Evaluations of Non-metallic Inclusionsin Ca-treated Steels and Their Effect on the Machinability

(1)

Doctoral Thesis in Materials Science and Engineering

Evaluations of Non-metallic Inclusions in Ca-treated Steels and Their Effect on the Machinability

HONGYING DU

Stockholm, Sweden 2021

kth royal institute of technology

(2)

in Ca-treated Steels and Their Effect on the Machinability

HONGYING DU

Doctoral Thesis in Materials Science and Engineering KTH Royal Institute of Technology

Stockholm, Sweden 2021

Academic Dissertation which, with due permission of the KTH Royal Institute of Technology, is submitted for public defence for the Degree of Doctor of Philosophy on Thursday, the 8th April 2021, at 10:00 a.m. in Green room, Osquars backe 31, södra tornet, plan 4, Stockholm

(3)

Printed by: Universitetsservice US-AB, Sweden 2021

(4)

To my beloved family

(5)

(6)

I

Abstract

In recent decades, a considerable development of steel with respect to the performance of steel has taken place, which also has resulted in large challenges to process these steel grades. Therefore, it is essential to make suitable modifications of non-metallic inclusions (NMI) in the steelmaking process and to have a good control of its characteristics to meet the target mechanical properties and to obtain a good machinability.

Based on a case of 316L stainless steel trials with a calcium modification to improve the machinability of steel, the influence and contribution of different NMIs on the machinability were discussed. First, based on the Thermo-Calc calculation results with respect to the appropriate range of Ca additions, steel is produced by an additional Ca treatment at the end of the ladle treatment. In order to evaluate non-metallic inclusions and their influence on machinability tests, steel samples were collected from rolled bars produced by the conventional production route (316R) and an experimental trial with Ca treatment (316Ca).

The metal chips generated during the machining test were also collected for the evaluation of chip breakability and NMIs characteristics after machining. In addition, the Electrolytic extraction (EE) technology is used to extract NMIs from steel and chip samples. Then, a three-dimensional (3- D) study is performed on the inclusions collected on a membrane filter using a scanning electron microscope (SEM) equipped with energy- dispersive X-ray spectroscopy (EDS). The morphology, size, number, frequency, and composition of non-metallic inclusions are studied. Four main types of inclusions were found in the 316Ca steel: Type I (elongated MnS), Type II (oxy-sulfides with hard oxide cores), Type III (soft elongated oxides), and Type IV (hard undeformed oxides).

The results show that the morphologies of NMIs in stainless steel chips

were significantly changed after cutting. Overall, three different main

shapes of NMIs were found: i) Group I having similar shapes, ii) Group II

stretched inclusions having very thin film-like (Group II-a) and fractured

stretched morphologies (Group II-b), and iii) Group III brittlely fractured

inclusions. The total areas of MnS and SO inclusions in the secondary

(7)

II

deformation zone of the chips were significantly increased (by up to 2-3 times) compared to that of the reference steel sample before the cutting test.

It was found that the morphologies of NMIs during machining depend on the location in chips, the workpiece material, as well as the applied cutting speed. This results in different temperatures and metal matrix deformation degrees during machining.

In addition, the chip breakability and chip tool contact length of the reference steel and the experimental steel were evaluated and compared with the characteristics of NMIs in the two steels. A new weight- measurement-based method was developed. The results show that the 316Ca steel generally has a better machinability compared to the reference 316R steel. However, the chip-tool contact length results show that the modification of NMIs for machinability improvements is only beneficial in some machining processes. The 316R steel was preferred at low cutting speeds, whereas the 316Ca steel was preferred at high cutting speeds. The different characteristics of NMI in the various cutting conditions and materials lead to different behaviors and functions of NMI during processing.

Finally, the possible application of PDA/OES in the steelmaking process was also evaluated. This online survey method developed in the industry during recent years provides a high possibility for implementing a rapid screening of the NMI content and shows the potential of establishing an online control of NMI during the processing of steel.

Keywords non-metallic inclusions, Ca treatment, stainless steel, machinability.

(8)

III

Sammanfattning

Under de senaste decennierna har en stor utveckling av ståls prestande skett, vilken lett till stora utmaningar gällande bearbetningen av dessa stålsorter.

Därför är det viktigt att göra en lämplig modifiering av icke-metalliska inneslutningar (NMI) vid ståltillverkningsprocessen och ha en god kontroll över dess egenskaper för att uppnå de önskade mekaniska egenskaperna och en god skärbarhet.

Baserat på ett fall innefattande ett 316L-försök i rostfritt stål med en kalciummodifiering för att förbättra stålets bearbetbarhet diskuterades påverkan av olika NMI på skärbarheten. Inledningsvis så användes Thermo-Calc-beräkningar för att bestämma ett lämpligt intervall för Ca- tillsats för stål som skulle produceras genom att utföra en kalciumbehandling vid slutet av skänkbehandlingen. För att utvärdera icke-metalliska inneslutningar och deras påverkan på skärbarhetstest, så samlades stålprover från valsade stänger som producerats genom användande av den konventionella produktionsvägen (316R) och ett försök med användande av en kalciumbehandling (316Ca). Metallspånorna som genererades under skärbarhetstesten samlades också in för utvärdering av spånbrytbarheten och NMI-egenskaperna efter den skärande bearbetningen. Dessutom så användes en elektrolytisk extraktionsteknik (EE) för att extrahera NMI från stål- och spånprover. Därefter utfördes en tredimensionell (3-D) studie på inneslutningar som samlats på ett membranfiltret efter filtrering, med användande av ett svepelektronmikroskop (SEM) utrustat med energidispersiv röntgenspektroskopi (EDS). Morfologin, storleken, antalet, frekvensen och sammansättningen hos de icke-metalliska inneslutningarna studerades.

Fyra huvudtyper av inneslutningar hittades i 316Ca stålet: typ I (långsträckta MnS), typ II (oxisulfider med hårda oxidkärnor), typ III (mjuka långsträckta oxider) och typ IV (hårda odeformerade oxider).

Resultaten visar att morfologin hos NMI i rostfria spån förändrades

markant efter en skärnande bearbetning. Sammanlagt upptäcktes tre olika

huvudtyper av NMI: i) Grupp I liknande former, ii) Grupp II sträckta

inneslutningar med mycket tunna filmliknande (Grupp II-a) och utsträckta

spruckna morfologier (Grupp II-b) och iii) Grupp III spröda inneslutningar.

