MicrostructureBased Kinetic Model of Austenitization of Steels

1
Microstructure-Based Kinetic Model of
Austenitization of Steels
Ph.D. Thesis
Gang Shi
ADVISORY COMMITTEE   CHAIR Prof. T.
Calvin Tszeng MEMBERS Prof. Philip Nash
Prof. Sheldon Mostovoy Prof.
Sudhakar Nair
Department of Mechanical, Materials Aerospace
Engineering Illinois Institute of
Technology November 13, 2003
2
Outline

• Chapter I. Introduction
• Chapter II. A Critical Review on Kinetics of
  Austenitization
• Chapter III. Extracting the Kinetic Data from
  Dilatometry Data
• Chapter IV. Experimentation
• Chapter V. Characterization of the Initial
  Pearlite Microstructures
• Chapter VI. Microstructure-Based Model
• Chapter VII. Model Based on FEM
• Chapter VIII. Conclusion
• Acknowledgement

3

• OBJECTIVE
• Characterize the kinetics of austenitization
• Establish a kinetic model of austenitization that
  accounts for the micro-structural features
• Develop a FEM-based model to simulate the
  austenitization process
• MOTIVATIONS
• The austenitization of steels remains poorly
  characterized
• Great need for more reliable and accurate models

4
Chapter IIA Critical Review on Kinetics of
Austenitization

• 2.1 Experimental Studies on Austenitization
  Phenomena
• Nucleation Grain Growth
• Kinetics
• 2.2 Modeling of Austenitization Kinetics
• Avramis Equation
• Cementite Dissolution
• 2.3 Microstructure-Based Kinetic Models of
  Austenitization
• Roosz Model Roo83
• Caballeros Model Cab01a

5
Nucleation at the Junction of Pearlite Colonies
G. A. Roberts and R. F. Mehl Rob43
6
Nucleation of Austenite from Fine Pearlite at Ac1
With increase in time at Ac1, austenite grains
grow and more nuclei appear
Austenite has grown in a finger-like fashion in
the direction of the lamellae
Nucleation Sites
T.G. Digges and S. J. Rosenberg Dig43
7
Nucleation of Austenite from Pearlite
Eutectoid steel T735C
G. A. Roberts and R. F. Mehl Rob43
8
Experiments Original Data Roo83

Preparing initial Microstructures
Austenitizing Experiments
9
Experiments Original Data Cab01a
Preparing initial Microstructures
Austenitizing Experiments
10
Existing Models

• Avramis Equation
• Cementite Disolution
• Speich Spe69
  Hillert Hil71
  AkbayAkb93Atkinson Atk95
• Numerical Models
• Inoue Ino87 2D finite difference model
• Srolovitz Sro85, Jacot Jac99 Monte-Carlo
  model

11
Schematic FEM Model

1. Jacot et al Jac98 developed a 2D FEM
model to estimate the grain growth rate, the
shape of the interface and the carbon
concentration field in austenitization. But the
model is only valid at low overheating (?Tlt5?C)
without considering the nucleation phenomena. In
our experiments on the gleeble, ?T3-18 ?C.
12
Pearlite Morphology

• The pearlite-to-austenite phase transformation is
  well established as a nucleation and grain growth
  process
• RooszRoo83 established a kinetic model of
  isothermal transformation depending on initial
  Pearlite Morphology in the eutectoid steel based
  on the microstructures, A, B, C and D
• Caballero Cab00a established a kinetic model of
  non-isothermal transformation based on the
  microstructures, M1,M2 and M3

13
Roosz Model
Avramis equation
? -4 ? -8
14
Comparison of Experimental Data with Roosz Model
Prediction of Roosz Model
Prediction of Improved Roosz Model Experimental
Data from Caballero Cab01a
15
Fitted Result of Experimental Data from Roo83
Cab01a
16
Chapter IIIExtracting Kinetic Data from
Dilatometry Data

• 3.1 Atomic Volume
• Mix Rule
• Lattice Parameters
• 3.2 Extracting Kinetics from Dilatation Data
• 3.3 Comparison of Lattice Parameters from
  Different Sources

