I have been under the impression that the p-elements used in ProMECHANICA means that you can use fewer elements than with a h-element code to get similar convergence, as the processor will increase the polynomial function defining the element shape in order to achieve convergence. Consequently, for plates where solid elements are used (rather than shells), I would expect that one p-element through the plate thickness should be adequate in most cases (I expect that there may be occassions where more elements could be required to capture certain details, or high strain gradients, etc ?), as the convergence would be equivalent to a model where more than one h-element was used through the thickness.
Is this correct? Please could the group confirm or correct these beliefs?