increased SiO2 contents of the slag. The deterioration of the desulfurization kinetics originates from the reduction of the depolymerization degree of molten slag structures and the enhancement of the diffusion resistance of structural units as the SiO2 content of slag melts increases [49]. The comparison of the desulfurization degree of the slag with different CaO contents (i.e., 70 wt% CaF2 –15wt%CaO–15wt%Al2 O3 , 80wt%CaF2 –20wt%CaO, and 70wt%CaF2 –30wt%Al2 O3 ) in the open air atmosphere ESR of three different grades of tool steel shows that 80wt%CaF2 –20wt%CaO slag is the most effective in the desulfurization via gas–slag reaction, and 70wt%CaF2 –15wt%CaO– 15wt%Al2 O3 slag gives the highest degree of desulfurization via slag–metal reaction [26]. Eissa and EI-Mohammadi [26] indicated that these differences in the desulfurization degree are attributed to the variations in the slag viscosity and interfacial tension. However, their explanation did not include the differences in the sulfide capacities of these slag systems and steel compositions, especially the oxygen contents of these three steels for desulfurization. These thermodynamic aspects can make a remarkable difference in the desulfurization degree of ESR. For the slag without CaO in its initial chemistry, desulfurization also occurs in open air atmosphere ESR [11, 50]. The sulfur content decreases from 0.039–0.0171 wt% in NiCrMoV alloy and from 0.011 to 0.0017 wt% in CrMoV alloy after ESR [11]. Liu et al. [11] claimed that CaO is generated through the chemical reaction expressed in Reaction (5.8), and subsequently reacts with sulfur in liquid steel for desulfurization, and the gasifying desulfurization ratio reaches 92.2%. 3(CaF2 ) + (Al2 O3 ) = 3(CaO) + 2{AlF3 }

(5.8)

However, Liu et al. [11] pointed out that the amount of CaO formed through the reaction described in Reaction (5.8) is quite small. Therefore, the desulfurization ability of the slag with CaO-free in its initial chemistry is virtually limited. With respect to the desulfurization by ESR when using the slag without CaO in its initial chemistry, more research is needed to quantify the amount of CaO generated via fluoride evaporation from slag melts and its contribution to the desulfurization. Desulfurization contributed by MnO and FeO, which is an inevitable component in the slag during ESR, should also be considered. The sources of FeO in ESR have been summarized in Reference [5]. The desulfurization abilities of the slag with different CaF2 , CaO, and Al2 O3 contents in the ESR of 21CrMoV5-7 steel in an argon atmosphere at a slight overpressure (0.12 MPa) are compared [37], as shown in Fig. 5.7. The chemical compositions of the slags are shown in Table 5.4 [37]. The desulfurization ability of the ESR almost remains unchanged as the CaO content increases from 20 to 40 wt% and as the CaF2 content simultaneously decreases from 60 to 20 wt%. As shown in Fig. 5.7, 20 wt% CaO is sufficient to keep a satisfactory sulfur removal until the end of ESR. Increasing the MgO content of the slag from 2 to 15 wt% and decreasing the CaO content from 19 to 12.5 wt% do not change the desulfurization ability of protective

5.3 Dependence of Desulfurization on the Processing Parameters of ESR

107

Fig. 5.7 Variation in the sulfur contents of remelted ingots for different slag systems [37]

Table 5.4 Chemical compositions of the slags used in ESR trials [37] wt% Slag No.

CaF2

CaO

Al2 O3

MgO

SiO2

FeO

S1

98.9

0.35

0.02

0

0.17

0.01

S2

79.12

9.666

9.33

1.016

0.196

0.028

S3

59.34

18.982

18.64

2.032

0.222

0.046

S4

39.56

28.298

27.95

3.048

0.248

0.064

S5

19.78

37.614

37.26

4.064

0.274

0.082

S6

0

46.93

46.57

5.08

0.3

0.1

argon atmosphere ESR because the desulfurization by ESR under this atmosphere is mainly based on CaS formation and slightly based on MgS formation [38]. The sulfur content is decreased from 0.0070 wt% in the electrode to 0.0010–0.0018 wt% in remelted steel. The chemical compositions of the studied slags in reference [38] are summarized in Table 5.5. The activities of slag components should be used to characterize the differences in the desulfurization capacities of slag melts with different chemistries, rather than their contents in slag melts. Table 5.5 Chemical compositions of the slags used in ESR trial [38] wt% Slag No.

CaF2

CaO

Al2 O3

MgO

SiO2

FeO

F1

59.34

18.98

18.64

2.03

0.22

0.046

F2

59.34

17.48

17.13

5.07

0.01

0.04

F3

59.34

15.05

14.69

9.98

0.03

0.04

F4

59.34

12.62

12.25

14.89

0.04

0.03

108

5 Desulfurization in Electroslag Remelting

For ESR of a steel electrode with a low sulfur content, slag chemistry virtually slightly affects the desulfurization during ESR. The sulfur content of X37CrMoV5-1 hot work tool steel decreases from 0.0013 wt% in the steel electrode to 0.0004–0.0008 wt% after ESR in open air atmosphere when the slag with varying CaF2 , CaO, Al2 O3 , and SiO2 contents is used [32]. The chemical compositions of the slag used in ESR trials are listed in Table 5.2. In these cases, the initial sulfur content of the steel electrode is very low. Therefore, the differences in the sulfur contents of remelted ingots are subjected to some uncertainties associated with the chemical analysis of the sulfur contents. For ultralow-sulfur steel and alloy production, more studies are needed to reveal the effect of slag compositions on the desulfurization during the ESR of the electrodes with initial low sulfur contents.

5.3.4 Deoxidation Schemes of ESR Desulfurization of liquid metal is influenced by the soluble oxygen content of liquid metal. The deoxidation of ESR is an indispensable condition to increase the desulfurization degree of liquid metal. This precondition and its related factors have been systematically assessed in Reference [5]. The continuous or periodic addition of deoxidizing agents to the slag pool of ESR is widely used for the deoxidation of ESR [13, 51–53]. For the ESR of an Fe–Al intermetallic alloy, the presence of a high aluminum content (15 wt%) of this alloy lowers the oxygen potential of given system during ESR. Consequently, sulfur content decreases from 0.0120 wt% in the electrode to 0.0032 wt% in the alloy after ESR [54]. The deoxidation of ESR should be performed to keep a low oxygen potential of molten slag (FeO and MnO contents) by adding deoxidizing agents because this practice is a prerequisite to successfully lower the sulfur content (S ≤ 0.0020 wt%) of 200t scale ingots, besides increasing the oxidability of the remelting atmosphere [55]. Decreasing the oxygen potential of molten slag by periodically adding aluminum shots (0.1% of the ingot weight) during the ESR of 15CDV6 steel has been verified to increase the desulfurization degree of ESR and eliminate the loss of silicon, manganese, and chromium in the steel [56]. For the ESR of steel with a low oxygen content (0.0018 wt%), the addition of different amounts of calcium and Al-based deoxidizing agents in protective atmosphere ESR fails to further reduce the oxygen content of remelted ingots. Consequently, calcium and Al-based deoxidizing agents addition exerts no influence on the sulfur content of remelted ingots (0.0016 wt% in all cases) [29].

5.3.5 Melting Rate of ESR Liquid metal films form on an electrode tip during ESR and thereafter accumulate as liquid metal droplets. This stage plays a predominant role in the desulfurization of

5.3 Dependence of Desulfurization on the Processing Parameters of ESR

109

liquid metal during ESR. The melting rate of ESR largely determines the thickness of a liquid metal film on electrode tip and the residence time of the liquid metal film and metal droplets at the electrode tip [20, 57]. In general, the desulfurization degree of ESR increases as the melting rate of ESR decreases. This observation has been verified in the ESR of Mn18Cr18N steel containing 0.0082–0.0110 wt% sulfur using the slag composed of 69wt%CaF2 –29wt%Al2 O3 –1wt%SiO2 –1wt%FeO [58] and the ESR of commercial-grade iron [6]. However, the mechanisms of these findings are not presented in these two studies [6, 58]. The present authors claim that the increase in the desulfurization degree originates from the increase in the residence time of liquid metal films at the electrode tip as the melting rate of ESR decreases. In the case of protective argon atmosphere electroslag rapid remelting (ESRR) of a steel electrode with 0.0008 wt% oxygen and 0.0026 wt% sulfur, reoxidation of liquid steel takes place during the ESRR, which results in the oxygen contents nearly double (0.0014–0.0017 wt%), and sulfur contents are between 0.0018 and 0.0021 wt% after ESRR. In these cases, the sulfur contents negligibly differ in the remelted steel produced at different melting rates (350, 400, 450, and 500 kg/h) [10]. This finding is different from those obtained by Mehrabi et al. [6] and Tang et al. [58]. From the viewpoint of desulfurization kinetics, increasing the melting rate of ESR lowers the residence time of liquid metal films at the electrode tip, and this reduction is unfavorable to the removal of sulfur from the liquid metal. The differences in the sulfur contents of remelted ingots with varying melting rates are negligible because the desulfurization is limited by the thermodynamics of chemical reactions in these cases, including the initial sulfur content and soluble oxygen content of liquid steel.

5.3.6 Electrical Parameters of ESR At the initial stage of ESR development, direct current (DC), along with AC power supply, is widely used [23, 59–61]. However, ESR is generally operated using high frequencies (50 or 60 Hz) of AC in production practices worldwide. ESR with a DC power supply has been studied, but previous studies mainly focused on the deoxidation of liquid steel and non-metallic inclusions during ESR with a DC power supply [23, 61–63]. Studies on the desulfurization during ESR with a DC power supply are extremely limited. The influences of the type and polarity of an electric current on the desulfurization have been studied through laboratory-scale open air atmosphere ESR trials of 30KH13 steel with 70wt%CaF2 –30wt%Al2 O3 slag [64]. Electric current and voltage are maintained at the same level (I = 0.8 kA, U = 46 V) for all power supply types. The desulfurization degree of ESR with an AC power supply is much higher than that of ESR with a DC of both polarities (see Fig. 5.8). The reversed polarity contributes a higher desulfurization degree of ESR than the straight polarity does. These results are in agreement with the observations reported by Zhang et al. [65] in the ESR of wood alloy with AC, DC of reversed polarity, and DC of straight polarity (see Fig. 5.9).

110

100 Desulfurization degree / %

Fig. 5.8 Influence of the type and polarity of electric current on the desulfurization degree of ESR [64]

5 Desulfurization in Electroslag Remelting

80 60 40 20 0

0.04 Sulfur content of the alloy / wt%

Fig. 5.9 Influence of the type and polarity of electric current on the sulfur contents of ESR ingots [65]

Alternating Straight-polarity Reversed-polarity Type and polarity of electric current

Electrode ESR ingot

0.03

0.02

0.01

0.00

Alternating Straight-polarity Reversed-polarity Type and polarity of electric current

In ESR with direct reversed polarity current, electrons move from a steel electrode (cathode) because of electrolytic processes, and sulfur is removed from liquid steel. However, during ESR with straight-polarity DC, S2– anions are attracted to the electrode (anode). These anions are adsorbed in liquid metal films at the electrode tip and transferred with metal droplets to liquid metal pool, leading to a high-sulfur content. Therefore, the desulfurization degree of ESR operated with straight-polarity DC is much lower than that of ESR with reversed polarity DC. Among these three power supply modes, the desulfurization degree of ESR with an AC power supply is maximal because of an increased area of the slag–metal interface caused by a developed electrocapillary vibration of the molten slag–liquid metal interface during ESR with an AC power supply [64, 65].

5.5 ESR for Sulfur-Bearing Steel Production

111

5.4 Desulfurization Associated with Sulfide Inclusion Evolution During ESR Sulfide inclusion (single-phased or sulfide phase in an oxide-sulfide complex inclusion) removal during ESR is achieved on the basis of the good desulfurization capability of ESR [62, 66]. The present authors’ previous study [31] shows that CaO– Al2 O3 –SiO2 –MgO inclusion acts as formation site for patch-type (Ca,Mn)S inclusion in consumable steel electrode. No sulfide inclusions are present in liquid metal pool, and all (Ca,Mn)S inclusions in the steel electrode are removed before liquid metal droplets collect in liquid metal pool during ESR. In addition, CaS inclusions adhering to CaO–Al2 O3 –MgO inclusions, together with isolated CaS-only inclusions, dissociate into liquid steel as soluble calcium and sulfur during liquid metal film formation and subsequent collection into droplets at the electrode tip during ESR [10]. This phenomenon is in accordance with the findings regarding the removal of (Mn,Cr)S and MnS inclusions during ESR [11, 13, 14, 29]. It is the trajectory that the original CaS, (Mn,Cr)S, and MnS inclusions in consumable steel electrodes are fully removed during ESR. The soluble sulfur in liquid steel that dissociates from original sulfide inclusions is removed by desulfurization, as expressed in Reaction (1), during ESR. Therefore, the desulfurization during ESR is closely associated with the removal of original sulfide inclusions. The factors that affect the removal of original sulfide inclusions influence the desulfurization of liquid steel during ESR. Furthermore, the desulfurization degree of ESR influences the generation of fresh sulfide inclusions during liquid steel solidification in ESR. In this process, a high desulfurization efficiency of ESR suppresses the generation of fresh sulfide inclusions.

5.5 ESR for Sulfur-Bearing Steel Production A high level of sulfur content is expected for some kinds of steel grades, such as freecutting steel, and some crankshaft steel. The uniformity of sulfur in these sulfurbearing remelted steels can hardly be successfully kept. Low-CaO or CaO-free slag, which has a low sulfide capacity, is usually used for ESR to prevent the loss of sulfur during ESR and ensure the uniformity of sulfur in remelted ingots. The optimized slag system for the ESR of sulfur-bearing 34CrNiMo6 steel is composed of 60wt%CaF2 – 30wr%Al2 O3 –5wt%MgO–5wt%SiO2 , which gives a uniform distribution of sulfur in ESR ingots (610 mm in diameter) [67]. A certain amount of FeS addition in slag is also one of the effective countermeasures to produce sulfur-bearing remelted steel in ESR practice. However, the inhomogeneity of sulfur in steel should be noted in this operation, and further studies should be conducted on related topics, such as sulfur transfer between molten slag and liquid steel and the role of slag compositions in sulfur transfer.

112

5 Desulfurization in Electroslag Remelting

In open air atmosphere ESR of the steel with 0.09 wt% sulfur when using the slag composed of 50wt%CaF2 –30wt%Al2 O3 –20wt%SiO2 , the sulfur contents of remelted ingots range from 0.066 to 0.075 wt%, which meets the requirement of the sulfur contents of AS136 steel (0.05–0.1 wt%). However, alumina and silicate inclusions in the steel are beyond the requirement of inclusion evaluated criterion [50]. By contrast, a uniform distribution of sulfur (0.05–0.061 wt%) in remelted ingots, an improved inclusion distribution and size are achieved when using the slag composed of 70wt%CaF2 –28wt%Al2 O3 –2wt%MgO in combination with the continuous addition of 4.5 wt% MgO to the molten slag pool during ESR to maintain a constant sulfur distribution ratio between liquid steel and molten slag [50]. This result indicates that CaO-free slag largely suppresses the decrease in the sulfur content during the ESR of sulfur-bearing steel, and SiO2 addition in CaO-free slag can further suppress the desulfurization ability of ESR. A ternary CaF2 –Al2 O3 –CaS system as the initial slag chemistry in combination with periodic CaS addition to molten slag pool during protective argon atmosphere ESR prevents the loss of sulfur in crankshaft steel [68]. The slag composed of 60wt%CaF2 –15wt%Al2 O3 –5wt%CaO–20wt%SiO2 slag with intentionally added 2.5 wt% CaS exhibits an excellent performance in open air atmosphere ESR to prevent the sulfur loss of sulfur-containing high-speed M35 steel (0.015–0.030 wt% sulfur) during ESR in open air atmosphere [17]. Sulfur contents at the different positions of remelted ingots remain nearly constant, and only a quite small amount of sulfur is lost in remelted ingots. For the slag without intentionally added CaS, a significant amount of sulfur is lost from the steel after open air atmosphere ESR [17]. CaO-free slag with intentionally added CaS has an excellent performance in producing sulfurbearing remelted steel because of the limited desulfurization capability of ESR. For sulfur-bearing steel production, the uniformity of sulfur in the remelted ingots remain a challenge, especially in producing heavy remelted ingots.

5.6 Summary A low initial sulfur level generally gives a low sulfur content of the remelted ingots, but this is not always the case. Low sulfur levels of the remelted ingots also depend on the slag compositions and the oxygen contents. Gasifying desulfurization remarkably contributes to the desulfurization by ESR. The reduction of the desulfurization caused by protective atmosphere is highly dependent on the sulfide capacities of slag. Deoxidation during ESR is a prerequisite for improving the desulfurization. Slag–metal and gas–slag reactions for the desulfurization during ESR not only affect each other but also have inherent contradictions. The desulfurization rates of slag–metal and gas–slag reactions are enhanced with the reduction of slag viscosity. The activity of CaO in slag is the main factor of the remarkable desulfurization by ESR. CaO-free slag in its initial chemistry can exhibit apparent desulfurization in the ESR. More work is needed to quantify the amount of CaO generated through fluoride evaporation from slag melts and its contribution to the desulfurization. Slag

References

113

chemistry exerts only a slight effect on the desulfurization capacity of the ESR of the electrode with a low initial sulfur content. The desulfurization capability of ESR increases as the melting rates of ESR decrease except when desulfurization is determined by reaction thermodynamics. The desulfurization degree of ESR with an AC power supply is much higher than that in the case of the DC of straight and reversed polarities. The desulfurization degree of ESR with straight-polarity DC is minimal. Desulfurization during ESR is associated with the elimination of sulfide inclusions and the generation of fresh sulfide inclusions. CaO-free slag with intentionally added CaS shows an excellent performance in producing sulfur-bearing remelted steel. Significant progress has been made in the desulfurization and a high-sulfur target for sulfur-bearing steel in ESR. However, the desulfurization by ESR at various partial pressures in different remelting atmospheres for an ultralow-sulfur target and achieving an accurate sulfur level in ESR for sulfur-bearing steel production are still in their infancies. Future work is needed to study the roles of different remelting atmospheres on the desulfurization by ESR when using the slag with different compositions and on the desulfurization capacity of ESR of low-sulfur electrodes. For sulfur-bearing steel production, the uniformity of sulfur in remelted ingots remains a challenge, especially heavy remelted ingots.

References 1. M. Morinaga, Y. Murata, R. Hashizume, and Y. Sawaragi, Remarkable improvement in steam oxidation resistance due to the presence of sulfur in high Cr ferrtic steels[J]. ISIJ Int., 2001, 41(3): 314–316. 2. T. Kizu and T. Urabe, Hot ductility of sulfur-containing low manganese mild steels at high strain rate[J]. ISIJ Int., 2009, 49(9): 1424–1431. 3. Q.L. Li, H.R. Zhang, M. Gao, J.P. Li, T.X. Tao, and H. Zhang, Mechanisms of reactive element Y on the purification of K4169 superalloy during vacuum induction melting[J]. Int. J. Miner. Metall. Mater., 2018, 25(6): 696–703. 4. X.F. Bai, Y.H. Sun, R.M. Chen, Y.M. Zhang, and Y.F. Cai, Formation and thermodynamics of CaS-bearing inclusions during Ca treatment in oil casting steels[J]. Int. J. Miner. Metall. Mater., 2019, 26(5): 573–587. 5. C.B. Shi, Deoxidation of electroslag remelting (ESR)—A review[J]. ISIJ Int., 2020, 60(6): 1083–1096. 6. K. Mehrabi, M.R. Rahimipour, and A. Shokuhfar, The effect of slag types and melting rate on electro-slag remelting (ESR) processing[J]. Int. J. Iron Steel Soc. Iran, 2005, 2(1): 37–42. 7. D. Hou, Z.H. Jiang, D.Y. Wang, T.P. Qu, J. Tian, and H.H. Wang, Transfer phenomenon study of slag–metal reaction during electroslag remelting process[J]. Proceedings on Liquid Metal Processing & Casting Conference 2019, Birmingham, UK, 2019, p. 139. 8. J. Hlineny and Z. Buzek, Desulphurization in electro–slag melting[J]. Hutnicke Listy, 1966, (8): 524. 9. T. Mattar, K. EI-Fawakhry, H. Haifa, and M. Eissa, Effect of nitrogen alloying on sulphur behaviour during ESR of AISI M41 steel[J]. Steel Res. Int., 2008, 79(9): 691–697. 10. C.B. Shi, D.L. Zheng, B.S. Guo, J. Li, and F. Jiang, Evolution of oxide-sulfide complex inclusions and its correlation with steel cleanliness during electroslag rapid remelting (ESRR) of tool steel[J]. Metall. Mater. Trans. B, 2018, 49(6): 3390–3402.

