Menu

Discussion Forum

New Discussion
Filter

Note: This discussion is about an older version of the COMSOL Multiphysics® software. The information provided may be out of date.

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.

Time Dependent Analysis

Posted Mar 12, 2012, 4:48 p.m. EDT Version 4.2a 8 Replies

Delwar
Send Private Message Flag post as spam

Please login with a confirmed email address before reporting spam

Dear Moderator and COMSOL Users,

I am doing a simulation using time dependent analysis of COMSOL Multiphysics. But as new user, I am confused about the physical settings of the model, is it correct or not? When I rerun(compute) the time dependent study, sometimes there are lot of error are generated and confused about the simulation result. Here In the model, I used five domain for the simulation purposes and I am looking to see the Pressures or displacement of piezoelectric with respect to time (range 0 to 10ms, step size 1us).
I would be thankful to you if you give me feedback or modifications regarding the topics. I have attached the model for your kind consideration.

Thanks
D Bhuyan
8 Replies Last Post Mar 19, 2012, 7:17 a.m. EDT
Ivar KJELBERG COMSOL Multiphysics(r) fan, retired, former "Senior Expert" at CSEM SA (CH)
Send Private Message Flag post as spam

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago Mar 13, 2012, 3:07 a.m. EDT
Hi

the "mess" you are seeing is a mixture of natural oscillations undersampled and aliasing with a not fine enough mesh compared to the time sequence and sample size.

First you can improve by turning OFF the inertial terms and stay in +Quasi static", at least unti lyou model is working and better tuned, its often difficult to get around all the physics in one go when starting, this is so for all of us ;).

I have a few other comments, you second PZT does not have any + electrode defined, you are hitting the device hard with a 10V step starting from initial conditions 0V this puts the device into oscillations when you use the inertial terms (rho*d^2_u_/dt^2 ones). Then you need to define some coordinate orientations as I do not expect the PZD are aligned with the main coordinate system, (add one per PZT as these are oreiented differently, as you need to add one PZT physics per device to use the different orientations

Then do not forget that if you want to simualte a rapid time hange, in full dynamic you must consider the speed of sound or waves in your material to get a mesh fine enough to sample that wave

And pls use the latest version forget old 4.0 (you will be less frustrated with the newer one)

--
Good luck
Ivar
Hi the "mess" you are seeing is a mixture of natural oscillations undersampled and aliasing with a not fine enough mesh compared to the time sequence and sample size. First you can improve by turning OFF the inertial terms and stay in +Quasi static", at least unti lyou model is working and better tuned, its often difficult to get around all the physics in one go when starting, this is so for all of us ;). I have a few other comments, you second PZT does not have any + electrode defined, you are hitting the device hard with a 10V step starting from initial conditions 0V this puts the device into oscillations when you use the inertial terms (rho*d^2_u_/dt^2 ones). Then you need to define some coordinate orientations as I do not expect the PZD are aligned with the main coordinate system, (add one per PZT as these are oreiented differently, as you need to add one PZT physics per device to use the different orientations Then do not forget that if you want to simualte a rapid time hange, in full dynamic you must consider the speed of sound or waves in your material to get a mesh fine enough to sample that wave And pls use the latest version forget old 4.0 (you will be less frustrated with the newer one) -- Good luck Ivar
Delwar
Send Private Message Flag post as spam

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago Mar 13, 2012, 11:05 a.m. EDT
Hello Mr. Ivar,
Thanks for your encouraging reply.
The previous model was in COMSOL 4.0a but now, I am using COMSOL version 4.1. I did not see the Quasi-Static for time dependent analysis. I think it behaves by default as a Quasi-static.
But how can I turn off the inertial terms? Please tell me how to define some coordinate orientations so that PZD are aligned with the main coordinate system?

I want to simulate the model for the time range(0 to 10ms) and step size 1 microsecond(1us). So what should be the proper mesh for this model? I rebuild the model in COMSOL 4.1and attached for your kind consideration

Regards
D Bhuyan
Hello Mr. Ivar, Thanks for your encouraging reply. The previous model was in COMSOL 4.0a but now, I am using COMSOL version 4.1. I did not see the Quasi-Static for time dependent analysis. I think it behaves by default as a Quasi-static. But how can I turn off the inertial terms? Please tell me how to define some coordinate orientations so that PZD are aligned with the main coordinate system? I want to simulate the model for the time range(0 to 10ms) and step size 1 microsecond(1us). So what should be the proper mesh for this model? I rebuild the model in COMSOL 4.1and attached for your kind consideration Regards D Bhuyan
Ivar KJELBERG COMSOL Multiphysics(r) fan, retired, former "Senior Expert" at CSEM SA (CH)
Send Private Message Flag post as spam

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago Mar 13, 2012, 4:28 p.m. EDT
Hi
the inertial terms/quasi static appears in the main physics node (if you do not see it you might need to turn on advanced features in the view section) at least I ave it in my 4.2a version. When you check the main physics node equation view you will see that the rho*d^2_u_/dt^2 is dropped in quasi static mode

