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.

Propagation of impulse in 3D block

Please login with a confirmed email address before reporting spam

Hi everyone, I am trying to simulate the measurement of an acceleration given by an impulsive force on a 3D block. I set up the time domain analysis looking at user manual and similar simulations in the forum. The impulse force is simulated using the gaussian function and I would like to measure the acceleration at a given point on the opposite face of the block. Looking at the 1D graphs of displacement, velocity or acceleration at the measurement point I cannot find the delay in the signals given by the propagation. Given the material parameters I would expect a delay between the signal at the application and measurement points of 0.01[s]. By the way, the setup in terms of geometry and physics looks correct to me, should I change something in the time domain solver, like relative tollerance? I attach the comsol project. Thank you.



3 Replies Last Post 07.10.2019, 16:50 GMT-4
Edgar J. Kaiser Certified Consultant

Please login with a confirmed email address before reporting spam

Posted: 5 years ago 07.10.2019, 13:45 GMT-4

With the given material data and the dimensions of the model I would expect a delay of 0.0001 s. You will need to set the solver to manual time stepping otherwise it may miss the impulse.

Cheers Edgar

-------------------
Edgar J. Kaiser
emPhys Physical Technology
www.emphys.com
With the given material data and the dimensions of the model I would expect a delay of 0.0001 s. You will need to set the solver to manual time stepping otherwise it may miss the impulse. Cheers Edgar

Henrik Sönnerlind COMSOL Employee

Please login with a confirmed email address before reporting spam

Posted: 5 years ago 07.10.2019, 14:47 GMT-4

Note also that the time step should not be significantly larger than the time it takes for a wave to travel across a typical element. Many solid materials have a wave speed of the order of 5000 m/s. So if a typical element size is 2 cm, then the time step should limited to 0.02/5000 = 4 ms.

Since the time marching method is implicit, the time step limit is not strict as it would be for an explicit method (which is unstable at too large time steps). Significantly larger larger time steps than the recommended may however distort the wave and change the arrival time.

Regards,
Henrik

-------------------
Henrik Sönnerlind
COMSOL
Note also that the time step should not be significantly larger than the time it takes for a wave to travel across a typical element. Many solid materials have a wave speed of the order of 5000 m/s. So if a typical element size is 2 cm, then the time step should limited to 0.02/5000 = 4 ms. Since the time marching method is implicit, the time step limit is not strict as it would be for an explicit method (which is unstable at too large time steps). Significantly larger larger time steps than the recommended may however distort the wave and change the arrival time. Regards, Henrik

Please login with a confirmed email address before reporting spam

Posted: 5 years ago 07.10.2019, 16:50 GMT-4
Updated: 5 years ago 07.10.2019, 12:50 GMT-4

Note also that the time step should not be significantly larger than the time it takes for a wave to travel across a typical element. Many solid materials have a wave speed of the order of 5000 m/s. So if a typical element size is 2 cm, then the time step should limited to 0.02/5000 = 4 ms.

Since the time marching method is implicit, the time step limit is not strict as it would be for an explicit method (which is unstable at too large time steps). Significantly larger larger time steps than the recommended may however distort the wave and change the arrival time.

Regards,
Henrik

I am doing a very analogous problem -- where a short cylinder of linearly viscoelastic material is supported by an identically-sized cylinder below it, with the top cylinder subjected to an impulse force. I have been having problems in getting Comsol 5.4 to accept the impulse force expression in the boundary condition that I'm trying to set up, but now it works fine when I use the setup as in M-Pexx, i.e., I input an applied force as F*an1(t[1/s]).

But here is my question (which must have a very simple answer): Why is the force expression in M-Pexx's BC is written as F*an1(t[1/s]) with the [1/s]? I know the BC expects force (Newtons), but isn't the function an1 already a normalized function in this problem?

>Note also that the time step should not be significantly larger than the time it takes for a wave to travel across a typical element. Many solid materials have a wave speed of the order of 5000 m/s. So if a typical element size is 2 cm, then the time step should limited to 0.02/5000 = 4 ms. > >Since the time marching method is implicit, the time step limit is not strict as it would be for an explicit method (which is unstable at too large time steps). Significantly larger larger time steps than the recommended may however distort the wave and change the arrival time. > >Regards, >Henrik I am doing a very analogous problem -- where a short cylinder of linearly viscoelastic material is supported by an identically-sized cylinder below it, with the top cylinder subjected to an impulse force. I have been having problems in getting Comsol 5.4 to accept the impulse force expression in the boundary condition that I'm trying to set up, but now it works fine when I use the setup as in M-Pexx, i.e., I input an applied force as F*an1(t[1/s]). But here is my question (which must have a very simple answer): Why is the force expression in M-Pexx's BC is written as F*an1(t[1/s]) with the [1/s]? I know the BC expects force (Newtons), but isn't the function an1 already a normalized function in this problem?

Note that while COMSOL employees may participate in the discussion forum, COMSOL® software users who are on-subscription should submit their questions via the Support Center for a more comprehensive response from the Technical Support team.