[zegtuhetmaar]

I am doing a multifield analysis (Ansys-CFX) on a linux server ia64.
When I start my analysis with command:

cfx5solve -def 'deffile' -batch -start-method "MPICH Local Parallel" -part 4

it will kill the analysis just after initializing:

MULTI FIELD ANALYSIS STEP OPTIONS
 TIME AT BEGINNING OF ANALYSIS. . . . . . .   0.0000    
 TIME AT END OF ANALYSIS. . . . . . . . . .   2.0000    
 TIME STEP NUMBER. . . . . . . . . . . . . . .   1
 STAGGER ITERATION NUMBER. . . . . . . . . . .   1.   CUM MF ITER =      1
                  FIELD INFORMATION FOR MULTIFIELD ANALYSIS
 CURRENT FIELD NAME FOR ANALYSIS. . . . . ANSYS
 TIME AT BEGINNING OF CURRENT ANALYSIS. . .   0.0000    
 TIME AT END OF CURRENT ANALYSIS. . . . . .  0.50000E-01
 THIRD SYNCHRONIZATION POINT OF MASTER CODE: 
 SERVE SLAVES THE TIME STEP CONTROL INFORMATION TO START A SOLVE
 FOURTH SYNCHRONIZATION POINT OF MASTER CODE: 
 LOAD TRANSFER AT THE BEGINNING OF SOLUTION
 *********************
 MF_send_Error
  Msg:     
 Read                                                                           
 
 
 
 *** FATAL ***                           CP =      43.848   TIME= 12:45:51
 Error during get surface loads: Total Force.  Error Message: Read.      
 ALL CURRENT ANSYS DATA WRITTEN TO FILE NAME= ansabort.db
  FOR POSSIBLE RESUME FROM THIS POINT

However, using a 32 bit machine I don't get this error:

MULTI FIELD ANALYSIS STEP OPTIONS
 TIME AT BEGINNING OF ANALYSIS. . . . . . .   0.0000    
 TIME AT END OF ANALYSIS. . . . . . . . . .   2.0000    
 TIME STEP NUMBER. . . . . . . . . . . . . . .   1
 STAGGER ITERATION NUMBER. . . . . . . . . . .   1.   CUM MF ITER =      1
                  FIELD INFORMATION FOR MULTIFIELD ANALYSIS
 CURRENT FIELD NAME FOR ANALYSIS. . . . . ANSYS
 TIME AT BEGINNING OF CURRENT ANALYSIS. . .   0.0000    
 TIME AT END OF CURRENT ANALYSIS. . . . . .  0.50000E-01
 THIRD SYNCHRONIZATION POINT OF MASTER CODE: 
 SERVE SLAVES THE TIME STEP CONTROL INFORMATION TO START A SOLVE
 FOURTH SYNCHRONIZATION POINT OF MASTER CODE: 
 LOAD TRANSFER AT THE BEGINNING OF SOLUTION
FORCE SUM FOR FIELD   1 ACROSS INTERFACE   1
RECEIVING FORCE FX SUM =  0.0000    
RECEIVING FORCE FY SUM =  0.0000    
RECEIVING FORCE FZ SUM =  0.0000

But this machine is way too slow.

I also get at the same moment a unaligned access warning.

Does anyone have any idea how to solve this problem? Thanks

[baserinia]

Hi, I don't know whether this is the answer to your question, but I usually use PVM rather than MPICH for local parallel runs on my AMD64 Linux machine.

[zegtuhetmaar]

Originally Posted by baserinia

Hi, I don't know whether this is the answer to your question, but I usually use PVM rather than MPICH for local parallel runs on my AMD64 Linux machine.

Thanks baserinia, but unfortunately it doens't work. Neither does HP MPI.

I hope somebody knows a solution?

[zegtuhetmaar]

Since the update to 12.1, CFX got a GUI and I can run my calculations from the interface.

[alinik]

hi

I am getting the same error"error during get surface loads"

How did you make it work?

[zegtuhetmaar]

just a "surface load" error could mean anything. Probably something wrong with the communication between different software. be more precise with your error

[alinik]

The text of the error that I am getting is exactly the same as the one that you were getting. what more do you need to know?
if you want to know about the problem, I can say that it is a two way FSI simulation for modeling an unsteady non-newtonian fluid flow inside a compliant vessel.

[acaba]

I'm not sure if this will solve your problem, but there was a service pack for Ansys 12.1 to enable 2-way FSI for 64 bit computers. Check out the ANSYS customer portal -> Download Software -> Standard Versions -> Product Updates

Aaron