Thanks Matej !
I’ll have a look into this.
Thanks a lot.
Dear all,
Just a quick update for the sake of completeness on this forum.
At first I was really focused on understanding why my mesh structure was done in a specific way by my custom meshing and why it wouldn’t work exactly the same way as in PrePoMax.
I created two meshes : one by myself with my algorithm and one with PrePoMax’s built-in mesher. I compared the area that had the weird behavior (c.f deformed final mesh in previous posts) in the initial position of the simulation (i.e initial mesh).
I noticed that such spikes formed in my mesh during my simulation mainly because of its orientation and because my mesh made some shapes that wouldn’t react well when deformed.
Initial situation: when looking at this specific node, we can guess that with the compression of the mandrel it’ll be pushed towards the outside, reacting in an exagerated, unrealistic way.
Mid situation: node has been pushed greatly and starts to deform. Nothing too alarming for now.
Final situation: Because node sticks out so much, it is being dragged greatly by the mandrel pull, deforms in an unrealistic manner and ends up crashing the simulation.
As, apart from changing my meshing parameters, I couldn’t do much in how the mesh was generated. I followed Matej’s advice and aimed to make the simulation work rather than reproducing PrePoMax’s exact mesh.
And it worked! After loooots of parameters tweaking I found out that reducing curvature safety and segments per edge was enough to make a mesh that’d work with this simulation !
Now it works perfectly with parametric optimization 
Thanks again for your help guys !
All the best.
1 Like