114

5 Desulfurization in Electroslag Remelting

11. Y. Liu, Z. Zhang, G.Q. Li, Q. Wang, L. Wang, and B.K. Li, Evolution of desulfurization and characterization of inclusions in dual alloy ingot processed by electroslag remelting[J]. Steel Res. Int., 2017, 88(1): 1700058. 12. K. Narita, T. Onoye, T. Ishii, and T. Kusamichi, A metallurgical study of oxide-base slag used for electroslag remelting[J]. Tetsu-to-Hagané, 1978, 64(10): 1568–1577. 13. C.B. Shi, X.C. Chen, H.J. Guo, Z.J. Zhu, and H. Ren, Assessment of oxygen control and its effect on inclusion characteristics during electroslag remelting of die steel[J]. Steel Res. Int., 2012, 83(5): 472–486. 14. C.B. Shi, W.T. Yu, H. Wang, J. Li, and M. Jiang, Simultaneous modification of alumina and MgO·Al2 O3 inclusions by calcium treatment during electroslag remelting of stainless tool steel[J]. Metall. Mater. Trans. B, 2017, 48(1): 146–161. 15. M. Detrois, P.D. Jablonski and J.A. Hawk, Evolution of tantalum content during vacuum induction melting and electroslag remelting of a novel martensitic steel[J]. Metall. Mater. Trans. B, 2019, 50(4): 1686–1695. 16. Q.K. Yang, P. Shen, D. Zhang, Y.X. Wu, and J.X. Fu, Analysis on composition and inclusions of ballpoint pen tip steel[J]. Int. J. Miner. Metall. Mater., 2018, 25(4): 420–428. 17. Z.B. Xie, Q.L. Shao, G.P. Zhang, K. Wang, and N. He, Research on quality of high speed steel M35 bearing sulfur remelting by acid slag[J]. Special Steel, 2019, 40(1): 27–30. 18. C.K. Cooper, D. Ghosh, D.A.R. Kay, and R.J. Pomfret, ESR reaction sites[C]. Iron and Steel Society of AIME, ed., The 28th Electric Furnace Conference, Warrendale, PA, 1970: 8. 19. P. Mellberg and H. Sandberg, On the mechanism of desulphurization during a.c. and d.c. electroslag remelting[J]. Scand. J. Metall., 1973, 2(3): 121–124. 20. W. Holzgruber, K. Petersen, and P. E. Schneider, Comparison of products and economics from large VAR, AC-ESR and DC-ESR ingots[C]. Transaction of the International Vacuum Metallurgy Conference, California, 1968: 499–523. 21. E. Plöckinger, Electroslag remelting—A modern tool in metallurgy[J]. J. Iron Steel Inst., 1973, 211(4): 533-541. 22. C.B. Shi, Behaviour and Control Technique of Oxygen and Inclusions during Protective Atmosphere Electroslag Remelting Process[D]. University of Science and Technology Beijing, Beijing, 2012: 95. 23. M. Kato, K. Hasegawa, S. Nomura, and M. Inouye, Transfer of oxygen and sulfur during direct current electroslag remelting[J]. Trans. Iron Steel Inst. Jpn., 1983, 23(7): 618–627. 24. Q. Wang, Z. He, G.Q. Li, B.K. Li, C.Y. Zhu, and P.J. Chen, Numerical investigation of desulfurization behavior in electroslag remelting process[J]. Int. J. Heat Mass Transfer, 2017, 104: 943–951. 25. D. Hou, Z.H. Jiang, Y.W. Dong, Y. Li, W. Gong, and F.B. Liu, Mass transfer model of desulfurization in the electroslag remelting process[J]. Metall. Mater. Trans. B, 2017, 48(3): 1885–1897. 26. M. Eissa and A. EI-Mohammadi, Effect of physical properties of slag on sulphur removal mechanism during ESR process[J]. Steel Res., 1998, 69(10-11): 413–417. 27. T.M.T. Mattar, A.M. Fathy, H.S.R. El-Faramawy, M.M. Eissa and K.A.R. El-Fawakhry, Optimization of desulphurization of tool steels during EAF and ESR processes[C]. Metal 2001, Ostrava, Czech Republic, 2001: 1–6. 28. G.M. Padki, M.S.N. Balasubramanian, K.M. Gupt, and P. Krishna Rao, Effect of ESR and carbon content on mechanical properties of a high-strength steel (15CDV6) for aerospace applications[J]. Ironmak. Steelmak., 1983, 10(4): 180–184. 29. C.B. Shi, X.C. Chen, H.J. Guo, Z.J. Zhu, and X.L. Sun, Control of MgO·Al2 O3 spinel inclusions during protective gas electroslag remelting of die steel[J]. Metall. Mater. Trans. B, 2013, 44(2): 378–389. 30. C.B. Shi and J.H. Park, Evolution of oxide inclusions in Si–Mn-killed steel during protective atmosphere electroslag remelting[J]. Metall. Mater. Trans. B, 2019, 50(3): 1139–1147. 31. C.B. Shi, H. Wang, and J. Li, Effects of reoxidation of liquid steel and slag composition on the chemistry evolution of inclusions during electroslag remelting[J]. Metall. Mater. Trans. B, 2018, 49(4): 1675–1689.

References

115

32. R.S.E. Schneider, M. Molnar, S. Gelder, G. Reiter, and C. Martinez, Effect of the slag composition and a protective atmosphere on chemical reactions and non-metallic inclusions during electro-slag remelting of a hot-work tool steel[J]. Steel Res. Int., 2018, 89(10): 1800161. 33. C.K. Cooper and D.A.R. Kay, Behaviour of sulphur during ac electroslag remelting[J]. J. Iron Steel Inst., 1970, 208(9): 856–859. 34. S. Ahmadi, H. Arabi, A. Shokuhfar, and A. Rezaei, Evaluation of the electroslag remelting process in medical grade of 316LC stainless steel[J]. J. Mater. Sci. Technol., 2009, 25(5): 592–596. 35. X.C. Chen, F. Wang, C.B. Shi, H. Ren, and D. Feng, Effect of electroslag remelting process on desulphurization of GH4169[J]. J. Iron Steel Res., 2012, 24(12): 11–16. 36. L.Z. Chang, X.F. Shi, and J.Q. Cong, Study on mechanism of oxygen increase and countermeasure to control oxygen content during electroslag remelting process[J]. Ironmak. Steelmak., 2014, 41(3): 182–186. 37. S. Radwitz, H. Scholz, B. Friedrich, and H. Franz, Influencing the electroslag remelting process by varying fluorine content of the utilized slag[C]. GDMB Society of Metallurgists and Miners, ed., Proceedings of European Metallurgical Conference 2015, Vol. 2, Düsseldorf, Germany, 2015: 887–896. 38. S. Radwitz, H. Scholz, B. Friedrich, and H. Franz, Process and refining characteristics of ESR using MgO containing slag systems[C]. International of Symposium on Liquid Metal Processing & Casting 2015, Leoben, Austria, 2015: 153–162. 39. C.P. Kang, F.B. Liu, Z.H. Jiang, J. Yu, and K. Chen, Study on desulfurization kinetics model of electroslag remelting under different atmosphere[C]. Proceedings of 2018 China Special Metallurgical Technology Meeting, Shenyang, 2018: 68–73. 40. A.S. Ballantyne, The capabilities and benefits of argon shrouding during electroslag remelting[C]. Medovar Memorial Symposium Proceeding, Kiev, Ukraine, 2001: 113–118. 41. G. Stein and J. Menzel, High pressure electroslag remelting—A new technology of steel refining[J]. Int. J. Mater. Prod. Technol., 1995, 10(3-6): 478–488. 42. P. Pant, P. Dahlmann, W. Schlump, and G. Stein, A new nitrogen alloying technique—A way to distinctly improve the properties of austenitic steel[J]. Steel Res., 1987, 58(1): 18–25. 43. J. Yu, F.B. Liu, H.B. Li, Z.H. Jiang, Y. Li, C.P. Kang, A. Wang, W.C. Zhang, and H. Feng, Numerical simulation and experimental investigation of nitrogen transfer mechanism from gas to liquid steel during pressurized electroslag remelting process[J]. Metall. Mater. Trans. B, 2019, 50(6): 3112–3124. 44. A. Sekiya, S. Nakayama, and T. Taketsuru, Deoxidation of stainless steel on vacuum electroslag remelting process[J]. Denki-Seiko, 1995, 66(1): 47–53. 45. X.C. Huang, B.K. Li, Z.Q. Liu, T.C. Jiang, Y.Y. Chai, and X.X. Wu, Oxygen transport behavior and characteristics of nonmetallic inclusions during vacuum electroslag remelting[J]. Vacuum, 2019, 164: 114–120. 46. Y. Liu, X.J. Wang, G.Q. Li, Q. Wang, Z. Zhang, and B.K. Li, Role of vacuum on cleanliness improvement of steel during electroslag remelting[J]. Vacuum, 2018, 154: 351–358. 47. S.J. Li, G.G. Cheng, Z.Q. Miao, L. Chen, and X.Y. Jiang, Effect of slag on oxide inclusions in carburized bearing steel during industrial electroslag remelting[J]. Int. J. Miner. Metall. Mater., 2019, 26(3): 291–300. 48. Y.W. Dong, Z.H. Jiang, Y.L. Cao, A. Yu, and D. Hou, Effect of slag on inclusions during electroslag remelting process of die steel[J]. Metall. Mater. Trans. B, 2014, 45(4): 1315–1324. 49. C.B. Shi, J. Li, J.W. Cho, F. Jiang, and I.H. Jung, Effect of SiO2 on the crystallization behaviors and in-mold performance of CaF2 –CaO–Al2 O3 slags for drawing-ingot-type electroslag remelting[J]. Metall. Mater. Trans. B, 2015, 46(5): 2110–2120. 50. D. Hou, D. Wang, T. Qu, J. Tian, and Z. Jiang, Sulfur content and inclusion control during electroslag remelting of die steel[J]. Chin. J. Eng., 2018, 40(Suppl. 1): 19–25. 51. F. Reyes-Carmona and A. Mitchell, Deoxidation of ESR slags[J]. ISIJ Int., 1992, 32(4): 529– 537. 52. A. Mitchell, F. Reyes-Carmona, and E. Samuelsson, The deoxidation of low-alloy steel ingots during ESR[J]. Trans. Iron Steel Inst. Jpn., 1984, 24(7): 547–556.

116

5 Desulfurization in Electroslag Remelting

53. S.F. Medina and A. Cores, Thermodynamic aspects in the manufacturing of microalloyed steels by the electroslag remelting process[J]. ISIJ Int., 1993, 33(12): 1244–1251. 54. R.G. Baligidad, U. Prakash, V.R. Rao, P.K. Rao, and N.B. Ballal, Electroslag remelting of Fe–28 at.%Al intermetallic alloy[J]. Ironmak. Steelmak., 1994, 21(4): 324-331. 55. D.L. Xiang, X.W. Zhu, and K.W. Wang, Production of ultralow sulphur big ESR ingots[J]. Iron Steel, 1993, 28(6): 21–25. 56. M. Chatterjee, M.S.N. Balasubramanian, K.M. Gupt, and P. K. Rao, Inoculation during electroslag remelting of 15CDV6 steel[J]. Ironmaking Steelmaking, 1990, 17(1): 38–42. 57. M.E. Fraser and A. Mitchell, Mass transfer in the electroslag process: Part 1 mass-transfer model[J]. Ironmak. Steelmak., 1976, 3(5): 279–287. 58. X.W. Tang, R. Zhu, C. Li, and W. Liu, Effect of ESR technology on desulfurization of high nitrogen steel[J]. J. Iron Steel Res., 2015, 27(6): 24–28. 59. Y. Kojima, M. Kato, T. Toyoda, and M. Inouye, The formation of liquid droplet and the behavior of oxygen in direct current electroslag remelting process[J]. Trans. Iron Steel Inst. Jpn., 1975, 15(8): 397–406. 60. M. Kawakami, K. Nagata, M. Yamamura, N. Sakata, Y. Miyashita, and K.S. Goto, Profiles of temperature, voltage and local heat generation in slag phase and metal pool of ESR unit under operation[J]. Tetsu-to-Hagané, 1977, 63(13): 2162–2171. 61. M. Kawakami, T. Takenaka, and M. Ishikawa, Electrode reactions in dc electroslag remelting of steel rod[J]. Ironmak. Steelmak., 2002, 29(4): 287–292. 62. A. Paar, R. Schneider, P. Zeller, G. Reiter, S. Paul, and P. Würzinger, Effect of electrical parameters on type and content of non-metallic inclusions after electro-slag-remelting[J]. Steel Res. Int., 2014, 85(4): 570–578. 63. A. Mitchell, Electrochemical aspects of the ESR process[C]. IOP Conf. Ser.: Mater. Sci. Eng., 2016, 143: 012001. 64. I.A. Aksenov, M.A. Matveeva, and I.V. Chumanov, Influence of the ESR parameters on the removal of sulfur[J]. Russ. Metall., 2019, 2019(6): 601–607. 65. J.W. Zhang, P.M. Guo, and Z.B. Li, Study on electro-capillary oscillation in ESR system[J]. Iron Steel, 2000, 35(5): 23–25. 66. W. Holzgruber and E. Plöckinger, Metallurgical and technological foundations of electroslag re-melting of steel[J]. Stahl Eisen, 1968, 88(12): 638–648. 67. Q. Ma, C.L. Sun, G.P. Feng, and X.L. Zhang, Process practice for improvement of B-type inclusions in ESR sulfur-bearing steel 34CrNiMo6[J]. Special Steel, 2017, 38: 39–41. 68. L.G. Wang, X.B Jia, Y.Q. Li, W.Y. Zhao, L.G. Bi, X. Yun, and C.J. Shang Guan, Study on the control of sulfur and aluminum element during electroslag remelting process[C]. Proceedings of 2016 China Special Metallurgical Technology Meeting, Dalian, 2016: 233–236.

Chapter 6

Sulfide and Nitride Inclusion Evolution During ESR

Abstract A good understanding of the formation and evolution of secondary inclusions, which form during the solidification of liquid steel, in the ESR process is quite necessary to control these inclusions in the liquid steel refining. This chapter presents the evolution of secondary inclusions during the ESR process based on trials and thermodynamic analysis. The driving force for the precipitation of secondary inclusions increases as solutes segregate during solidification of liquid steel and the solubility of oxides and sulfides in steel diminishes as temperature decreases. This chapter ascertains the evolution and removal of (Mn,Cr)S inclusions, MnS inclusions, sulfide in oxide-sulfide complex inclusions, AlN and TiN inclusions. The findings provide useful information for studying other types of secondary inclusions evolution in the ESR process.

6.1 Sulfide Inclusions The trials of protective atmosphere ESR of S136 and H13 tool steel were conducted using the pre-melted slag (60mass% CaF2 , 20mass% CaO, and 20mass% Al2 O3 ). The chemical compositions of S136 and H13 tool steel used as the consumable electrodes in the ESR trials are listed in Table 6.1. The detailed ESR experimental procedure has been described elsewhere [1]. For microscopic examination of the inclusions in steel, two-dimensional determination, i.e., the direct observation on polished surface by SEM-EDS, and threedimensional extracted inclusions determination. For two-dimensional determination, the steel samples that were cut from the electrode and each ESR ingot were prepared for SEM-EDS determination. Electrolytic extraction technique was employed to obtain the inclusions in the consumable electrode. The steel sample (Diameter: 12 mm, Length: 100 mm) cut from the electrode was electrolyzed in organic solution used as electrolyte under anode current density of no greater than 100 mA/cm2 . Stainless steel slice was used as cathode. The temperature of electrolyte was kept in the range of 268–278 K. After metal matrix dissolution, the non-dissolved inclusions were elutriated with ethanol, and then collected on a thin copper foil. The extracted

© Metallurgical Industry Press 2023 C. Shi et al., Electroslag Remelting Towards Clean Steel, https://doi.org/10.1007/978-981-99-3257-3_6

117

118

6 Sulfide and Nitride Inclusion Evolution During ESR

Table 6.1 Chemical compositions of consumable steel electrode (mass%) C

Mn

Si

P

S

S136 0.39 0.43 0.26 0.020 0.018 H13

0.36 0.36 0.95 0.013 0.0018

Cr

Ni

Cu

13.37 0.10 0.04

Mo

V

sol. Al

N

0.21 0.34 0.079 –

5.17 0.16 0.066 1.48 0.99 0.1

T.O 0.0089

0.010 0.0018

inclusions were then analyzed for the size, microstructure and elemental compositions by SEM-EDS. In addition, some of the extracted inclusions were prepared for the determination of their cross section by SEM-EDS. A thin copper foil inlaid with the extracted inclusions was fixed with epoxy resin before the cross section was polished. A gold film was sprayed onto the surface of the extracted inclusions before SEM-EDS measurement. Figure 6.1 shows the SEM-EDS analysis results of typical sulfide inclusions in the S136 consumable steel electrode by two-dimensional observation. The sulfide inclusions observed in the electrode are (Mn,Cr)-containing sulfide inclusions and oxide-sulfide duplex inclusions of Al2 O3 core surrounded by a sulfide ring of (Mn,Cr)S. Inclusion electrolytic extraction in combination with SEM-EDS analysis were performed in order to further identify the derivation of element Cr (from inclusion itself or metal matrix containing 13.37mass% Cr) and inclusion microstructure in the electrode. Figure 6.2 presents the three-dimensional morphology and EDS results of typical extracted sulfide inclusions in S136 consumable electrode. It is clear that the morphology of sulfide inclusions in the electrode is irregular, near-spherical or rod-like shape. In order to analyze the inner microstructure of the extracted threedimensional inclusions, the cross sections of some extracted inclusions were determined by SEM-EDS as shown in Fig. 6.2d–f. It is confirmed that the sulfide inclusions are pure sulfide inclusions composed of Mn, Cr and S or the duplex inclusions of Al2 O3 core surrounded with an outer (Mn,Cr)S ring. These sulfide inclusions generally have large size. While after ESR and P-ESR refining process, only single-phased Al2 O3 inclusions with the size of about 1 µm were observed in ESR ingots produced

Fig. 6.1 SEM-EDS results of typical sulfide inclusions in S136 consumable steel electrode

6.1 Sulfide Inclusions

119

under different operating conditions. No sulfide inclusions were found in the as-cast ingots. Figure 6.3 presents the sulfide inclusions observed in H13 tool steel used as consumable electrode in P-ESR experiments. All observed sulfide inclusions by SEM-EDS are associated with oxide inclusions. The sulfide inclusions were identified as MnS. According to SEM-EDS analysis, MgO·Al2 O3 spinels are the only oxide inclusions in H13 consumable electrode. No isolated sulfide inclusions were found. After P-ESR refining, the sulfur content was reduced to 16 ppm in the ingots. It was confirmed by SEM-EDS determination that no MnS inclusions in ESR ingots were found, although MgO·Al2 O3 spinel, CaO–MgO–Al2 O3 and CaO–Al2 O3 inclusions existed in ESR ingots. It indicates that all original MnS inclusions were removed in P-ESR refining process.

Fig. 6.2 Typical extracted sulfide inclusions in S136 consumable steel electrode: a–c threedimensional morphology of inclusions, d–f cross section of typical extracted inclusions

Fig. 6.3 Typical original sulfide inclusions in consumable H13 steel electrode

120

6 Sulfide and Nitride Inclusion Evolution During ESR

It is concluded from SEM-EDS analysis that all single-phased sulfide inclusions and the outer sulfide layer of the complex inclusions in S136 consumable electrode steel are composed of Mn, Cr and S. No MnS and CrS inclusions were found in the steel electrode. It should be noted that the atomic ratios of Mn, Cr and S slightly vary in different sulfide inclusions, and the atomic ratio of (Mn+Cr)/S in each detected inclusion is almost the same value (i.e., 1:1) at every analyzed point. The atomic ratios of Mn/Cr are in a very narrow range (0.4:0.6–0.6:0.4), and 0.5:0.5 in most of the sulfide inclusions. Therefore, the thermodynamic analysis of (Mn,Cr)S inclusions evolution was made based on the evaluation of (Mn0.4 ,Cr0.6 )S, (Mn0.5 ,Cr0.5 )S and (Mn0.6 ,Cr0.4 )S inclusions. The chemical reaction for the generation of (Mn0.5 ,Cr0.5 )S inclusions in liquid steel can be expressed as 1/2[Mn] + 1/2[Cr] + [S] = (Mn0.5 , Cr0.5 )S

(6.1)

The standard Gibbs free energy change of Reaction (6.1) △G o1 can be obtained by combining standard Gibbs free energy change for Reactions (6.2), (6.3) and (6.4) [Mn] + [S] = (MnS) △G Θ 12 = 168822 + 98.87T [2] (J/mol)

(6.2)

[Cr] + [S] = (CrS) △G Θ 13 = 86696 + 79.496T [3, 4] (J/mol)

(6.3)

1/2(MnS) + 1/2(CrS) = (Mn0.5 , Cr0.5 )S

(6.4)

The standard Gibbs free energy change △G Θ 1 can be expressed as Θ Θ Θ Θ △G Θ 1 = △G 4 + 1/2△G 2 + 1/2△G 3 = △G 4 − 127759 + 89.183T (J/mol) (6.5)

The standard Gibbs free energy change for (Mn0.4 ,Cr0.6 )S and (Mn0.6 ,Cr0.4 )S inclusions formation can be calculated by combining the following reactions 2/5(MnS) + 3/5(CrS) = (Mn0.4 , Cr0.6 )S

(6.6)

3/5(MnS) + 2/5(CrS) = (Mn0.6 , Cr0.4 )S

(6.7)

The liquidus temperature Tliq and solidus temperature Tsol of the studied steel calculated by corresponding formulas are 1747 and 1619 K [1], respectively. Due to the lack of thermodynamic data of (Mn,Cr)S formation reported in literatures, the standard Gibbs free energy change of Reactions (6.4), (6.6) and (6.7)△G o4 △G o6 △G o7 was calculated by software FactSage 6.2. The values of △G o4 , △G o6 and △G o7 were calculated to be −18.6 kJ/mol, −20.0 kJ/mol and −16.6 kJ/mol at 1747 K, as well as −17.3 kJ/mol, −18.7 kJ/mol and −15.5 kJ/mol at 1619 K, respectively.