(9)

IV

De totala ytorna för MnS- och SO-inneslutningarna i spånens sekundära deformationszon ökade signifikant (upp till 2-3 gånger) jämfört med det resultaten för referensstålprovet taget före skärbarhetstestet. Det visade sig att morfologin hos NMI under bearbetning beror på lokaliseringen av inneslutningen i spånan, provmaterialet samt den applicerade skärhastigheten. Detta resulterar i olika temperaturer och olika deformationsgrader av metallmatrisen under den skärande bearbetningen.

Dessutom utvärderades spånbrytbarheten och kontaktlängden mellan spånan och verktyget för referensstålet och experimentstålet och resultaten jämfördes med karakteristiken hos NMI i de två stålen. En ny viktmätningsbaserade metod utvecklades. Resultaten visar att 316Ca- stålet i allmänhet har en bättre skärbarhet jämfört med referens 316R-stålet.

Däremot så visar resultaten gällande kontaktlängden mellan spånan och verktyget att modifieringen av NMI för förbättringar av skärbarheten endast är fördelaktig i vissa bearbetningsprocesser. Specifikt, så är en användning av ett 316R-stål att föredra vid låga skärhastigheter, medan stålet 316Ca föredrogs vid höga skärhastigheter. De olika egenskaperna hos NMI för olika bearbetningsförhållanden och material leder till olika beteenden och funktioner hos NMI under skärande bearbetning.

Slutligen så utvärderades den möjliga tillämpningen av PDA/OES i ståltillverkningsprocessen. Denna online-undersökningsmetod som utvecklats i branschen under de senaste åren erbjuder en stor möjlighet att åstadkomma en snabb bestämning av NMI-innehållet och visar en potential att kunna utveckla möjligheten för en kontroll av NMI under stålprocessen.

Nyckelord icke-metalliska inneslutningar, kalciumbehandling, rostfritt stål,

skärbarhet.

(10)

V

Acknowledgments

First of all, I want to express my deepest and most immense gratitude to Docent Andrey Karasev and Professor Pär Jönsson for their guidance and supervision with patience during my whole PhD life. I have learned a lot of academic knowledge and scientific methods of carrying research procedures from Andrey. I am also powered by Pär’s open and welcoming support and encouragement to me.

Special thanks to Thomas Björk, Nils Stavlid, Simon Lövquist, for their help and discussion during the project Machinopt. Annika Yang and Arne Bengtson are appreciated for their support and discussion in the project INCONTROL.

I also cherish the unforgettable experience of working with all the considerate and dedicated members in the unit of process. Thanks for the four and a half years with happiness. I would also give my thanks to Wenli Long, who has given me SEM guidance and support, and shared the fantastic lunchtimes.

I would like to acknowledge the financial support of the CSC (China Scholarship Council for my PhD study in Sweden. Jernkontoret, Stiftelsen Prytziska Fonden, Bergshögskolans Jubileumsfond, and Stiftelsen Skandinaviska Malm are appreciated for the financial support for my extended study and my attendance at the conference.

Finally, I want to thank my family for their full support, encouragement, and endless love during my whole life. Without the support from my parents, my man Chencheng, and my cats Maja and Sunday, I cannot have gone through the hard times.

Hongying Du

Stockholm, 2021

(11)

VI

Supplements

This thesis is based on the following supplements:

Supplement I:

Du, H.; Karasev, A.; Sundqvist, O.; Jönsson, P.G. Modification of Non- Metallic Inclusions in Stainless Steel by Addition of CaSi. Metals 2019, 9, 74.

Supplement II:

Du, H.; Karasev, A.; Björk, T.; Lövquist, S.; Jönsson, P.G. Assessment of Chip Breakability during Turning of Stainless Steels Based on Weight Distributions of Chips. Metals 2020, 10, 675.

Supplement III:

Du, H.; Karasev, A.V.; Jönsson, P.G. Three-dimensional investigations of non-metallic inclusions in stainless steels before and after machining. ISIJ International, accept, 2021.

Supplement IV:

Du, H.; Karasev, A.V.; Jönsson, P.G. Influence of non-metallic inclusions in 316L stainless steels on machining using different cutting speeds.

Submitted manuscript Supplement V:

Du, H.; Yang, A.; Karasev, A.V.; Jönsson, P.G. The characterization of non-metallic inclusions in low-alloyed steels by using PDA/OES and off- line investigation methods.

Submitted manuscript

(12)

VII Contribution to the supplements:

I: Performed all of the literature survey, experimental work, analysis, and major part of writing

II: Performed all of the literature survey, the major part of the experimental work, analysis, and major part of writing

III: Performed all of the literature survey, experimental work, analysis and major part of writing

IV: Performed all of the literature survey, the major part of the experimental work, analysis, and major part of writing

V: Performed all of the literature survey, the major part of the experimental work, analysis, and major part of writing

Part of the work presented at the conferences:

Oral presentation in “ISIJ-VDEh-JK Symposium” (Stockholm, Sweden, 12-13 June, 2017)

Oral presentation in “SPS 2018” (Stockholm, Sweden, 16-18 May, 2018) Oral presentation in “10th International Conference on CLEEN STEEL”

(Budapest, Hungary, 18 - 20 September, 2018)

Poster presentation in “MOLTEN 2021” (Seoul, Korea, 21-25 Feb, 2021)

(13)

VIII

Abstract... I Sammanfattning ... III Acknowledgments ... V Supplements ... VI

Chapter 1. Introduction ... 1

1.1. Research Background ... 1

1.2. Investigation of NMIs ... 3

1.3. Machinability ... 4

1.4. Objectives and Overview of This Work ... 5

Chapter 2. Methodology and Experimental ... 9

2.1. Thermodynamic Calculations ... 9

2.2. Materials ... 10

2.3. Machining Tests ... 12

2.4. Characterization of NMIs ... 14

Chapter 3. Results and Discussion ... 19

3.1. Thermodynamic Calculations ... 19

3.2. Online Control Technology for Process Parameters Improvement ... 22

3.3. The Effect of Ca Treatment on the Characteristics of NMIs in 316L Stainless Steel ... 27

3.4. The Characteristics of NMIs in 316L Stainless Steel Chips After Machining ... 36

3.5. The effect of Ca treatment on the machinability of 316L stainless steel ... 50

Chapter 4. Concluding Discussion ... 61

Chapter 5. Conclusions ... 63

(14)

IX

Chapter 6. Sustainability and Recommendations for Future Work .. 67

6.1. Sustainability Considerations ... 67

6.2. Recommendation for Future Work ... 67

References ... 69

(15)

(16)

1 Chapter 1. Introduction

1.1. Research Background

During the last six decades, steelmakers have developed a considerable amount of new steel grades with low levels of impurities by optimizing production processes, to remain competitive in the steel industry. By keeping the impurities at a low-level in the steel, these new steel grades obtain excellent mechanical properties. However, these extremely high- quality steels lead to production problems, due to their decreased machining performance. Because of the increased tool wear, difficult- break chips, and the significantly reduced tool life, the consumption of time, energy, and money during the manufacture of these new grades of steels have increased [1, 2]. Thus, it is imperative to balance a favorable machinability and desired mechanical properties.