17
Lattice ParametersThe Mixture Rule
The atomic volume Ferrite Cementite Pear
lite Austenite
where and are the atomic volume and
volume fraction of phase i T is the
temperature ? is the carbon content.
18
Lattice Parameters Data from Onink Oni96

• Ferrite Onink Oni93

800KltTlt1200K

• Cementite Stuart Stu66

• Austenite Onink Oni93

1000KltTlt1250K, 0.0005lt?lt0.0365
Verified by Reed Ree98
Used by Li Li00
Kop Kop01
19
Extracting Kinetic Data
20
Lattice Parameters the Computer Program
21
Comparison of Lattice Parameters
0.8wt, Heating rate 0.05K/s, And98
Comparison of the dilation curve of eutectoid
steel
The parameter data from Onink Oni96 is more
reliable than data from other sources.
22
Chapter IVExperimentation

• 4.1 Materials and Specimens
• AISI 1080
• ?6mm and ?10mm Gleeble specimens
• 4.2 Design of Experiments
• Preparing Initial Pearlite Microstructures
• Isothermal Austenitization Process
• Creep at High Temperatures
• 4.3 The Procedure and Result of Dilatometry
  Experiments

23
?10mm Specimens, 1080 Steel
Composition C 0.77 Mn0.8 Si 0.24 Cu 0.23 P
0.015 S 0.025 Ni 0.08 Cr 0.032 V
0.032 Mo0.017 Sn 0.015 Cb 0.002 N 0.002
24
Isothermal Transformation on Gleeble3500
Preparing initial Microstructures
Austenitizing Experiments
25
Thermal Processes for Microstructure S2
1080-10-11
1080-10-05
SEM
SEM
?00.123 ?m ap3.25 ?m
interlemellar spacing
?00.135 ?m Edge Length of Pearlite Colonies
ap3.46 ?m
26
Efforts to Reduce the Creep at High Temperature
A Typical Temperature Profile and Stress of ?6mm
Specimen
A Typical Temperature Profile and Stress of ?10mm
Specimen
27
Dilatation Curves from Isothermal Transformation
28
Dilatation Curves from Isothermal Transformation
29
Austenite Volume Fraction vs. Time
1050?C,30minn
1000?C ,5min

850?C, 30min
675 ?C
650 ?C
30
Chapter VCharacterization of the Initial
Pearlite Microstructures

• 5.1 Dilatometry
• 5.2 Optical Light Microscope
• 5.3 SEM
• 5.4 Measurement of Morphological Parameters

31
Dilatation During Austenite Decomposition
32
Comparison of Optical Light Micrographs
1050?C,30minn
1000?C ,5min

850?C, 30min
675 ?C
650 ?C
33
Comparison of SEM Light Micrographs
1050?C,30minn
1000?C ,5min
850?C, 30min
675 ?C
650 ?C
34
Characterization of the Morphology Parameters

• Measure the interlamellar spacing
• 1) Put a circular test grid of diameter dc on
  the micrograph.
• 2) Count the number of intersections, n, of
  lamellae of carbine with the test grid.
• 3) The mean random spacing
  
• 
  
   
  where, M is the magnification of the micrograph.
• 4) The true interlamellar spacing ?0,
  according to Saltykov, is
  
  
  
• 5) At least 50 fields should be counted.
• Measure the edge length or grain size
• Similar procedure developed by Roosz 1983.
• The relation between the edge length and the
  ASTM grain size number m can be expressed as

