Learning Forum

[Swapnil]

Hello, 

I need to run a two-way FSI simulation on the HPC cluster. In the submission script, I am not able to give an execution path for Ansys mechanical (transient structural) for 2023R2 version. I have mentioned the execution path for system coupling and fluent below.

module load ansys/2023r2

/lfs/sware/ansys2023r2/ansys_inc/v232/SystemCoupling/bin/systemcoupling -R syscplg.py &

/lfs/sware/ansys2023r2/ansys_inc/v232/fluent/bin/fluent 3d -g -mpi=intel -t40 -cnf=${PBS_NODEFILE} -scport=${HOSTPORT[0]} -schost=${HOSTPORT[1]} -scname="${JOB[4]}"-i input.jou> output.FLUENT &

ansys2023r2 -b -scport=${HOSTPORT[0]} -schost=${HOSTPORT[1]} -scname="${JOB[2]}" -i Mech_FSI_A1.dat -o output.Ansys -np20

[Rahul]

Hey Swapnil,

To run system coupling in batch mode, you just need system coupling path to be specified. Starting the System Coupling CLI (ansys.com)
Also refer Using Parallel Processing Capabilities (ansys.com) user guide. 

[Swapnil]

I tried giving the system coupling path and I am facing the following error now.

application-specific initialization failed: Can't find a usable init.tcl in the following directories: 

    /home2/polyflowbuilds/buildagents/milpolybld01/work/1210bc2a0080eb69/Tcl-tk/8_5_11/linx64/Release/tcltk/lib/tcl8.5 /usr/lib/tcl8.5 /lib/tcl8.5 /usr/library /library /tcl8.5.11/library /tcl8.5.11/library

 

This probably means that Tcl wasn't installed properly.

 

Log file:

MPI Option Selected: intel

-------------------------------------------------------------------------------

 

+-----------------------------------------------------------------------------+

| Errors were found during data model validation.                             |

+-----------------------------------------------------------------------------+

Error: To begin a coupled analysis setup, define at least two coupling         

    participants, one interface, and one data transfer.                        

 

Traceback (most recent call last):

  File "PyLib/main/Controller.py", line 204, in

  File "PyLib/main/Controller.py", line 199, in _run

  File "PyLib/main/Controller.py", line 115, in _executeScript

  File "PyLib/kernel/commands/__init__.py", line 34, in readScriptFile

  File "PyLib/kernel/commands/CommandManager.py", line 203, in readScriptFile

  File "Scycpl.py", line 2, in

    Solve()

  File "PyLib/kernel/commands/CommandDefinition.py", line 74, in func

  File "PyLib/kernel/commands/__init__.py", line 30, in executeCommand

  File "PyLib/kernel/commands/CommandManager.py", line 165, in executeCommand

  File "PyLib/cosimulation/externalinterface/core/solver.py", line 137, in execute

  File "PyLib/cosimulation/solver/__init__.py", line 123, in solve

  File "PyLib/cosimulation/solver/__init__.py", line 499, in __checkSetup

  File "PyLib/cosimulation/datamodelrules/__init__.py", line 270, in raiseValidationError

RuntimeError: Errors were found during data model validation.

Shutting down System Coupling compute node processes.

[Rahul]

Do you have all the required files for two-way FSI? System coupling settings file should have participants defined.

[Swapnil]

Yes, I have added all the required participants (Fluent Transient structural) and I have added interface and data transfers as well. I think the issue is with the partitioning algorithm.

I need to run the simulation on HPC (distributed memory type) cluster with 20 cores for fluent and 4 cores for transient structural. I have checked the manual but I am not clear about which partition algorithm should I use and how to specify the additional arguments and parallel fraction. 

I am using the following submission script:

#!/bin/bash

#PBS -o logfile.log

#PBS -e errorfile.err

#PBS -l walltime=40:00:00

#PBS -l select=1:ncpus=24

tpdir=echo $PBS_JOBID | cut -f 1 -d .

tempdir=$HOME/scratch/job$tpdir

mkdir -p $tempdir

cd $tempdir

cp -R $PBS_O_WORKDIR/* .

module load ansys2023r2

/lfs/sware/ansys2023r2/ansys_inc/v232/SystemCoupling/bin/systemcoupling -R Scycpl.py

mv ../job$tpdir $PBS_O_WORKDIR/.

 

could you please suggest the required changes?

[Rahul]

Core count or fraction of compute resources to be allocated to the participant. Used to partition resources across coupling participants that are running in parallel.

Accepts real values greater than or equal to 0. Default value is 1.

For cluster with 20 cores: Fluent fraction = 1 => 20 cores; Transient fraction = 0.2 => 4 cores.

With SharedAllocateMachines partitioning algorithm , If all the fractions are between 0.0 and 1.0, then each participant will run on exactly the specified fraction of the total core count.

For more detailed information on how these fractions are applied, see Resource Allocation Fractions in the System Coupling User's Guide.

Shown in the GUI as the Coupling Participant | Execution Control | Parallel Fraction setting.

[Swapnil]

I tried to simulate the case in HPC cluster but system coupling  log file is written with the following lines while MAPDL log file is blank. Is it because of the unavailability of liscence?

Awaiting connections from coupling participants...
Elapsed time: 5 min. Still awaiting connection from: MAPDL-2. If you feel the participant(s) should already have connected, check participant logs and/or abort the run.
Elapsed time: 10 min. Still awaiting connection from: MAPDL-2. If you feel the participant(s) should already have connected, check participant logs and/or abort the run.