One of the most critical aspects to improve machinability is the

modification of non-metallic inclusions (NMIs), since NMIs have a

significant influence on both mechanical properties and machinability of

steels. Generally, NMIs and especially the big ones, are considered to be

harmful to the mechanical properties of the steel by reducing the toughness,

fatigue life, and corrosive resistance, which may increase the failure

probability of the final product [3, 4]. For example, the presence of MnS

inclusions will result in an anisotropy of mechanical properties due to its

elongation. They can also provide the origins for fatigue cracks and

corrosion. [1, 5] NMIs can also cause problems during the production

process. It is reported [6, 7] that Al

2

O

3

inclusions in aluminum-killed steels

can easily aggregate to form big clusters, which sometimes reach sizes of

up to more than 100 µm. They can result in serious nozzle clogging

problems during the continuous casting operation, which significantly

increases the production costs. On the other hand, because of the different

properties of NMIs, such as hardness and thermal expansions, the NMIs

will have different effects (negative or positive) on the machinability

characteristics of materials (such as tool life, tool wear, chip breakability,

etc.). [8-12] Even similar steels with similar mechanical properties but

different non-metallic inclusions (NMIs) show different machinabilities

(17)

Introduction

2 according to previous reports [9-19]. In general, rigid inclusions (such as Al

2

O

3

, TiCN) are considered to be harmful to the machinability, since these hard particles will scratch the tool surface during machining and thereby decrease the tool life dramatically [1, 15, 16]. On the contrary, it is reported that some elements, such as S, Ca, Cu, and Bi, can decrease the cutting force, chip thickness, rake wear, and flank wear, helping to extend the tool life and to improve the cutting surface quality of the metal. [1, 8-15, 20-24]

Therefore, they improve the steel machinability and lower the consumption of power and manufacturing cost. The roles of machinability-preferred inclusions can be summarized primarily in two ways [1, 10, 11, 17, 24-26]:

(1) as a source of stress concentration effects, which is beneficial in reducing the cutting force and promoting the chip breakability and (2) forming a constant tool protective layer or working as a solidus lubricant in the contacting zone of the cutting tools and workpieces, which help to reduce the abrasive, adhesive and chemical wears of the tool and extend the tool life. Thus, modifications and control of the NMI characteristics in the steel during steel manufacturing are necessary.

Ca treatment is reported to be a feasible way to modify NMIs to improve the machinability of steel. [9, 20, 22, 23, 27]. For instance, the rigid Al

2

O

3

inclusions in the melt were found to become smaller compact spherical

calcium–aluminate (CaO–Al

2

O

3

) inclusions by using calcium treatment

[28, 29]. Moreover, Bletton et al. [23] found that the Ca-modified

inclusions like anorthite (CaO·Al

2

O

3

·2SiO

2

) and gehlenite

(2CaO·Al

2

O

3

·SiO

2

) can form a protective lubricant layer between the

workpiece and cutting tools during machining. Ånmark et al. [22] also

found that by the formation of a barrier composed of (Mn,Ca)S and

(Ca,Al)(O,S) on the tool, the Ca-treated carburizing steel resulted in a

significantly improved tool life during hard-turning operations. However,

it is reported that the modifications of NMIs took effect only within a

specific cutting condition range [30-32]. The reason is that the

characteristics and properties of NMIs can be different at different

machining conditions. As a result, the influence of non-metallic inclusions

on the machinability will depend on the cutting situations.

(18)

3 1.2. Investigation of NMIs

To get a quick response during the steelmaking process, pulse distribution analysis with optical emissions spectrometer (PDA/OES) is one online investigation method to determine the NMI characteristics, which has been developed in the industry in recent years. It was first a novel technology in spark optical emissions spectrometer that J.Ono [33] came up with in 1978, which was the integration of the sequence of single sparks followed by a statistical analysis of the spark data, termed as PDA. First, this method was used to improve the precision of the OES results of the bulk composition.

However, it was soon discovered that the data obtained from PDA also contain information about NMIs. It gives a high possibility for a rapid screening of the NMIs content during the steelmaking process since the measurements can be obtained in only seconds. [34]

To get detailed information about NMIs, conventionally, NMIs are investigated in two dimensions (2D) on a polished cross-section of a metal sample using a light optical microscope (LOM) or a scanning electron microscope (SEM). Previous researchers [35-38] have established the drawbacks of this conventional method with respect to that NMIs having complex morphologies (such as largely deformed MnS) and big size clusters. They have suggested three-dimensional (3D) investigations based on electrolytic extraction (EE) for more reliable measurements of size and morphology. After the electrolytic dissolution of the steel matrix, the liquid can be filtrated to collect the insoluble NMIs on a filter. Then, these inclusions can be investigated as 3D objects on a surface of a film filter using, for example, SEM.

Various studies have been carried out to evaluate the effect of NMIs on the

machinability using different approaches and methods. Some researchers

[13, 14, 16, 17] focused on the machinability results of different steels and

compared the typical compositions of NMIs in those steels. Moreover, by

investigating the remaining elements on the surface and the cross-section

of the cutting tools after the machining process, a layer containing similar

compositions as those of NMIs has been described by several researchers

[9, 18, 19, 22-24, 39]. Besides, by using a transmission electron

microscopy (TEM), the structure and composition of the built-up layer on

coated tools have also been studied by Larsson et al. [19].

(19)

Introduction

4 Most of the previously mentioned studies did not focus on a more in-depth research on the chips obtained from machining with respect to the NMI characteristics. Only a few publications have attempted to directly investigate the NMIs in the chips after machining [17, 32]. It was found to be challenging to investigate significantly deformed thin NMIs (exceedingly soft inclusions like MnS) near the cutting zone on polished chip surfaces even by using conventional two-dimensional (2D) methods and using the precise SEM method [32].

1.3. Machinability

Machinability is a complex concept that needs varus parameters and factors to describe. Several parameters, such as tool wear (TW), tool life (TL), cutting forces (CF), chip characteristics (CC), and surface roughness (SR), have been utilized to evaluate the workpiece material machinability [1]. Among them, the chip breakability presents the easiness to remove metal. For instance, if long composite chips are obtained during machining, the possibility of chips getting stuck in the rotating cutting tool or the chuck will be increased. This, in turn, is negative for the production efficiency.

