[navarag]

Hello everyone,

I'm thrilled to be part of this forum and excited to discuss my current project with you. I'm currently working on a fluent project that involves using a User-Defined Function (UDF). Initially, I ran the simulation successfully on my local system for 10,000 iterations. However, in order to simulate more timesteps, I need to use a High-Performance Computing (HPC) system.

Unfortunately, when I tried running the simulation on the HPC system, I encountered the following error:

```
Updating mesh at time level N...
================================================== ============================

Node 4: Process 3316853: Received signal SIGSEGV.

================================================== ============================

================================================== ============================

Node 17: Process 3280829: Received signal SIGSEGV.

================================================== ============================

================================================== ============================

Node 31: Process 3280846: Received signal SIGSEGV.

================================================== ============================

===============Message from the Cortex Process================================

Fatal error in one of the compute processes.
free(): corrupted unsorted chunks

================================================== ============================
free(): corrupted unsorted chunks
free(): corrupted unsorted chunks
free(): corrupted unsorted chunks
free(): corrupted unsorted chunks
free(): corrupted unsorted chunks
free(): corrupted unsorted chunks
```

I am seeking assistance in resolving this issue. Any help or guidance would be greatly appreciated.
P.S- Its a 2d analysis of the door opening due to suction pressure is given below. Note that the X axis is towards right, y axis toward top and z axis towards front(if it counts), while door should open in clockwise direction towards left


Thank you in advance.

[AlexanderZ]

ansys fluent customization manual,
look for parallelization chapter:
global summation
and
node_to_host command

[navarag]

Originally Posted by AlexanderZ

ansys fluent customization manual,
look for parallelization chapter:
global summation
and
node_to_host command

Hey Alexander,

Thanks for the feedback.
The problem was with cas file and redoing it eliminated the problem.
Now,
I have doubt in regards with running udf in HPC.
Currently I copy the CAS, libdif, udf code ,default journal file and sub file in HPC server and run, is it the correct method or is there something else to do with journal file to get udf.


I have attached the journal file along with this reply.


Thanks

[navarag]

Got the answer. Just need to copy libudf and .c file alomg with other sub and journal file with cas.