35
Correlation of the Computer Program

Table C1 Summary of the results

(?m)
According to Caballero Cab01 ?00.1950.030?m
36
Measuring Interlamellar Spacing
37
Measuring Interlamellar Spacing
38
Measuring Edge Length of Pearlite
Colonies--Examples
B-10-03 ap 21.2? m
S1-06-08 ap 5.25? m
A-08-04 ap 11.3? m
C-09-03 ap 9.50? m
S2-05-03 ap 3.05? m
D-07-06 ap 2.73? m
39
Measuring of Morphological Parameters
Dilatation during austenite decomposition
40
Chapter VIMicrostructure-Based Kinetic Model
6.1 Fitting the Kinetic Data with the Avramis
Equation 6.2 Isothermal Transformation Diagrams
6.3 Transformation Diagrams of Continuous
Heating with Constant Heating Rates 6.4
Dependence of Austenite Transformation on Initial
Pearlite Morphological Parameters
41
Fitting with Avramis Equation
Microstructure A, B, C, D, S1 and S2 Using
the Computer Program
Avramis Equation
Average Error
N2743
42
Fitting the Constant K
?o0.203 ?m, ap10.3 ?m
43
Best Fitting for Each Microstructure
44
Fitting Activation Energy
?o0.222 ?m, ap17.6 ?m ?o0.212 ?m,
ap5.70 ?m ?o0.203 ?m, ap10.3 ?m
?o0.127 ?m, ap10.4 ?m ?o0.123 ?m,
ap3.25 ?m ?o0.122 ?m, ap2.86 ?m
45
Summary of Fitting Result
46
Comparison of Fitting Result with Experimental
Data
?o0.222 ?m, ap17.6 ?m ?o0.212 ?m,
ap5.70 ?m ?o0.203 ?m, ap10.3 ?m
?o0.127 ?m, ap10.4 ?m ?o0.123 ?m,
ap3.25 ?m ?o0.122 ?m, ap2.86 ?m
Average Error 4.2
47
Isothermal Transformation Diagrams
?o0.222 ?m, ap17.6 ?m ?o0.212 ?m,
ap5.70 ?m ?o0.203 ?m, ap10.3 ?m
?o0.127 ?m, ap10.4 ?m ?o0.123 ?m,
ap3.25 ?m ?o0.122 ?m, ap2.86 ?m
48
Comparison of Isothermal Transformation from
Different Microstructures
49
Transformation Diagrams of Continuous Heating
?o0.222 ?m, ap17.6 ?m ?o0.212 ?m,
ap5.70 ?m ?o0.203 ?m, ap10.3 ?m
?o0.127 ?m, ap10.4 ?m ?o0.123 ?m,
ap3.25 ?m ?o0.122 ?m, ap2.86 ?m
50
Dependence of Isothermal Transformation on
Initial Pearlite Morphological Parameters
? -0.6947 ? -1.440
51
A-08-04
52
D-07-08
53
Chapter VIIModel Based on Finite Element Method
7.1 FEM Algorithm 7.2 A Diffusion Controlled
Moving Boundary Problem 7.3 Initial
Microstructure and Auto Meshing 7.4 Prediction
of Nucleation
54
Model Based on FEM

• The image of initial pearlite microstructure.
• Temperature profile
• Elastoplastic properties
• Thermal expansion coefficients
• Interface tension of each phase
• Stress and strain relation
• Diffusion coefficient of each phase
• Initial stress field

• Challenges
• Moving boundary
• Complex geometry

55
A Diffusion Controlled Moving Boundary Problem
Modeling of the Austenitization in a Lamellar
Pearlite Microstructure of Eutectoid Steels at
High Temperatures

• Pearlite is quickly heated up and hold on gt913?C
  .
• Ferrite phase transforms to austenite in a short
  period
• of time without any cementite being dissolved
• Carbon atoms diffuse from cementite to ferrite

56
Mathematical Model at High Temperatures
at at at
Normalized parameters
57
FEM Mesh
58
FEM Cementite Dissolution
59
FEM Cementite Dissolution
60
FEM Carbon Homogenization
61
Calculated TTT Diagram
To determine the carbon homogenization time th at
temperature 930?C for pearlite with interlamellar
spacing ?0 1?m. Using temperature T930?C and
the ?h curve, ?h ?0.43. Thus,
62
Comparison
Comparison of carbon profile at ?d
Comparison of different models
63
FE Mesh over a Complex Pearlite Microstructure
Image from SEM
Identify boundaries
Multiphase in colors
64
Stress and Strain Fields
65
Strain Energy
66
Nucleation (Schematic)
67
Nucleation
Prediction of Nucleation at Junction of
Hypoeutectoid Steel with Pearlite Area Surrounded
by Ferrite
68
Conclusion

