Problem with multiframe restart of two-way fsi coupled problem

solomedic · July 26, 2011, 23:57

You said it,it is impossible to do harmonic analysis in coupled field simulation?
can you throw some light on this statement?
I want to have some recent restart point whenever the solution is stopped in unusual manner like power trips off,it shows you some error and you want to correct that and then resart.
Thanks Dimone for your patience and help

solomedic · July 27, 2011, 04:08

hi dimoine, as you may know that I am doing two way fsi in workbench coupling CFX and Mechanical Module of Ansys. but as per part of my project I have to do two way fsi in ansys Apdl and CFX as well, so kindly give me some tips how to start with, a very general question but try to answer as APDL is not my field so I am starting with a general question

Dimone · July 27, 2011, 04:16

The only thing I can suggest, start with corresponding section of ansys apdl manual (MFX analysis). It cannot be described in forum post.

lingdeer · December 18, 2011, 02:03

Hi

My 2way FSI simulation get killed and I need to restart them.

I added
MFOU,1
MFRC,10,2

to backup my ANSYS results.

And I does see
ANSYS.r001, ANSYS.r002 being written up to the time step it get killed.

However, one thing very weird is, in the .ansys folder, I can only find .db file instead of .rdb file.

Which is really strange, as for the same case same setting, I have both .db and .rdb written in other runs.

I now proceed to restart using my only .db file to restart, and also the .trn of the lastest time step.

I wondered what I could possibly be wrong at the beginning that it didnt write a .rdb file? (I really have no clue.) The run is smooth and the rXXn file are written.

What can I do to prevent this from happening?
And is it ok if I used the .db file?

lingdeer · December 18, 2011, 02:05

In addition, in case .db wouldn't work, what I can do to 'save' my run back with the .rXXn files I have?

Also, if I backup every 10 time steps. What should I do to restart if it crash at the middle? (MFRC, 10,2)
e.g. I crashed at the 56th timestep
I m thinking of:
restart with the .rdb, and I supposed it to be last written at the 50th timestep?
and select 50_full.trn to restart from the 50th timestep?

Correct me if I am wrong.

Thanks!!!

Dimone · December 20, 2011, 07:43

As I remember, I've turned on some general (not MFX) restart option for .rdb file to be written (sorry, but I do not remember now - look at the "Restart solution" section of manual for general structural analysis. MFRS is either needed). Although, the fact, that in some runs .rdb is written and in other it is not - is really strange for me.

In you example you can restart only from 50 step. IMHO usually it is more useful to write rXX files every step with restriction on their count (they are cycled through numbers in this case). Although, it may be slower, as more write file operations are performed (I did not check this).

lingdeer · December 20, 2011, 11:10

Thanks Dimone.

I added RESCONTROL and MFRSTART and they didnt work.
I really didnt know what did I do differently for this run to make it not writting that file...

December 18, 2011, 02:05		#25
lingdeer Member Lingdeer Join Date: Jul 2011 Posts: 49 Rep Power: 15	In addition, in case .db wouldn't work, what I can do to 'save' my run back with the .rXXn files I have? Also, if I backup every 10 time steps. What should I do to restart if it crash at the middle? (MFRC, 10,2) e.g. I crashed at the 56th timestep I m thinking of: restart with the .rdb, and I supposed it to be last written at the 50th timestep? and select 50_full.trn to restart from the 50th timestep? Correct me if I am wrong. Thanks!!! Last edited by lingdeer; December 18, 2011 at 14:40.

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Coupled FSI using ANSYS-FLOTRAN	yusun	Main CFD Forum	2	January 10, 2010 04:57
Problem with Coupled Solver!	Parag Gadgil	Main CFD Forum	2	December 20, 2007 14:02
Looking for a open FEM code to solve a FSI problem	Esto	Main CFD Forum	3	June 15, 2006 12:02
problem on DPM coupled with CFD	harry	Main CFD Forum	0	June 2, 2005 07:12
Help: Flow-Response Dynamic Coupled problem	wavefront	Main CFD Forum	3	September 26, 2001 11:28

July 26, 2011, 23:57		#21
solomedic Member Muhammad Amir Join Date: May 2011 Posts: 31 Rep Power: 15	You said it,it is impossible to do harmonic analysis in coupled field simulation? can you throw some light on this statement? I want to have some recent restart point whenever the solution is stopped in unusual manner like power trips off,it shows you some error and you want to correct that and then resart. Thanks Dimone for your patience and help

July 27, 2011, 04:08		#22
solomedic Member Muhammad Amir Join Date: May 2011 Posts: 31 Rep Power: 15	hi dimoine, as you may know that I am doing two way fsi in workbench coupling CFX and Mechanical Module of Ansys. but as per part of my project I have to do two way fsi in ansys Apdl and CFX as well, so kindly give me some tips how to start with, a very general question but try to answer as APDL is not my field so I am starting with a general question

July 27, 2011, 04:16		#23
Dimone New Member Dmitry Semikin Join Date: Dec 2010 Posts: 16 Rep Power: 15	The only thing I can suggest, start with corresponding section of ansys apdl manual (MFX analysis). It cannot be described in forum post.

December 18, 2011, 02:03		#24
lingdeer Member Lingdeer Join Date: Jul 2011 Posts: 49 Rep Power: 15	Hi My 2way FSI simulation get killed and I need to restart them. I added MFOU,1 MFRC,10,2 to backup my ANSYS results. And I does see ANSYS.r001, ANSYS.r002 being written up to the time step it get killed. However, one thing very weird is, in the .ansys folder, I can only find .db file instead of .rdb file. Which is really strange, as for the same case same setting, I have both .db and .rdb written in other runs. I now proceed to restart using my only .db file to restart, and also the .trn of the lastest time step. I wondered what I could possibly be wrong at the beginning that it didnt write a .rdb file? (I really have no clue.) The run is smooth and the rXXn file are written. What can I do to prevent this from happening? And is it ok if I used the .db file?

December 20, 2011, 07:43		#26
Dimone New Member Dmitry Semikin Join Date: Dec 2010 Posts: 16 Rep Power: 15	As I remember, I've turned on some general (not MFX) restart option for .rdb file to be written (sorry, but I do not remember now - look at the "Restart solution" section of manual for general structural analysis. MFRS is either needed). Although, the fact, that in some runs .rdb is written and in other it is not - is really strange for me. In you example you can restart only from 50 step. IMHO usually it is more useful to write rXX files every step with restriction on their count (they are cycled through numbers in this case). Although, it may be slower, as more write file operations are performed (I did not check this).

December 20, 2011, 11:10		#27
lingdeer Member Lingdeer Join Date: Jul 2011 Posts: 49 Rep Power: 15	Thanks Dimone. I added RESCONTROL and MFRSTART and they didnt work. I really didnt know what did I do differently for this run to make it not writting that file...