Thus, short, broken corkscrew, or spiral chips are preferred to obtain during the machining operation [40].

One traditional method to evaluate the chip breakability applied in the

industry today is to use chip breaking curves. The experienced operators

determine the transition point of cutting parameters by visual inspections,

where the chip form turns range from an unacceptable to an acceptable

form. Another method of defining the chip breakability is to use a chip

chart, which is supplied from the cutting tool companies to give operation

suggestions to customers. [40] To make a chip chart, chips are collected

from different cutting parameters and are then photographed. Thereafter,

the pictures of the chips are classified into a chart according to cutting

parameters. Besides these most common methods, there are also some

more complex evaluation methods to determine the chip breakability, such

as the fuzzy rule-based system, which tries to describe the breakability

performance[41]. However, these methods are not applicable enough to be

used for a quantitatively sensitive comparison of similar steel grades (such

as NMI-modified steels and reference steels) or for the optimization of the

(20)

5 chip breakability when using different cutting parameters. Therefore, more systematic investigations are necessary to obtain quantitatively accurate assessments of the chip breakability of steels.

The tool-chip contact length (L

t

) is also an essential factor in the machining.

In an orthogonal cutting process, the chip is produced in the shearing zone and moves along the rake face of the tool until it curves off or breaks up.[42]

Several different machining aspects can affect the tool-chip contact length, such as cutting parameters, tool geometry, tool material, coating material, cutting fluid, workpiece diameter, workpiece material, etc. [43] For instance, it is reported that L

t

is a positive correlation to the cutting depth but a negative correlation to cutting speed.[43-45] Conversely, the tool- chip contact length could also affect or directly relate to several other essential aspects in the cutting process, such as the tool life, chip form, tool temperature, cutting forces, tool stability, surface finish, and energy consumption. Previous work has found a proportional relationship between the contact length and the chip compression ratio [44], the radius of chip curvature [46], and tool wear [47]. Thus, investigations of the tool-chip contact length will provide a lot of information about the steels’ machining performance.

1.4. Objectives and Overview of This Work

The focus of this thesis is to optimize steel during industrial production by using different methods to investigate the characteristics of NMIs (such as size, number, composition, and morphology) in steel samples for different specific purposes. New investigation technologies such as the electrolytic extraction and PDA/OES were used in the study. Overall, this thesis is based on five supplements, as shown in Figure 1.1 and Table 1.1. The main focus of this study is to evaluate the NMIs characteristics in steels and their link to the steel quality and especially to the machinability and castability.

Within this scope, the application of PDA/OES methods for a Ca treatment

control during steel making was conducted in Supplement V. Moreover,

the characteristics of the non-metallic inclusions in commercial 316L steel

products without Ca treatment (316R) and with Ca treatment (316Ca)

during the steelmaking process were determined by using 3D

investigations of inclusions extracted using the EE method (Supplement I).

(21)

Introduction

6 Furthermore, the non-metallic inclusions in different zones of the chips after machining are systematically investigated and compared to each other by using the electrolytic extraction method in Supplement III. The characteristics of NMIs in different stainless steels when using different cutting speeds cutting were evaluated in Supplement IV. The tool-chip contact lengths at different cutting speeds were also compared to the determined characteristics of NMIs. A sensitive and quantitative evaluation method of the chip breakability based on the chip weights was developed and tested for 316R and 316Ca steels in Supplement II. The results were compared with the obtained results using the standard chip breakability definition method.

Figure 1.1 Overview of the structure of the appended supplements

(22)

7

Table 1.1 Overview of the main topics, objectives, and parameters used in the thesis.

Suppl Study Objective Approach Parameters

I Modification of Non-Metallic Inclusions in Stainless Steel by Addition of CaSi

Effect of Ca treatment on characteristics of NMIs in 316L stainless steel

Analysis of NMIs in 316R and 316Ca steels using EE

Classification, Composition, Size and number

II Assessment of Chip Breakability during Turning of Stainless Steels Based on Weight Distributions of Chips

Method to differentiate chip

breakability of similar steels

Chip weight measurements

Chip shape, Chip weight, IWchip index

III Three-dimensional investigation of non- metallic inclusions in stainless steel before and after machining

Evaluation of NMIs in a 316Ca steel before and after machining

Application of EE on different zones of chips

Classification, Average area, Total area per mm³

IV The behavior and role of non-metallic inclusion during machining when using different cutting parameters

Evaluation of NMIs in 316R and 316Ca steels when using different cutting speeds

Application of EE on different chips

Classification, Average area, Total area per mm³,

Contact length, Chip thickness V The characterization

of non-metallic inclusions in low- alloyed steels by using PDA/OES, INCA-Feature, and electrolytic extraction.

Possible application of PDA/OES to control NMI modification by Ca

Analysis of NMIs using PDA/OES, INCA-Feature and EE

Composition, B-factor, Ratio of oxides

(23)

(24)

9 Chapter 2. Methodology and Experimental

2.1. Thermodynamic Calculations

To get a prediction of the type and composition of the non-metallic inclusions existing in the melt, thermodynamic calculations of precipitated non-metallic inclusion phases in different grades of steels were carried out by using the Thermo-Calc software [48]. Themo-Calc is based on the CAPHAD methodology (Computer coupling of phase diagrams and thermochemistry). The mechanism is to obtain a consistent description, described by using the Gibbs free energy of the phase diagram and the thermodynamic properties of the system. Using Themo-Calc, a reliable set of stable phases in the given system at the equilibrium stage, and their thermodynamic properties can be predicted.

For 316L stainless steel, the SSUB5 database [49] was used in the calculation of a typical composition of 316L stainless steel as follows (in mass %): 0.014% C, 16.7% Cr, 11.2% Ni, 2.0% Mo, 1.8% Mn, 0.5% Si, 0.002-0.003% Al, 0.001-0.004% Ca and 0.005% O. The compositions and the amounts of formed stable phases in this steel were calculated depending on the different Al and Ca contents.

For the low-alloyed steel, the slag solution phase from the SLAG4 [50]

database was appended to the TCFE9 [51] database to consider liquid complex oxides in the melt. To simplify the thermodynamic system, only the following ten main elements from the steel composition were considered in the calculation: Ca, Mg, O, Al, C, Si, Mn, Cr, S, and Fe. The thermodynamic balance was calculated at different Ca levels (3.3 ppm and 8.5 ppm) at a temperature of 1550 ℃, which is in the normal range of melt temperatures used during ladle treatment. Since the contents of Mg and O were not determined in the steels, only a rough estimate could be made.

