Hi all,
I have generated a model and tried to export it as a volume mesh. I believe most of the users who tried the volume mesh option had to deal with element distortion or analysis warnings in ABAQUS due to bad quality of the elements around the tows. So I tried changing the merge tolerance based on previous recommendation and it did improve the current issue but this option it changes the cross section of the tows which I think it does effect the analysis later. I might be wrong about this, So if there is anyone who has to deal with the volume mesh issue, please share you input on it here.
Also, I would like to know what does merge tolerance do exactly?
Thanks
Matt
Volume Mesh: Element distortion
Moderators: Martin, Developers
Re: Volume Mesh: Element distortion
Hi Matt,
The merge tolerance specifies the size of gap between yarns which will be ignored when the mesh is generated. If, for example, a merge tolerance of 0.001 is specified then any area of matrix less than that size will be ignored and the volume mesh will be generated with the elements for the two yarns next to each other (with no matrix inbetween). This will therefore have the effect of slightly increasing the yarn cross-sections. You could try increasing the seed size which would have the effect of increasing the mesh density which might improve the meshing in the small gaps.
It's a tricky problem I'm afraid which, given the nature of textile geometries, isn't easy to solve (as evidenced by the number of publications trying to come up with methods to solve the problem!).
Best wishes,
Louise
The merge tolerance specifies the size of gap between yarns which will be ignored when the mesh is generated. If, for example, a merge tolerance of 0.001 is specified then any area of matrix less than that size will be ignored and the volume mesh will be generated with the elements for the two yarns next to each other (with no matrix inbetween). This will therefore have the effect of slightly increasing the yarn cross-sections. You could try increasing the seed size which would have the effect of increasing the mesh density which might improve the meshing in the small gaps.
It's a tricky problem I'm afraid which, given the nature of textile geometries, isn't easy to solve (as evidenced by the number of publications trying to come up with methods to solve the problem!).
Best wishes,
Louise