6.1 Sulfide Inclusions

121

The thermodynamic condition for (Mn0.5 ,Cr0.5 )S precipitation in liquid phase can be expressed as follows 1

1

1

1

1

1

2 2 2 aCr a S = lg f Mn lg aMn [%Mn] 2 f Cr2 [%Cr] 2 f S [%S] =

△G Θ 1 19.144T

(6.8)

where f Mn , f Cr and f S are the activity coefficient of dissolved manganese, chromium and sulfur in liquid steel, respectively, and can be expressed by the following formulas [5] lg f i = j

j

) ∑( j j ei [% j] + ri [% j]2

(6.9)

where ei and ri are the first-order and second-order interaction parameters, respectively, which have been summarized in Reference [1] for calculating f Mn , f Cr and fS. The stability diagram of (Mn0.4 ,Cr0.6 )S, (Mn0.5 ,Cr0.5 )S and (Mn0.6 ,Cr0.4 )S inclusions precipitation in S136 tool steel containing 13.37mass% of Cr is developed as shown in Fig. 6.4a along with the experimentally analyzed Mn and S contents in the electrode steel and ESR ingots (as points e-0, e-1, e-2, e-3 and e-4 in order). The dashed line and solid line shown in Fig. 6.4a were calculated by combining Eqs. (6.8) and (6.9) as well as the related thermodynamic data, respectively. The chemical reaction for the formation of MnS inclusions in liquid steel can be expressed as Reaction (6.2). Figure 6.4b shows the stability diagram of MnS inclusions precipitation in H13 tool steel along with the experimentally analyzed Mn and S contents in the electrode and ingots (as points e-0, e-1, e-2, e-3, e-4, e-5 and e-6 in order). It can be seen from Fig. 6.4 that the experimentally analyzed solubility products are greatly lower than the equilibrium values for (Mn,Cr)S and MnS precipitation. It indicates that (Mn0.4 ,Cr0.6 )S, (Mn0.5 ,Cr0.5 )S and (Mn0.6 ,Cr0.4 )S inclusions are unable to precipitate with Mn, Cr and S contents of points e-0, e-1, e-2, e-3 and e-4, MnS inclusions are unable to precipitate with Mn and S contents of points e-0, e-1, e-2, e-3, e-4, e-5 and e-6 above the solidus temperature. It is concluded that the precipitation of (Mn,Cr)S and MnS inclusions in the electrode steel is due to the segregation of [Mn], [S] and/or [Cr] during liquid steel solidification. The Gibbs free energy change for the chemical reactions of (Mn0.5 ,Cr0.5 )S, (Mn0.4 ,Cr0.6 )S and (Mn0.6 ,Cr0.4 )S inclusions formation in S136 steel is calculated as 58.6 kJ/mol, 57.2 kJ/mol and 60.5 kJ/mol at 1747 K, as well as MnS inclusions formation in H13 steel as 110.86 kJ/mol at 1753 K, respectively. Considering that the temperature of the liquid metal film at the electrode tip is greatly close to the liquidus temperature of the steel electrode, the thermodynamic calculation indicates that the dissociation of (Mn0.4 ,Cr0.6 )S, (Mn0.5 ,Cr0.5 )S and (Mn0.6 ,Cr0.4 )S inclusions in S136 tool steel, and MnS inclusions in H13 tool steel would take place in liquid metal phase during liquid metal film formation, and thereafter collection into a droplet at the electrode tip in ESR refining process.

122

6 Sulfide and Nitride Inclusion Evolution During ESR

Fig. 6.4 Stability diagram of sulfide inclusion precipitation in the steel: a (Mn,Cr)S in S136 steel b MnS in H13 steel

100

(a) e-0 e-1 e-2 e-3 e-4

80

[%Mn]

60

T liq

40

T so l

(Mn0.4,Cr0.6)S (Mn0.5,Cr0.5)S

=1747 K =1619 K (Mn0.6,Cr0.4)S

20 0 0.00

100

0.06 [%S]

0.09

0.12

(b)

80

[%Mn]

0.03

60 40

e-0 e-1 e-2 e-3 e-4 e-5 e-6

Tliq =1753 K

Tsol =1635 K

20 0 0.000

0.004

0.008 [%S]

0.012

0.016

Therefore, the removal of sulfide inclusions during ESR process involves the following two steps: (i) sulfide inclusions dissociate into dissolved sulfide-forming elements in liquid steel during liquid metal film formation at electrode tip, and (ii) the sulfur in liquid phase [S] is removed according to reaction (CaO) + [S] = [O] + (CaS) at the electrode tip for reduction of sulfur content in liquid steel. SEM-EDS observation confirmed that no sulfide inclusions were presented in S136 and H13 remelted ingots. It is expected that the sulfur remaining in ESR ingot is in the form of tiny sulfide inclusions which can hardly be examined by SEM-EDS. In addition, the low degree of segregation of [Mn], [Cr], and/or [S] at the solidifying front during liquid steel solidification (due to the fact that small quantity of liquid steel solidifies at the bottom of shallow liquid metal pool in a short time during ESR process) and the great reduction of sulfur from original 180 to 12 ppm in ESR process are also unfavorable to the generation of sulfide inclusions, especially large sulfide inclusions.

6.2 Sulfide in Oxide–Sulfide Complex Inclusions

123

6.2 Sulfide in Oxide–Sulfide Complex Inclusions 6.2.1 From Original Attached State to Patch-Type and Shell-Type Sulfide The consumable electrode of tool steel was produced as follows: BOF → ladle furnace (LF) refining → Ruhrstahl-Heraeus (RH) refining → continuous casting. The billets produced by the continuous casting were used as consumable electrodes in the industrial-scale electroslag rapid remelting (ESRR) trials. ESRR is a novel technology in the family of ESR technologies. The schematic diagram of ESRR apparatus is present in Fig. 6.5. ESRR has some novel design principles and exclusive devices. For example, unlike standard ESR, the ESRR unit is exclusively equipped with a current-carrying T-shaped mold and an embedded graphite ring in the upper mold. The major amount of imposed current in ESRR is taken through the graphite ring. The oxide scale on the steel electrode surface was basically removed mechanically prior to ESRR plant trials. The chemical composition of the consumable electrode is shown in Table 6.2. The pre-melted slag (30.4 mass% CaF2 , 28.7 mass% CaO, 30.7 mass% Al2 O3 , 2.5 mass% MgO, 6.7 mass% SiO2 , and others (FeO < 0.4 mass%, S < 0.02 mass%, TiO2 < 0.03 mass%)) was roasted at 973 K (700 °C) for 8 h prior to ESRR plant trials. These remelting trials were conducted under Ar gas protective atmosphere throughout the ESRR process. The steel electrodes were remelted at four different melting rates (i.e., 350, 400, 450, and 500 kg/h) in individual ESRR trial. (1) Inclusions in the Steel before and after ESRR Refining The sulfur content was lowered by ESRR from 0.0026 mass% in the electrode to 0.0018 mass%–0.0021 mass% in the ingots. The desulfurization efficiency by ESRR Fig. 6.5 Schematic diagram of ESRR apparatus

124

6 Sulfide and Nitride Inclusion Evolution During ESR

Table 6.2 Chemical composition of the consumable steel electrode (mass%) C

Si

Mn

Cr

V

Mo

Ca

Mg

Al

O

S

N

0.41

1.06

0.36

5.17

0.96

1.27

0.0017

0.0004

0.012

0.0008

0.0026

0.0062

Fig. 6.6 Element mappings of a typical inclusion in the consumable steel electrode: a complex inclusion of CaO–Al2 O3 –MgO partially wrapped by CaS

is very limited. According to the SEM-EDS determination, the inclusions in the steel electrode are oxide-sulfide type of CaS adhering to CaO–Al2 O3 –MgO inclusion (containing a small amount of SiO2 in some cases), together with occasionally observed isolated CaS-only inclusions. The inclusions observed in the steel electrode are about 2 µm. Figure 6.6 presents the element mappings of a typical inclusion in the steel electrode. The SEM backscattered electron (BSE) images and EDS results of the typical inclusions are shown in Fig. 6.7. The area fraction of sulfide phase seems to be much larger than the fraction of the oxide in the oxide-sulfide inclusion. The inclusions in the remelted ingots contain more or less sulfide phases. These inclusions are in the form of oxide-sulfide complex type. The sulfide phase is identified as CaS. From the morphological feature of CaS in the oxide-sulfide complex inclusions, these complex inclusions could be categorized into two types: (Type I) oxide associated with patch-type CaS (some CaS is embedded in the oxide portion of the complex inclusion, see examples in Figs. 6.8 and 6.9), and (Type II) oxide accompanied by shell-type CaS (a very thin CaS shell, see an example in Fig. 6.10). The CaS patch-type oxide inclusions were identified as CaO–Al2 O3 –SiO2 –MgO oxide containing a very small amount of sulfur, as shown in Fig. 6.9. According to the EDS analysis, no appreciable sulfur was detected in the oxide portion of CaO–Al2 O3 – SiO2 –MgO associated with a shell-type CaS (see EDS spectrum shown in Fig. 6.10). According to the morphological feature of CaS, it can be seen that the area fraction of CaS in the oxide-sulfide complex inclusions remarkably decreased compared with that in the original inclusions in the steel electrode. In the remelted billets,

6.2 Sulfide in Oxide–Sulfide Complex Inclusions

9

125

cps/eV Acquisition 5259

8

0.04atom% Si

7

(c)

6 5 Al

4

Fe

Ca

S

O

3 S Mn Cr Mg

2 1

Fe Ca

Cr

Si

Mn

0 1

2

3

4

5

6

7

8

9

10

keV

cps/eV

Acquisition 5231

9 S

8 7

(d)

Ca

6

Fe

5 4 3

S

2 1

Ca

Cr Fe O

Mg

Al

Cr

Si

0 1

2

3

4

5

6

7

8

9

10

keV

Fig. 6.7 Typical inclusions observed in the consumable steel electrode. (EDS spectra shown in (c) and (d) correspond to the oxide and sulfide phase in the inclusion shown in (a), respectively)

126

6 Sulfide and Nitride Inclusion Evolution During ESR

Fig. 6.8 Element mappings of oxide-sulfide complex inclusions observed in the remelted ingots. (patch-type CaS associated with CaO–Al2 O3 –SiO2 –MgO inclusion)

the observed oxide-sulfide complex inclusions exhibit a spherical morphology, indicating that these inclusions contain a certain amount of liquid phase. (2) Evolution Mechanism of Sulfide in Oxide-Sulfide Complex Inclusions during ESR Process The area fraction of CaS in the oxide-sulfide complex inclusions considerably decreased in the steel after ESRR. Meanwhile, the morphology of CaS in the oxidesulfide inclusions changed from attached state to patch-type or shell-type state. These differences could originate from the different formation pathways of CaS. Thermodynamic calculation was performed to evaluate the evolution mechanism of CaS inclusions. The temperature of liquid metal film at the electrode tip in ESR process is close to the liquidus temperature of steel electrode, [6, 7] and its superheat could hardly exceed 20–30 °C [7–9]. Thus, the temperature of the liquid metal film in the current ESRR refining process was taken as the liquidus temperature of the studied steel. The liquidus temperature Tliq and solidus temperature Tsol of the studied steel were calculated with Thermo-Calc software (TCFE7 database) to be 1748 K (1475 °C) and 1649 K (1376 °C), respectively.

6.2 Sulfide in Oxide–Sulfide Complex Inclusions

127

cps/eV Acquisition 5355

8 Al

7

(e)

6

Ca

5

Fe

4 3

S

2 1

Cr V O

Mg Si

Fe

S

Ca

0

1

2

Cr

V

3

4

5

6

7

8

9

10

keV

Fig. 6.9 CaO–Al2 O3 –SiO2 –MgO inclusion associated with patch-type CaS in the remelted ingots: (EDS spectra shown in (e) correspond to EDS point analysis at the center of the inclusions shown in (b)). Arrows indicate some of patch-type CaS in a complex inclusion

22

cps/eV Acquisition

Ca

20

(b)

Al

18 16 14 12 10 8 6

Mg

4 2 0

O

Si

Fe Ca

Fe

1

2

3

4

5

6

7

8

9

10

keV

Fig. 6.10 Element mappings of shell-type CaS associated with CaO–Al2 O3 –SiO2 –MgO inclusion in the remelted ingot. (EDS spectrum in (b) correspond to the point analysis at the center of the inclusion shown in (a))

128

6 Sulfide and Nitride Inclusion Evolution During ESR

The Gibbs free energy change for the reaction (CaS)inclusion = [Ca] + [S] in the steel electrode was calculated to be −34.5 kJ/mol at 1748 K (1475 °C), by using the standard Gibbs free energy change (△G o = 530,900−116.2 T [10] (J/mol)), firstorder interaction parameters summarized in Reference [11], and available secondC Si O = 0.012, rCa = 0.0009, rCa = 3.6 × 106 , order interaction parameters as: [3, 12] rCa Ca,O Al C Si S rCa = 0.0007, rCa = 2.9 × 106 , rS = 0.0058, rS = 0.0017, rS = −0.0009, and rSAl = 0.0009. In the case of a higher temperature at the electrode tip, the calculated value of the Gibbs free energy change is smaller. The measured concentration of the total calcium in steel includes the dissolved calcium (free calcium) and the insoluble calcium combined as inclusions. In the thermodynamic calculation for Gibbs free energy change, the concentration of soluble calcium in the steel should be used. According to the average composition of oxide inclusions and the measured oxygen content in the steel electrode, the concentration of insoluble calcium was calculated to be 6 ppm. Meanwhile, it is reasonable to expect that the concentration of insoluble calcium combined as CaS inclusions is considerably large (considering 26 ppm sulfur in the steel electrode). Several previous studies also demonstrated that the dissolved calcium in steel was estimated to be a few parts per million at most [13–15]. Hence, the concentration of the soluble calcium is taken as 3 ppm in the above thermodynamic calculation. The current thermodynamic analysis indicated that the dissociation of original CaS inclusions into soluble calcium and sulfur in liquid steel took place during the liquid metal film formation and subsequent collection into droplets at the electrode tip during ESRR. This should be responsible for the removal of the original CaS inclusions in the ESRR process, as supported by the experimental observations. This is in accordance with the previous findings regarding the removal of (Mn,Cr)S and MnS inclusions during ESR [4, 6, 16]. After ESRR of the steel, two types of CaS inclusions are generated in the remelted billets, i.e., patch-type and shell-type CaS associated with an oxide inclusion, which are unlike the presence of CaS inclusions in the steel electrode. The generation of CaS inclusions according to the direct reaction between dissolved calcium and sulfur in liquid steel ([Ca] + [S] = (CaS)inclusion ) could not take place before the starting of liquid steel solidification in mold because the necessary thermodynamic limit could not be reached, as confirmed by the thermodynamic calculation as above approach. The solubility of CaS in liquid calcium aluminates is extremely low, as determined by other researchers [17–20]. For example, at mass%CaO/mass%Al2 O3 = 1.6, the solubility of CaS in calcium aluminates melt is about 4.8 mass% at 1873 K [19]. The sulfide capacity of oxide inclusions is dependent on the chemistry of the inclusion and temperature. In the case of a lower temperature, the solubility of CaS in calcium aluminate melts is lower [17–20]. During ESRR of liquid steel, sulfur distributed between liquid steel and liquid CaO–Al2 O3 –SiO2 –MgO inclusions. As the temperature decreased in liquid metal pool during ESRR, liquid sulfide-containing oxide inclusions were gradually cooled and solidified, resulting in the decrease in the sulfur solubility in CaO–Al2 O3 –SiO2 –MgO inclusions and subsequently the precipitation of CaS on the surface of the existing oxide from the oxide-sulfide inclusion

6.2 Sulfide in Oxide–Sulfide Complex Inclusions

129

matrix, which eventually contributed to the generation of patch-type CaS around CaO–Al2 O3 –SiO2 –MgO inclusion. In the remelted billets, the compositions of the oxide inclusions with shell-type CaS were located in the fully liquid region [< 1773 K (1500 °C)], as ascertained in Sect. 4.8. Many of these oxide inclusions have a lower melting temperature [< 1673 K (1400 °C)]. Shell-type CaS formation was thought to be due to a reaction between CaO in the liquid oxide inclusions and dissolved aluminum and sulfur in liquid steel, as expressed by reaction (6.1). 3(CaO)inclusion + 3[S] + 2[Al] = 3(CaS)inclusion + (Al2 O3 )inclusion △G Θ = −859500 + 288.9T [10] (J/mol) K=

3 3 · aAl2 O3 · aAl2 O3 aCaS aCaS = 3 3 3 2 aCaO · aS · aAl aCaO · ( f S [%S])3 · ( f Al [%Al])2

(6.10)

(6.11)

where K is the equilibrium constant. aCaS , aAl2 O3 and aCaO are the activities of CaS, Al2 O3 and CaO, respectively. f S and f Al are the activity coefficients of dissolved sulfur and aluminum in liquid steel, respectively. In the Reaction (6.10), (CaO)inclusion and (Al2 O3 )inclusion indicate CaO and Al2 O3 in the oxide inclusion. The activities of CaO and Al2 O3 relative to pure solid standard states in the oxide inclusions associated with shell-type CaS were estimated with FactSage 7.1 (FToxid database) based on their average composition. The calculated activities of CaO and Al2 O3 at different temperatures are present in Fig. 6.11. The activity of CaS was taken as unity due to its extremely small solubility in the liquid calcium aluminates system [21]. In order to determine the thermodynamic driving force for Reaction (6.10) during the cooling and solidification of liquid steel, the microsegregation of aluminum and sulfur was predicted with the modified Clyne-Kurz model [22–24], in which the equilibrium partition coefficient (k i ) and diffusion coefficient in solid steel were cited from Reference [25]. The secondary dendrite arm spacing (SDAS) of the as-cast remelted billet was experimentally determined to be 117 µm. The calculated concentrations of sulfur and aluminum in interdendritic liquid with the progress of solidification are shown in Fig. 6.12, as well as the variation of the temperature of residual liquid steel estimated by the equation reported in Reference [26]. The dependence of the thermodynamic driving force (the Gibbs free energy change) for the reaction between CaO in the oxide inclusions and the dissolved aluminum and sulfur in liquid steel on the temperature and solid fraction is present in Fig. 6.13. It can be seen that the reaction could not take place thermodynamically before the starting of liquid steel solidification. As shown in Fig. 6.13, the thermodynamic driving force progressively increases with the increase in the solid fraction due to increasing concentrations of enriched aluminum and sulfur. The thermodynamic calculation suggested that Reaction (6.10) did take place until the solid fraction exceeded the critical value of 0.34 during the ESRR process. During the solidification of liquid steel in the ESRR process, the enriched sulfur and aluminum in the residual liquid steel diffused to the interface of CaO–Al2 O3 –SiO2 –MgO inclusion

6 Sulfide and Nitride Inclusion Evolution During ESR Activity of component in oxide inclusion

130 0.45

CaO Al2O3

0.36 0.27 0.18 0.09

1650

1700

1750 1800 Temperature (K)

1850

1900

Fig. 6.11 Activities of CaO and Al2 O3 relative to pure solid standard states in the oxide inclusions with shell-type CaS (average composition) at different temperatures estimated with FactSage 7.1

1800 [S] [Al] T

0.06

1770 1740

0.04 1710 0.02

Temperature (K)

Concentration (mass%)

0.08

1680 1650

0.00 0.0

0.2

0.4 0.6 Solid fraction

0.8

1.0

Fig. 6.12 Variation of the sulfur and aluminum concentration in the interdendritic liquid with the progress of liquid steel solidification predicted by the modified Clyne–Kurz equation [22–24] and the temperature at the solid–liquid interface (Color Fig. online)

and liquid steel at the bottom of the mold, and subsequently reacted with CaO in this oxide inclusion. As the progress of Reaction (6.10), dissolved sulfur and aluminum constantly diffused through the product layer, and CaO in the reaction layer was consumed. Once CaS layer formed, the mass transfer of sulfur and aluminum to the reaction site through the product slayer became rather slow due to the low diffusivity.