• The austenitization kinetics from pearlite
  microstructures in eutectoid steels has been
  investigated in the present study.
• Six different initial pearlite microstructures
  are obtained in carefully controlled isothermal
  processes carried on Gleeble 3500
  Thermo-Mechanical Simulator.
• The dilatometry data during the isothermal
  austenitization processes have been recorded.
  Kinetic data of phase transformation have been
  extracted from the dilatometry data.
• The transformation time of pearlite to austenite
  at a given isothermal temperature varies with
  initial pearlite microstructures with a factor of
  10.
• The initial pearlite microstructures have been
  characterized by using dilatometry, optical light
  microscopy and scanning electron microscopy
  (SEM).
• The morphological parameters, i.e., the
  interlamellar spacing and edge length of pearlite
  colonies, are measured by using a computer
  program developed in the project.
• The interlamellar spacing of pearlite colonies
  mainly depends on the austenite decomposition
  temperature during preparing the pearlite. The
  average edge length of pearlite colonies depends
  on both the austenitizing temperature and holding
  time.

69
Conclusion (continued)

• A technique to kinetic data and establish
  isothermal kinetic model of austenitization by
  using a dilatometer on the Gleeble3500 has been
  developed in the present project
• The kinetic data fit very well with Avramis
  Equation
• 
  
• The fitting results are
• Rate constant n1.0
• Activation energy
• Constant b exp(174.29), exp(175.41),
  exp(175.50), exp(175.83), exp(176.58) and
  exp(176.65) for initial pearlite microstructures,
  B, S1, A, C, S2 and D, respectively
• The average error is 4.2
• Isothermal transformation diagrams for six
  different initial pearlite microstructures have
  been constructed
• Transformation diagrams of constant heating-rate
  for six different initial pearlite
  microstructures have been constructed

70
Conclusion (continued)

• A microstructure-based kinetic model of
  austenitization has been established. The kinetic
  data are consistent with morphological parameters
  except microstructure A.
• Techniques have been developed for a FEM-based
  model of austenitization process considering the
  thermodynamics, diffusion and strain energy.
• Algorithm and a computer program have been
  developed to convert the SEM image to
  computational domains that contain different
  phases and automatically generate the FEM mesh.
• A model based on finite element analysis of
  isothermal austenitization in a lamellar pearlite
  microstructure of eutectoid steel at temperatures
  higher than 913?C also has been developed. The
  model considers both cementite dissolution and
  carbon homogenization.
• The dependence of diffusion coefficient of carbon
  in austenite on the temperature and carbon
  concentration has been taken into account, which
  results in about 35 difference for the cementite
  dissolution times compared to calculation based
  on a constant diffusion coefficient.

71
Conclusion (continued)

• The relations of time, temperature and
  transformation as well as distribution of carbon
  concentration are obtained in the form of a TTT
  diagram in the temperature range between 913 ?C
  and 1148 ?C
• The cementite dissolution time and carbon
  homogenization time are proportional to the
  square of half the interlamellar spacing of
  pearlite in the temperature range between 913 ?C
  and 1148 ?C.
• The FEM-based model has been employed to predict
  the nucleation site during austenitization.
• According to the preliminary calculation that
  involves a simple microstructure consisting of
  cementite in ferrite matrix.
• For a pearlite microstructure, nuclei will
  appear at the junction of pearlite colonies For
  hypoeutectoid steel with pearlite surrounded with
  ferrite, nuclei will occur at several points on
  the boundaries between pearlite and ferrite, and
  at the junction of pearlite colonies.

