[chiwonkim]

hello,

Some problem drives me crazy.

i tried to solve oscillating plate of speaker.(transient)

Can i impose normal velocity(not tangential) on stationary wall boundary?

if not, please allow me to know which method is appropriate

i already used 'morper' but the magnitude of displacement is too high relative to mesh size, so after several iteration/time step my solver diverged, generating negative volume mesh.

my 2D, axisymmeric model is attached.

Any help can save my life.

thank you

[abdul099]

No, you can't set up a normal velocity on a wall without except by use of the morpher. All other models will not work since you need to change the topology of your mesh (position of the oscillating membrane relative to the housing of the speaker etc.)

Monitor the mesh quality, and when the mesh starts to get bad, remesh your geometry. Also consider to reduce the time step size. It's better (with respect to mesh quality) to morph several small steps instead of morphing one huge step.
When the oscillation is a constrained motion and therefore the position of the oscillating plate is defined for every time step, you might create several meshes and replace the existing mesh instead of remeshing every period.

[chiwonkim]

Thank you for your help.

So you meen that my computer has a lot of work to do....
i will try remeshing, but i wonder star ccm+ can regenerate 2d mesh because original mesh profile is generated from 'Ansys'.

anyway, thank you again

[Henry Arrigo]

I have been struggling with the same problem and I could not solve the problem yet. I checked almost everything but I could not find the cause. I thought maybe the Morpher solver does not function very well in star-ccm+. http://www.cfd-online.com/Forums/sta...e-morpher.html