Learning Forum

[Ben_Ben]

Hi,

I am getting different simulation results for the same input file when I change the command line for my HPC. To show the differing results I had composed a test simulation with radial Cylindrical BCs applied on the surface of cylinder. 

I currently use the following command to run LS-Dyna MPP sims on my old HPC:

lsdyna -dis -np 24 -lsdynampp -dp I=test_simulation.k memory=700m

And the simulation works fine:



However I now need to use an LS-Dyna executable for my new HPC so I use the following command:

mpirun -np 24 /users/mn17b2m/LSD/solver/ls-dyna_mpp_d_R15_0_2_x64_centos79_ifort190_avx512_openmpi405 s=INTFOR i=test_simulation.k memory=700m

and I get 0 displacement:


So I thought I should check my old HPC but with the same executable as the new HPC:

mpirun -np 24 /home/home01/mn17b2m/LSD/ls-dyna_mpp_d_R15_0_2_x64_centos79_ifort190_avx512_openmpi405/ls-dyna_mpp_d_R15_0_2_x64_centos79_ifort190_avx512_openmpi405 i=test_simulation.k memory=700m

result:



I have tried several other executables different versions and different instruction sets. All give the same output. Other BCs work but not the BC that I use for my simulation.

For Context this is how I prescribe the BCs (example for one node:
*BOUNDARY_PRESCRIBED_MOTION_NODE
$#     nid       dof       vad      lcid        sf       vid     death     birth
    216188         4         2         3       1.0     158421.00000E28       0.0

If anyone has any suggestions I'd be very grateful.

Cheers, 
B

[igandiko]

What solver are you calling using this command: lsdyna -dis -np 24 -lsdynampp -dp I=test_simulation.k memory=700m? Please check the banner in mes0000 file. 

I would compare the d3hsp files of good vs. bad run to see if there are any differences in how the solver is interpreting the input. Any chance there is a spacing issue in the *boundary_<> card? For example, in line 2 of the card, adding an extra space in the front may shift death time to be read as "21.00000E2" instead of "1.00000E28" and birth time as "8" instead of 0.0 which mean prescribed motion will not kick in until time=8. Maybe read the input file in LS-PrePost and save it in the format suitable for R15 (File > save as > save keyword as > output version= v971_R15). 

-------------------
Introducing Ansys Assistant a 24/7 AI-powered virtual assistant for active Ansys academic customers.

[Ben_Ben]

Hi Igandiko,

I am using R13 for the lsdyna command. 

Thanks for your suggestion! Will try that and update you. 

Cheers, 
B