Hallo,
I am working with the MICRESS calculation on gamma-alpha transformation in DP steel with TC interface. As in the system there are only austenite and ferrite, the cementite phase will be suspended by ThermoCalc. Therefore the max. ferrite fraction will be calculated from the lever rule between Ae1 and Ae3 lines (like eq. . My case is the isothermal transformation at 680°C and I found in a case that the predicted ferrite fraction is 15% higher than the experiment. (So far I take into account only 2D transformation and didn't take the correction factor for 3D.)
This reminds me on Hultgren extrapolation, by which the max. ferrite fraction is calculated by the extended Acm line below the A1 temperature (attached Hultgren). This was used also by Han (attached) and fitted his experiment very well. In Fig. 3 it is therefore clear that the max ferrite fraction under A1 temp. is reducing with the reducing temperatures. I know that this is not the calculation for the 'real' equilibrium. But as our simulation is concerning the kinetics, does it really make sense to use the max ferrite fraction calculated from the real equilibrium which is far from the kinetics.
thanks,
nokkikku
Hultgren's extrapolation for calculating max. ferrite
Hultgren's extrapolation for calculating max. ferrite
- Attachments
-
- Pages from Militzer_3D_phasefield_Aus_Ferrite.pdf
- see Eq.10-11
- (45.35 KiB) Downloaded 293 times
-
- Han(2001)_Cooling_Phase_Trans_trip.pdf
- see Fig.3 max. ferrite fraction (calc. from Hultgren's extrapolation)
- (335.3 KiB) Downloaded 337 times
-
- Hultgren.pdf
- part of original paper
- (100.97 KiB) Downloaded 389 times
Re: Hultgren's extrapolation for calculating max. ferrite
Dear nokkikku,
the theories and approaches which you attached to your post all seem quite complex for me, and I am not expert on gamma-alpha transformation in steels - it would be great if some of our members like e.g. people from the Militzer group could share their experiences here!
From my understanding, the maximum ferrite fraction which you refer to is the limit for long simulation times, where the ferrite fraction should converge to. As substitutional elements are assumed not to move, it is calculated from a pseudo-binary para-equilibrium phase diagram. I do not understand how it is done in the references you attached, because it is not described there...
This value of the maximum ferrite fraction has nothing to do with the 3D correction which I proposed, because this is only a kinetic correction and should not change the long time limit!
It depends strongly on the substitutional elements how close you get equilibrium, because diffusion of these elements is very slow. Even if they are not unequally distributed (e.g. due to solidification), they must be redistributed (assuming NPLE), and a certain value of the driving force is needed for that. As soon as the driving force for the transformation falls below this value, the transformation stops, before reaching equilibrium and the full predicted ferrite fraction.
Bernd
the theories and approaches which you attached to your post all seem quite complex for me, and I am not expert on gamma-alpha transformation in steels - it would be great if some of our members like e.g. people from the Militzer group could share their experiences here!
From my understanding, the maximum ferrite fraction which you refer to is the limit for long simulation times, where the ferrite fraction should converge to. As substitutional elements are assumed not to move, it is calculated from a pseudo-binary para-equilibrium phase diagram. I do not understand how it is done in the references you attached, because it is not described there...
This value of the maximum ferrite fraction has nothing to do with the 3D correction which I proposed, because this is only a kinetic correction and should not change the long time limit!
It depends strongly on the substitutional elements how close you get equilibrium, because diffusion of these elements is very slow. Even if they are not unequally distributed (e.g. due to solidification), they must be redistributed (assuming NPLE), and a certain value of the driving force is needed for that. As soon as the driving force for the transformation falls below this value, the transformation stops, before reaching equilibrium and the full predicted ferrite fraction.
Bernd