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.
The difference between "low reflecting boundary" and "spring foundation"
Posted Jul 6, 2017, 11:13 p.m. EDT 2 Replies
Please login with a confirmed email address before reporting spam
I encountered this problem when I was using the structural mechanics module, frequency domain. The "low-reflecting boundary" was used to absorb outgoing elastic waves. Then I used the "spring foundation" to substitute it. The spring stiffness kA was set to be zero and the damping constant dA was set to be rho*(cs+cp)/2, which is identical to the definition in "low reflection boundary". The results obtained by both methods are similar, but not the same.
So I was wandering, what caused this difference.
So I was wandering, what caused this difference.
2 Replies Last Post Jul 7, 2017, 11:29 a.m. EDT
Please login with a confirmed email address before reporting spam
Posted:
7 years ago
Hi,
This should, as you assume, give the same result, assuming that all other settings are the same. Can you upload your model?
Regards,
Henrik
This should, as you assume, give the same result, assuming that all other settings are the same. Can you upload your model?
Regards,
Henrik
Please login with a confirmed email address before reporting spam
Posted:
7 years ago
The low reflecting boundary condition imposes a force proportional to velocity on the surface. The proportionality constant is the acoustic impedance, the product of mass density and velocity. There is actually no damping; the boundary condition basically terminates the surface in the acoustic impedance.
The complication is that in a bulk solid there are two velocities, the transverse and longitudinal velocities. So the Comsol default is to use an average of the transverse and longitudinal velocities. This means that the termination is not ideal for EITHER of the two types of waves.
Sometimes you know the direction of arrival of the waves. In such a case it can be better to customize the absorbing boundary condition to apply different forces in transverse and normal directions. (I have done this in the past because earlier versions of Comsol did not have an absorbing boundary condition. All you need to do is specify a boundary load on the surface proportional to the velocity).
I learned that there are articles in the literature about more sophisticated and better absorbing boundary conditions. What I have described is the simplest approach.
But the takeaway is (1) absorbing boundary conditions are imperfect and (2) results will depend on the type of wave and the arrival direction.
D.W. Greve
DWGreve Consulting
The complication is that in a bulk solid there are two velocities, the transverse and longitudinal velocities. So the Comsol default is to use an average of the transverse and longitudinal velocities. This means that the termination is not ideal for EITHER of the two types of waves.
Sometimes you know the direction of arrival of the waves. In such a case it can be better to customize the absorbing boundary condition to apply different forces in transverse and normal directions. (I have done this in the past because earlier versions of Comsol did not have an absorbing boundary condition. All you need to do is specify a boundary load on the surface proportional to the velocity).
I learned that there are articles in the literature about more sophisticated and better absorbing boundary conditions. What I have described is the simplest approach.
But the takeaway is (1) absorbing boundary conditions are imperfect and (2) results will depend on the type of wave and the arrival direction.
D.W. Greve
DWGreve Consulting