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.
Simulating a solid cylinder with a transient displacement point source to get phase velocity from the dispersion curve
Posted Apr 13, 2015, 5:44 a.m. EDT Acoustics & Vibrations, Geometry, Mesh, Modeling Tools & Definitions, Parameters, Variables, & Functions, Results & Visualization, Studies & Solvers, Structural Mechanics 4 Replies
Please login with a confirmed email address before reporting spam
I have been trying to simulate a cylindrical model using structural mechanics(time dependent) with a transient displacement point source (0.00005*sin(2*3.14*3000*t)*e^(-((t-t0)^2)/(2*(sigma^2))) [gaussian pulse] in y direction. I have applied a low reflecting boundary condition at the ends of the cylinder. The time step for the simulation is 1e-6 and I have been using a 'finer' physics defined mesh.
radius=10mm
length=1500mm
I have multiple points(and not directly probes) added to the model which would act as sensors to measure the displacement in y direction.
With the data generated from the simulation when I plot the dispersion curve for the phase velocity vs the frequency, I get varying plots with the change in the location of the sensors.
Please let me know what am I missing while modelling, which is leading to such an erroneous results.
Please login with a confirmed email address before reporting spam
Please login with a confirmed email address before reporting spam
Attachments:
Please login with a confirmed email address before reporting spam
first of all your model seems to lack a material, which would actually be essential to check if your settings for mesh and and time steps are reasonable. I have taken the freedom to set it to steel as a standard.
The main issue however is probably that you have set (or accepted the default) in the time-dependent solver: time-stepping = free. This would allow Comsol to set nuerical time steps as it wants (thereby oversampling your vibration) and returning only interpolated values. I recommend here "intermediate" (see the manual for the meaning).
An even better way would be to constrain the step size by tolerance criteria, depending on your expected results. This approach is illustrated in ch.comsol.com/model/bracket-transient-analysis-10314
specifically ch.comsol.com/model/download/198659/models.sme.bracket_transient.pdf
Note also that you may want to excite your part by a force rather than a prescribed displacement, because the latter actually puts constarints on the deformation even after the pulse hase decayed.
Please find attached a running model - if further discussion on the model is needed please create a regular support case here: ch.comsol.com/support
Best regards,
Sven
Attachments:
Please login with a confirmed email address before reporting spam
I have my time steps as 'strict' in my simulations, with Low reflecting boundary condition. Is it possible to get no reflections from the ends, since even with such a BC, the reflections that are there significantly changes the results.
I have attached my model for you to get a better understanding of what I am actually talking about. I have tried to simulate the model with a point load at the source.
Attachments: