Meshing strategy for merged instances

[ajroman]

I am having some trouble meshing a merged part and would like to hear of a few strategies on how to approach/solve my issue. I am trying to make a unit cell model of a 3D weave composite structure. I first created the unit cell’s fiber architecture by compressing a previous model and taking its deformed orphan mesh and converted it into a geometric part using Solidworks. I then created a solid geometric block with the same perimeter dimensions as the compressed fiber model and merged the two geometries while retaining their intersecting boundaries. But due to the complexity of the compressed fiber model, I am unable to mesh the entire merged unit cell. I have also included a few pictures. Any assistance on this matter would be much appreciated. Thank you.

 

[rstupplebeen]

I would try to clean up the geometry as best you can in SolidWorks. After that in Abaqus virtual topology is your best option. It might be worth trying to tackle a significantly smaller section first to develop methodologies. In SolidWorks for the fibers I suggest trying to sweep an elliptical profile for the fiber geometry. If you post the SolidWorks file I will try to give it a look. I hope this helps.

Rob Stupplebeen

 

[corus]

You might also try meshing the weaves and outer block separately so removing the restriction of connecting nodes between regions. You'd have to have two separate parts for that. Virtual topology might work but may be tedious to find every line that needs removing. If the outer block won't mesh then try playing about with the seeds to see if that helps. You'd then have to tie all the surfaces between the weave and outer block to ensure continuity of the structure. Wouldn't a simple model with fudged material properties from tests be a better option?

Tata