MICRESS Forum

Dear all,

I would like to ask if anyone could help me create a script for the following case, as I am new to Micress (took part in the Micress online training last August but haven't had much chance to use the software since then).

Goal: 2D-grain growth simulation to evaluate resulting grain sizes after specific heat treatments

Heat treatment: room temperature -> heating to 1300 °C with 0,333 K/s -> holding 1300 °C for 1 h -> cooling to RT with 0,333 K/s

Material: low alloyed steel DC04 (1.0338): C = 0,08 wt.% Mn = 0,4 wt.%; carbide formation was not detected in the samples (so it can be neglected in the simulation); average initial grain size = 33 µm (median initial grain size = 23 µm)


Using the training files, I created an isothermal heat-treatment simulation at 1300 °C (with no phase transformations) for an unspecified material composition. The initial simulation cell number was set to x and y = 1000, with a cell dimension (grid spacing in micrometers) of 1.5 (which, to my limited understanding, should result in a 2D area of 1,5 x 1,5 mm).
But not surprisingly, the results of the grain growth simulation do not reflect the observed microstructure. The calculation is just too simple.

I tried wrapping my head around building a script for the case of the whole oven treatment (with heating and the phase transformations from alpha to gamma upon heating and back from gamma to alpha upon cooling), but I am struggling hard, as I am too unfamiliar with integrating both the phase transformations and the temperature curves.

That's why I would like to ask if anyone on the forum can help me build my desired case (hopefully in a way that makes it easy to change the heat-treatment parameters)? Or even simpler, if anyone already has such a script and could share it with me?

Thanks in advance!
Best regards, Fabian

Dear Fabian,

Welcome to the MICRESS Forum!

If I understood correctly, the main process you want to simulate is the phase transformation of ferrrite to austenite upon heating, starting from a single-phase ferrite initial microstructure, followed by grain coarsening of the austenite grains at 1300°C, and then cooling with an opposite phase transformation from austenite to ferrite. As this process involves the phase transformation of an alloy (Fe-Mn-C), concentration coupling is required, as well as adding the 3 elements and considered phases in the simulation setup.

Fortunately, there is already a MCRESS application example (A004_Gamma_Alpha_TQ_dri) which represents a perfect starting point for setting up such a simulation. It already uses the elements and phases which you need, defines an initial grain microstructure, and has a preset of parameters for the required phase-interactions:

1/1: Grain coarsening of austenite
1/2: Phase transformation between austenite and ferrite assuming nple-conditions
2/2: Grain coarsening of ferrite

Furthermore, it contains 3 nucleation conditions for the ferrite phase at triple junctions, interfaces, and in the bulk of austenite grains, which is a good starting point at least for the backward phase transformation on cooling.

In order to modify the input file to your process conditions, you need to:

- change the alloy composition to your needs,

- implement the temperature-time curve including heating, holding and cooling by using a "time_dependent" temperature trend (either obtained from a text file or written directly into the MICRESS input file). To avoid potential but completely unnecessary numerical issues, I would recommend not really to begin and end at room temperature but at a higher temperature (perhaps just before the start of the first and after the end of the second phase transformation, respectively). At lower temperatures, diffusivity is extremely small, and nothing will happen anyway.

- symmetrically amend 3 further nucleation types for formation of austenite phase out of ferrite,

- change the grains of the initial microstructure from austenite to ferrite ("Initial Microstructure"),

- and adapt the output time-steps and the total simulation time so that it fits to the temperature-time curve

Now, the simulation of your process should, at least in principle, work more or less according to your specification. However, the two most important tweaks need still to be done:

1.) Adjust physical parameters using values from literature and/or by calibrating to match the experimental findings. This concerns especially the interfacial energies, the interface mobilities of the 1/1 and 2/2-interfaces, and the nucleation conditions. Do not trust preset parameters!

2.) Adjust numerical parameters, especially grid resolution, but also checking intervals, updating intervals, etc., to your boundary conditions (heating rate, cooling rate, physical parameters)

This should be all (if I did not miss something...).
Give it a try!

Bernd