Hello everybody,

I did the tutorial of a piezoelectric micropump described in "Multi-field Analysis Using Code Coupling" and I have two questions:

1) When I change the fluid to water, the simulation returns an error of negative volume element. Could anybody explain it?

2) I tried to excitate the piezoelectric ceramic in its frequency resonance applying a sinusoidal voltage input, but the model doesn't change its behavior, no matter what frequency. Is it possible to do it?

Thanks!

1) The negative-volume indicates a folded-mesh (collapsed cell faces). This may happen if your mesh suddenly moves a particular location with unphysical values. There can be several reasons for this, e.g. unconverged solution-fields at the time of mesh motion, messed-up boundary conditions or physical properties, very large time-steps, bad-value for mesh-diffusion coeffcient (or mesh smoothing coeffcient). Check all of these one by one for your case.

Are you coupling CFX with ANSYS? If yes please let me know where one can find this tutorial. I want to see how one can apply votlage input in a structual analysis. Do you have some special elements for electrodynamics equations? I guess your second question is not for CFX users rather for ANSYS FEA users.

There is a tutorial in ANSYS' help, under the topic 'Coupled-Field Analysis Guide' (here: "Coupled-Field Guide -> Chapter 4. Multi-field Analysis Using Code Coupling -> 4.4. Example Simulation of a Piezoelectric Actuated Micro-Pump"), that explain how to model a piezoelectric pump using CFX and ANSYS.

About the first question, I agree with you about the reasons to cause negative volume elements, but I still don't understand what is wrong in this particular example. All I did was to change the fluid to water. All the rest (mesh, boundary conditions, time-steps, ...) is the same as the example.

What you're looking for is element SOLID5 (structural and voltage dofs).