h-Elements vs. p-Elements - Facts or Myths?

Solution methods involving two different type of Finite Elements are discussed here... the traditional h-Element and a more recently preferred (by some), p-Element. The p-Element has been available in ANSYS® since revision 5.1, and the h-Element has been the standard since day one. There are advantages and disadvantages to using each, and some of them follow.

Today’s trend by software developers is to implement integrated analysis tools that use either or both types of elements into CAD systems. This allows Design Engineers to do basic analysis on their parts and assemblies during the early stages of the design process, a distinct advantage that allows earlier product introduction to market and elimination of costly prototype iterations.

These same developers have done their best to convince users that their tools are a panacea to meeting all their analysis needs. In some cases this may be true, but many times capabilities are not fully explained or understood, leaving the user searching for answers. Let’s focus on the pros and cons.

The most basic issues relate to accuracy and speed. Unlike the h-Method that requires users to vary meshing parameters to obtain desired accuracy, the p-Method uses a constant mesh (usually coarser than an h-Element mesh). The p-Method then uses a variable polynomial level to obtain the results (such as displacements, stresses or strains) to achieve a user-specified accuracy. In each case there is at least one unknown variable, so there really is not one clear winner in general. A real winning combination is a program that allows the user to make a selection between the two based on the application... a clear advantage for ANSYS users who have that choice. Other advantages(+) and disadvantages (-) are:

+ Faster, more accurate, easier to use + Solution times known in advance
+ No remeshing for greater accuracy + Adaptive Meshing to automatically achieve desired accuracy
+ Good for fatigue and fracture where local accuracy is required mesh  + SmartMeshing in ANSYS minimizes poor mesh design
+ Local or global error estimates + Can be used for dynamics, CFD, Coupled Field, Magnetics
- Used only for linear structural and nonlinear solutions applications - Mesh refinement for precise accuracy control could be tedious
- Requires significantly greater computational   resources

And so, the battle rages on. Who will win? Look into ANSYS where you’re given the "Power of Choice" to solve basic or complex analysis and give our technical experts a call if you would like to explore the above comparisons in more detail.

[Support] [Home]