I don't understand the question, did you take this model as base but draw one in 3D instead?

Hi, differences are substantial.

2D simulations allow for the imposition of simmetries that arise from the system under investigation. Simmetry should be intended not geometrically but from a fluidynamic point of view. Variables should expect to spread simmetrically under the imposition of symmetry. Of course, 2D simulations require less computational burden and also less BCs but these should be carefully implemented in order to give a realistic description of your system.

3D simulations are employed if a completely detailed modeling is required since no simmetry hypothesis are contemplated. The system is modeled as a whole and this will require the imposition of more BCs. The correspondence with the real geometry may help to impose the right BCs that is usually not easy in 2D "simplified" approach.

I hope I clarified some aspects.

If you need help just contact me.

Paolo

Thanks for your answer @Paolo Mocellin. I really appreciate  your help.  Actually I tried to draw 3D modeling of cantilever as a energy harvester but there is result as i expected. I 'm beginner in comsol.