Fig. 6.13 Thermodynamic driving force for the reaction between CaO in the oxide inclusion and dissolved aluminum and sulfur in liquid steel against the temperature and various solid fractions

Gibbs free energy change (kJ/mol)

6.2 Sulfide in Oxide–Sulfide Complex Inclusions

150

131

Temperature (K) 1875 1825 1775

100 Tliq =1748 K

50 0 -50 -100 -150

0.0

0.2 0.4 0.6 0.8 Solid fraction

1.0

6.2.2 From Original Patch-Type to Shell-Type Sulfide H13 tool steel was remelted for this study. The detailed experimental procedure has been described in Sect. 3.5.5.2. The inclusions in the steel before ESR refining are oxide-sulfide type of patch-type (Ca,Mn)S adhering to CaO–Al2 O3 –SiO2 –MgO inclusion. According to attempts varying different acceleration voltages and measurement time, SEM-EDS analysis could not clarify whether any Mn, Ca and S did exist or not in different parts of these complex micro-inclusions. EPMA line scanning and element mappings of oxide-sulfide inclusions confirm that Mn and S did only appear on the edges of an oxide-sulfide complex inclusion, and Ca was present in both oxide and sulfide phases. The EPMA line scanning and element mappings of the oxide-sulfide inclusions are shown in Figs. 4.3 and 4.4 (in Sect. 4.4), respectively. According to CaS-MnS binary phase diagram, [27] the sulfide phase in these inclusions was considered to be (Ca,Mn)S solid solution. All oxide inclusions are accompanied with (Ca,Mn)S inclusions as embedded form. Although SEM-EDS only offers a two-dimensional analysis and oxide inclusions are not always present on the polished surface of steel samples, EDS measurement with a collection time over 100 s did reveal the presence of oxide inside the sulfide. No single-phased sulfide inclusions were found in the consumable electrode. The morphology of these dual-phased inclusions in H13 steel is quite similar with the observations showing patch-type MnS precipitation on Mn-silicate inclusion in Si-Mn deoxidized steel, [28, 29] but different from the morphologies of oxysulfide inclusions of sulfide + Al2 O3 and sulfide + MgO·Al2 O3 [6, 16, 28]. The oxide inclusions in the liquid metal pool of different ESR heats were identified as CaO–Al2 O3 –SiO2 –MgO, CaO–Al2 O3 –MgO, and a small amount of MgAl2 O4 . No sulfide inclusions are present in the samples collected from liquid metal pool. Al2 O3 inclusions were occasionally observed in the liquid metal pool.

132

6 Sulfide and Nitride Inclusion Evolution During ESR

Three types of oxide inclusions were found in remelted ingots, i.e., CaO–Al2 O3 – SiO2 –MgO, CaO–Al2 O3 –MgO and MgAl2 O4 . No CaO–Al2 O3 –SiO2 –MgO inclusions with the compositions similar to those in consumable electrode were observed in ESR ingots. Nearly half proportion of CaO–Al2 O3 –SiO2 –MgO inclusions was invariably associated with CaS as a poor discontinuous shell (see Figs. 3.14 and 4.6 for examples). Unlike the sulfides in consumable steel electrode, no patch-type sulfide inclusions associated with oxide inclusions were observed in remelted ingots. The presence of patch-type (Ca,Mn)S adhering to CaO–Al2 O3 –SiO2 –MgO inclusion in consumable steel electrode was attributed to the fact that (i) oxide inclusions were in the form of liquid state at the moment of sulfide inclusions formation, (ii) sulfur distributed between liquid steel and liquid oxide inclusions. The sulfide capacity of oxide inclusions is dependent on the chemistry of the inclusion and temperature. As the temperature decreased, liquid sulfide-containing oxide inclusions gradually solidified, contributing to the decrease in sulfur solubility in the oxide inclusions. Consequently, CaS precipitated on the surface of oxide from inclusion matrix, accompanying with Mn transfer into the sulfide phase, eventually resulting in the formation of patch-type (Ca,Mn)S around oxide inclusion. The experimental determination of inclusions in liquid metal pool and remelted ingots revealed that these (Ca,Mn)S inclusions were fully removed before liquid metal droplets had collected in the liquid metal pool during ESR. The inclusion characterization shows that no sulfide inclusions were present in the liquid metal pool, whereas CaS exhibited as a poor ring around CaO–Al2 O3 – SiO2 –MgO inclusion in the remelted ingots. The calculation of both equilibrium and non-equilibrium sulfide inclusion precipitation predicted with Thermo-Calc software (TCFE7 database, using Scheil-Gulliver model) showed that no CaS inclusions formed during solidification and subsequent cooling of liquid steel, even though assuming that the measured calcium in steel existed as dissolved form completely. It was therefore concluded that CaS precipitated from oxide inclusion matrix during the transformation of original CaO–Al2 O3 –SiO2 –MgO inclusions from liquid state to solid state CaO–Al2 O3 –SiO2 –MgO inclusions (as illustrated in Sect. 3.5.5.2) due to the decrease in the sulfur solubility in oxide inclusions.

6.3 Nitride Inclusions In high-Al steel, transverse cracking normally generate during casting caused by AlN inclusions because considerable AlN inclusions form in the steel [30]. The aluminum content in NAK80 tool steel is about 1 mass%. For the high-Al content, generally AlN is the typical inclusion in the steel. Protective atmosphere ESR trials were conducted to produce NAK80 tool steel. The pre-melted slag (60mass% CaF2 , 20mass% CaO, and 20mass% Al2 O3 ) was used in each ESR trial. The chemical composition of NAK80 tool steel used as the consumable electrodes is listed in Table 6.3. Figure 6.14a shows the typical AlN inclusions in NAK80 tool steel electrode detected by SEM-EDS. The typical AlN inclusions observed in the ingots are

6.3 Nitride Inclusions

133

Table 6.3 Chemical composition of consumable NAK80 steel electrode (mass%) C

Mn

Si

P

S

Cr

Ni

Cu

Mo

sol. Al

N

T.O

0.07

1.70

0.14

0.014

0.002

0.16

3.04

1.09

0.34

0.84

0.0037

0.0034

Fig. 6.14 Typical AlN inclusions in NAK80 steel: a electrode, b–d ESR ingots

presented in Fig. 6.14b–d. The morphology of all observed AlN inclusions in the electrode and ESR ingots is quadrangle with clear angularities, except for some clusters of AlN inclusions in the ingot. The size of most AlN inclusions is about 2 µm. It was confirmed by SEM-EDS analysis that all AlN inclusions in NAK80 tool steel before and after ESR refining are single-phased precipitates. According to the chemical reaction of AlN formation as shown in Eq. (6.12) and the relevant thermodynamic parameters, Gibbs free energy change for Reaction (6.12) was calculated as 28.35 kJ/mol at 1772 K, (solidus temperature Tsol 1748 K and liquidus temperature Tliq 1772 K) [31], for remelting of NAK80 steel. The result suggests that AlN inclusions would dissociate in liquid metal phase during liquid metal film formation, and thereafter collection into a droplet at the electrode tip. [Al] + [N] = (AlN) △GΘ 9 = −245990 + 107.59T[32](J/mol)

(6.12)

The thermodynamic condition for AlN inclusion precipitation in liquid steel can be expressed as follows: lg aAl aN = lg f Al [%Al] f N [%N] = 5.620 −

12849.457 T

(6.13)

where f Al and f N are the activity coefficients of dissolved aluminum and nitrogen in liquid steel, respectively, and can be calculated with Eq. (6.9). The stability diagram of AlN precipitation is developed as shown in Fig. 6.15 along with the experimentally analyzed Al and N contents in the electrode and ESR ingots (as points e-0, e-1, e-2 and e-3 in order). The dashed line and solid line shown in Fig. 6.15 were calculated by combining Eqs. (6.9) and (6.13) as well as the thermodynamic data given in Reference [31], respectively. It can be seen that the experimentally analyzed solubility products are greatly lower than the equilibrium values for AlN precipitation. The thermodynamic calculation indicates that AlN

134

6 Sulfide and Nitride Inclusion Evolution During ESR

Fig. 6.15 Stability diagram of AlN inclusion formation in NAK80 steel

0.04

[%N]

0.03

e-0 e-1 e-2 e-3

Tliq =1772 K

0.02 Tsol =1748 K

0.01

0.00 0.6

0.7

0.8

0.9

1.0

1.1

[%Al]

inclusions are unable to precipitate in liquid steel with Al and N contents of points e-0, e-1, e-2 and e-3 above the solidus temperature and liquidus temperature of the steel. When the Al content in liquid steel is 0.84 mass%, the critical N content for AlN inclusion precipitation at liquidus temperature 1772 K is 0.025 mass% calculated by using Eq. (6.13). It can be seen that the calculated lowest N content for AlN formation in liquid metal pool is much greater than that in the studied steel. Therefore, it is impossible to precipitate AlN inclusion in liquid steel at liquidus temperature or above. When Al and N contents are 0.84 mass% and 0.0032 mass% in liquid steel, the calculated critical temperature for AlN precipitation is 1577 K, which is much lower than the solidus temperature 1748 K. This result indicates that AlN inclusion is unable to precipitate in liquid metal pool, or at the stage of metal droplets falling through the slag pool, or during the formation of droplets at electrode tip in ESR process based on thermodynamic equilibrium analysis. Meanwhile, thermodynamic calculation indicates that even if there are AlN inclusions in consumable electrode, these AlN inclusions still can decompose at the above-mentioned three stages. Fu et al. [33] reported that most of the original inclusions in consumable electrode could be removed at the stage of metal droplets falling through the slag pool and formation of droplets on electrode tip during ESR process. The inclusions in final ESR ingot are the newly-formed inclusions during the solidification of liquid steel in water-cooled mold. Similar results have been reported by Li et al. [34] and Kay et al. [35] based on the experimental work. A few inclusions observed in E1, E2, and E3 are expected to be the original inclusions in consumable electrode. With the formation of liquid metal droplets at the electrode tip, liquid steel begins to solidify at the bottom of liquid metal pool in water-cooled mold during the ESR process. During the cooling of liquid steel in water-cooled mold from liquid metal pool temperature to liquidus temperature, and then to solidus temperature, the solute is rejected into interdendritic liquid phase, which results in the enrichment of Al and N in residual liquid steel between solid steel dendritic arms at solidifying front because of the difference in solubility of solutes between liquid and solid phases [25]. When the product of [%Al]×[%N] in interdendritic liquid steel exceeds the equilibrium

6.3 Nitride Inclusions

135

value for AlN precipitation at certain temperature, the chemical Reaction (6.12) for AlN formation may occur. Because complete diffusion of solutes in solid phases is impossible, the following equation can be used to calculate the solute content in residual liquid phase during solidification by assuming no diffusion in solid phase [25, 36]. CL = C0 (1 − f S )(ki −1)

(6.14)

where CL is the concentration of solute in residual liquid phase during solidification, C0 is the initial concentration of solute in liquid steel, f S is the solid fraction, and ki is the equilibrium partition ratio of solute i between liquid and solid phases. During the cooling of liquid steel, Al and N contents in liquid phase can be calculated as follows: [%Al] = [%Al]0 (1 − f S )(kAl −1)

(6.15)

[%N] = [%N]0 (1 − f S )(kN −1)

(6.16)

where kAl and kN are equal to 0.6 and 0.27, respectively[30, 37]. The relationship between the product of [%Al]×[%N] and solid fraction f S is shown in Fig. 6.16. The calculated critical value of the product of [%Al]×[%N] for AlN precipitation at liquidus temperature using Eq. (6.13) is also shown in Fig. 6.16. It can be observed that the product of [%Al]×[%N] increases with increasing solid fraction f S . Al and N in the residual liquid phase between dendritic arms would enrich at solidifying front with the increase of solid fraction during solidification. When the product of [%Al]×[%N] exceeds the calculated critical value for AlN precipitation at liquidus temperature, AlN would precipitate in residual liquid phase at solidifying front. The critical value of solid fraction f S for AlN inclusions precipitation in the NAK80 steel containing 0.84 mass% Al and 0.0032 mass% N, 0.81 mass% Al and 0.0033 mass% N, 1.05% Al and 0.0032 mass% N was calculated to be 0.840, 0.841 and 0.800, respectively, as shown in Fig. 6.16a–c. AlN inclusions form when the solid fraction f S exceeds these critical values during liquid steel solidification. It is proposed that the AlN inclusions removal in the ESR process is the dissolution of AlN inclusions into soluble aluminum and nitrogen taking place in liquid steel at the electrode tip during liquid metal film formation, and thereafter collection into a droplet. Directional solidification proceeds from the bottom of shallow liquid metal pool during ESR process. The growth directions of the dendrites are mostly perpendicular to the bottom of the liquid metal pool. With the growth of dendrites, dendrites will be interconnected with each other, and then the residual liquid steel among interconnected dendrites will be closed up. Under this condition, the diffusion of solutes in interdendritic liquid steel was hindered, which would induce the enrichment of Al and N at solidifying front. Thereafter, AlN inclusions would precipitate at solidifying front caused by the enrichment of Al and N. The AlN inclusions formed in interdendritic liquid steel are nearly unable to float out, which are closed

136

6 Sulfide and Nitride Inclusion Evolution During ESR 0.10

(a)

(b)

0.08 [%Al]x[%N]

[%Al]x[%N]

0.08

0.10 Calculated value from fs Calculated critical value from Tliq

0.06 0.04

0.06 0.04 AlN precipitation

AlN precipitation

0.02

0.02

no AlN precipitation

no AlN precipitation fs=0.840

0.00 0.0

0.10

(c)

0.08 [%Al]x[%N]

Calculated value from fs Calculated critical value from Tliq

0.2

0.4 0.6 0.8 Solid fraction, fs

fs=0.841

0.00

1.0

0.0

0.2

0.4 0.6 0.8 Solid fraction, fs

1.0

Calculated value from fs Calculated critical value from Tliq

0.06 0.04 AlN precipitation

0.02 no AlN precipitation fs=0.800

0.00 0.0

0.2

0.4 0.6 0.8 Solid fraction, fs

1.0

Fig. 6.16 Relationship between AlN precipitation in NAK80 steel and liquid steel solid fraction

up by interconnected dendrites. Thereafter, some AlN clusters form among dendrites because fine precipitates collide with each other, which has been confirmed by SEM analysis as shown in Fig. 6.14. Elimination of inclusions formed at solidifying front during solidification by floating up is almost impossible [25, 35], especially during the ESR process [35, 38], due to its relatively rapid cooling rate. Under this condition, the precipitated AlN inclusions at solidifying front remain in the solidified ingot.

6.4 Summary It is the microsegregation of sulfide-forming elements and nitride-forming elements during liquid steel solidification that result in the precipitation of sulfide inclusions in the consumable steel electrode, and AlN inclusions in high-Al consumable electrode and ESR ingot steel. The removal of (Mn,Cr)S, MnS and AlN inclusions during ESR process results from the dissolution of sulfide and AlN inclusions into dissolved

References

137

sulfide-forming and nitride-forming elements in liquid steel taking place at the electrode tip during liquid metal film formation, and thereafter collection into a droplet at the electrode tip. The absence of (Mn,Cr)S and MnS inclusions in ESR ingots is expected to be due to the low degree of segregation of [Mn], [Cr], and/or [S] during liquid steel solidification and the great reduction of sulfur content in ESR process. The shell-type CaS in the oxide-sulfide complex inclusion generated as a result of the reaction between CaO in the liquid CaO–Al2 O3 –SiO2 –MgO inclusion and dissolved aluminum and sulfur in liquid steel until the solid fraction of liquid steel exceeded 0.34 during the ESRR process. The formation of patch-type CaS in the oxide-sulfide complex inclusion originated from the decreasing solubility of CaS in the CaO–Al2 O3 –SiO2 –MgO inclusion melt during the cooling of liquid steel in ESRR process. The original patch-type (Ca,Mn)S inclusion adhering to CaO–Al2 O3 –SiO2 – MgO inclusion was fully removed before liquid metal droplets collected in the liquid metal pool, whereas CaS appearing as a poor ring around CaO–Al2 O3 –SiO2 –MgO inclusion precipitated from the oxide inclusion matrix during the transformation of the original CaO–Al2 O3 –SiO2 –MgO inclusions from liquid- to solid-state CaO– Al2 O3 –SiO2 –MgO inclusions due to the decrease in the sulfur solubility in oxide inclusions. Some of the detected AlN inclusions in the ingot could be the original AlN inclusions which have not been removed in ESR process. However, there is no method to distinguish the original inclusions from the existing AlN inclusions in the remelted ingots.

References 1. Doostmohammadi H, Jönsson P G, Komenda J, et al. Inclusion characteristics of bearing steel in a runner after ingot casting[J]. Steel Res. Int., 2010, 81(2): 142–149. 2. Turkdogan E T. Physical chemistry of high temperature technology[M]. Academic Press Inc., 1980, 8–12. 3. Sigworth G K, Elliott J F. The thermodynamics of liquid dilute iron alloys[J]. Met. Sci., 1974, 8(1): 298–310. 4. Shi C B, Chen X C, Guo H J, et al. Assessment of oxygen control and its effect on inclusion characteristics during electroslag remelting of die steel[J]. Steel Res. Int., 2012, 83(5): 472–486. 5. Wagner C. Thermodynamics of Alloys[M]. Cambridge: Addison-Wesley Press, 1952, 51. 6. Shi C B, Chen X C, Guo H J, et al. Control of MgO·Al2 O3 spinel inclusions during protective gas electroslag remelting of die steel[J]. Metall. Mater. Trans. B, 2013, 44(2): 378–389. 7. Wei J H, Mitchell A. Changes in composition during A.C. ESR-I. theoretical development[J]. Acta Metall. Sin., 1984, 20(5): 261–279. (In Chinese). 8. Fraser M E, Mitchell A. Mass transfer in the electroslag process: part 1 mass-transfer model[J]. Ironmak. Steelmak., 1976, 3(5): 279–287. 9. Mitchell A, Szekely J, Elliott J F. Application of mathematical modelling to the ESR process[C]. Electroslag refining: proceedings of a conference on electroslag refining[M]. London: The Iron and Steel Institute, 1973, 3–15. 10. Suito H, Inoue R. Thermodynamics on control of inclusions composition in ultraclean steels[J]. ISIJ Int., 1996, 36(5): 528–536. 11. Shi C B, Zheng D L, Guo B S, et al. Evolution of oxide–sulfide complex inclusions and its correlation with steel cleanliness during electroslag rapid remelting (ESRR) of tool steel[J]. Metall. Mater. Trans. B, 2018, 49(6): 3390–3402.