Here, the values 10ppm O and 1ppm Mg were used in the Thermo-Calc

calculations. The compositions of the other elements were according to the

chemical composition determinations provided by the company: 0.05%Al,

0.19%C, 0.50%Si, 1.32%Mn, 0.25%Cr, 0.0003%S. A single equilibrium

of the system was achieved in the Thermo-Calc calculations. In addition,

the existing stable phases for systems with different Ca contents were

(25)

Methodology and Experimental

10 obtained. All equilibrium calculations were made at a temperature of 1873K and a pressure of 1 bar.

2.2. Materials

2.2.1. Stainless Steel Tests

For the purpose of machinability improvements of industrial 316L stainless steels, two steel samples produced without (reference heat from conventional technology—316R) and with Ca addition (experimental heat—316Ca) were evaluated in this study. Their compositions are given in Table 2.1. The production route of 316L stainless steel in the company includes scrap melting in an electric arc furnace (EAF), decarburization in an argon oxygen decarburization (AOD) converter, ladle treatment in a ladle furnace (LF), and continuous casting (CC). The reference sample 316R was produced by using the above process route without using any special Ca treatment. The experimental heat 316Ca includes a CaSi wire addition at the end of the ladle treatment. Hot-rolled bars (Ø121 mm) after continuous casting were obtained to evaluate non-metallic inclusions and to perform machining tests. Samples for evaluating non-metallic inclusions were cut from different zones of the cross-section (surface – sample LY, middle – sample LR and center – sample LC) of the bar, as shown in Figure 2.1.

Table 2.1. Chemical compositions of samples of the reference steel (316R) and the Ca- treated steel (316Ca).[mass %]

Sample C Si Mn Cr Ni Mo N Al S O Ca

316R 0.02 0.38 1.60 16.82 11.18 2.02 0.059 0.004 0.007 0.0020 - 316Ca 0.01 0.46 1.58 16.86 11.14 2.09 0.060 0.004 0.009 0.0059 0.0028

(26)

11

Figure 2.1. Steel sample positions (left) from bars of 316L+Ca stainless steels used for electrolytic extractions (EE). Typical sample sizes are shown (right) together with the rolling direction.

2.2.2. Low-alloyed Steel Tests

The contents of the main elements in the two grades of steel used for PDA/OES determinations are given in Table 2.2. Figure 2.2 shows the steelmaking process from the LD converter (basic oxygen furnace), Thyssen Niederrhein (TN) desulfurization process, ladle with vacuum degassing in the chamber, finally to a tundish for casting. The stages of sampling are also marked. The lollipop samples taken from the melt during the steel process are shown in Figure 2.3. The marked zone is the area used for PDA/OES and INCA-Feature studies. The same area was cut and prepared to be used in electrolytic extraction.

Table 2.2 Contents of main elements in steel samples (mass%)

Grade C Si Mn Cr Al Ca S

A 0.13 0.25 1.3 0.40 0.06 <0.003 0.002

B 0.07 0.05 1.3 0.09 0.04 <0.003 0.002

Figure 2.2 A schematic illustration of the steelmaking process

(27)

Methodology and Experimental

12

Figure 2.3. Illustration of lollipop sample taken from the melt, including the zone used for studies of NMIs

2.3. Machining Tests

Two different cutting tools were used in this study to evaluate the machinability of 316L stainless steels. CNMG120408-MM 2025 cemented carbide inserts, with CVD coatings (Ti(C,N)+Al2O3+TiN), from Sandvik Coromant, were used for the chip breakability testings. Also, TPUN 160304 2025 inserts with chemical vapor deposition (CVD) coating by Ti(C,N)+Al

2

O

3

+TiN were used for chip-tool contact length tests.

2.3.1. Chip Breakability Tests for stainless steels

The chip breakability tests included three parts: 1) chip breaking curves, 2)

chip charts, and 3) chip weight measurements. The longitudinal turnings

(Figure 2.4) of 316R and 316Ca steel bars were performed at two different

cutting speeds (130 m/min and 180 m/min) and using a flood coolant. The

first 6 operations were conducted to create chip breaking curves with six

cutting depths given in Table 2.3 by visual inspections. Also, the feed rates

where chips changed from long chips to short chips were recorded. Then,

the chip breaking curves were drawn by plotting the cutting depth (a

p

)

versus the feed rate (f

n

). In total, 30 operations were performed to make

chip charts according to the cutting parameters given in Table 2.3. Chips

from each operation were collected and photographed. Figure 2.5 shows

some typical chips obtained during turning. The 30 pictures were placed in

the chart with a

p

versus fn, with all acceptable sets of chips marked.

(28)

13

Figure 2.4. Schematic illustration of longitudinal turning test

Table 2.3. The cutting data used to create chip breakability curves and chip charts.

Feed rates (fn) [mm/rev]

Cutting depths (ap) [mm]

Cutting speed [m/min]

Chip breaking curves 0 → 0.5 (continuous) 0.5, 0.75, 1, 2, 3, 4 130, 180 Chip chart 0.15, 0.2, 0.3, 0.4, 0.5 0.5, 0.75, 1, 2, 3, 4 130, 180

Figure 2.5. Typical chip shapes obtained from different cutting parameters

Finally, the chips in the chip charts with similar sizes and shapes were compared by using weight measurements. More than 100 chips were randomly collected from each set of cutting parameters and weighted individually. Thereafter, the chips were divided into three groups according to their weights and shapes. Also, the frequency of chips in each group was calculated and compared.

2.3.2. Chip-tool Contact Length Tests

The orthogonal cutting processes were performed to study the chip-tool

contact length with a feed rate (f

n

) of 0.25 mm/rev and a cutting depth (a

p

)

of 3 mm at the cutting speeds (V

c

) of 100 and 250 m/min without using any

cutting fluid (Figure 2.6(a)). The rake face of the tool was investigated by

using light optical microscopy (LOM). A typical picture of the rake face of

(29)

Methodology and Experimental

14 the tool with information of the tool-chip contact length (L

t

) after the orthogonal cutting process is shown in Figure 2.6 (b). The length from the cutting edge to the edge of the first region (L

c1

), the second region (L

t

), and the third region (L

t2

) were measured on LOM photographs. The length of the second region was calculated by using the following equation:

L

c2

= L

t

- L

c1

(1)

Typical long curled chips obtained from the above cutting process were used to investigate non-metallic inclusions in different zones of chips by using the electrolytic extraction method.

(a) (b)

Figure 2.6. (a) Schematic diagram of tool action; (b) Tool-chip contact length of 316R stainless steel at Vc = 100 m/min.

