Discussion Closed This discussion was created more than 6 months ago and has been closed. To start a new discussion with a link back to this one, click here.
Acoustic-structure boundary among sold and fluid in pressure pulse generation
Posted Feb 15, 2018, 4:20 a.m. EST Structural & Acoustics, Acoustics & Vibrations Version 5.3 1 Reply
Please login with a confirmed email address before reporting spam
Dear all, I'm trying to simulate the pressure generated by a thin film that is expanded by temperature variation. In order to take into account the physics at the interface between the film and fluid domain where the pressue propagates (water in this case), I have used the acoustic-structure boundary multyphysics option. The model seems to work but I can't understand the numerical version of boundary condition imposed.
In particular, analyzing the condition expressed in the "equation" of
multyphsics -> acoustic-structure boundary
it seems to me that boundary condition studied in the book(i.e. continuity of the displacement and equality among stress tensor contracted with the normal to the surface and pressure) are applied. But if I plot, as a function of time, at the interface, the acceleration in solid (solid.u_ttZ) and the acceleration of pressure domain (actd.az) they don't match. Since the displacement is the same, I expect the acceleration too is the same. Analogously, the stress tensor in the direction of the normal to the surface (solid.sl33 or solid.sz) doesn't match with the pressure p at the interface.
Is there a particular reason why? I made some mistake (maybe mesh not enough refined)? Thanks
Please login with a confirmed email address before reporting spam
Hi Diego,
The stress on the solid side and the velocity on the Acoustics side are not the actual degrees of freedom (the displacements and the pressure, respectively, are).
So what you see is the same as if you add a pure boundary load in Solid Mechanics: The stress does only match the applied load approximately, since it is computed from the derivatives of the displacements in the element. Similarly, there is only continuity between two elements in the mesh in terms of displacements and not in terms of stresses. The same goes for pressure/velocity in the case of acoustics.
The differences decrease with mesh refinement. For second order shape functions (which is the default in many physics interfaces), the error in 'fluxes' which are gradients of the degree of freedom, is roughly proportional to the element size.
Regards,
Henrik
-------------------Henrik Sönnerlind
COMSOL