issues about thermodynamics and their coupling to MICRESS
-
zhubq
- Posts: 84
- Joined: Mon Jun 22, 2009 7:33 pm
Post
by zhubq » Tue Aug 13, 2013 3:11 am
Bernd wrote:Hi Ben,
yes, that's it! Did you try already?
Bernd
Hi Bernd,
not yet.
We got MICRESS and TCS servers installed on two computers, and are figuring out how they can communicate.
Ben
-
zhubq
- Posts: 84
- Joined: Mon Jun 22, 2009 7:33 pm
Post
by zhubq » Tue Sep 10, 2013 11:35 pm
Hi Bernd,
Eventually, our TQS version is running well. Thanks a lot for help from your group.
But I got a warning about GES5 file format incompatible with TQ library.
I generated the file with TCCS. It should be no problem.
Could you please double check it?
Thank you.
Ben
-
Attachments
-
- GES.rar
- (8.01 KiB) Downloaded 499 times
-
- 1d_per_fer1043.in
- (20.44 KiB) Downloaded 452 times
-
ralph
- Posts: 167
- Joined: Wed Apr 27, 2011 4:42 pm
- anti_bot: 333
Post
by ralph » Wed Sep 11, 2013 4:46 pm
Hi Ben,
unfortunately, the GES5 file format is also architecture dependent (32/64 bit).
You attached a file made with a 32 bit version of TCCS. It works with MICRESS_TQS_x32.
Ralph
-
ralph
- Posts: 167
- Joined: Wed Apr 27, 2011 4:42 pm
- anti_bot: 333
Post
by ralph » Wed Sep 11, 2013 4:55 pm
All the MICRESS versions can be found in the MICRESS.../bin folder for a normal installation.
For Linux, the easiest way to start another version is to copy the executable (MICRESS_TQS_x32) and matching TQ library (lib32/libtq32_S.so) to a new folder. Just rename the library to libtq.so and add the folder to the LD_LIBRARY_PATH.
Ralph
-
zhubq
- Posts: 84
- Joined: Mon Jun 22, 2009 7:33 pm
Post
by zhubq » Thu Sep 12, 2013 5:20 am
Hi Ralph,
Thank you. Now everything is OK.
Does MICRESS_TQS 6_001 version work well with paraTQ?
I got wrong thermodynamic data: for example, c0 is not consistent with what's obtained from TC.
later on, error 604 in QPARAD CALFUN. Total concentration greater than 1.
Here is part of my setting.
If I changed all components as "normal", the results are correct.
I also attach my files in case you may check them.
Thank you.
Ben
*******************************************************************************************************************
Please specify the redistribution behaviour of each component:
# Format: forward [backward]
# Options: nple para paraTQ normal ATC [mob_corr|verbose]
# Component 1 C
normal
# Component 2 MN
paraTQ
# Component 3 CR
paraTQ
# Component 4 SI
paraTQ
*******************************************************************************************************************
-
Attachments
-
- 1d_per_fer2.rar
- (12.76 KiB) Downloaded 465 times
-
Bernd
- Posts: 1506
- Joined: Mon Jun 23, 2008 9:29 pm
Post
by Bernd » Fri Sep 13, 2013 9:19 pm
Hi Ben,
I still had no time to check your example, but just a quick guess: It could be related to the periodic boundary (EW) conditions in 1D - I fear this would not work, because each cell is its own neighbor. But it could also be something else...
I guess that you compared the c0 with the para-equilibrium model in Thermo-Calc.The values for the substitutional elements should be quite similar in both phases (not for C).
Bernd
-
zhubq
- Posts: 84
- Joined: Mon Jun 22, 2009 7:33 pm
Post
by zhubq » Tue Sep 17, 2013 12:22 am
Hi Bernd,
The periodic boundary condition has no problem for 1d. Now I got correct results. I guess the problem is in the input of initial concentrations.
But for some reasons, I got correct results now.
Thank you.
Ben