72
Summary of Contributions

• Techniques to establish a microstructure-base
  kinetic model have been developed by using a
  high-speed dilatometer on the Gleeble 3500.
• The kinetic data obtained in the experiments are
  presented in the form of diagrams of austenite
  volume fraction vs. time at different isothermal
  temperatures for six initial pearlite
  microstructures.
• Based on the fitted result, isothermal
  transformation diagrams (Figure 6.6-6.11) and
  transformation diagrams of continuous heating
  with constant heating rate (Figure 6.14-6.19)
  for six different initial pearlite
  microstructures have been constructed.
• A microstructure-based kinetic model of
  austenitization has been established
• A computer program to characterize pearlite
  microstructures has been developed with
  correlation of results from other researchers.
  The six initial pearlite microstructures have
  been characterized.

73
Summary of Contributions (continued)

• Techniques have been developed for a FEM-based
  model of austenitization process considering
  thermodynamics, diffusion and mechanics.
• A computer program has been developed to convert
  the SEM image into FEM mesh.
• A model based on finite element analysis of
  isothermal austenitization process in a lamellar
  pearlite microstructure of eutectoid steel at
  temperatures higher than 913?C also has been
  developed.
• The FEM-based model has been employed to predict
  the nucleation site during austenitization.

74
Future Research

• In the present study, our efforts are focused on
  kinetics of austenitization from initial pearlite
  microstructure in eutectoid steels. Only Fe-C
  system has been considered. Other elements in the
  commercial steels may strongly affect
  austenitization processes.
• According to our experimental results, the
  transformation of pearlite to austenite is very
  fast at high temperature. Predicting the time of
  carbon homogenization during the austenitization
  process is imperative. In Chapter VII, efforts
  have been made to develop a FEM-based model to
  predict time for both cementite dissolution and
  carbon homogenization. Basic techniques have been
  developed with preliminary result. But, huge
  works remain there to achieve such a general
  model.
• Difficulties also exist in the characterization
  of initial pearlite. Our computer program cannot
  automatically identify the boundaries of pearlite
  colonies.
• As mentioned in Chapter IV, even though the
  thermal and mechanical processes on the Gleeble
  3500 are well controlled by the computer program,
  some uncertain factors exist. Efforts have been
  made to minimize the effects of such factors to
  obtain the kinetic data in the present study.
  But, explanations of some phenomenon are left
  open. One significant phenomenon is that the
  dilatation curves and kinetic of austenitization
  measured are affected by the interval time
  between two thermal cycles.

75
Acknowledgement

• My hearty thanks first goes to my dear wife,
  Yun, for her support, encouragement and
  understanding during my study. Without her
  support, this work would have never been done.
• I would like to express my sincere thanks to my
  adviser, Prof. Tszeng, and the Ph.D. committee
  members, Dr. Nash, Dr. Mostovoy and Dr. Nair, for
  their guidance and support during the course of
  my research for their valuable time and efforts
  to make this research a success.
• The financial support from Thermal Processing
  Technology Center and MMAE Department is highly
  appreciated.
• I would like to express my thanks to Dr.
  Caballero in CENIM-CSIC, Spain, for providing her
  valuable Ph.D. Thesis.
• Dr. Chen did all the SEM operations. His efforts
  and contribution to this project is gratefully
  acknowledged. I would like to thank Russell
  Janota for helping me in the Thermal Processing
  Technology Laboratory. I would also like to
  thank Johnson Craig in the MMAE Machine Shop for
  machining the Gleeble specimens.

76
Acknowledgement (Continued)

• I appreciate the helpful discussions and
  suggestions for using the Gleeble 3500 from Dr.
  Chen, Dynamic Systems Inc.
• I also appreciate the valuable time of professor
  James Dabbert in Humanities Department, and Ms.
  Reeta Roy and Veronia Seizys in CAC Writing
  Center for reading my thesis.
• I would like to express thanks to all my friends
  in Thermal Processing Technology Center, IIT, for
  their helps.
• I am also very grateful to my dad, my
  parents-in-law, brothers and my daughter Heling
  for their patience, understanding and support. I
  would like to express the great happiness coming
  from my baby son, Andrew.

77
Thank you!