7372
Mechanical
- May 11, 2004
- 5
Hi,
I try to compare results given by beams theory and CATIA GPS ones. The goal for me was to look for a real convergence criterion to use it on biggest problems with GPS and GAS. Normally with adaptativity box we can locally refine meshing to "improve" results but we had to give a goal of local error (based on the stress gradient into each element). However in practice the refinement of meshing make the maximum stress exploded with a 10 or more important factor... Then, I try to look for a method to stop meshing rafinement when CATIA results was conform to theory: Convergence of energie, Convergence of Error criterion, Convergence of Maximum Von Mises Stress, Influence of initial elements sized (compare with beam dimension) on convergence criterion... Unfortunately I don't fine any real method to say that the result was good.
Then, my questions are:
1)
-Is any body know how to use correctely GPS and GAS?
-Does a method exist? Or GAS and GPS has just been created to do nice images with requirement need of form or experience results. Is every body use it like that:
-Make a theoretical calculation with forms to fine maximum stress.
-Fine element size who give the same maximum stress than theoretical one.
-Admire the stress repartition...lol!
2)
Another solution consists to give us rules of meshing. For instance for a beam with rectangular section put 3 parabolic elements on hight and 2 on thickness. I try that but it doesn't give me real good results...
-Is anybody has ever do a form on that topic?
3)
I try deferent model: 3D, 2D and 1D elements. There isn't any problem with displacements and modal analysis but the problem stay the stress analysis...
-Could anybody say me why?
To conclude, I am disappointed by the tool: Adaptative Box... I don't find any information about its use, and in particulary from Dassault System. If you can help me...
Thank you.
Vince from Paris
I try to compare results given by beams theory and CATIA GPS ones. The goal for me was to look for a real convergence criterion to use it on biggest problems with GPS and GAS. Normally with adaptativity box we can locally refine meshing to "improve" results but we had to give a goal of local error (based on the stress gradient into each element). However in practice the refinement of meshing make the maximum stress exploded with a 10 or more important factor... Then, I try to look for a method to stop meshing rafinement when CATIA results was conform to theory: Convergence of energie, Convergence of Error criterion, Convergence of Maximum Von Mises Stress, Influence of initial elements sized (compare with beam dimension) on convergence criterion... Unfortunately I don't fine any real method to say that the result was good.
Then, my questions are:
1)
-Is any body know how to use correctely GPS and GAS?
-Does a method exist? Or GAS and GPS has just been created to do nice images with requirement need of form or experience results. Is every body use it like that:
-Make a theoretical calculation with forms to fine maximum stress.
-Fine element size who give the same maximum stress than theoretical one.
-Admire the stress repartition...lol!
2)
Another solution consists to give us rules of meshing. For instance for a beam with rectangular section put 3 parabolic elements on hight and 2 on thickness. I try that but it doesn't give me real good results...
-Is anybody has ever do a form on that topic?
3)
I try deferent model: 3D, 2D and 1D elements. There isn't any problem with displacements and modal analysis but the problem stay the stress analysis...
-Could anybody say me why?
To conclude, I am disappointed by the tool: Adaptative Box... I don't find any information about its use, and in particulary from Dassault System. If you can help me...
Thank you.
Vince from Paris