Boundary Conditions

dendritic solidification, eutectics, peritectics,....
Bernd
Posts: 1074
Joined: Mon Jun 23, 2008 9:29 pm

Re: Boundary Conditions

Post by Bernd » Fri Jul 06, 2018 4:00 pm

OK, I said it is "not impossible".

I first need to find out how difficult it would be...

Bernd

CharMIC
Posts: 196
Joined: Mon Aug 21, 2017 5:01 pm
anti_bot: 333

Re: Boundary Conditions

Post by CharMIC » Wed Jul 18, 2018 8:49 am

Hej Bernd,

The reason Why I ask for this is that We are investigating the effect of undercooling on the solidification velocity. We have simulated steady sate case with constant undercooling with moving frame option. But now we want to link this to thermal profiles taken from the melt pool of Electron Beam Melting simulation. In order to get the correct tip radius, we need very finer mesh. But if that mesh is used for the entire melt pool height, then the computational time is too large.

BR
Chamara

Bernd
Posts: 1074
Joined: Mon Jun 23, 2008 9:29 pm

Re: Boundary Conditions

Post by Bernd » Wed Jul 18, 2018 9:09 am

Hej Chamara,

I am out of office this week. Once I am back, I will check out how difficult it is, and - if not too complicated - try to implement it.

If I forget, please remind me :)

Bernd

CharMIC
Posts: 196
Joined: Mon Aug 21, 2017 5:01 pm
anti_bot: 333

Re: Boundary Conditions

Post by CharMIC » Tue Jul 31, 2018 2:25 pm

Hej Bernd,

This is just to remind you about the implementation :)

BR
Chamara

Bernd
Posts: 1074
Joined: Mon Jun 23, 2008 9:29 pm

Re: Boundary Conditions

Post by Bernd » Tue Jul 31, 2018 6:21 pm

Hej Chamara,

Thanks for reminding me :?

I just checked our code and found that it should be possible to make the changes without causing too many difficulties. However, I am short in time, and therefore I would like to propose you a "deal": I would make the changes as soon as possible and send you the modified executable, but leave the testing to you! I would ask you to develop a small, fast and simple test case (e.g. one with a set of linear profiles with constant T-offset, so that it will do the same as if one just would use a constant gradient and cooling rate) which we could use later in our automated testbed.

For technical reasons, we would send you a modified beta-version of MICRESS (which has already other new developments and, by the way, meanwhile contains quite a few improvements over MICRESS 6.4...). If necessary we can create a "correct" 6.4 for you later after the modifications passed your tests.

What do you think?

CharMIC
Posts: 196
Joined: Mon Aug 21, 2017 5:01 pm
anti_bot: 333

Re: Boundary Conditions

Post by CharMIC » Wed Aug 01, 2018 8:55 am

Hej Bernd,

Your deal sounds good. I can do the testing. However We have some issues with our MICRESS installed server. So I am not able to run any simulation until our IT department fix it. Hope fully they do it before end of August.

Hope this is okay with you.

BR
Chamara

Bernd
Posts: 1074
Joined: Mon Jun 23, 2008 9:29 pm

Re: Boundary Conditions

Post by Bernd » Wed Aug 01, 2018 11:29 pm

Great!

Hope you are operational soon!

CharMIC
Posts: 196
Joined: Mon Aug 21, 2017 5:01 pm
anti_bot: 333

Re: Boundary Conditions

Post by CharMIC » Mon Sep 03, 2018 9:32 am

Hi Bernd,

Do we expect some difference in a unit cell simulation if we use periodic BC vs Symmetric BC?

BR
Chamara

Bernd
Posts: 1074
Joined: Mon Jun 23, 2008 9:29 pm

Re: Boundary Conditions

Post by Bernd » Mon Sep 03, 2018 10:10 am

Dear Chamara,

of course yes, because periodic and symmetric boundary conditions are not the same...


However, you are right that if you use a unit cell which consists of fully periodic "units" like e.g. a full 2D-dendrite cross section in a regular cubic dendrite array, then both cases should be identical. But the symmetric boundary condition in this case would allow you to chose a smaller "unit cell" with exactly the same results (1/4 dendrite in 2D).

Bernd

Post Reply