2.4. Characterization of NMIs 2.4.1. Electrolytic Extraction Method

The electrolytic extraction was performed to evaluate the characteristics of non-metallic inclusions (such as number, size, composition, and morphology). During the extraction, the steel matrix was dissolved with the help of an electrical current, whereas non-metallic inclusions did not dissolve in the selected electrolytes. The electrolytic extraction process was carried out using two different non-aqueous electrolytes, namely a 10%

AA (10% acetylacetone–1% tetramethylammonium chloride–methanol) and a 2% TEA (2% triethanolamine–1% tetramethylammonium chloride–

methanol) electrolyte. Except for the studied metal surface, all other sides

were covered to ensure that only the desired zone was extracted. The

electrolytic dissolutions were run with the following electric parameters

(30)

15 settings depending on steel grades and sample sizes: a voltage of 2.9–4.3 V, an electric current of 32–70 mA, and an electric charge from 80 to 1000 coulombs.

After extraction, the undissolved non-metallic inclusions were collected by filtration with a membrane polycarbonate film filter with an open-pore size of 0.4 µm. The inclusions were then investigated using a scanning electron microscope (SEM) and using the back-scattered electrons (BSE) mode. By using the energy-dispersive X-ray spectroscopy (EDS) included in the SEM equipment, the compositions of the extracted inclusions and their different phases were analyzed individually. A summary of the main parameters of the electrolytic extractions and SEM investigations of non- metallic inclusions in different samples are given in Table 2.4. Figure 2.7 shows a schematic diagram of the EE equipment and a typical elongated MnS investigated on the surface of a film filter by using SEM.

(a) (b)

Figure 2.7. (a) Schematic diagram of the equipment used for electrolytic extraction (EE);

(b) a typical elongated MnS inclusion observed after EE.

Table 2.4. Main parameters used in the electrolytic extractions and SEM investigations of non-metallic inclusions (NMIs) in different samples.

Steel

Grade Sample Vc

[m/min]

Ddis [µm]

Afil

[mm²] Wdis

[g]

Aobs

[mm²]

Number observed, n

Size Range [µm]

316R 316R 188 1200 0.1563 0.898 435 2–98

316R-L 100 27 80 0.0177 0.125 218 2-24

316R-H 250 33 80 0.0199 0.045 126 2-23

316Ca 316Ca 85 1200 0.0935 0.898 180 3–124

316Ca-L 100 31 80 0.0215 0.090 153 2-32

316Ca-H1 250 65 1200 0.0369 0.675 96 2-49

316Ca-H2 250 50 80 0.0329 0.089 219 2-44

316Ca-H3 250 34 80 0.0258 0.056 142 2-32

316Ca-T 250 47 80 0.0191 0.045 100 2-46

(31)

Methodology and Experimental

16 The length (l), width (w), and surface area (A

NMI

) of each individual investigated NMI were measured on the SEM images. The aspect ratio value (AR) and equivalent size (D

e

) of elongated inclusions were calculated by using Equations (2) and (3), and by assuming that the inclusions were long ellipsoids:

AR= l/w (2)

3 2

 

D

e

w l (3)

The number of inclusions per unit volume of steel (N

v

) and volume fraction (f

v

) of different NMIs in steel was estimated by using Equation (4) and (5) below:







v

dis obs

fil

N n

W A

A

(4)

3

( ) ( )

1 1

( )

= 6

/ π



 



 



ⁿ ^{incl i}



ⁿ ^{e i}

i i fil

v

me dis obs

V D

f A

V w A

(5)

where n is the number of observed NMIs, A

obs

is the total observed area of the film filter, A

fil

is the whole filtration area of a film filter (=80 or 1200 mm

³

depending on selected filters), W

dis

is the weight loss of the metal sample during EE, and ρ is the metal density (=0.0078 g/mm

³

).

Furthermore, l

i

and w

i

are the length and width of the i-th NMI, respectively.

To quantitatively evaluate the degree of deformation of NMIs after machining, the surface area (A

NMI

) of every individual NMI was also measured automatically on the photos obtained by SEM after EE using the Image-J image analysis software. Then, the total surface area per unit volume (A

v

) and the average surface area of investigated inclusions (A

ave

) of different types were calculated by using the following equations:

( )









^{NMI Type i}

v Type i

dis obs

fil

A A

W A

A

(6)

(32)

17 (7)

where is the area of each NMI of Type i observed on the filter, and is the number of Type i inclusions investigated. When calculating , two sides of one NMI were considered as the total surface area of this inclusion.

2.4.2. INCA-feature Method

INCA-Feature is a high-performance feature detection, analysis, and classification method. The size and composition of non-metallic inclusions in a cross-section of a steel sample can be detected automatically by this method.

The B-factor of an element i is defined as the mass fraction of that element bound to non-metallic inclusions. The evaluations of B-factors for different elements (Al, Ca, Mg, Si, S, etc.) based on INCA-Feature results were calculated using the following equation:

(8) where A is the total area assessed; a

j

is the area of inclusion j (mm

²

); c

ⁱj

is the mass fraction of element i in the inclusion j; N is the number of inclusions. The factor 0.5 is the (average) density ratio between inclusions (3.9g/cm

³

) and the steel matrix (7.8g/cm

³

).

The total assessed area (A) for these samples varied from 87 mm

²

to 132 mm

²

.

2.4.3. PDA/OES Method

The principle of the PDA/OES method is shown in Figure 2.8. Here, sparks caused by an electrode excite atoms and ions in the discharge plasma during the ablation of the material. Therefore, the elevated electron will jump back to a stable level and emit light with particular wavelengths for different elements. These light intensities could be analyzed by photomultipliers after a diffraction grating in the spectrometer. Based on

( )

NMI Type i ave Type i

Type i

A A





n

( )

NMI Type i

A

(Type i )

n

( )

NMI Type i

A

(33)

Methodology and Experimental

18 the light intensities and special calibration functions, the mass percentages of different elements are determined in the computer. The results of a PDA/OES determination can give estimations such as the inclusion size, total oxygen content, insoluble part, inclusion density.[52-56]

Figure 2.8 Schematic illustration of the principle of the PDA/OES method

To develop the application of PDA/OES during steelmaking production,

ten samples from different stages of the steelmaking process were taken

and investigated using the ARL iSpark spectrometer for optimization of

sampling and the analytical procedure during production of different steel

grades, as shown in Figure 2.2. Two to three spots were measured on each

sample. For each spot, 3000 sparks (78ng ablated mass per spark) were

performed for analysis. The first 500 sparks were discarded to avoid the

effect of any contamination on the sample surface. Specifically, the B-

factors of Ca and Al were obtained and compared. The average

compositions of inclusions (here only consider inclusions with a size

smaller than 15 µm) calculated based on the results obtained from EE,

PDA/OES, and INCA-Feature were also compared.

