Fellows,

My 3D numerical wave tank (built in COMSOL) doesn't work right, hopefully someone here can shed some light.

first question, Q: can anyone point me to a good source for the proper moving mesh displacement/velocity components in terms of fluid velocities?

The imposed moving mesh (on the 'free-surface') does not show waves propagating down the tank but they remain trapped (in time) within the first row of mesh elements.

/\_________ --> wave direction | | | | ------------

I have a gravity force on the domain, non-slip conditions on the end, bottom and side wall and a velocity profile on the left hand side straight from linear wave theory.

2^nd Q: Has anyone observed this? 3^rd Q: Why would a moving mesh be so stiff? 4^th Q: Does anyone have a better way?

Alot of questions, yes. Thanks for any help you can offer!

Piotr

Have you perhaps created a 'numeric' shock in the first cells? Your boundary/conditions may/or may not, actually be predicting a shock & locking your wave in that position.

Some solvers can cause this problem with wave-based simulations. Try changing your solver type, & reduce element dimension. Also, make sure you have the correct time-step to capture the wave flow.

mw...

<www.adthermtech.com/wordpress3>