Mid Surface extracting in Hypermesh(Tricks & Tips)

5 reasons why you should use a mid-surface shell mesh for thin-walled parts

It happens all the time. Too often we still see companies that trust on simulations being performed with a solid-mesh, while the geometry is thin walled, like with sheet-metal, extruded parts or plastic parts. If you ask the people why they didn’t use a shell mesh, most of the time they will tell you it will take too much time to generate a mid-surface shell model, or they think it is not feasible capturing the necessary detail in the analysis, so they decided to just ignore the rules.

Below we have listed why creating a mid-surface shell model is still favourable.

1. Accuracy of the end result can be way off with solid mesh elements 
   Whether you have followed a basic course from a CAD-integrated FEA software, or from a stand-alone High-End FEA software, they will all teach you to create shell meshes for thin-walled parts. Even if you setup a simple thin-walled beam and let it suffer under bending, you already see differences in both stress and displacement for a solid mesh versus a mid-surface shell mesh. Particularly when the solid mesh approach is a course mesh, the effects are even bigger. When you refine the solid mesh, the results will become more accurate, but you will need quite some iterations to have mesh convergence.
   
2. The solid mesh approach, which needs 3 to 4 elements across the thickness, results in huge amount of elements and long running times 
   If you do decide to take the solid mesh approach, please be aware that you need a minimum of 3 to 4 elements across the thickness to capture all bending and stiffness effects to generate an accurate solution. For products that don’t look like a simple beam, it can potentially create huge number of elements, resulting in large running times. If you think of injection moulded parts with ribs, draft angles and fillets and other types of typical geometry, the necessary amount of elements will be enormously and therefor the running times will be huge. We have seen differences for injection moulded parts with a shell approach that take 1 hour to run, while with a proper solid mesh including 3 to 4 elements across the thickness the running time could potentially be somewhere between 5 to 10 days per simulation run.
   
3. Meshing nasty geometry with solid mesh elements needs preparation too 
   For a lot of the standard meshing techniques, the solid mesh needs to be as small as the smallest detailed geometry in the model. We receive a lot of files from customers that contain sliver faces, short edges, surfaces that do not align, etc. Simply pushing the mesh button won’t do the trick here since the system will most likely tell you it is unable to mesh. So you need to clean-up the geometry anyway, which also for a solid part can be quite a hassle. Sometimes switching to a curvature based mesher can do a part of the trick, but you can potentially lose the necessary detail. Since you need to clean up this geometry anyway, why not define a mid-surface model from the original model?
   
4. Post processing of huge amount of solid element simulations takes ages
   It seems to be trivial, but often overlooked. The time that you will need for running these huge amount of element simulations, also contributes to the potentially frustrating post processing of the end results. Ever needed to wait for more than 30 minute to generate one simple stress plot, or even a few hours to generate a single animation? Or what about non-linear simulations with more than 100 time steps and the results of those? Once you have been in that situation of generating huge amount of data, you will hopefully try to prevent that in other projects.
   
5. The shell mesh approach creates a fast running accurate simulations
   Since the amount of elements is less for the shell approach, because the wall thickness is captured as a mathematical value instead of actually modelling the thickness, there will be less equations to solve. This results in simulations that run in minutes or an hour instead of days to weeks. That also means that a shell mesh approach can be easily used to perform iterative simulations. Another benefit is that changing the wall thickness in your simulations with the shell approach are easy, and you will see the results of those changes in a very short time.