(34)

19 Chapter 3. Results and Discussion

In this chapter, the results and discussion in the current work are summarized in the following sections. In order to predict the optimum Ca addition range, ThermoCalc was applied to evaluate the possible non- metallic inclusion phases at the equilibrium state in the melt, as reported in section 3.1. An evaluation of the application of the PDA/OES technique in the process control is reported in section 3.2 and compared to thermodynamic calculation results (Supplement V). Also, a study of Ca addition based on thermodynamic calculation was performed for 316L stainless steels to study how the machinability could be improved. The characteristics of NMIs in 316R and 316Ca stainless steel before and after machining (Supplements I, III and IV) are summarized in sections 3.3 and 3.4, respectively. Finally, the effect of a Ca treatment on machinability (Supplements II and IV) is reported in section 3.5.

3.1. Thermodynamic Calculations

In this chapter, thermodynamic predictions of stable inclusion compositions and amounts are made using Thermo-Calc, which is useful for the development of new steel grades. The content of elements in steel, such as Ca, Si, Al, and O, will have significant effects on the inclusion composition. Based on the calculation results, an optimum range of Ca contents for the particular purpose of NMIs control was obtained.

3.1.1. Thermodynamic Calculation Results for Low-alloyed Steel

For the sample taken before Ca addition at 1550 ℃, there are three

thermodynamically stable phases (solid circles in Figure 3.1), namely a

liquid complex oxides phase (62%Al

2

O

3

-34%CaO-4%MgO), CA2

(CaO·2Al

2

O

3

), and AM (Al

2

O

3

·MgO). For the sample taken after Ca

treatment, only a liquid complex oxide (hollow circles, 48%CaO-

47%Al

2

O

3

-3%MgO-2%SiO

2

, as shown in Figure 3.1) is the

thermodynamically stable phase. The results show that, from the aspect of

thermodynamics, a Ca addition results in that only liquid oxides inclusions

will exist in the system. Thus, a Ca modification would be successful. This

(35)

Results and Discussion

20 result matches the results from the experimental characterization of the real production samples (discussed in Chapter 3.3) to some degree.

Figure 3.1: The stable phases obtained from thermodynamic calculations at different Ca contents.

3.1.2. Thermodynamic Calculation Results for 316L Stainless Steel According to previous results [9, 23], oxides with compositions close to gehlenite and anorthite are preferred in this study to improve the steel machinability. The composition of possibly stable inclusions in 316L stainless steel with the given composition range of steel (0.014% C, 16.7%

Cr, 11.2% Ni, 2.0% Mo, 1.8% Mn, 0.5% Si, 0.002-0.003% Al, 0.001-0.004%

Ca and 0.005%O) is shown in the ternary phase diagram of Figure 3.2.

With different ratios of Al and Ca contents, the inclusion composition

would change along the direction from pure Al

2

O

3

or Al

2

O

3

∙MnO

inclusions to 2MnO∙SiO

2

inclusions. However, it does not seem possible

to obtain Anorthite (C2SA) in this composition range. Only a large amount

of gehlenite (2CAS) could be obtained in the given composition ranges,

based on equilibrium calculations using this database. The mass

percentages of different non-metallic inclusion phases in the 316L stainless

steels at different contents of Ca (0.001-0.004%) and Al (0.002-0.003%)

are shown in Figure 3.3. When the steel has a 0.003 mass% Ca, the

(36)

21 maximum amount of gehlenite, which can increase the machinability of this grade steel[8], can be obtained for both Al contents. Based on the obtained results, a content range of 0.002-0.003% Al, 0.003% Ca was suggested as suitable to use in industrial trials.

Figure 3.2. Predicted oxide non-metallic inclusions in 316L steel with different contents of Ca and Al, based on Thermo-Calc calculations by using SSUB5 database.

(a) (b)

Figure 3.3. Calculated results of stable phases in an equilibrium state for the following contents of Al: a) 0.002 and b) 0.003 mass% at 1873 K based on Thermo-Calc calculations by using SSUB5 database.

(37)

Results and Discussion

22 3.2. Online Control Technology for Process Parameters Improvement

In this chapter, the Ca-rich NMIs in low-alloy steels are evaluated for the modification of Al

2

O

3

inclusions and clusters that leads to clogging problems during casting.

3.2.1. Samples Qualities for PDA/OES Analysis

Ten lollipop samples from different stages of the steelmaking process for Grade A (given in Table 2.2) were taken and investigated by using the PDA/OES technique. Two to three measurements were performed on each sample. Based on the algorithm used by the software, the B-factors of Ca and Al were obtained. The values of B

Ca

and B

Al

are shown in Figure 3.4.

It shows that the relative standard deviations (RSD) of the B-factors of Ca and Al are < 20% for most samples. However, a few exceptions show very large RSD values, particularly for Ca (samples 4 and 9). This fact is probably due to the sampling itself (for instance, samples contained slag droplets) or the sampling preparation. Thus, an extra burn or even a new sample will be suggested for a good quality of inclusion evaluation. In this study, the abnormal extra-large results were removed from calculations for the S4 and S9 samples. The acceptable values for these samples were shown by stars in the figures.

(a) (b)

Figure 3.4: B-factors and standard deviations of Ca (a) and Al (b) in samples taken from different stages of the steelmaking obtained by using PDA/OES technique.

(38)

23 3.2.2. Comparison of Ca Evaluation by Using INCA-Feature, EE+SEM , and PDA/OES methods

The compositions of typical inclusions (Al

2

O

3

-MgO and Al

2

O

3

-MgO- CaO-CaS) investigated by using INCA-Feature and SEM after EE (EE+SEM) are shown in Figure 3.5 for samples of Grade A before (S4) and after (S6 and S8) Ca addition. Both results show that the prominent inclusions changed from the Al

2

O

3

-MgO phase in sample S4 (before a Ca addition) to the liquid phase (the grey area) of Al

2

O

3

-CaO-CaS in sample S6 and S8 (after a Ca addition). However, the composition determined by SEM after EE shows a higher content of CaO (around 5-10% on average for different samples), compared to the result from INCA-Feature determinations. The different investigation methods cause this difference in the results. On the one hand, when investigating inclusions on the cross- section of metals (2D method, INCA-Feature), the Al dissolved in the metal matrix will increase the Al signal of inclusions. On the other hand, when a whole large inclusion is investigated on the filter using EDS (3D method, EE+SEM), mainly the outer layer, which can have a higher CaO content, is detected.