--
Good luck
Ivar
Hi the inertial terms/quasi static appears in the main physics node (if you do not see it you might need to turn on advanced features in the view section) at least I ave it in my 4.2a version. When you check the main physics node equation view you will see that the rho*d^2_u_/dt^2 is dropped in quasi static mode -- Good luck Ivar
Delwar
Send Private Message Flag post as spam

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago Mar 14, 2012, 12:17 p.m. EDT
Hello Mr. Ivar,
Thanks for your reply. I did not find the Quasi-Static in Comsol 4.1 but I found it in COMSOL 4.2a and replaced the model in 4.2a. But when I compute the study, it shows error(Failed to find consistent initial values). I have attached the model for your kind consideration. If you have time, can you please check this one?


Thanks
D Bhuyan

Hello Mr. Ivar, Thanks for your reply. I did not find the Quasi-Static in Comsol 4.1 but I found it in COMSOL 4.2a and replaced the model in 4.2a. But when I compute the study, it shows error(Failed to find consistent initial values). I have attached the model for your kind consideration. If you have time, can you please check this one? Thanks D Bhuyan
Ivar KJELBERG COMSOL Multiphysics(r) fan, retired, former "Senior Expert" at CSEM SA (CH)
Send Private Message Flag post as spam

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago Mar 14, 2012, 2:33 p.m. EDT
Hi

have you cehced that you have enough BC's for all physics, I see no "fixed" part so your model is flaoting in free air, that makes the life too tough for the solver it integrates away, adda fixed constraint and it will work, I'm sure ;)

--
Good luck
Ivar
Hi have you cehced that you have enough BC's for all physics, I see no "fixed" part so your model is flaoting in free air, that makes the life too tough for the solver it integrates away, adda fixed constraint and it will work, I'm sure ;) -- Good luck Ivar
Delwar
Send Private Message Flag post as spam

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago Mar 15, 2012, 9:34 a.m. EDT
Dear Mr. Ivar,
Thanks a lot for your response. Now it is working. But the problem is, the simulationn results(1D plot for pressure) shows constant value, it is not changing with time. But, I want to see the pressure/displacement in the material with respect to time(for example 0 to 5ms, stepsize 1us). I have attached the model for your kind consideration.

Regards
D Bhuyan
Dear Mr. Ivar, Thanks a lot for your response. Now it is working. But the problem is, the simulationn results(1D plot for pressure) shows constant value, it is not changing with time. But, I want to see the pressure/displacement in the material with respect to time(for example 0 to 5ms, stepsize 1us). I have attached the model for your kind consideration. Regards D Bhuyan
Ivar KJELBERG COMSOL Multiphysics(r) fan, retired, former "Senior Expert" at CSEM SA (CH)
Send Private Message Flag post as spam

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago Mar 19, 2012, 3:52 a.m. EDT
Hi

strange, once I deleted your solver (for 2.model) and recreated them with default COMSOL settings, then changed the frequency domain sweep to range(9000,100,10000) to catch the first eigenfrequency, then changed the "Piezo material model Coordinate System" to Y-Z plane and looked at point "9" (because you have fixed point 3) I get pothe a pressure change (assymetric) and a displacement peak up in my scan

--
Good luck
Ivar
Hi strange, once I deleted your solver (for 2.model) and recreated them with default COMSOL settings, then changed the frequency domain sweep to range(9000,100,10000) to catch the first eigenfrequency, then changed the "Piezo material model Coordinate System" to Y-Z plane and looked at point "9" (because you have fixed point 3) I get pothe a pressure change (assymetric) and a displacement peak up in my scan -- Good luck Ivar
Delwar
Send Private Message Flag post as spam

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago Mar 19, 2012, 7:17 a.m. EDT
Dear Mr. Ivar,
Thanks a lot. Your previous feedback was frequency domain analysis. I did it according to your instructions and it's working fine for frequency domain anlysis. Now, I can easily observe the impedance analysis of the piezo.

For your kind consideration, I am more interested to see the time dependance behaviour of piezo. So, I am trying to see the simulation results for time dependent analysis (for example range 0 to 5ms, step size 1us). When I compute the time dependent study, it is working but the pressure or displacement in the Ti does not changing with respect time(for 1D plot). If you have time can you please look at my model.
Dear Mr. Ivar, Thanks a lot. Your previous feedback was frequency domain analysis. I did it according to your instructions and it's working fine for frequency domain anlysis. Now, I can easily observe the impedance analysis of the piezo. For your kind consideration, I am more interested to see the time dependance behaviour of piezo. So, I am trying to see the simulation results for time dependent analysis (for example range 0 to 5ms, step size 1us). When I compute the time dependent study, it is working but the pressure or displacement in the Ti does not changing with respect time(for 1D plot). If you have time can you please look at my model.

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.

Suggested Content