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.
How to implement 2 piezo parts with different polling directions in 4.0a?
Posted 25 août 2010, 16:16 UTC−4 Structural Mechanics Version 5.0 3 Replies
Please login with a confirmed email address before reporting spam
I am trying to convert to 4.0a and I'm having trouble building a model I made in 3.5. Two Piezo plates (two domains), one polled in the Z direction other in X direction. Currently, I'm adding a second Piezo Material Model in 4.0 and having it override the second domain. Then adding my two coordinate systems to each piezo material model. But this doesn't work when solving the eigenvalue I get garbage for displacement (almost looks like the solver is getting overlaying or two sets of coefficients), I also get warnings in the solver. Its not terribly descriptive so I'm not sure whats its referring to. In 3.5 I can do this quite simply, I typically solve for eigenvalue, then use that value to run frequency domain solution for impedance or such.
"The relative error is greater than the relative tolerance.
Relative residual 8.8e-011
Relative error 1e+002"
Thanks,
Mike
"The relative error is greater than the relative tolerance.
Relative residual 8.8e-011
Relative error 1e+002"
Thanks,
Mike
3 Replies Last Post 27 oct. 2015, 02:04 UTC−4
Please login with a confirmed email address before reporting spam
Posted:
2 decades ago
Hi
I can only confirm that I see too often "the relative error..." appearing, but at least in your case it is correctly stated (the relation is really bigger) I have often seen inconsitency in the values, still I do neither not know exactly what it means, nor how to avoid, but mostly my results are still coherent and as expected.
However, concerning the coupled analysis you have (if I understand you correctly) I wounder if it could be linked that you have (perhaps cannot really tell 100%) 2 physics with 2 solver states, or more, (i.e. eigenmode or harmonic for the driver and stationary for the receiver) but you have not managed (or COMSOL has not "understood" correctly how) to segregate the two. If my hypothessi is correct, then you should check the physics setting, and replace the "automatic solver selection" (close to the bottom of the window) by the user defined specific solver type for each(all) physics.
In V4, COMSOL does a certain number of hypothesis for multi physics - multi solver cases (in automatic mode), but does not always get it as we want it. Then we must decide which physics and which solver case shall be linked together.
I hit this issue some months ago I beleive with a pre-stressed eigenmode analysis in structural, and its not yet very documented, or just that I havent seen it
Hope this helps, if not "support" should be able to give you a better reply
--
Good luck
Ivar
I can only confirm that I see too often "the relative error..." appearing, but at least in your case it is correctly stated (the relation is really bigger) I have often seen inconsitency in the values, still I do neither not know exactly what it means, nor how to avoid, but mostly my results are still coherent and as expected.
However, concerning the coupled analysis you have (if I understand you correctly) I wounder if it could be linked that you have (perhaps cannot really tell 100%) 2 physics with 2 solver states, or more, (i.e. eigenmode or harmonic for the driver and stationary for the receiver) but you have not managed (or COMSOL has not "understood" correctly how) to segregate the two. If my hypothessi is correct, then you should check the physics setting, and replace the "automatic solver selection" (close to the bottom of the window) by the user defined specific solver type for each(all) physics.
In V4, COMSOL does a certain number of hypothesis for multi physics - multi solver cases (in automatic mode), but does not always get it as we want it. Then we must decide which physics and which solver case shall be linked together.
I hit this issue some months ago I beleive with a pre-stressed eigenmode analysis in structural, and its not yet very documented, or just that I havent seen it
Hope this helps, if not "support" should be able to give you a better reply
--
Good luck
Ivar
Please login with a confirmed email address before reporting spam
Posted:
2 decades ago
Thanks for the reply. I will look into this. I'll let you know how I make out.
Thanks-
Mike B.
Thanks-
Mike B.
Please login with a confirmed email address before reporting spam
Posted:
1 decade ago
Hi
i am also facing same problem can u give me some hint or solution to solve it??
i am also facing same problem can u give me some hint or solution to solve it??