cranisch
Mechanical
- Jan 25, 2007
- 25
Hello everyone,
I got the following issue with Abaqus:
I have got a Static Linear Pertubation Step with a lot of loadcases and a quite big model. I need to determine the displacement of a lot of nodes for each loadcase. This should be as fast as possbile and as robust as possible, since I have to give away the tool to other colleagues.
1. My first approach was to use the python scripting interface to get all results out of the .odb-file. It works but takes very much time. And furthermore, I don't want much that my colleagues are able to manipulate the script. The python script is therefore not very useful.
2. The second approach was to use the C++ interface of abaqus to access the .odb-file and get the data. That works and is also a little bit faster. But it took me a hell of time to get the script compiled and it is not a kind of easy to handle, since we are not that experienced with C++.
3. My third thought was to write out the result to a .fil file and use Fortran to read it. We have a lot of experience with Fortran inhouse and we want also to proceed with the result data in Fortran. So, I could combine further process with obtaining the results. Unfortunately Abaqus tells me, that I can not use the *FILE NODE command with a Loadcase-Step. So, it does not create a .fil-file at all.
4. The fourth idea was to use the .dat-file and use *PRINT NODE during the solving process. Unfortunately the numerical results are not in a sufficient precision in the .dat-file so that I can not use it.
I heard of implementing Fortran scripts to the solver, so that writes result data to a user defined file and format during the solving process. This sounds interesting, but I firstly don't know how and if it could work and secondly if it is robust to version changes.
Does anyone have a clue how to solve this issue by a quickest and most robust process?
I got the following issue with Abaqus:
I have got a Static Linear Pertubation Step with a lot of loadcases and a quite big model. I need to determine the displacement of a lot of nodes for each loadcase. This should be as fast as possbile and as robust as possible, since I have to give away the tool to other colleagues.
1. My first approach was to use the python scripting interface to get all results out of the .odb-file. It works but takes very much time. And furthermore, I don't want much that my colleagues are able to manipulate the script. The python script is therefore not very useful.
2. The second approach was to use the C++ interface of abaqus to access the .odb-file and get the data. That works and is also a little bit faster. But it took me a hell of time to get the script compiled and it is not a kind of easy to handle, since we are not that experienced with C++.
3. My third thought was to write out the result to a .fil file and use Fortran to read it. We have a lot of experience with Fortran inhouse and we want also to proceed with the result data in Fortran. So, I could combine further process with obtaining the results. Unfortunately Abaqus tells me, that I can not use the *FILE NODE command with a Loadcase-Step. So, it does not create a .fil-file at all.
4. The fourth idea was to use the .dat-file and use *PRINT NODE during the solving process. Unfortunately the numerical results are not in a sufficient precision in the .dat-file so that I can not use it.
I heard of implementing Fortran scripts to the solver, so that writes result data to a user defined file and format during the solving process. This sounds interesting, but I firstly don't know how and if it could work and secondly if it is robust to version changes.
Does anyone have a clue how to solve this issue by a quickest and most robust process?