INCA-Feature EE+SEM

S4

(a) (b)

S6

(c) (d)

S8

(e) (f)

(39)

Results and Discussion

24

INCA-Feature EE+SEM

Figure 3.5. Main compositions of oxide inclusions in samples before (S4) and after (S6 and S8) a Ca treatment plotted in a CaO-Al2O3-MgO ternary phase diagram determined by using INCA and EE+SEM methods.

B-factor (corresponding to the insoluble part of an element, here only consider inclusions with a size smaller than 15 µm) was used in this study to compare the results determined by INCA-Feature and PDA/OES. Close B

Ca

values were obtained from these two methods for samples S4 (before the Ca treatment), S6 (10 minutes after Ca treatment), and S8 (20 minutes after Ca treatment). An increase of total Ca contents in inclusions can be clearly found with time after a Ca addition, as shown in Figure 3.6a. This is shown both in the results from both the INCA and PDA/OES determinations. On the contrary, B-factor values for Al show a decreasing trend after a Ca addition.

INCA-Feature EE+SEM

S4

(a) (b)

S6

(c) (d)

S8

(e) (f)

INCA-Feature EE+SEM

S4

(a) (b)

S6

(c) (d)

S8

(e) (f)

(40)

25

(a) (b)

Figure 3.6. The B-factors of Ca (a) and Al (b) in different samples (S4, S6, and S8) determined by using INCA and PDA/OES methods.

3.2.3. Process Control of Ca Treatment Based on PDA/OES determinations

As shown in Figure 3.7, based on the B-factors of Al, Ca, and Mg for grade A obtained from the PDA/OES determinations, the average compositions of NMI in each sample are calculated and plotted in a ternary diagram. The average chemical compositions of inclusions observed by using the INCA- Feature method in the S4, S6, and S8 samples were also shown as solid circles. The grey zone corresponds to the liquid inclusion zone at 1600

^o

C. The dashed line shows a possible semi-liquid inclusion zone, in which inclusions can contain a solid core and a liquid (or semi-liquid) outer layer.

A clear tendency of the change of the average composition could be found

after a vacuum treatment (S4) and a Ca addition (S6-S8). A good

correlation was obtained for the S4 and S6 samples when using two

different methods, whereas a 25% difference in the %CaO content was

found for the S8 sample. However, the average compositions of sample S8

determined by these two methods are both in or near the liquid zone. Thus,

the inclusion characteristics in the steel melt can be corrected based on the

B-factors from PDA/OES determinations. If the point in the ternary

diagram (Figure 3.7) is outside of the optimum zone, technical parameters

(such as Ca addition, stirring intensity and time, etc.) can be corrected

online to optimize inclusion compositions. However, for an even better

understanding and definition of the optimum zone, more investigations of

NMI’s are still needed.

(41)

Results and Discussion

26

Figure 3.7. The average compositions of non-metallic inclusions in samples taken from different stages of steelmaking in a (CaO+CaS)–Al2O3–MgO ternary phase diagram.

3.2.4. Other Possibilities of PDA/OES Applications - Clogging Problem PDA/OES investigations of samples from the tundish from several trials with the same grade were also evaluated with respect to the clogging problem, as shown in Figure 3.8. Here 51 trials from the steel grade B (2 trials having clogging problem during casting) were selected and compared.

For each trial, two samples from the tundish were collected: one sample

from the early stage (after casting ~25 tons) and one sample from the late-

stage (~25-ton before the casting finish). The ratio %Al

2

O

3

/

(%Al

2

O

3

+%CaO), defined as clogging risk index, was calculated for each

sample based on the values of the B-factors of Ca and Al obtained from

the PDA/OES results. As shown in Figure 3.8, this ratio for NMIs in

problematic trials (circle marks) has a much lower content of CaO in the

alumina inclusion. For this steel grade, the ratio of Al

2

O

3

/(CaO+Al

2

O

3

)

higher than 0.75 (dash line) would be a risk zone. In the investigated trials,

25% of heats’ first samples are in the risk zone. Furthermore, only 10% of

heats have the second samples in the risk zone. In total, only 8% of the

(42)

27 heats (4 heats) have both samples taken from tundish being in the risk zone.

Among them, 50% of the heats (4% in total) resulted in a clogging problem.

Therefore, by determining the B-factor of Ca and Al in steel samples taken from the tundish using the PDA/OES method, it has been shown that an early warning for a possible clogging can be obtained, allowing for corrective measures to be taken.

Figure 3.8. The clogging risk index calculated from the PDA/OES results in the tundish for different trials.

3.3. The Effect of Ca Treatment on the Characteristics of NMIs in 316L Stainless Steel

3.3.1. The Effect of Different Electrolytes on the Ca Treated Steel

According to the results reported by Inoue et al.[57], it needs to be noticed

that the NMIs with CaO-containing phases have the possibility to be

dissolved during electrolytic extraction in some strongly corrosive

solutions. This, in turn, may lead to an inaccurate result. Depending on the

concentration of CaO in the inclusions, 2% TEA (milder one) and 10% AA

(43)

Results and Discussion

28 (stronger one) electrolytes have been suggested for the extraction of inclusions containing CaO–Al

2

O

3

and CaO–SiO

2

. Therefore, both electrolytes were used to extract CaO–Al

2

O

3

–SiO

2

inclusions in a Ca- treated 316L steel. The results obtained from different electrolytes are compared in this section.

The contents of CaO, the ratios of CaO/SiO

2

, and CaO/Al

2

O

3

in different oxide NMIs extracted by using these two electrolytes are shown in Figure 3.9 as a function of on the inclusion length. The dotted lines in this figure correspond to the average values for all analyzed oxide inclusions.

The results show that no apparent differences could be found between the compositions of the inclusions extracted by using the 2% TEA and 10%

AA electrolytes. All the investigated inclusions have a CaO content smaller than 50%. Moreover, the morphologies and sizes of the observed inclusions after extraction in the two different electrolytes were also similar. Thus, there is no indication of a dissolving problem of CaO- containing phases in NMIs. Therefore, it can be safely suggested both electrolytes are suitable to extract (Ca, Al, Si) oxides in the 316Ca steel grade. Thus, the 10% AA electrolyte was chosen to perform EE for all other samples.

(a) (b) (c)

Figure 3.9. CaO contents (a) and ratios of %CaO/%SiO2 (b) and %CaO/%Al2O3 (c) in oxide inclusions in 316Ca samples after extraction by using 2% TEA and 10% AA electrolytes.