138

6 Sulfide and Nitride Inclusion Evolution During ESR

12. Ohta H, Suito H. Activities in CaO-SiO2 -Al2 O3 slags and deoxidation equilibria of Si and Al[J]. Metall. Mater. Trans. B, 1996, 27(6): 943–953. 13. Kang Y J, Li F, Morita K, et al. Mechanism study on the formation of liquid calcium aluminate inclusion from MgO·Al2 O3 spinel[J]. Steel Res. Int., 2006, 77(11): 785–792. 14. Verma N, Pistorius P C, Fruehan R J, et al. Transient inclusion evolution during modification of alumina inclusions by calcium in liquid steel: part II. results and discussion[J]. Metall. Mater. Trans. B, 2011, 42(4): 720–729. 15. Yoshioka T, Shimamura Y, Karasev A, et al. Mechanism of a CaS formation in an Al-killed high-S containing steel during a secondary refining process without a Ca-treatment[J]. Steel Res. Int., 2017, 88(10): 1700147. 16. Shi C B, Yu W T, Wang H, et al. Simultaneous modification of alumina and MgO·Al2 O3 inclusions by calcium treatment during electroslag remelting of stainless tool steel[J]. Metall. Mater. Trans. B, 2017, 48(1): 146–161. 17. Kor G J W, Richardson F D. Sulfur in lime-alumina mixtures[J]. J. Iron Steel Inst., 1968, 206: 700–704. 18. Sharma R A, Richardson F D. Activities in lime-alumina melts[J]. J. Iron Steel Inst., 1961, 198: 386–390. 19. Ozturk B, Turkdogan E T. Equilibrium sulfur distribution between molten calcium aluminate and steel. Part 1. Calcium sulfide-calcium oxide-aluminum oxide melts equilibrated with liquid iron containing aluminum and sulfur[J]. Met. Sci., 1984, 18(6): 299–305. 20. Fukaya H, Miki T. Phase equilibrium between CaO·Al2 O3 saturated molten CaO–Al2 O3 –MnO and (Ca,Mn)S solid solution[J]. ISIJ Int., 2011, 51(12): 2007–2011. 21. Fujisawa T, Inoue S, Takagi S, et al. Solubility of CaS in the molten CaO-Al2 O3 -CaS slags and the equilibrium between the slags and molten steel[J]. Tetsu-to-Hagané, 1985, 71(7): 839–845. 22. Clyne T W, Kurz W. Solute redistribution during solidification with rapid solid state diffusion[J]. Metall. Trans. A, 1981, 12(6): 965–971. 23. Voller V R, Beckermann C. A unified model of microsegregation and coarsening[J]. Metall. Mater. Trans. A, 1999, 30(8): 2183–2189. 24. Won Y, Thomas B G. Simple model of microsegregation during solidification of steels[J]. Metall. Mater. Trans. A, 2001, 32(7): 1755–1767. 25. Choudhary S K, Ghosh A. Mathematical model for prediction of composition of inclusions formed during solidification of liquid steel[J]. ISIJ Int., 2009, 49(12): 1819–1827 26. Ma Z, Janke D. Characteristics of oxide precipitation and growth during solidification of deoxidized steel[J]. ISIJ Int., 1998, 38(1): 46–52. 27. Piao R, Lee H G, Kang Y B. Activity measurement of the CaS–MnS sulfide solid solution and thermodynamic modeling of the CaO–MnO–Al2 O3 –CaS–MnS–Al2 S3 system[J]. ISIJ Int., 2013, 53(12): 2132–2141. 28. Wakoh M, Sawai T, Mizoguchi S: Effect of S content on the MnS precipitation in steel with oxide nuclei[J]. ISIJ Int., 1996, 36(8): 1014–1021. 29. Kim H S, Lee H G, Oh K S. MnS precipitation in association with manganese silicate inclusions in Si/Mn deoxidized steel[J]. Metall. Mater. Trans. A, 2001, 32(6): 1519–1525. 30. Yin H B. Inclusion Characterization and thermodynamics for high-Al advanced high-strength steels[J]. Iron Steel Technol., 2006, 25(6): 64–73. 31. Shi C B, Chen X C, Guo H J. Characteristics of inclusions in high-Al steel during electroslag remelting process[J]. Int. J. Miner. Metall. Mater., 2012, 19(4): 295–302. 32. Wada H, Pehlke R D. Nitrogen solubility and aluminum nitride precipitation in liquid iron, Fe–Cr, Fe–Cr–Ni, and Fe–Cr–Ni–Mo alloys[J]. Metall. Mater. Trans. B, 1978, 9(3): 441–448. 33. Fu J, Zhu J. Change of oxide inclusions during electroslag remelting process[J], Acta Metall. Sin., 1964, 7(3): 250–262. (in Chinese) 34. Li Z B, Zhou W H, Li Y D. Mechanism of removal of non-metallic inclusions in the ESR process[J], Iron Steel, 1980, 15(1): 20–26. (in Chinese) 35. Kay D A R, Pomfret R J. Removal of oxide inclusions during ac electroslag remelting[J]. J. Iron Steel Inst., 1971, 209(12): 962–965

References

139

36. Kurz W, Fisher D J. Fundamentals of Solidification[M]. 3rd Ed, Trans Tech Publications, Switzerland, 1992, 280. 37. Turkdogan E T. Fundamentals of Steelmaking[M]. London: The Institute of Materials, 1996, 298. 38. Mitchell A. Oxide inclusion behavior during consumable electrode remelting[J]. Ironmaking Steelmaking, 1974, 1(3): 172–179.

Chapter 7

Evolution of Original Oxide Inclusions During ESR

Abstract The characteristics of the original inclusions in consumable electrode affect the evolution of the inclusions during ESR process and the characteristics of the inclusions in ESR ingot. Comparing the change of the inclusions in size, distribution, morphology, and chemical composition before and after ESR is a very effective method to assess steel cleanliness and the inclusion evolution mechanism during the ESR. This chapter presents the earlier work on the sites of oxide inclusion removal during ESR. The evolution of Al2 O3 , MgO·Al2 O3 and Al2 O3 –Ti2 O3 inclusions during the ESR is discussed. The transformation of calcium aluminate inclusions during the ESR is typically assessed. The more information on calcium aluminate inclusion evolution during the ESR is present in Chaps. 4 and 9. The generation of manganese silicate inclusions in the steel is critically discussed. The evolution of manganese silicate inclusions in the steel is described in Chap. 8. The crucial factors that affecting inclusion characteristics in ESR are presented and discussed in this chapter.

The process of consumable electrode manufacturing (such as ladle furnace (LF), vacuum degassing (VD), and Ruhrstahl-Heraeus (RH) refining) for ESR would give rise to the inclusions with different compositions and sizes, depending on liquid steel composition, slag chemistry, deoxidation schemes and activities of inclusionforming species (typically soluble oxygen, aluminum, calcium, magnesium, silicon, sulfur, and nitrogen), heating scheme, vacuum and stirring states. Different deoxidation schemes (deoxidizing agent chemistry, and its addition amount and timing) are employed for liquid steel deoxidation according to the requirements of individual steel grade when manufacturing the consumable electrode for ESR. Non-metallic inclusions with different chemical compositions basically experience various evolution trajectories during ESR. Furthermore, the removal degree (number proportion) of the original inclusions during ESR is also largely dependent on the compositions and sizes of these inclusions, and other parameters (liquid metal and slag compositions, melting rate and capacity of ESR, etc.) have more or less influences on the original inclusion removal.

© Metallurgical Industry Press 2023 C. Shi et al., Electroslag Remelting Towards Clean Steel, https://doi.org/10.1007/978-981-99-3257-3_7

141

142

7 Evolution of Original Oxide Inclusions During ESR

7.1 Sites of Oxide Inclusion Removal During ESR Over the past decades, the removal of inclusions in the ESR process has been the subject of many investigations. In the early years of booming ESR (1960s–1980s), the studies on non-metallic inclusions have been focused on revealing the removal sites (electrode tip, molten slag pool and liquid metal pool) of oxide inclusions during the ESR process. The early researchers claimed that the inclusions removal during ESR process is the result of inclusion floating up in liquid metal pool, and then absorbed by molten slag [1, 2]. However, this viewpoint ignores the reactions of inclusionsteel-slag system in the ESR process. The results from examining the steel samples taken from the electrode, metal droplets at electrode tip, dropping metal droplets, liquid metal pool and remelted ingot in Ref. [3] demonstrate that almost all original oxide inclusions in the steel electrode are removed before they enter into the liquid metal pool during the ESR process and most of the original oxide inclusions are removed at the electrode tip during liquid metal films formation and their collection into metal droplets. Li et al. [4] proposed that the removal of inclusions occurred mainly at the electrode tip during liquid metal droplets formation in ESR of bearing steel, most of the original inclusions in electrode were removed at the electrode tip, and inclusion removal in liquid metal pool plays a minor role. Evseyev and Filippov also demonstrated that the majority of the oxide inclusions in electrode was removed at the electrode tip during ESR [5]. The mechanism of inclusion removal proposed by Kay and Pomfret [6] is that oxide inclusions dissolve in the liquid metal at the electrode tip and that oxygen and deoxidant are transferred to slag by slag-metal reactions taking place at the tip interface, metal droplet-slag interface and slag-metal pool interface. In contrast, Mitchell [7] claimed that the physical dissolution of inclusions is a minor part of overall inclusion removal. The present authors claim that the contribution degree of the dissolution of inclusions to overall inclusion removal differs according to the types of the inclusions, examples of sulfide inclusion removal through its full dissolution will be discussed in Sect. 6.1. Removal of inclusions by flotation from bulk liquid in the liquid metal pool during ESR is always considered unlikely [6–8]. It is the adhesion and absorption of oxide inclusions by molten slag, because the thickness of liquid metal film is very small at the electrode tip, that contributes to the removal of oxide inclusions in the ESR process [8]. Up to now, it has been widely accepted that non-metallic inclusion removal in the ESR process takes place predominantly at the stage of liquid metal films formation and subsequent collection into liquid metal droplets at the electrode tip, whereas the stage when metal droplets pass through the slag pool and the process in the liquid metal pool contribute in an extremely small manner (does not play an important role).

7.2 Al2 O3 and MgO·Al2 O3 Inclusions

143

7.2 Al2 O3 and MgO·Al2 O3 Inclusions Aluminum is the most widely used deoxidizing agent for liquid steel deoxidation in the steelmaking, consequently generating endogenous Al2 O3 and MgO·Al2 O3 inclusions in aluminum-killed steel. These two types of oxide inclusions are very common in the consumable steel electrode for ESR. In the present article, MgO·Al2 O3 inclusion represents both MgO·Al2 O3 spinel and low-MgO-containing MgO·Al2 O3 inclusion (with composition in the spinel + alumina two-phase region, namely the mixture of Al2 O3 with very small amounts of MgO·Al2 O3 spinel [9]). As for original Al2 O3 inclusions from the electrode, they would be converted to MgO·Al2 O3 inclusions in liquid steel after soluble magnesium pickup during the ESR. During the ESR without Mg-containing alloy addition, steel-slag reaction is the only potential source of soluble magnesium pickup in liquid steel. If the reduction of MgO from the slag by soluble species in liquid steel could take place, the dissolved magnesium originating from this source would be transported into the liquid steel and modify Al2 O3 inclusions to MgO·Al2 O3 inclusions. Nevertheless, commercial slag for ESR contains low MgO content (much smaller than 5 mass%) and high Al2 O3 content (higher than 20 mass%), and the reduction reaction between MgO in the slag and soluble specie in liquid steel has to be critically assessed. The reaction between soluble aluminum in liquid steel and MgO in the slag is expressed as follows:   2[Al] + 3(MgO) = 3 Mg + (Al2 O3 ) G  = 980685 − 328.486T [10, 11] (J/mol) (7.1)

Figure 7.1 exhibits the Gibbs free energy change for the reaction between MgO in the slag and the dissolved aluminum in liquid steel against the soluble aluminum content in liquid steel. Slag A1 (30.4 mass% CaF2 , 28.7 mass% CaO, 30.7 mass% Al2 O3 , 2.5 mass% MgO, 6.7 mass% SiO2 ) is a typical slag used in ESR industrial production. Slag B1 (32.9 mass% CaF2 , 28.7 mass% CaO, 30.7 mass% Al2 O3 , 5 mass% MgO, 1.7 mass% SiO2 ) is designed in order to assess the reaction (7.1) at upper limit of MgO content. A tool steel with the chemical composition of 0.41 mass% C–1.06 mass% Si–0.36 mass% Mn–5.17 mass% Cr–0.96 mass% V–1.27 mass% Mo–0.0017 mass% Ca–0.0002 mass% Mg–0.0008 mass% O is used in the thermodynamic calculation. In consideration of possible higher temperatures of molten slag, 1800 °C is also employed for comparison as shown in Fig. 7.1. The Gibbs free energy change for Reaction (7.1) was calculated in combination with the formula for calculating the activity coefficients [12] and the reported interaction parameters [10, 13] as well as the estimated activities of slag components relative to pure solid standard states at 1600 and 1800 °C with FactSage 7.2 (FToxid database). It can be learned from Fig. 7.1 that Reaction (7.1) would occur toward the righthand side only when the MgO content of the slag is higher than 5 mass% and soluble aluminum content is higher than 0.075 mass% at 1800 °C, but this is not case in the normal ESR practice, otherwise the reaction could not take place thermodynamically during the ESR. It suggests that steel-slag reaction cannot provide magnesium pickup

Fig. 7.1 Gibbs free energy change for the reaction between MgO in the slag and soluble aluminum in liquid steel against the soluble aluminum content of liquid steel and the MgO content of the slag

7 Evolution of Original Oxide Inclusions During ESR

Gibbs free energy change/(kJ·mol -1)

144

200

1600 oC 1600 oC 1800 oC 1800 oC

150

Slag A1 Slag B1 Slag A1 Slag B1

100 50 0 -50

0.00 0.04 0.08 0.12 0.16 Soluble aluminum content/mass%

in Al-killed steel, let alone lower aluminum steel and the temperatures lower than 1800 °C. For this case, Al2 O3 inclusions cannot be modified to MgO·Al2 O3 inclusions during the ESR process. Figure 7.2 shows examples of Al2 O3 inclusions in S136 consumable steel electrode with SEM-EDS two-dimensional observation. Since the scanned area is larger than the observed inclusion itself when the size of the inclusion is extremely small, the elements Fe and Cr in the matrix are inevitably detected by EDS except for the atoms in the inclusion. In addition, the SEM-EDS technique cannot quantitatively determine the lightest element such as oxygen, the measured value of oxygen content in inclusion can only be regarded as a reference value. The inclusions observed in the electrode are single-phased Al2 O3 inclusions, the large inclusions in the form of Al2 O3 core surrounded by an outer sulfide layer, besides large (Mn,Cr)S, as shown in Fig. 7.2. The SEM image together with element mappings of a typical inclusion in the form of two-layer structure in the electrode is shown in Fig. 7.3. It is confirmed from these element mappings that the complex

Fig. 7.2 Typical Al2 O3 inclusions in the consumable electrode

7.2 Al2 O3 and MgO·Al2 O3 Inclusions

145

inclusion is composed of Al2 O3 core surrounded by an outer sulfide layer. It is believed that Al2 O3 inclusions act as the sites for sulfide generation. The morphology of Al2 O3 particles in the electrode acted as the core of the complex inclusions is either near-spherical or irregular, and these Al2 O3 particles are 2–4 µm in the size. The three-dimensional morphologies and the compositions of the electrolytically extracted Al2 O3 inclusions were analyzed by SEM–EDS, as shown in Fig. 7.4. It can be observed that the extracted Al2 O3 inclusions are near-spherical shape or clusters, some of which are the clusters composed of small spherical Al2 O3 inclusions. Figure 7.5 shows the morphology and compositions of typical inclusions observed in each ESR ingot. It should be noted that all oxide inclusions in the as-cast ingots are identified as single-phased Al2 O3 inclusions. These as-cast ingots are produced by the ESR in protective argon gas atmosphere and/or deoxidating agent addition, as described in Reference [14]. Neither sulfide inclusion nor the inclusion in the form of

Fig. 7.3 SEM image and element mappings of a typical inclusion in the form of two-layer structure in consumable electrode

Fig. 7.4 SEM micrographs of the three-dimensional morphology of the electrolytically extracted Al2 O3 inclusions in the consumable electrode

146

7 Evolution of Original Oxide Inclusions During ESR

Fig. 7.5 SEM images and EDS results of typical inclusions observed in as-cast ingots: a–c E1, d–f E2, g–i E3, and j–l E4

Al2 O3 core surrounded by an outer sulfide layer, which are the dominating inclusions in the electrode, is observed in each ingot. Almost all observed Al2 O3 inclusions are 1 µm in size, and a few Al2 O3 inclusions larger than 2 µm are found by SEM. There is no obvious difference in the size of inclusions observed in different ESR ingot. The types of the oxide inclusions do not change before and after ESR. The morphology of Al2 O3 inclusions is influenced by steel conditions during deoxidation, such as the soluble oxygen content of the liquid steel before deoxidation, supersaturation degree of the liquid steel following deoxidation, and the presence of any pre-existing inclusions in the steel prior to deoxidation [15, 16]. Previous studies [15, 17] show that Al2 O3 clusters are more likely to be formed at high supersaturation of oxygen and aluminum than that at low supersaturation. In the present study, it can be seen by comparing the morphology of the typical inclusions observed in ESR ingots that Al2 O3 clusters were observed only in the ingot E4, rather than in ingots E1, E2 and E3. It may be attributed to the higher supersaturation of oxygen and aluminum in comparison with that in ingots E1, E2 and E3. In addition, it is believed by comparing the compositions of inclusions observed in the electrode and ingots that all the (Mn,Cr)S inclusions and the outer sulfide layer in complex inclusions were completely removed during the ESR process. Figure 7.6 shows an overview of the inclusion distributions in the S136 consumable electrode and each ESR ingot observed by optical observation. As shown in

7.2 Al2 O3 and MgO·Al2 O3 Inclusions

147

Fig. 7.6 Optical microscope images of inclusions in the consumable electrode and each ESR ingot: a consumable electrode b E1, c E2, d E3, and e E4

Fig. 7.6a, the distribution of inclusions in the electrode (i.e., sample Elec.) is very uneven, and the size varies greatly. Large inclusions were occasionally observed in the electrode. After ESR process, however, the distributions of inclusions in all ESR ingots are relatively even, as shown in Fig. 7.6a–e. No local aggregated distribution of inclusions was observed. Table 7.1 shows the characteristics of inclusions detected by image analyzer in ESR ingots and the electrode. As shown in Table 7.1, the maximum equivalent diameter dmax of inclusions detected in ingots E1, E2, E3 and E4 is 16.1 µm, 15.8 µm, 18.8 µm and 19.7 µm, respectively, as well as 18.6 µm in the sample Elec. The number of observed inclusions per mm2 in each ESR ingot and the electrode is also listed in Table 7.1. With increasing the oxygen content in ESR ingot, the number of observed inclusions per mm2 increases, as shown in Table 7.1. It can also be seen that: (i) in the case of ESR-1 (with oxygen content of 12 ppm), the number of inclusions per mm2 is the least, (ii) in the ingot ESR-4 with oxygen content of 0.0033%, which contains the higher oxygen content than other ESR ingots, the number of inclusions per mm2 is the most, (iii) the number of inclusions found in the electrode per mm2 is far more than that in each ESR ingot. However, it should be noted that it is not comprehensive to represent the inclusions content solely by the total amount of the observed inclusions. The ratio of the sum of cross-sectional area of all the inclusions to the total area of the observed 50 view fields, Sa , is also an indispensable index. As can be seen in Table 7.1, the value of Sa is 0.230% in the electrode, 0.068% in ESR-1, 0.095% in ESR-2, 0.094% in ESR-3 and 0.118% in ESR-4, respectively. Although the different oxygen control operating conditions were employed in ESR experiments, it is clear that the ratio is

148

7 Evolution of Original Oxide Inclusions During ESR

Table 7.1 Statistical results of inclusions in the electrode and ESR ingots Sample No.

Number of observed inclusions per mm2

dmax , µm

Sa , %

Elec

188

18.6

0.230

E1

89

16.1

0.068

E2

117

15.8

0.095

E3

112

18.8

0.094

E4

119

19.7

0.118

significantly lowered after ESR in comparison with that in the electrode. It is also clear that the higher oxygen content in ESR ingot gives rise to higher value of Sa . Figure 7.7 shows the size distributions of inclusions observed in the consumable electrode and each ESR ingot. It is clear from Fig. 7.7 that the inclusions smaller than 2 µm in the electrode in size take up about 50% in the relative fraction of the number, secondly the inclusions of 2–4 µm (nearly 30%), and followed by the inclusions of 4–12 µm in size (about 20%). In addition, the inclusions larger than 12 µm also account for a small proportion. As shown in Fig. 7.7, the majority of inclusions in each ESR ingot are smaller than 2 µm in size in spite of the different ESR conditions are adopted. It is clear from Fig. 7.7 that the number of inclusions in each sample decreases with increasing the inclusion size. In the case of samples E1, E2 and E3, about 70% of the inclusions are smaller than 2 µm, and the inclusions smaller than 4 µm take up about 90% of the total inclusions in each sample. As shown in Fig. 7.7, the inclusions with the size of 2–6 µm take up a larger proportion of the total inclusions in sample E4 in comparison with that in samples E1, E2 and E3. The inclusions smaller than 2 µm in sample E4 account for less than 60% of the total inclusions. The results from comparison Fig. 7.7 Size distributions of inclusions observed in the consumable electrode and each ESR ingot

7.2 Al2 O3 and MgO·Al2 O3 Inclusions

149

Fig. 7.8 Typical original inclusions in the consumable electrode

suggest that, with the higher oxygen content in the ingot, the large inclusions content would be relatively higher. In a separate study, the evolution of Al2 O3 and MgO·Al2 O3 inclusions is investigated. According to SEM-EDS determination, the oxide inclusions in the consumable electrode are either Al2 O3 or MgO·Al2 O3 inclusions. Figure 7.8 presents the SEM back-scattered electron (BSE) images and EDS results of typical inclusions in the electrode. All observed oxide inclusions are irregular in morphology, and mostly about 1–4 µm in size. After electroslag remelting, all oxide inclusions in the ingot were identified as alumina and MgO·Al2 O3 . The BSE images and EDS results of typical inclusions in the ingot are shown in Fig. 7.9. Most of these oxide inclusions serve as nucleation site for nitride (Ti,V)N formation. Pure alumina and MgO·Al2 O3 inclusions were just occasionally found. No single-phased nitrides (without oxide inclusion core) were observed. To reveal the inclusion evolution, the liquid steel samples collected from liquid metal pool during the P-ESR were determined by SEM-EDS. Examples of the inclusions are shown in Fig. 7.10. The inclusions are Al2 O3 and MgO·Al2 O3 in liquid metal pool. It could be noted that the types of inclusions are not changed during the ESR process until in ESR ingot. Al2 O3 and MgO·Al2 O3 inclusions are stable in their compositions during the ESR without intentional alloy addition. The original Al2 O3 and MgO·Al2 O3 inclusions in the electrode that have not been removed during the ESR would survive until in as-cast ingot [14, 18]. The results from industrial trials on martensitic stainless steel

Fig. 7.9 Typical original inclusions in the steel ingot

150

7 Evolution of Original Oxide Inclusions During ESR

Fig. 7.10 Typical inclusions in the sample collected from liquid metal pool during the P-ESR

in Ref. [19] indicate that the MgO·Al2 O3 inclusions from the electrode would survive in the ESR process according to the low amount of MgO in the slag (about 3 mass%) and the lack of refractory in the ESR. In view of the serious detriment of Al2 O3 and MgO·Al2 O3 inclusions to the steel, besides maximizing Al2 O3 and MgO·Al2 O3 inclusions removal from steel, modifying these inclusions to calcium aluminates with low-melting-temperature by calcium treatment is an alternative countermeasure to minimize their detriments to steel. In the case where MgO·Al2 O3 spinel is the only type of oxide inclusions in the steel electrode, all original MgO·Al2 O3 spinels in the electrode (except for that has been removed in the P-ESR process) are modified to mainly CaO–MgO–Al2 O3 and some CaO–Al2 O3 inclusions during protective argon gas atmosphere ESR combined with proper amount of online calcium addition, both of which have a low-meltingtemperature and homogeneous compositions [20]. More detailed description and discussion are presented in Chap. 9. Simultaneous modification of alumina and MgO·Al2 O3 inclusions and its possible extent have been ascertained by Shi et al. [13] It shows that calcium treatment modifies all MgO·Al2 O3 and alumina inclusions that have not been removed in the P-ESR process to liquid/partially liquid CaO–Al2 O3 –(MgO) with uniformly distributed elements, in addition to a small proportion of partially modified inclusions of a CaO–MgO–Al2 O3 core surrounded by an outer liquid CaO–Al2 O3 . It has been confirmed that online calcium addition during the P-ESR process did indeed modify all alumina to liquid/partially CaO–Al2 O3 inclusions with homogeneous compositions, as well as modified MgO·Al2 O3 inclusions to calcium aluminate inclusions. A detailed discussion is presented in Chap. 9.

