From liquid to solid and to liquid again

[CharMIC]

Hej

I have done a solidification simulation of Nickel alloy where I start from a liquid and cool down to temperature where everything become solid and reheat to a temperature above liquids temperature. The cooling rate and the heating rate are almost same.

But during the reheating I don't get any melting or incipient melting. Microstructure stay at complete solid. The cooling down and heating up was simulated in a single simulation using 2 connecting points under the boundary conditions

Any reason why I donot see any melt in the simulation?
Do I need to define a seed for the liquid phase to see melting.

BR
Chamara

[Bernd]

Dear Chamara,

Exactly!

The liquid phase is just a phase like the others, and without nucleation it won't appear. The only chance is when it did not solidify completely, then remelting could occur without nucleation.

Just a tip: Use "add_to_grain" with option "new set" or "grain_number 0" in order not to get different liquid grains - they would have grain boundaries when they join...

Bernd

[CharMIC]

Hej Bernd

As always Many thanks

BR
Chamara

[CharMIC]

Hej Bernd

I am some what confuse with defining nucleation data for the liquid phase.

I have gamma and laves phase in my solidified structure. So how should define the nucleation sites for liquid during the reheating. Any tip or advice?

BR
Chamara

[CharMIC]

Also how should I define the undercooling during reheating. Should it be negative undercooling since its heating simulation?

BR
Chamara

[deepumaj1]

Hi Chamara,
The undercooling should always be a positive value, whether the simulation is heating or cooling.

As for the nucleation site, you may play with the 'reference phase' and 'substrate phase' in nucleation data to put the liquid nuclei where you want to.

Deepu

[Bernd]

Hi,

I fully agree with what Deepu said. I just want to add that especially those places should be taken into account for nucleation which solidified last, because typically they will melt first. This probably is the [gr]gamma[/gr]-Laves interface.

Another hint: When using "add_to_grain", the small-grain model typically does not work as properly as normal, because the model uses the highest fraction of all seeds (including already existing entities) with this grain number. In case that nuclei do not grow properly, it typically helps to set the grain radius not to 0 but to a somewhat bigger value r<Δx. Then, the initial liquid fraction is bigger than 2*phMin, and it does not disappear so easily.

Bernd

[CharMIC]

Thanks Deepu and Bernd.

[CharMIC]

Another question..

I am running the cooling and heating in the same simulation. I have two nucleation for Laves and Liquid phase.

I get the following error during the simulation.

Any idea about the problem. Is it related to defining nucleation in the same simulation and it create problems. Do I need to first do a solidification simulation just nucleating laves and then use restart option and define liquid nucleation during heating?

BR
Chamara

[Bernd]

Hej Chamara,

sorry, I somehow missed your post...

The problem you show should have nothing to do with checking simultaneously nucleation for Liquid and Laves. I would need to see the entire input file find other possible reasons...

However I have two general tips:

- If you would nucleate two different phases with exactly identical seed types including identical check frequency, it would turn out that seeds of both phases appear always at the same place - this is probably not what you want. My personal method to avoid that is to specify different odd nucleation intervals (e.g. 3.1368 s and 2.93845 s etc.) and/or different random seeds.
- You could use the nucleation intervals "from file" in case you wish to start checking for Liquid only after heating started.

Bernd