Hello,

I am modelling a double-lap bolted joint in a composite material and have encountered two major issues. First, the model exhibits significantly higher stiffness than the experimental results. Second, it suffers from severe element distortion during the analysis.

Initially, I used the built-in Hashin damage model in Abaqus. However, excessive element distortion prevented the analysis from completing, and element deletion was not activating as expected. To address this, I switched to a VUMAT implementation. Unfortunately, the distortion became even more severe, although I was able to delete elements using a specified strain criterion.

Element deletion allowed the simulation to run, but the resulting load drops were excessive. While mesh refinement improved this behaviour, the response remained highly unstable and strongly mesh-dependent, even with energy-based damage evolution..

I have read that using a coarser mesh can sometimes reduce distortion problems. Although this seems to slightly improve the behaviour when using the built-in Hashin model, it does not provide any benefit in the VUMAT implementation.

My questions are:

1. Are there alternative methods for controlling element distortion without relying on element deletion?

2. What factors could be responsible for the model's excessive stiffness? I have already reduced the material properties considerably, but this has had little effect on the predicted stiffness.

   Thank you