7.3 Calcium Aluminate Inclusions In the aluminum-deoxidized steel containing a few parts per million of calcium, calcium aluminate inclusions are inevitably generated in the steel. Calcium aluminate inclusions are a common type of inclusions in the steel produced by ESR, besides Al2 O3 and MgO·Al2 O3 inclusions. Unlike Al2 O3 and MgO·Al2 O3 inclusions

7.3 Calcium Aluminate Inclusions

151

in the steel electrode, calcium aluminate inclusions generally take part in the chemical reactions among slag-metal-inclusion, resulting in the change in the chemical compositions of original calcium aluminate inclusions. The findings in Chap. 4 show that the oxide inclusions in the steel before PESR are liquid CaO–Al2 O3 –SiO2 –MgO, which are mostly 3–8 µm in size. Three types of oxide inclusions are present in both liquid metal pool and remelted ingots, i.e., CaO–Al2 O3 –MgO, CaO–Al2 O3 –SiO2 –MgO, and MgAl2 O4 inclusions (about 1.5 µm in size). In the case of the remelting using the slag with 12 mass% SiO2 , CaO-Al2 O3 -MgO inclusions invariably contain approximately 5 mass% SiO2 . Most of these calcium aluminate inclusions are 2–6 µm in size. CaO–Al2 O3 –SiO2 –MgO inclusions (type I) originate from the reduction of SiO2 from the original oxide inclusions in consumable electrode by soluble aluminum in liquid steel during the ESR. CaO–Al2 O3 –MgO, MgAl2 O4 , and CaO–Al2 O3 –MgO–SiO2 (type II) inclusions are generated by the reactions taking place inside liquid steel in liquid metal pool as reoxidation products. Except for the soluble species that are already contained in the steel, the species that are brought from alloy additions into liquid steel during ESR could also modify original oxide inclusion composition. Wang et al. [21] reported that original CaO– SiO2 –MgO–Al2 O3 (with average composition of 13.6 mass%, 7.6 mass%, 2.6 mass%, 35.2 mass%) in the H13 steel electrode were transformed to CaO–MgO– Al2 O3 in liquid metal pool by the reduction of SiO2 from the inclusion by soluble aluminum during protective atmosphere ESR. Meanwhile, the MgO content in this type of oxide inclusions is increased to 10.3 mass% in liquid metal pool and 21.1 mass% in the remelted ingot. It is due to the modification reaction after 94.2 mass% Al–4 mass% Mg alloy addition during ESR, as expressed in Eq. (7.2) [21]. 

 Mg + (xCaO · yMgO · zAl2 O3 ) = (xCaO · (y + 1)MgO · (z − 1/3)Al2 O3 ) + 2/3[Al]

(7.2) The results from the ESR trials on a tool steel electrode, in which the inclusions are mainly CaO–MgO–Al2 O3 –SiO2 + (CaS), MgO–Al2 O3 –SiO2 and Al2 O3 –SiO2 , show that almost all these original inclusions are removed during the ESR [22]. In a separate work, industrial-scale protective atmosphere ESR trials on bearing steel G20CrNi2Mo with a low oxygen content (0.0012 mass%) show that all original oxide inclusions (CaO–MgO–Al2 O3 (with melting temperature lower than 1600 °C) and the complex inclusions of MgO·Al2 O3 core surrounded by an outer CaO–Al2 O3 layer) are removed in two ways, i.e., most of them are absorbed by the molten slag, and the others are dissolved into liquid steel [23]. Even elimination of these original calcium aluminate inclusions is the case, these complex inclusions of MgO·Al2 O3 core surrounded by an outer CaO–Al2 O3 layer cannot be removed by dissolution into liquid steel. More research is needed to reveal the evolution of calcium aluminate inclusions with different compositions during ESR, and ascertain the steel and slag composition conditions for maximal removal of these inclusions.

152

7 Evolution of Original Oxide Inclusions During ESR

7.4 Manganese Silicate Inclusions For Si–Mn–killed steel, some of the manganese silicate deoxidation product remains in the steel as oxide inclusions. In the case where the liquid steel is in contact with Al2 O3 -containing slag, the Al2 O3 concentration in the manganese silicate deoxidation product would increase, whereas MnO and SiO2 would decrease. The driving force for the reactions of MnO–SiO2 –Al2 O3 inclusion formation is the difference in oxide activities between the slag and oxide inclusions (which results in different concentrations of dissolved oxygen and aluminum at the steel–slag interface and in the bulk steel, among other effects). If the concentration ratio of dissolved manganese, silicon to aluminum in the steel is in the correct range, these inclusions would be liquid MnO–SiO2 –Al2 O3 . Liquid MnO–SiO2 –Al2 O3 is a common type of oxide inclusion in the steel before ESR. Although the change in the inclusion chemistry before and after ESR has been widely studied, there is a lack of studies on the evolution of liquid oxide inclusions in steel, such as MnO–SiO2 –Al2 O3 inclusions, during ESR process. The evolution of MnO–SiO2 –Al2 O3 inclusions and generation of Al2 O3 and MgAl2 O4 inclusions in Si–Mn deoxidized steel during protective atmosphere electroslag remelting (P-ESR) will be ascertained through monitoring the change in the transient inclusion in combination with thermodynamic calculation in Chap. 8.

7.5 Role of Processing Parameters of ESR on Inclusions 7.5.1 Deoxidation Schemes of ESR The deoxidation of ESR and deoxidation-related processing parameters have been discussed in Chap. 3. The intention of deoxidizing agent addition in the ESR is to deoxidize the molten slag. In parallel, it is inevitable in most cases that deoxidizing agent would penetrate into the liquid steel and directly deoxidize of liquid steel [14]. The interactions of atmosphere-slag-metal-inclusion system determine the deoxidation during the ESR of steel. The slag with a high oxygen potential would transfer oxygen into liquid steel, resulting in the generation of new oxide inclusions during the ESR process. The study by the present authors did indeed show that new CaO– Al2 O3 –MgO, MgO·Al2 O3 , and CaO–Al2 O3 –MgO–SiO2 inclusions are formed by the reactions taking place inside liquid metal pool as reoxidation products [24]. In addition, transformation of original semiliquid CaO–Al2 O3 –MgO inclusions to liquid CaO–Al2 O3 –MgO–SiO2 inclusions as a result of the reaction between these original inclusions, the dissolved oxygen supplied from the FeO in slag and concerned elements in liquid steel has been confirmed in protective Ar gas atmosphere electroslag rapid remelting (ESRR) of the steel with a low oxygen concentration (0.0008 mass%) [25]. It is still a challenge to prevent oxygen introduction from the atmosphere and FeO in the slag even though protective inert atmosphere is employed in the ESR

7.5 Role of Processing Parameters of ESR on Inclusions

153

production. In the family of ESR technologies, vacuum electroslag remelting exhibits a particular advantage in prevention of oxygen introduction. The ESR refining in vacuum atmosphere (0.01 MPa) brings lower oxygen content of H13 steel in comparison with argon atmosphere ESR (0.1 MPa) because of carbon deoxidization in vacuum atmosphere [26], which consequently lowers the size and number of inclusions. Slag deoxidation of ESR not only determines the oxygen level of liquid steel, but also influences the concentrations of oxide inclusion-forming species in liquid steel. Electroslag remelting of 4340 steel using 50% CaF2 –40% CaO–10% Al2 O3 slag was performed by Reyes-Carmona and Mitchell [27], showing that Ca-deoxidizing agent addition above critical addition rate 10 kg/t gives a low FeO activity of the slag, high aluminum content and aluminates in remelted ingot caused by the reaction between calcium and Al2 O3 in the slag. In the case of Si-Ca and Al deoxidizing agents for the deoxidation of 316L stainless steel during the ESR, the inclusions in the remelted ingot are MgO·Al2 O3 , CaO–MgO–Al2 O3 , MgO·Al2 O3 core surrounded by a CaO–Al2 O3 layer. However, CaO–MgO–Al2 O3 –MnO and CaO–Al2 O3 –SiO2 inclusions are generated because of Si–Ca and Al–Si–Mn alloy addition for the deoxidation during the ESR, besides MgO·Al2 O3 and CaO–MgO–Al2 O3 [28]. It is because Al–Si–Mn alloy addition introduces extra Mn and Si into liquid steel for oxide inclusion formation. Al and Ca–Si are most commonly used deoxidizing agents for the deoxidation in the ESR. Some of the species from the deoxidizing agent pickup more or less in liquid steel could hardly be avoided during the slag deoxidation of ESR, as well as an increase in other alloying elements in liquid steel through steel-slag reaction because of excessive deoxidizing agent addition [14, 20, 29–32]. These species would modify oxide inclusion composition during the ESR through steel-inclusion reaction for many cases. The slag deoxidation of ESR using Al–Mg alloy demonstrated that the aluminum content increases from 0.0003 mass% in the electrode to 0.41 mass% in remelted ingot after 94.2 mass% Al–4 mass% Mg alloy addition for deoxidation of liquid steel during ESR [33]. Consequently, the calcium content is increased because of the reduction of CaO from the slag by aluminum (0.41 mass%), and fresh CaO– Al2 O3 –MgO and CaO–Al2 O3 inclusions in liquid steel form leading to an increase in the proportions of CaO–Al2 O3 –MgO and CaO–Al2 O3 inclusions and the CaO content of oxide inclusions [33]. An increase in the Al2 O3 content in the complex oxide inclusions as a result of the increase in the aluminum content of liquid steel because of excessive addition rate of deoxidizing agent aluminum during the ESR has been demonstrated in Ref. [27] Nevertheless, this is not always the case. In electroslag remelting of H13 tool steel with MgO·Al2 O3 spinels as the only oxide inclusions, significant soluble aluminum pickup in liquid steel did not cause the modification of the composition and size of the original oxide inclusion that have not been removed during the ESR [20]. A same conclusion was drawn for electroslag remelting of a corrosion resistant die steel, in which Al2 O3 is the only type of oxide inclusion [14].

154

7 Evolution of Original Oxide Inclusions During ESR

7.5.2 Slag Composition The elimination of oxide inclusions during the ESR is attributed to their dissolution into the slag phase, except for that dissociated in its individual chemical species into liquid steel. The loss of tantalum during protective argon atmosphere ESR of a Tacontaining martensitic steel is the focus of the study in Ref. [34], and Ta-containing precipitated phase and the number density of inclusions were identified as shown in Fig. 7.11 in combination with slag chemistry analysis, which demonstrates that 95% reduction in number density of Ta2 O5 inclusions during the ESR results in 25% loss of Ta and an increase of the Ta2 O5 content by 1.66 mass% in the slag. Slag chemistry generally exerts a pronounced influence not only on the inclusion removal, but also on the modification of oxide inclusion composition in the ESR process. Note that the role of slag chemistry on the modification of oxide inclusion composition is accomplished through steel-slag reactions and steel-inclusion reactions. The driving force for steel-inclusion reaction is the difference in oxide activities between the slag and oxide inclusions. It is the steel-slag reaction, by affecting the element content in liquid steel, that contributes to the compositional modification of oxide inclusions during ESR. The MgO content in MgO·Al2 O3 inclusions in remelted ingot is found to increase with increasing the CaO content of the slag from 8 mass% to 20 mass% for ESR [35]. It originates from that the activity of SiO2 in slag is

Fig. 7.11 Element mappings of a Ta2 O5 inclusion by electron probe microanalysis (a) and number density of inclusions in martensitic steel CPJ7Z and CPJ7AC produced by vacuum induction melting (VIM) and ESR (b) [34]

7.5 Role of Processing Parameters of ESR on Inclusions

155

lowered by increasing the CaO content of the slag, leading to an increase in the Mg content of liquid steel according to the steel-slag reaction [Si] + 2(MgO) = (SiO2 ) + 2[Mg]. ESR slags generally are CaF2 –CaO–Al2 O3 -based system with minor additions of MgO, TiO2 and/or SiO2 to tailor the slag for the specific remelting requirements. The slag generally is required to minimize the SiO2 content for ESR of alloy and steel. However, it has been demonstrated that a certain amount of SiO2 has to be added in the slag for ESR of some varieties of steels in order to prevent loss (or pickup) of silicon and aluminum in liquid steel and poor surface quality of as-cast remelted ingot [36, 37]. The laboratory-scale ESR trials on low-alloy steel using the slag (31–55 mass% CaF2 , 0–30 mass% Al2 O3 , 15–46 mass% CaO) with different SiO2 contents (0– 23 mass%) show that the slag with low SiO2 content bring alumina or low-calcium aluminates as predominant inclusion type in the remelted ingot, and the SiO2 content of the oxide inclusions increases as the SiO2 content of the slag is increased, giving rise to aluminosilicate inclusions [38]. It is indeed the eventual direction that the evolution of oxide inclusion chemistry will come close to the chemical composition of the slag [39, 40], because steel-slag reaction and steel-inclusion reaction would continue towards equilibrium state until the ratios of the activities of the species are equal in the slag and oxide inclusions. The variation of some components contents of the slag could restrain steel-slag reaction, which would retard steel-inclusion reaction for oxide inclusion composition evolution. It has been demonstrated in Ref. [24] that the SiO2 content in liquid CaO– Al2 O3 –SiO2 –MgO inclusions is obviously lowered because of the reduction of SiO2 from the original oxide inclusions by dissolved Al in liquid steel during the ESR, whereas this reduction of the SiO2 content from the oxide inclusions become less marked originating from considerably decreasing Al pickup in liquid steel caused by increasing SiO2 content in the slag which retards the steel-slag reaction. The slag composed of 70 mass% CaF2 –30 mass% Al2 O3 is a typical ESR-type slag, which has been widely used for several decades in the ESR industrial production. The study by Dong et al. [22] shows that almost all inclusions are Al2 O3 in the ingot remelted under the 70 mass% CaF2 –30 mass% Al2 O3 slag, whereas the inclusions are MgO·Al2 O3 spinels when using the multi-component slag (50 mass% CaF2 –20 mass% CaO–20 mass% Al2 O3 –5 mass% SiO2 –5 mass% MgO) in the ESR, nearly all original inclusions (CaO–MgO–Al2 O3 –SiO2 + (CaS), MgO–Al2 O3 –SiO2 , and Al2 O3 –SiO2 ) are removed during ESR in these two cases. The classification of nonmetallic inclusions in the electrode and the remelted ingots is shown in Fig. 7.12 [22]. These two slag systems do not show a noticeable difference in the capacities for removal of original non-metallic inclusions. In the ESR production practice, Al2 O3 is an indispensable component in almost all commercial slag systems (basically around 30 mass% in its content). When Al2 O3 in slag could exert an effect on the oxide inclusion composition, it would change the contents of the species in liquid metal first (usually soluble aluminum, silicon and oxygen) by slag-steel reaction, and then indirectly modify inclusion composition by steel-inclusion reaction. The results in Ref. [41] indicates that almost all of the

156

7 Evolution of Original Oxide Inclusions During ESR

Fig. 7.12 Classification of the non-metallic inclusions in the electrode and remelted ingots. a the ingot remelted using the slag composed of 50 mass% CaF2 –20 mass% CaO–20 mass% Al2 O3 –5 mass% SiO2 –5 mass% MgO, b the ingot remelted using the slag composed of 70 mass% CaF2 –30 mass% Al2 O3 , respectively [22]

detected inclusions are Al2 O3 -type with minor contents of MgO or CaO in the ingot remelted with the 68 mass% CaF2 –30 mass% Al2 O3 –2 mass% SiO2 slag. Protective nitrogen gas atmosphere ESR trials of bearing steel 100Cr6 with 0.0008 mass% of oxygen, in which the oxide inclusions are low–MgO–containing Al2 O3 – SiO2 –MgO and MgO–SiO2 , were performed by Schneider et al. [42, 43] in order to reveal the effect of the Al2 O3 content in the slag on inclusions. In the case of remelting with Al2 O3 -free slag, the inclusions in the ingot are MgO-enriched SiO2 – MgO. Almost all inclusions are Al2 O3 in the ingot produced by ESR using the slag containing 33 mass% of Al2 O3 , low-Al2 O3 -containing slag (6 mass%) gave MA spinel inclusions in the ingot. It is attributed to the increased aluminum and oxygen contents of the steel, mainly resulting from the decomposition of Al2 O3 caused by the increase in the Al2 O3 content of the slag [42, 43]. Note that the oxygen content of steel is increased by twice to eight times after ESR in the above work by Schneider et al. [42, 43]. Therefore, reoxidation of liquid steel could also modify the compositions of these inclusions. It should be stressed that the decomposition of Al2 O3 from the ESR-type slag could take place or not is a critical issue. In ESR of the bearing steel G20CrNi2Mo in which the oxide inclusions are CaO–MgO–Al2 O3 (most of which are larger than 5 µm and in low-meltingtemperature region (< 1600 °C)) of CaO–MgO–Al2 O3 phase diagram and MgO– Al2 O3 surrounded by an outer CaO–Al2 O3 layer. The results show that all these

7.5 Role of Processing Parameters of ESR on Inclusions

157

original oxide inclusions have been removed in ESR. In the case of the slag with lower CaO and MgO contents (4.42 mass% and 0.1 mass%), most of the inclusions are Al2 O3 , the others are Al2 O3 -riched CaO–Al2 O3 and MgO–Al2 O3 in the remelted ingot. In contrast, in the remelted ingots produced by ESR using the slag with higher CaO and MgO contents, the inclusions are MgO–Al2 O3 , CaO–Al2 O3 and CaO– MgO–Al2 O3 , in which the CaO and MgO contents are higher arising from increased soluble calcium and magnesium contents in liquid steel [44]. Electroslag remelting of Q235B steel with low Al content (< 0.004 mass%) using the slag composed of 30 mass% CaF2 –44 mass% Al2 O3 –20 mass% CaO–6 mass% MgO shows that the oxide inclusions changes from SiO2 –MnO in the electrode to Al2 O3 –MnO during the ESR process resulting from the transfer of aluminum from the slag to liquid steel [45]. As stated in Ref. [45], it is the constant transfer of aluminum from the slag to the steel during ESR process that results in the increase in the aluminum content of the steel and the reaction between MnO and SiO2 in inclusions and the soluble aluminum in liquid steel to generate Al2 O3 in the inclusions. Herein, it is critical to assess the aluminum transfer from the slag to liquid steel according to steel-slag reaction. For electroslag remelting of FGH96 superalloy with MgO–Al2 O3 as the only oxide inclusions, the oxide inclusions in the remelted ingots are transformed from MgO– Al2 O3 to MgO–Al2 O3 –Ce2 O3 and Al2 O3 –(69–73 mass%) Ce2 O3 inclusions with increasing the CeO2 content from 0 to 5 mass% and decreasing the Al2 O3 content from 20 to 1mass% in the slag. The change in the oxide inclusion composition is attributed to the reduction of MgO from the MgO–Al2 O3 by soluble Ce in liquid metal, caused by the increased Ce content in the liquid metal originating from the reduction of CeO2 from the slag by soluble aluminum in liquid metal [18]. As discussed above, the variation of slag chemistry frequently induces the change in the chemical compositions of oxide inclusions in the ESR. But this is not always the case. Zhou et al. [46] experimentally proved that although increasing the SiO2 content from 0 to 25 mass% in the slag for ESR of Al-killed bearing steel, almost all oxide inclusions always are Al2 O3 in the ingots. The electroslag remelting of Fe– 25Ni–15Cr–2.8Ti alloy indicates that increasing the TiO2 contents from 5.66 mass% to 17.69 mass% in the slag has no influence on the oxide inclusion compositions in remelted ingots [47].

