I think it will be good to show the nature of your 3D model. This way, those that can help will be able to tailor their suggestions for your problem.

It depends generally on the 3D model. Does the model have really same features and sections? Did surface meshing work and tetrahedral meshing didn't work? What are the portions where mesh failed? Could you try to reduce the element size of the mesh or perform refinement across any faces? Could you share your 3D model alongside the mesh figure showing portions where meshing failed? I could give further solutions if you could share the figures.

Typically this means that your geometry is to complex for meshing. You can avoid this, by partitioning your model. In your case you should partitioning at the connection of the trusses.

The question for me is, why do you use tetrahedrons and not beams?

Hi @Aiman. Are you using workbench or APDL?

There are various reasons for a tetrahedron mesh to fail which I mention some of them in the following:

1. Not appropriate mesh size

2. Having high aspect ratio which affect the complexity of the meh

3. Complex geometry which requires considering different zoning for mesh

4. Having limit on number of nods and elements (which is kind of obvious!)

Also, I agree with Christian Willberg , why you are not simulating the jacket structure using appropriate beam element. Using TET mesh can impose a lot of issues such as the one you are facing right now and even if you manage to get the mesh done, I'm not sure about the accuracy of your simulation results.

If I want to give you an example, this is like you choose to use a TET mesh for meshing a thin plate instead of using a shell element which can impose a significant error in you analysis and affect final results.

If you don't see the shape of your beam in ANSYS predefined shapes, there is an option for creating your own beam section.

Hope this answer helps!!