1. Development of computational microstructure model Aarne Pohjonen, PhD, Computational materials science, Postdoctoral researcher, University of Oulu, Aarne.Pohjonen@oulu.fi

2. I have implemented input files reading and interaction with user interface for Juha Pyykkönen’s code, currently in use at SSAB User interface developed at VTT (Vesa Kyllönen) Prior work: Cooling model by Juha Pyykkönen

3. Computational model for microstructure evolution in hot rolling and subsequent cooling developed by A. Pohjonen

4. Simulate plastic deformation in rolling process and subsequent cooling, which both will be coupled with microstructure model Differential equation solver (FEM): Plasticity, heat conduction. Solve the equations for time t Microstructure model Calculate, for example: -Local grain size change -Local dislocation density -Phase changes in cooling Correction: Change parameters of the differential equations according to microstructure model for next timestep Output local stress, strain, strain gradient, temperature, etc. needed by microstructure model Increase time, t = t + Δt Simulations and literature knowledge to obtain parameters to model Coupled plasticity/heat concuction/microstructure model (Pohjonen)

5. Set up initial values for physical quantities Differential equation solver Solve temperature distribution, T(x,t), from the non-linear heat conduction equation Phase transformation model Calculate physical quantities for the next timestep Calculate fractions of phases transformed during timestep (ferrite, pearlite, bainite, martensite) Increase time Finite element solver implemented in FORTRAN 90 language Implemented in FORTRAN 90 language Cooling model (Pohjonen)

6. Model comparison to analytical result Time-dependent linear heat equation: ( Conduction of Heat in Solids, Carslaw, Jaeger, Oxford University Press) X: simulation result, Lines: analytical solutions Ready: Heat equation solver

7. Model comparisons to analytical results Steady state case 1. Dirichlet boundary conditions 2. Heat transfer boundary conditions 1. Dirichlet boundary conditions: T(0)=0, T(2)=3.0 f(x)=1+x, k(x)=1+x/2 Solution: 2. Heat transfer boundary conditions: X: simulation result, Lines: analytical solutions a=0, b = 2.0, T ext1 =0.0, T ext2 =3.0 h 1 = 10 = h 2 1. Dirichlet 2. Heat transfer Ready: Heat equation solver

8. Based on regression modelling, source: Computer assisted modelling of metallurgical aspects of hot deformation and transformation of steels (J. Herman, B. Thomas, U.Lotter ) Gives an estimate, which can be further adjusted according to fractional CCT diagram The model can be adjusted exactly for given steels, provided experimental data is available (fractional CCT + Final fractions of phases transformed) Ready: Phase transformation model

9. Phase transformation simulation: TH25 simulation vs. experimental Experimental values from EU report: Computer assisted modelling of metallurgical aspects of hot deformation and transformation of steels (Phase 2) ISBN 92-828-5295-4 (J-C. Herman, B. Donnay, A. Schmitz, U. Lotter, R. Grossterlinden)

10. Phase transformation simulation:CRM39 simulation vs. experimental Experimental values from EU report: Computer assisted modelling of metallurgical aspects of hot deformation and transformation of steels (Phase 2) ISBN 92-828-5295-4 (J-C. Herman, B. Donnay, A. Schmitz, U. Lotter, R. Grossterlinden)

11. Phase transformation simulation:TH25 final phase fractions, simulation vs. experimental Experimental values from EU report: Computer assisted modelling of metallurgical aspects of hot deformation and transformation of steels (Phase 2) ISBN 92-828-5295-4 (J-C. Herman, B. Donnay, A. Schmitz, U. Lotter, R. Grossterlinden)

12. Phase transformation simulation:CRM39 final phase fractions, simulation vs experimental Experimental values from EU report: Computer assisted modelling of metallurgical aspects of hot deformation and transformation of steels (Phase 2) ISBN 92-828-5295-4 (J-C. Herman, B. Donnay, A. Schmitz, U. Lotter, R. Grossterlinden)

13. Pulsed water flows Heat conduction and phase transformation simulation: - Simulation setting Calculate: 1. Time dependent temperature distribution 2. Phases transformed as function of time - on surface - on ¼ thickness

14. Heat conduction and phase transformation simulation: - Time dependent temperature distribution Pulsed water flows

15. Heat conduction and phase transformation simulation: - Temperature at surface - Temperature at ¼ thickness - Fractions of ferrite and bainite formed on surface and on ¼ thickness

16. A Experimental data required to run the model as fitted to the original data: 1.Steel composition and austenite grain size 2.Final rolling temperature 1.Fractional CCT diagram for linear cooling rates 2.Transformation kinetics (transformed austenite as function of temperature/time) 3. Final phase fractions after cooling to room temperature B Experimental data required to fit the model: Experimental data required for the model

17. Effective activation energy of transformation start Nucleation and growth described by avrami eq.

18. Effect of altering composition to the effective activation energy of transformation start Ferrite

19. Bainite Effect of altering composition to the effective activation energy of transformation start

20. Effect of austenite deformation To the transformation start

21. Gleeble physical simulation (J. Uusitalo) FEM numerical simulation with Abaqus CAE (A. Pohjonen, J. Ilmola, O. Leinonen) Plasticity simulations

22. Microstructure evolution during and after plastic deformation - Anisotropic plasticity - Recrystallization - Dislocation substructures Effect on phase transformation during cooling Effect of hot forming to phase transformation

23. Graphical user interface for defining cooling path and calculating phases formed Graphical user interface for estimating required water to cool steel strip to desired temperatures in cooling system (Joni Paananen Bachelor Thesis, supervised by A. Pohjonen) Additional tools developed

24. Graphical user interface for defining cooling path and calculating phases formed Additional tools developed: TemperatureCTemperatureC Time (s)

25. Graphical user interface for estimating required water to cool steel strip to desired temperatures in cooling system (Joni Paananen Bachelor Thesis, supervisesd by A. Pohjonen) Additional tools developed: TemperatureCTemperatureC Distance (m)

26. Phase transformation model Transformation start Scheil’s additivity rule Kinetics Avrami type function Fraction transformed n, a, b, c are parameters determined from Experimental CCT (inversion of CCT to TTT) Umemoto 1983 determined from experimental CCT (inversion of CCT to TTT)

27. Model fitted to 960QC / deformed at 850 °C simulation vs. experiment

28. 2d/3d coupled heat conduction / phase transformation simulations for any geometry (preliminary results) PT model coupled with Elmer FEM

29. Simulate plastic deformation in rolling process and subsequent cooling, which both will be coupled with microstructure model Differential equation solver (FEM): Plasticity, heat conduction. Solve the equations for time t Microstructure model Calculate, for example: -Local grain size change -Local dislocation density -Phase changes in cooling Correction: Change parameters of the differential equations according to microstructure model for next timestep Output local stress, strain, strain gradient, temperature, etc. needed by microstructure model Increase time, t = t + Δt Simulations and literature knowledge to obtain parameters to model Coupled plasticity/heat concuction/microstructure model (Pohjonen)