7.5.3 Melting Rate of ESR Liquid metal films form at the electrode tip and then collect into liquid metal droplets during ESR. Melting rate of the electrode in ESR (usually termed as melting rate of ESR) largely determines the thickness of liquid metal films and their residence time at the electrode tip [48, 49], as well as pool depth, the local solidification time and solidification rate of liquid metal [50–52], consequently determines the residence time of inclusions in liquid metal. Non-metallic inclusions removal in the ESR process takes place predominantly at the stage of liquid metal films formation

158

7 Evolution of Original Oxide Inclusions During ESR

and subsequent collection into liquid metal droplets at the electrode tip, as well as liquid metal pool (more or less contribution). Therefore, it seems that increasing the melting rates of ESR is unfavorable to inclusion removal. The removal effectiveness of CaO–Al2 O3 –SiO2 inclusions in 316LC stainless steel during laboratory-scale ESR using 33.3 mass% CaF2 –33.3 mass% Al2 O3 –33.3 mass% CaO has been compared in terms of different melting rates of the electrode. The amounts of inclusions and oxygen in the ingot increases with increasing the melting rate of ESR from 0.7, 0.9 to 1.2 kg/min. Ahmadi et al. [51] attributed this increase in the inclusion amounts to the increase in the speed of liquid metal droplets passing through the slag pool as the melting rates were increased, resulting in a decrease in the elimination of oxide inclusions. However, this deduction is open to question. The effect of three different melting rates (at an interval of 100 kg/h) of protective atmosphere ESR on the inclusions in X12CrNiMoV12-3 steel deoxidized with Al was studied by Korp et al. [53] for producing a ingot with diameter of 750 mm. Melting rates have no influence on the types of oxide inclusions, which are Al2 O3 and oxysulphide (main types), calcium aluminate, MgO·Al2 O3 and MnO·Al2 O3 before and after ESR. The increase in the melting rates of ESR results in an increase in the inclusion area proportion. Korp et al. [53] attributed the increase in the inclusion amount to an increase in the dissolved oxygen content of liquid steel as a result of the dissolution of oxide inclusions at electrode tip and possible source of SiO2 dissolution from the slag, which is caused by the increase in the process temperature in ESR increases as increasing the melting rates. However, this deduction is lack of thermodynamic calculation for assessing the dissolution of original oxide inclusions and SiO2 from the slag. The variation of the melting rates of ESR does not necessarily make a difference in inclusions in the ESR and ingot, even if in a large-scale change. Melting rates (350, 400, 450, and 500 kg/h) of protective atmosphere electroslag rapid remelting exert a negligible effect not only on the steel cleanliness and the removal efficiency of CaO–Al2 O3 –MgO (27.5–40.3 mass% CaO, 36.4–52.0 mass% Al2 O3 , and 14.9– 19.6 mass% MgO (a small amount (< 2 mass%) of SiO2 is detected in some cases) inclusions, but also on the inclusion size distribution [25]. The contact angle of liquid CaO–Al2 O3 –SiO2 –MgO inclusions in liquid steel is much smaller than 90° [54]. The negligible effect of melting rates on these inclusions removal is expected to originate from such low contact angle which deteriorates the removal tendency of oxide inclusions from liquid steel in electroslag rapid remelting process.

7.5.4 Electrical Parameters of ESR In addition to the above factors discussed in Sect. 7.6, the power supply (alternating current, direct current, and frequency) could also play a role in the removal and evolution of non-metallic inclusions in the ESR to some extent. During ESR of Fe–Ni alloy and plain carbon steel, the dissolved aluminum and oxygen contents

7.5 Role of Processing Parameters of ESR on Inclusions

159

in liquid metal continuously increase due to inevitable electrochemical reactions in the ESR with a DC power supply, consequently giving rise to alumina inclusions formation in the liquid metal pool [55]. The increase in both total oxygen content and inclusion amount is more pronounced in Fe–Ni ingot. An increase in the amount of oxide inclusions in the hot work tool steel after ESR with lower frequency AC power supply, DC of reversed polarity, DC of straight polarity was also reported by Paar et al. [56] on laboratory-scale experimental work. It should be stressed that these variations are dependent on the electrode material and the slag chemistry. Different frequencies (2.5, 5, 7 5 and 50 Hz) of AC power supply of the laboratoryscale ESR experimentally proved that lower frequency AC power supply gave rise to a higher amount of inclusions in the remelted ingots because electrolytic reactions occur probably in the slag bath and the oxygen produced by electrolyzing in the slag bath enters the liquid metal pool [57]. The laboratory-scale ESR with low frequencies of alternating current (1, 3 and 4.5 Hz) trials shows a similar finding, in which MgO·Al2 O3 and large calcium aluminates are oxide inclusion type in the H11 tool steel electrode [58]. Figure 7.13 presents a comparison of the content of different types of inclusions and the changes in oxygen content. The remelting at the frequency of 1 Hz leads not only to a significant increase in oxygen content, but also to a significant increase in oxide inclusion amount, both of which are much higher than that in the electrode. Increasing the frequency to 4.5 Hz leads to a reduction of total inclusions, which correlates to the slightly reduced oxygen content. The amount of oxide inclusions is much lower than that at lower frequencies but higher than that in the electrode. The larger amount of oxide inclusions is attributed to the electrolytic absorption of oxygen at lower frequency of ESR. It is evident that the composition of oxide inclusions changes towards higher Al2 O3 content in the ESR with the frequency of 1 Hz, whereas the inclusion composition shifts towards a higher MgO content (spinel type) and CaO rich oxides at higher frequency [58]. It could be due to the reaction between soluble aluminum and the increased soluble oxygen in liquid steel contributed by electrolytic absorption of oxygen. Electroslag remelting production practices are widely operated with high frequency (50 or 60 Hz) of alternating current around the world. Regarding the role of power supply (alternating current, direct current, and frequency) on the steel cleanliness with respect to deoxidation and inclusions, the existing references are extremely limited. Despite these recent advances in low frequency ESR, there are still issues that will require more research in the future.

160

7 Evolution of Original Oxide Inclusions During ESR

Fig. 7.13 Amount of non-metallic inclusions and oxygen content in ingots produced by ESR with different power supply frequencies [58]

7.6 Newly-Formed Inclusions in Remelted Ingot Fresh oxide inclusions could form at different stages of ESR, but fresh sulfide and nitride inclusions generally are generated during cooling and solidification of liquid metal in ESR. The newly-formed oxide inclusions could be the products of the reaction between liquid steel and original inclusions from the electrode. In this section, the newly-formed oxide inclusions originated from this route will not be discussed. The compositions of the fresh oxide inclusions formed in liquid steel during ESR are dependent on the activity levels of oxygen and other species and the reaction kinetics condition [59]. Previous studies show that remelted ingot has same types of fresh inclusions as that in the liquid metal pool [24, 60], which is an indication of the absence of the formation of new types of inclusions even if fresh inclusions are formed during liquid steel solidification. No inclusions, which are relics from the electrode, were found in the liquid metal pool [60]. Al2 O3 and MgO·Al2 O3 inclusions are formed in the liquid metal pool as a result of the reactions between alloying elements and the dissolved oxygen that dissociated from MnO–SiO2 –Al2 O3 inclusions in liquid steel. This experimental observation of oxide inclusions in the liquid metal pool is supported by thermodynamic calculation with FactSage (see Fig. 7.14) [60]. MgO·Al2 O3 spinels and Al2 O3 inclusions readily form in liquid steel at the temperatures much higher than the liquidus temperature of the steel as shown in Fig. 7.14. MgO·Al2 O3 spinels and Al2 O3 inclusions formed during cooling and solidification of liquid steel take up only a very small proportion of total oxide

7.6 Newly-Formed Inclusions in Remelted Ingot

161

0.008

0.006

Liquidus of steel

Solidus of steel

Amount of Inclusions (mass%)

0.010

Al2O3 0.004

0.002

0.000 1200

MnS MgO·Al2O3 spinel

1400

1600

1800

2000

Temperature (K) Fig. 7.14 Inclusions formation in the steel as a function of the temperature calculated with FactSage 7.2 [60]

inclusions in remelted ingot. This proportion is mainly dependent on the activities of soluble oxygen and other inclusion-forming elements in liquid steel. Fresh oxide inclusions would form since the soluble oxygen is supplied into liquid steel. This soluble oxygen could originate from not only the dissociation of original oxide inclusions into liquid steel, but also the reoxidation of liquid steel. Recent work in Ref. [24, 47] has confirmed the generation of new oxide inclusions in liquid metal pool caused by reoxidation of liquid steel during ESR. It has been noted in Ref. [24, 47] that the newly-formed Al2 O3 and MgO·Al2 O3 inclusions observed in liquid metal pool are spherical in their morphology. Spherical morphology of Al2 O3 and MgO·Al2 O3 inclusions form at high supersaturation of [O] and [Al] [15, 16]. It suggests that the dissociation of original oxide inclusions and reoxidation of liquid steel during ESR give rise to a high soluble oxygen level of liquid steel. In parallel, as a result of the reoxidation of liquid metal during ESR, two types of fresh oxide inclusions formed during ESR of Fe–25Ni–15Cr alloy in both liquid metal pool and remelted ingot, i.e., MgO·Al2 O3 spinels and complex inclusion of MgO·Al2 O3 spinel with an outer Ti2 O3 -rich layer [47]. The element mappings of typical inclusions observed in the liquid metal pool are illustrated in Fig. 7.15. MgO·Al2 O3 inclusions precipitate in the liquid metal pool during the ESR process according to Eq. (7.3).   2[Al] + Mg + 4[O] = (MgO · Al2 O3 )inclusion

(7.3)

162

7 Evolution of Original Oxide Inclusions During ESR

Fig. 7.15 SEM-EDS element mappings of typical inclusions observed in the sample sampled from the liquid metal pool during ESR: a MgO·Al2 O3 inclusion, b MgO·Al2 O3 inclusion with an outer MgO–Al2 O3 –Ti2 O3 layer [47]

As for the generation of MgO·Al2 O3 inclusions with an outer Ti2 O3 -rich layer, it is the reaction between soluble titanium in liquid alloy and MgO·Al2 O3 inclusion that forms an outer Ti2 O3 -rich layer on unreacted MgO·Al2 O3 inclusion core [21]. During industrial-scale protective argon gas atmosphere ESR of the bearing steel G20CrNi2Mo with low oxygen content (0.0012 mass%), reoxidation of liquid steel and oxide inclusion removal bring about an increase in the oxygen content up to 0.0020 mass% in the remelted ingot [23]. During this ESR process, all original oxide inclusions have been removed, as discussed in Sect. 7.3. Al2 O3 inclusions smaller than 2 µm are generated at the slag/metal interface of electrode tip according to steel-slag reaction as expressed in Eq. (7.4) [23] 2[Al] + 3(FeO) = (Al2 O3 ) + 3[Fe]

(7.4)

The thermodynamic calculation with FactSage 7.0 indicates that low-MgO MgO– Al2 O3 , Al2 O3 -based CaO–Al2 O3 –(MgO) and CaO–Al2 O3 inclusions form in liquid steel during the ESR, which has been supported by the experimentally observed inclusions in remelted ingot [23]. It could be reasonably assumed that the reoxidation of liquid steel during ESR has contributed the modified of oxide inclusion composition. The liquid metal droplets pass through the molten slag pool and reach the liquid metal pool. The liquid metal solidifies directionally at the bottom of liquid metal pool and then builds as-cast ingot in a water-cooled mold. During cooling and solidification of liquid steel, the concentrations of non-metallic inclusion-forming elements gradually approach to saturation for inclusion formation. Fresh non-metallic inclusions would form once the oversaturation of these elements is reached at solidification front at the bottom of liquid metal pool. The source of inclusions in the remelted ingot formed in this way has been demonstrated for Al2 O3 [14, 41, 60], MgO·Al2 O3 [60], AlN [26, 62], MnS [61] and TiN [26, 39, 47]. Even if fresh inclusion removal by floating up in liquid metal pool is possible, the removal in this pathway brings only an inappreciable contribution to overall inclusion removal, elimination of the inclusions formed during solidification of liquid steel in

7.7 Illustration of Inclusion Removal and Fresh Inclusion Generation …

163

the ESR is almost impossible [3, 6, 7]. In most cases, these fresh inclusions significantly differ from the original inclusions in the electrode with regard to composition and size. These newly-formed inclusions usually have small size. Fu et al. [3] Kay and Pomfret [6], and Mitchell [7], claimed that most even almost all of the inclusions in remelted ingot are the newly-formed inclusions during the solidification of liquid steel in water-cooled mold. However, this is not always the case. Persson and Mitchell [63] reported that 50% of the inclusions in remelted ingot are relics from the electrode without a change in the ESR process. In the ESR of H13 steel electrodes deliberately containing a high content of several-millimeter-sized large oxide inclusions, only about 13.66% of these inclusions were removed during the ESR [64]. The sources of the inclusions in remelted ingot are summarized in Sect. 7.7. It is still a challenge to distinguish the fresh inclusions from the inclusions observed in remelted ingots.

7.7 Illustration of Inclusion Removal and Fresh Inclusion Generation During ESR In this section, the oxide inclusion removal and fresh inclusion generation during the ESR process is summarized. Non-metallic inclusions in liquid steel are removed during the ESR process in two ways (one or both work for a certain ESR case), i.e., (1) absorbing them into molten slag and (2) being dissociated in their individual chemical species into liquid steel. Inclusion removal into molten slag involves three steps: (i) transport to the steel/slag interface, (ii) separation into the slag phase across the slag/steel interface, (iii) dissolution in the slag. Inclusion removal into molten slag phase is the ultimate strategy for inclusion control of ESR. Among these three stages, step (i) is the rate-limiting step for oxide inclusion removal during ESR [65]. Inclusion removal takes place at three process stages of ESR, which do, more or less, contribute to the overall inclusion removal. In the liquid metal film formation and its collection into droplet process at the electrode tip (stage I), many of the original oxide inclusions are rejected to the slag/liquid metal interface. Most if not all of these oxide inclusions exposed the steel/slag interface would separate into slag and substantially dissolve in molten slag. The original oxide inclusions that have not been removed at stage I remain in the metal droplets falling from the electrode. The other two stages are that movement of the metal droplets passing through the molten slag pool (stage II) and gather of liquid metal in liquid metal pool (stage III). It is suggested that inclusion removal by flotation up in liquid metal pool is possible, but does not play an important role with regard to the removal of non-metallic inclusions in ESR. The inclusion removal and fresh inclusion generation during the ESR process is schematically summarized in Fig. 7.15. For some kinds of oxide and sulfide inclusions, dissolution of the inclusions in its individual chemical species into liquid steel

164

7 Evolution of Original Oxide Inclusions During ESR

Fig. 7.16 Schematic illustration of inclusion removal and fresh inclusion generation during the ESR process

would occur. Inclusion removal in this trajectory is completed before the liquid metal droplets collect in the liquid metal pool [13, 14, 20, 25, 60]. Fresh oxide inclusions could be generated by the reactions taking place inside liquid steel as reoxidation products at stages I, II and III (see Fig. 7.16). Fresh oxide inclusions could also be products of the reactions between the soluble oxygen that arise from the dissociation of original oxide inclusions and other dissolved species in liquid steel [O] + [M] = (MO). Furthermore, stage IV, precipitation of new nonmetallic inclusions during cooling and solidification of liquid steel in ESR is another source. The behaviors of TiN, low-melting-temperature CaO–Al2 O3 –SiO2 –MgO and MnO–SiO2 –Al2 O3 inclusions on the surface of liquid steel have been studied with in situ confocal laser scanning microscopy [33, 66, 67], so as to identify possible physical mechanisms of these inclusions removal based on in situ disappearance of these inclusions on the surface of liquid steel. Combining with this physical mechanism, more study is needed to clarify these types of inclusions removal in the ESR process and examine its thermodynamic and kinetic mechanisms in terms of steel composition. The non-metallic inclusions in the remelted ingot would originate from one or more of the following routes: (1) the original inclusions that have not been removed in the ESR process, (2) for many cases [25], the original inclusions that have not been removed in the ESR process would experince composition change caused by steel-inclusion reaction in the ESR, (3) reaction products of alloying elements and the dissolved oxygen that dissociated from original oxide inclusions and/or that caused by reoxidation [60], (4) the inclusions newly-formed during cooling and solidification of liquid steel in the ESR [60].

References

165

7.8 Summary Non-metallic inclusions with different compositions and sizes from the electrode basically experience various evolution trajectories during the ESR, including absorption by slag, composition modification by steel-inclusion reaction, and/or dissociation in its individual chemical species into liquid steel. The evolution of these inclusions is also dependent on one or more of other processing parameters, including liquid metal composition, slag chemistry, deoxidation scheme, melting rate, electrical parameters, etc. These evolution trajectories determine the removal efficiency of inclusions in the ESR, and the compistion and size of inclusions (the relics from the electrode, modified composition of original inclusions or newly-formed inclusions) in remelted ingot. Dexidation of ESR is quitely needed because of its significant contribution to reduction of inclusion amount and retardation of new inclusion formation and modification of oxide inclusion composition. Melting rates of ESR have only a limited effect on inclusion removal. Nevertheless, more research is needed to assess the effect of meltig rate on the modification of oxide inclusion compostion, as well as the formation and remeoval of new inclusions. Generation of new inclusions during ESR are inevitable. Fresh oxide inclusions could form at different stages of ESR, but fresh sulfide and nitride inclusions generally are generated during cooling and solidification of liquid metal in ESR. The contribution of oxide inclusions formed at the different stages to the overall oxide inclusions in remelted ingot is dependent on the activities of species in liqudi steel at diffetent time scales. More attention has to be paid to the reoxidation of liquid steel during the ESR. It is inevitable during the ESR of low oxygen steel, even if protective inert atmosphere is employed throught the ESR process. The reoxidation during the ESR results in not only an increase in the amount of inclusions through caused by new oxide inclusion formation, but also modification of oxide inclusion compostion. More work is still needed to ascertain the role of the reoxidation of liquid steel during ESR on the inclusions possessing different compositions and sizes together with its association with the compositions steel and slag.

References 1. Klyuyev M M, Tonilin F F. Journal of Ferrous Metallurgy, 1962, 1: 78–80. (in Russian) 2. O. Madono. Production of aluminum–iron base alloys by electroslag melting[C]. Proceedings of the 2nd International Symposium on Electroslag Remelting Technology, Pittsburgh, USA, 1969, 23–25. 3. Fu J, Zhu J. Change of oxide inclusions during electroslag remelting process[J]. Acta Metall. Sin., 1964, 7(3): 250–262. (in Chinese) 4. Li Z B, Li Y D, Ye Y W, et al. Effect of process parameters on inclusion removal of ball-bearing steel by electroslag remelting[J]. Iron Steel, 1966, (1): 20–24. (in Chinese) 5. Evseyev P P, Filippov A F. Izvest. Akad. Nauk SSSR, Metally, 1968, 3: 41–43. 6. Kay D A R, Pomfret R J. Removal of oxide inclusions during ac electroslag remelting[J]. J. Iron Steel Inst., 1971, 209(12): 962–965.

