Inlet boundary condition for acoustic wave

[lzhaok6]

Dear all,

I want to simulate the propagation of an acoustic wave with a spherical wave source outside my computational domain using Finite Element Method. The governing equation is a linearized acoustic equation with velocity potential as the primary variable ( the secondary variable is thus velocity). In order to let the wave propagate into the domain, I might need to add an impedance condition on the boundary by setting the predicted fluid particle velocity on the boundary nodes (natural boundary condition). However, the discretization of FEM only allows the normal component of the velocity be prescribed on the boundary (divergence theorem).

My question is:
Is it physical to enforce only normal component of fluid particle velocity on the boundary?
Would it decrease the wave magnitude or distort the wave?
If so, is there any way to circumvent that?

Thank you all very much for your attention.

 

[hacksaw]

Both Longitudinal and Transverse waves exists, but the for liquids and gases, the transverse wave impart little force to the solid boundary, and resulting in reflection of the wave. The longitudinal waves exert pressure and momentum forces on the surface as your FEA package descibes.

 

[lzhaok6]

Thanks for your reply, hacksaw. I might not make my question clear. The boundary mentioned in the question is actually the outer boundary of the fluid. By adding an impedance boundary condition on that boundary, the wave could pass from the infinite acoustic domain into the computational domain. I already assume that it is a longitudinal wave in that the fluid particle displacement is in the direction of wave propagation.

 

[hacksaw]

That's how I see it, transverse "acoustic", longitudinal "pressure" waves. The transverse is only significant if the material is characterized by an elastic behavior in the transverse.

 

[scheah]

In ANSYS, you could define a monopole outside of the computation domain. Here's an example. The help document says it applies incident pressure to the exterior boundary (link).


Best regards,
Sze Kwan (Jason) Cheah