Nitriding

solid-solid phase transformations, influence of stresses and strains
Post Reply
Basti
Posts: 3
Joined: Mon Aug 01, 2022 4:18 pm
anti_bot: 333

Nitriding

Post by Basti » Mon Aug 01, 2022 5:03 pm

Hello everyone.

I have the task of simulating the microstructural changes during nitriding of austenitic and duplex Fe-alloys. We have knowledge of how N effect the material. Now we want to set up a simulation supporting our statement before doing further nitriding experiments and microstructural investigations.

On first research I did not find any topics or micress examples with those keywords: gas-phase, nitriding or carburizing. Do you have some input for me: How to set up a nitriding simulation in micress (TC-cuppled)?

I know we need to work with simplifications along this nitriding process. My first attempt would be starting from a higher N concentration in the matrix. But this is in conflict with the phase constitution, phase fractions and of course the microstructure evolution of new phases like precipitates.

Feel free to comment, your help is welcome.


P.S. I only have access to the TC diffusion database in micress, not the TC DICTRA module ...

Bernd
Posts: 1504
Joined: Mon Jun 23, 2008 9:29 pm

Re: Nitriding

Post by Bernd » Mon Aug 01, 2022 7:01 pm

Dear Basti,

Welcome to the MICRESS forum!

Up to now I do not know of anybody who made a MICRESS simulation of Nitriding before. Nevertheless, I already have been thinking about this application before, and it appears very interesting and also promising, especially because all the required thermodynamic knowledge seems already to be included in the TCFE database!

However, one can approach the challenge on different levels of simplification/approximation, and if you are not yet very experienced in using the MICRESS software I would strongly advise to start at the most simple level which still makes sense.

It would of course have been easier if already an input file (driving file) for nitriding existed. So you need to build up your system yourself, i.e. decide on which elements and phases are needed, and create the corresponding .GES5-file. Here also I would recommend to start from the simplest possible system, e.g. an austenitic steel consisting of a ternary Fe-(M)-N or a quaternary Fe-(M)-C-N which forms the N-containing precipitate phase being most relevant for you. You can check out a suitable system with Thermo-Calc that also should be close to your alloys of interest.

With MICRESS you never start completely from scratch, because there are too many parameters and options which you don't know beforehand and which you normally even don't need to touch. I would e.g. take the training example T02_21_FeCMn_E_GammaAlphaIsotherm_2D_TQ_dri a a starting point. Of course, you need to adapt the element and phases, and also the process conditions (temperature, time scale, etc.). I a first step, I would include N just with a fix amount (as you also proposed), and elaborate the physical and numerical conditions for the nitride phase to form. Later you can introduce nitrogen by using a fixed boundary condition on one side of the domain, (or even by putting a gas phase in contact with the correct C:H ratio for acetylene...).

At the beginning you can just stick to the initial austenite grain microstructure provided by the T02_21 example. Later, you may want to include a more realistic initial microstructure, which you can either synthezise by placing initial phases and grains, by reading from experimental metallographic images, or by simulating the production process itself with MICRESS...

This way, step by step, you can build up a more and more elaborate simulation model and get a sound feeling about the nitriding process.

Please don't hesitate to ask if you have more questions or if you need help.

Bernd

Basti
Posts: 3
Joined: Mon Aug 01, 2022 4:18 pm
anti_bot: 333

Re: Nitriding

Post by Basti » Tue Aug 02, 2022 12:05 pm

Dear Bernd,

I appreciate your interest in this topic and would like to discuss it with you.

As you say:
Later you can introduce nitrogen by using a fixed boundary condition on one side of the domain, (or even by putting a gas phase in contact with the correct C:H ratio for acetylene...).
This step-by-step strategy describe what I am looking for. Especially to introduce boundary conditions or even better a gas phase is quit challenging for me. In Particular with little to no examples and documentation.

To apply a conenctration boundary conditions, first I have to understand how it works. I looked up in the Micress User Guide Vol 2: Running MICRESS - Chapter 3.2.19: Boundary Conditions. On page 52 and 53 they give explanations to the available conditions. In my understanding it will affect the set-up concentration field. Therefore its nots able to apply a different concentration field on the domain boundary itself but with 1D-fixed it´s possible to have a constant concentration field.
Maybe you have some important details I am missing?

The second solution would be the master way. If you have any hints how to simulate a gas phase I would like to read it. The only mention of a gas phase in this manual is page 87 chapter 4.12.1: pressure conditions for TQ.

greetings Basti

Bernd
Posts: 1504
Joined: Mon Jun 23, 2008 9:29 pm

Re: Nitriding

Post by Bernd » Tue Aug 02, 2022 2:55 pm

Dear Basti,

Boundary conditions are defined in the upper part of the MICRESS input file (after section "Model") in form of a string of single characters, which stand for the different condition types:

# Boundary conditions for concentration field in each direction
# Options: i (insulation) s (symmetric) p (periodic/wrap-around) g (gradient) f (fixed)
# Sequence: W E B T
ssif


Once you use the "f" character for "fixed", you will be asked later down in section "Process conditions" for the value(s) (to which the boundary is fixed). If applied to the concentration field, the choice is automatically taken for all elements, i.e. you fix the complete alloy composition at the specified boundary. In the case above, the "fixed" condition is chosen for the top boundary in a 2D-simulation, i.e. the composition at the upper side of the simulation domain is set to this composition. In your case you would fix the upper boundary to a composition which is the nominal + N, nitrogen will diffuse into the domain from above, just as if there was a nitriding gas in contact with the top. You can externally calculate a suitable N-composition using Thermo-Calc either based on literature information on the N-activity or by equilibrium calculations which include the gas phase.

The smarter (but also more complex) way, of course, would be to include the gas-phase into the MICRESS simulation. Including a gas-phase is quite straightforward as long as it acts only as boundary condition (i.e. we are not interested in the correct volume and shape). We did that for a pure oxygen phase and Reactive Air Brazing some time ago:

M.Apel, G.Laschet, B.Böttger and R.Berger, Phase Field Modeling of Microstructure Formation, DSC Curves, and Thermal Expansion for Ag-Cu Brazing Fillers Under Reactive Air Brazing Conditions, Advanced Engineering Materials (2014)
DOI: 10.1002/adem.201400101

R.Berger, M.Apel and B.Böttger, Phase Field Modeling Applied to Reactive Air Brazing: Investigating Reaction Kinetics with Focus on Oxygen Exchange, Advanced Engineering Materials (2014)
DOI: 10.1002/adem.201400103

As far as I know, the GAS phase in the Thermo-Calc Steel database contains acetylene as a species, so that the gas equilibria should be calculated correctly, and the nitrogen (and H, C) composition at the boundary to the metal would automatically be determined depending on pressure, temperature and composition of the gas. Of course, the gas phase elements (C,H,N) need to be included in the simulation and thus also in the metallic phase. For defining the pressure of the gas you can use the pressure condition for TQ which you mentioned (if you don't use standard pressure).

Bernd

Basti
Posts: 3
Joined: Mon Aug 01, 2022 4:18 pm
anti_bot: 333

Re: Nitriding

Post by Basti » Wed Aug 03, 2022 3:44 pm

Thanks for your kind help, Bernd. This was very helpfull. I will now work on the implementation. Surely this will take some time.

greetings Basti

Post Reply