166

7 Evolution of Original Oxide Inclusions During ESR

7. Mitchell A. Oxide inclusion behaviour during consumable electrode remelting[J]. Ironmak. Steelmak., 1974, 1(3): 172–179. 8. Baligidad R G, Prakash U, Rao V R, et al. Electroslag remelting of Fe-28 at.-%Al intermetallic alloy[J]. Ironmak. Steelmak., 1994, 21(4): 324–331. 9. Verma N, Pistorius P C, Fruehan R J, et al. Calcium modification of spinel inclusions in aluminum-killed steel: reaction steps[J]. Metall. Mater. Trans. B, 2012, 43(4): 830–840. 10. Suito H, Inoue R. Thermodynamics on control of inclusions composition in ultra-clean steels[J]. ISIJ Int., 1996, 36(5): 528–536. 11. Andersson M A T, Jönsson P G, and Nzotta M M. Application of the sulphide capacity concept on high-basicity ladle slags used in bearing-steel production[J]. ISIJ Int., 1999, 39(11): 1140– 1149. 12. Wagner C. Thermodynamics of Alloys[M]. Addison-Wesley Press, Cambridge, 1952, p. 51. 13. Shi C B, Yu W T, Wang H, et al. Simultaneous modification of alumina and MgO·Al2 O3 inclusions by calcium treatment during electroslag remelting of stainless tool steel[J]. Metall. Mater. Trans. B, 2017, 48(1): 146–161. 14. Shi C B, Chen X C, Guo H J, et al. Assessment of oxygen control and its effect on inclusion characteristics during electroslag remelting of die steel[J]. Steel Res. Int., 2012, 83(5): 472–486. 15. Dekkers R, Blanpain B, Wollants P. Crystal growth in liquid steel during secondary metallurgy[J]. Metall. Mater. Trans. B, 2003, 34(2): 161–171. 16. Adaba O, O’malley R J, Bartlett L N. Effect of pre-existing inclusions on the size and morphology of Al2 O3 single stage deoxidation with varying levels of supersaturation[J]. Metall. Mater. Trans. B, 2020, 51(4): 1679–1686. 17. Van Ende M A, Guo M X, Proost J, et al. Formation and morphology of Al2 O3 inclusions at the onset of liquid fe deoxidation by Al addition[J]. ISIJ Int., 2011, 51(1): 27–34. 18. Gao X Y, Zhang L, Qu X H, et al. Effect of CeO2 addition in the slag on inclusions of FGH96 superalloy during electroslag remelting[J]. Metall. Res. Technol., 2020, 117(5): 501–509. 19. Persson E S, Karasev A, Mitchell A, et al. Origin of the inclusions in production-scale electrodes, ESR ingots, and PESR ingots in a martensitic stainless steel[J]. Metals, 2020, 10(12): 1620–1635. 20. Shi C B., Chen X C, Guo H J, et al. Control of MgO·Al2 O3 spinel inclusions during protective gas electroslag remelting of die steel[J]. Metall. Mater. Trans. B, 2013, 44(2), 378–389. 21. Wang H, Shi C M, Li J, et al. Evolution of CaO–MgO–Al2 O3 –CaS–(SiO2 ) inclusions in H13 die steel during electroslag remelting process[J]. Ironmak. Steelmak., 2018, 45(3): 6–16. 22. Dong Y W, Jiang Z H, Cao Y L, et al. Effect of slag on inclusions during electroslag remelting process of die steel[J]. Metall. Mater. Trans. B, 2014, 45(4): 1315–1324. 23. Li S J, Cheng G G, Miao Z Q, et al. Evolution of oxide inclusions in G20CrNi2Mo carburized bearing steel during industrial electroslag remelting[J]. ISIJ Int., 2018, 58(10): 1781–1790. 24. Shi C B, Wang H, Li J. Effects of reoxidation of liquid steel and slag composition on the chemistry evolution of inclusions during electroslag remelting[J]. Metall. Mater. Trans. B, 2018, 49(4): 1675–1689. 25. Shi C B, Zheng D L, Guo B S, et al. Evolution of oxide–sulfide complex inclusions and its correlation with steel cleanliness during electroslag rapid remelting (ESRR) of tool steel[J]. Metall. Mater.Trans. B, 2018, 49(6): 3390–3402. 26. Huang X C, Li B K, Liu Z Q, et al. Oxygen transport behavior and characteristics of nonmetallic inclusions[J]. Vacuum, 2019, 164: 114–120. 27. Reyes-Carmona F, Mitchell A. Deoxidation of ESR Slags[J]. ISIJ Int., 1992, 32(4): 529–537. 28. Yang S F, Li J S, Zhang L F, et al. Clean steel technologies for the production of 316L stainless steel ultrafine wires[J]. Iron & Steel Technology, 2012, 9(1): 47–55. 29. Wen P J, Xiang D L. Research on deoxidization technology of electroslag remelting process for heavy ingot with high quality[J]. Heavy Cast. Forg., 2016, (1): 23–25. (in Chinese) 30. Shi C B, Chen X C, Guo H J. Steel cleanliness control during protective gas electroslag remelting of h13 die steel[C]. AISTech 2012 Proc., Association for Iron & Steel Technology, Warrendale, PA, 2012, 947–957.

References

167

31. Medina S F, Cores A. Thermodynamic aspects steels in the manufacturing of microalloyed steels by the electroslag remelting process[J]. ISIJ Int., 1993, 33(12): 1244–1251. 32. Pateisky G, Biele H, and Fleischer H J. The reactions of titanium and silicon with Al2 O3 -CaOCaF2 slags in the ESR process[J]. J. Vac. Sci. Technol., 1972, 9(6): 1318–1322. 33. Liu W J, Li J, Wang H, et al. The Effect of Al-Mg deoxidation on the cleanliness of steel during the electroslag remelting process[J]. Steel Res. Int., 2019, 90(10): 1900185. 34. Detrois M, Jablonski P D, and Hawk J A. Evolution of tantalum content during vacuum induction melting and electroslag remelting of a novel martensitic steel[J]. Metall. Mater. Trans. B, 2019, 50(4): 1686–1695. 35. Hou D, Wang D Y, Jiang Z H, et al. Investigation on slag–metal-inclusion multiphase reactions during electroslag remelting of die steel[J]. Metall. Mater. Trans. B. 2021, 52(1), 478–493. 36. Shi C B, Li J, Cho J W, et al. Effect of SiO2 on the crystallization behaviors and in-mold performance of CaF2 -CaO-Al2 O3 slags for drawing-ingot-type electroslag remelting[J]. Metall. Mater. Trans. B, 2015, 46(5): 2110–2120. 37. Kim D S, Lee G J, Lee M B, et al. Manufacturing of 9CrMoCoB steel of large ingot with homogeneity by ESR process[C]. Proc. 2015 Int. Symp. on Liquid Metal Processing and Casting, The Minerals, Metals and Materials Society, Warrendale, PA, 2015, 43–52. 38. Mitchell A, Reyes-Carmona F, and Samuelsson E. The deoxidation of low-alloy steel ingots during ESR [J]. Trans. Iron Steel Inst. Jpn., 1984, 24(7): 547–556. 39. Burja J, Tehovnik F, Godec M, et al. Effect of electroslag remelting on the non-metallic inclusions in H11 tool steel[J]. J. Min. Metall. Sect. B-Metall., 2018, 54(1): 51–57. 40. Kang Y B, Lee H. G. Inclusions chemistry for Mn/Si deoxidized steels: thermodynamic predictions and experimental confirmations[J]. ISIJ Int., 2004, 44(6): 1006. 41. Schneider R S E, Molnar M, Gelder S, et al. Effect of the slag composition and a protective atmosphere on chemical reactions and non-metallic inclusions during electro-slag remelting of a hot-work tool steel[J]. Steel Res. Int., 2018, 89(10): 1800161. 42. Schneider R S E, Molnar M, Klösch G, et al. Effect of the Al2 O3 content in the slag on the remelting behavior of a bearing steel[J]. Steel Res. int., 2020, 91(12): 2000241. 43. Schneider R S E, Molnar M, Klösch G, et al. Effect of the Al2 O3 content in the slag on the chemical reactions and nonmetallic inclusions during electroslag remelting[J]. Metall. Mater. Trans. B, 2020, 51(5): 1904–1911. 44. Li S J, Cheng G G, Miao Z Q, et al. Effect of slag on oxide inclusions in carburized bearing steel during industrial electroslag remelting[J]. Int. J. Miner. Metall. Mater., 2019 26(3): 291–300. 45. Wen T J, Ren Q, Zhang L F, et al. Evolution of nonmetallic inclusions during the electroslag remelting process[J]. Steel Res. int., 2021, 92(6): 2000629. 46. Zhou D G, Wang C S, Qian M, et al. Improving bearing steel inclusion through Ca-Si deoxidation and acid-slag ESR[J]. Iron Steel, 1994, 29(7): 25–28. (in Chinese) 47. Zheng D L, Li J, Shi C B, et al. Evolution of TiN and oxide inclusions in Ti-containing Fe-25Ni-15Cr alloy during electroslag remelting[J]. ISIJ Int., 2020, 60(8): 1577–1585. 48. Holzgruber W, Petersen K, and Schneider P E. Comparison of products and economics from large VAR, AC-ESR and DC-ESR ingots[C]. Transaction of the International Vacuum Metallurgy Conference, California, US, 1968, 499–522. 49. Fraser M E, Mitchell A. Mass transfer in the electroslag process: part 1 mass-transfer model[J]. Ironmak. Steelmak., 1976, 3(5): 279–287. 50. Hernandez-Morales B, Mitchell A. Review of mathematical models of fluid flow, heat transfer, and mass transfer in electroslag remelting process[J]. Ironmak. Steelmak., 1999, 26(6): 423– 438. 51. Ahmadi S, Arabi H, Shokuhfar A, et al. J. Mater. Sci. Technol., 2009, 25(5): 592–596. 52. Shi C B, Zheng X, Yang Z B, et al. Effect of melting rate of electroslag rapid remelting (ESRR) on the microstructure and carbides in a hot work tool steel[J]. Met. Mater. Int., 2021, 27(9): 3603–3616. 53. Korp J C, Kubin M. Influence of the melting rate on the Properties of non-metallic Inclusions during electroslag remelting under Protective gas atmosphere[C]. INTECO Remelting & Forging Symposium Shanghai 2010—Proceedings, Shanghai, China, 2010, 73–88.

168

7 Evolution of Original Oxide Inclusions During ESR

54. Yoshioka T, Nakahata K , Kawamura T, et al. Factors to determine inclusion compositions in molten steel during the secondary refining process of case-hardening steel[J]. ISIJ Int., 2016, 56(11): 1973–1981. 55. Kawakami M, Takenaka T, and Ishikawa M. Electrode reactions in DC electroslag remelting of steel rod[J]. Ironmak. Steelmak., 2002, 29(4): 287–292. 56. Paar A, Schneider R, Zeller P, et al. Effect of Electrical Parameters on Type and Content of Non-Metallic Inclusions after Electro-Slag-Remelting[J]. Steel Res. Int., 2014, 85(4): 570–578. 57. Chang L Z, Shi X F, Yang H S, et al. Effect of Low Frequency AC Power Supply During Electroslag Remelting on Qualities of Alloy Steel[J]. J. Iron Steel Res., 2009, 16(1): 7–11. 58. Schneider R, Mülleder M, Zeller P, et al. Effects of low frequency alternating currents on the electro-slag remelting process[C]. International of Symposium on Liquid Metal Processing & Casting 2015, Leoben, Austria, 2015, 23–31. 59. Kolpishon E Y, Ivanova M V, Afanas’ev S Y, et al. Effect of Aluminum on the Content of Oxygen and Nonmetallic Inclusions in 12Kh18AG18Sh Steel[J]. Russian Metallurgy (Metally), 2007, 2007(7): 591–593. 60. Shi C B, Park J H. Evolution of Oxide Inclusions in Si-Mn-Killed Steel During Protective Atmosphere Electroslag Remelting[J]. Metall. Mater. Trans. B, 2019, 50(3): 1139–1147. 61. Qi Y F, Li J, Shi C B, et al. Precipitation and growth of MnS inclusion in an austenitic hot-work die steel during ESR solidification process[J]. Metall. Res. Technol., 2019, 116(3): 322–333. 62. C. B. Shi, X. C. Chen and H. J. Guo. Int. J. Miner. Metall. Mater.,19 (2012), 295. 63. Persson E S, Mitchell A. Differences in inclusion morphology between ESR remelted steel with and without tracer in the slag[C]. Proc. Liquid Metal Processing and Casting Conf. 2017, The Minerals, Metals & Materials Society, Pittsburgh, PA, 2017, 373–379. 64. Du G, Li J, Wang Z B. Effect of initial large-sized inclusion content on inclusion removal during electroslag remelting of H13 die steel[J]. Ironmak. Steelmak., 2018, 45(10): 919–23. 65. Li Z B, Zhang J W, Che X Q. Control of content and composition of nonmetallic inclusion in ESR steel[J]. J. Iron Steel Res. 1997, 9(2): 7–12. (in Chinese). 66. Yang L, Cheng G G, Li S J, et al. Generation Mechanism of TiN Inclusion for GCr15SiMn during Electroslag Remelting Process[J]. ISIJ Int., 2015, 55(9): 1901–1905. 67. Wang H, Li J, Shi C B, et al. Formation and Evolution of Non-Metallic Inclusions in Calcium Treatment H13 Steel during Electroslag Remelting Process[J]. ISIJ Int., 2019, 59(5): 828–838.

Chapter 8

Evolution of Oxide Inclusions in Si–Mn-Killed Steel During ESR

Abstract This chapter presents the evolution of oxide inclusions in Si–Mn deoxidized steel during protective atmosphere electroslag remelting (P-ESR), and the effect of slag compositions on the oxide inclusion evolution. The evolution mechanism of oxide inclusions is proposed through monitoring the change in the transient inclusion in combination with thermodynamic calculation. The slag-steel reactions during P-ESR which could be a potential source of soluble oxygen pickup in liquid steel are considered. The generation of fresh MgAl2 O4 and Al2 O3 inclusions in the liquid metal pool and during solidification of liquid steel in the P-ESR process is evaluated.

8.1 Background Electroslag remelting (ESR) has superior ability to remove non-metallic inclusions from liquid metal. Like secondary refining of liquid steel (such as RH and LF refining), complete removal of oxide inclusions in liquid steel during ESR is not possible. An alternative effort is to properly control the chemistry of inclusions in order to minimize the detrimental effects of residual inclusions to the steel. Kolpishon et al. [1] reported that the chemistry of oxide inclusions was dependent on the aluminum content in high-Cr steel, and the increase in aluminum content above 0.03 mass% led to AlN inclusion formation in electroslag remelted steel. Schneider et al. [2] reported that the slag with different CaO and SiO2 contents exerted a pronounced effect on the relative amounts and size distribution of MgO·Al2 O3 spinel, calciummagnesia-aluminate and sulfide inclusions in the remelted ingots. It is demonstrated that the slag composition, reoxidation of liquid steel and calcium addition during ESR exert significant effects on the chemistry of oxide inclusions in the remelted steel [3–5]. Like the oxide inclusion control during ESR process, the chemistry modification of oxide inclusions during liquid steel refining (such as LF, VD, and RH refining) for consumable steel electrode manufacturing for ESR likewise plays an important role in practice. In order to improve the steel cleanliness, different kinds of deoxidizing

© Metallurgical Industry Press 2023 C. Shi et al., Electroslag Remelting Towards Clean Steel, https://doi.org/10.1007/978-981-99-3257-3_8

169

170

8 Evolution of Oxide Inclusions in Si–Mn-Killed Steel During ESR

agents are generally employed for liquid steel deoxidation according to the requirements of individual steel grade when producing steel electrode for ESR, which consequently generates different types of inclusions in the steel electrode, such as Al2 O3 , MgO·Al2 O3 and various calcium aluminate inclusions. Different types of inclusions basically experience various evolution trajectories during ESR, as demonstrated in the previous studies regarding Al2 O3 , MgO·Al2 O3 , nitride, sulfide and calcium aluminate inclusions [3–9]. The nitride and sulfide inclusions in the consumable steel electrode were removed by dissolving as single soluble element into liquid steel during ESR process [3, 4, 6–8]. Shi et al. [8] found that some of CaO–Al2 O3 –MgO inclusions in the consumable steel electrode were transformed to liquid CaO–Al2 O3 –SiO2 –MgO inclusions through reacting with the dissolved oxygen supplied from the reoxidation of liquid steel and concerned elements in liquid steel during ESR. For Si–Mn deoxidized steel, the oxide inclusions are manganese silicates which are liquid at steelmaking temperatures in many cases [10–13]. Although the change in the inclusion chemistry before and after ESR has been widely studied [3–8, 14–17], there is a lack of studies on the evolution of liquid oxide inclusions in steel, such as MnO– SiO2 –Al2 O3 inclusions, during ESR process. The evolution of MnO–SiO2 –Al2 O3 inclusions in the steel during ESR process is ascertained in this chapter.

8.2 P-ESR Trials of Si–Mn Deoxidized Steel The chemical composition of the consumable steel electrode is shown in Table 8.1. The oxide scale on the steel electrode surface was thoroughly removed through finish turning prior to P-ESR trials. The chemical compositions of the pre-melted slag used in the P-ESR experiments are presented in Table 8.2. The mass ratio of CaO/Al2 O3 for the different slag was kept at 0.88. The SiO2 content in the slag ranged from 1.9 to 9.0 mass%. The slag with 1.9 mass% SiO2 was prepared without additional SiO2 . Purified pre-melted slag, which has lower oxygen potential than conventional pre-melted slag, with three different SiO2 contents designated as F1, F2, and F3 was used in P-ESR trials T1, T2, and T3, respectively. Purified pre-melted slag was roasted at 973 K (700 °C) in an electrical resistance furnace for 8 h to remove the moisture in the slag prior to P-ESR trials. The remelting was conducted in protective argon gas atmosphere at a gas flow rate of 20 NL/min throughout the ESR process. High purity Ar gas (99.99% purity) was passed through a gas-drying system in order to minimize water vapor prior to introduction into the gas protective cap of P-ESR equipment. A steel sample Table 8.1 Chemical composition of the consumable steel electrode (mass%) C

Si

Mn

Cr

Mo

V

Ca

Mg

Al

O

S

0.39

0.75

0.32

4.93

1.17

0.82

< 0.0002

0.0002

0.0032

0.0074

0.0029

8.3 Inclusions in Consumable Steel Electrode

171

Table 8.2 Chemical compositions of the slag used in P-ESR trials (mass%) Slag No.

CaF2

CaO

Al2 O3

MgO

SiO2

F1

29.3

30.5

34.5

3.8

1.9

F2

28.1

29.2

33.1

3.6

6.0

F3

27.2

28.3

32.0

3.5

9.0

was taken from the liquid metal pool in the mold during P-ESR refining using a vacuum sampling tube that was made of quartz (6 mm in inner diameter), followed by quenching in water. The as-cast ingots produced in trials T1, T2 and T3 were designated as ESR-1, ESR-2 and ESR-3, respectively. The steel samples were cut from mid-height of the remelted ingots at the mid-radius position for chemical analysis. The contents of soluble Al, Mg and Si in the remelted ingots were measured by the inductively coupled plasma atomic emission spectroscopy (ICP-AES). The total oxygen and sulfur contents in the steel were measured by the inert gas fusion-infrared absorptiometry and the combustion-infrared absorption technique, respectively. Metallographic samples were taken from the steel electrode and mid-height of each as-cast remelted ingot at the mid-radius position. These steel samples along with the steel samples collected from the liquid metal pool by vacuum sampling tubes were mechanically ground and polished using diamond paste. The inclusions exposed on the cross section of the polished steel sample were analyzed using scanning electron microscope (SEM, FEI Quanta-250; FEI Corp., Hillsboro, OR) equipped with energy dispersive X-ray spectrometer (EDS, XFlash 5030; Bruker, Germany). Furthermore, an automated SEM (EVO18, ZEISS, Germany) equipped with EDS (X-MaxN , Oxford Instruments, U.K.) and INCA software (Oxford Instruments, U.K.) analysis was performed to characterize inclusions at an area of 17 mm2 on the polished cross section of steel samples. The size of inclusions was calculated with the equivalent circle diameter (ECD) and those inclusions with an apparent diameter smaller than 1 µm were neglected.

8.3 Inclusions in Consumable Steel Electrode The oxide inclusions in the consumable steel electrode were identified as ternary type MnO–SiO2 –Al2 O3 . An example of the inclusions observed in the steel electrode is shown in Fig. 8.1 (SEM image and EDS element mappings). The inclusions are spherical in morphology, indicating that these inclusions contain a certain amount of liquid phase. The inclusions are mostly 2–5 µm in size (accounting for 59.2% in the relative fraction of the number), followed by the inclusions in the size range of 5–8 µm (20.4% in the relative fraction). The inclusions smaller than 2 µm and larger than 8 µm take up 13.0% and 7.4% of the total inclusions, respectively.

172

8 Evolution of Oxide Inclusions in Si–Mn-Killed Steel During ESR

Fig. 8.1 SEM image and EDS element mappings of a typical inclusion observed in the steel electrode: MnO–SiO2 –Al2 O3

The compositions of the detected inclusions were depicted on ternary MnO– SiO2 –Al2 O3 phase diagram, as shown in Fig. 8.2. In view of the uncertainties in the analysis of oxygen content by EDS, only the contents of metallic elements were used to determine the compositions of inclusions. The MnO–SiO2 –Al2 O3 phase diagram was calculated with FactSage 7.2 (FToxid database). The compositions of almost all oxide inclusions were located in the low-melting-temperature region [