Thanks Tsambani Madam. The link provided by you is regarding python scripting. But, I am talking about fortran user subroutines.

Hi

If i understood correctly your question, then this might help: https://groups.yahoo.com/neo/groups/Abaqus/conversations/messages/32114

I had similar problem just few months ago.

Best,

Mihkel

Thanks Mihkel Kõrgesaar. I was not a member in the above mentioned Yahoo Group, so it took some time to gain access and view it. The problem is similar and it seems that it will work for me. Thanks once again.

Did you guys ever manage to solve this? I am trying to write out to a file but with multiple CPUs get a read-access error...

Dear Harnoor,

Most recently I used subroutine vexternaldb which abaqus recommends for data sharing. You open the file in there, then you dont have any issues. Tested it out myself. Here is the description.

http://abaqus-users.1086179.n5.nabble.com/Question-about-UMAT-td4662.html

Hi Mihkel, thanks for the quick reply. That indeed does look like a robust solution. In the meantime I implemented something simpler; I basically wanted to extract the internal element and material point numbers per element - these change depending on the number of cpus used; so I basically write a unique file per cpu. Previously the error arose since more than 1 cpu was attempting to write to a single file. The key was to use VGETRANK (thread_ID) to dynamically generate a filename. Cheers.