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.
microcantilever analysis with sinusoidal
Posted 14.11.2010, 00:57 GMT-5 5 Replies
Please login with a confirmed email address before reporting spam
Good day..
I'm new to comsol, and I'm trying to do an analysis with a sinusoidal wave as input to a microcantilever. I followed the example of a mems microcantilever in 2D in which the down part of the microcantilever is the positive electrode and then there's an air gap and then the ground electrode.
The example runs very well. In the example it's applied a DC sweep from 1 to 6.3 volts, but I want to enter a sinusoidal wave instead of a DC voltage.
I typed into the solver parameters -> parametric the function V*sin(2*pi*f0) but at the moment of solving the system it gives me error 7096 parametric list must be strictly monotone, and I'd investigated and this is as I understand, because comsol doesn't support for example 1,2,3,4,3,2,1 values, and because the sinusoidal is throwing values up and down, this could be the reason of my error.
Now, if i want to enter a sinusoidal voltage to my system and observe how the microcantilver moves with the frequency of the sinusoidal, how can I do that ??
i hope someone can help...
Thanks !!
I'm new to comsol, and I'm trying to do an analysis with a sinusoidal wave as input to a microcantilever. I followed the example of a mems microcantilever in 2D in which the down part of the microcantilever is the positive electrode and then there's an air gap and then the ground electrode.
The example runs very well. In the example it's applied a DC sweep from 1 to 6.3 volts, but I want to enter a sinusoidal wave instead of a DC voltage.
I typed into the solver parameters -> parametric the function V*sin(2*pi*f0) but at the moment of solving the system it gives me error 7096 parametric list must be strictly monotone, and I'd investigated and this is as I understand, because comsol doesn't support for example 1,2,3,4,3,2,1 values, and because the sinusoidal is throwing values up and down, this could be the reason of my error.
Now, if i want to enter a sinusoidal voltage to my system and observe how the microcantilver moves with the frequency of the sinusoidal, how can I do that ??
i hope someone can help...
Thanks !!
5 Replies Last Post 14.11.2010, 15:31 GMT-5
Please login with a confirmed email address before reporting spam
Posted:
1 decade ago
Hi
COMSOL has several methods to solve time dependent problems, these are either a true time sequency, or by a harmonic sweep (assumed sinusoidal excitation with an given amplitude and possibly phase). The time solver takes quite some solver steps, hence processing time to cover several sinus oscillations, while the harmonic solver does this in one go.
Now any parametric sweep,in COMSOL, requires a monotously increasing/decreasing parameter (for the first one) so you should set as parameter the time 1,2,3,4,5 ... and then in your voltage BC use V0"sin(2*pi*f*t).
But its easier to use a harmonic solver, then you will not "see" the oscillation
--
Good luck
Ivar
COMSOL has several methods to solve time dependent problems, these are either a true time sequency, or by a harmonic sweep (assumed sinusoidal excitation with an given amplitude and possibly phase). The time solver takes quite some solver steps, hence processing time to cover several sinus oscillations, while the harmonic solver does this in one go.
Now any parametric sweep,in COMSOL, requires a monotously increasing/decreasing parameter (for the first one) so you should set as parameter the time 1,2,3,4,5 ... and then in your voltage BC use V0"sin(2*pi*f*t).
But its easier to use a harmonic solver, then you will not "see" the oscillation
--
Good luck
Ivar
Please login with a confirmed email address before reporting spam
Posted:
1 decade ago
Hi
I'm new to Comsol as well and would like to model a mems electrostatic switch as well. So, could you please direct me to where you found this example?
Thanks!
I'm new to Comsol as well and would like to model a mems electrostatic switch as well. So, could you please direct me to where you found this example?
Thanks!
Please login with a confirmed email address before reporting spam
Posted:
1 decade ago
hi shak !.
I just open the file named memsmodlib.pdf
I find it in the installation directory->doc->mems
I just open the file named memsmodlib.pdf
I find it in the installation directory->doc->mems
Please login with a confirmed email address before reporting spam
Posted:
1 decade ago
Hi Ivar !
Thanks for your answer, i get a little bit of new knowledge for comsol =).
No i was searching the way to do the harmonic sweep you just recommend, but i don't find the way, and I was asking me if you know a basic tutorial for this type of analysis, just to learn how comsol works doing this calculation ( i don't need complex things, just basics so i can understand how comsol works with this type of calculation).
I try to do another model, this time in 3D but using frequency response, but I don't see where to enter the sinusoidal =s, and i observe that the variable that is evaluated is freq_smsld, and I think i get a little bit lost with this.
Or if it is not very difficult and you can explain in here step by step it would be useful too...
I hope you can help me ...
thanks !
Thanks for your answer, i get a little bit of new knowledge for comsol =).
No i was searching the way to do the harmonic sweep you just recommend, but i don't find the way, and I was asking me if you know a basic tutorial for this type of analysis, just to learn how comsol works doing this calculation ( i don't need complex things, just basics so i can understand how comsol works with this type of calculation).
I try to do another model, this time in 3D but using frequency response, but I don't see where to enter the sinusoidal =s, and i observe that the variable that is evaluated is freq_smsld, and I think i get a little bit lost with this.
Or if it is not very difficult and you can explain in here step by step it would be useful too...
I hope you can help me ...
thanks !
Please login with a confirmed email address before reporting spam
Posted:
1 decade ago
Hi
OK I have to get back to 3.5a convention then. 1) frequency sweep is a harmonic solution type, so you are on the right way. 2) Indeed Comsol does not "tell explicitely" what it's sweeping, it's as many things with COMSOL it's implicit. They explain it once in the doc and expect us to understand, I have still often doubt's allthough I have been using COMSOL for a few years now.
The main word is read the doc, and as its rather dense, each word counts, expect to read the doc several times, then pass over , in detail the exercices from the doc (and the model in the model library).
Then in the doc, and in the 3.5 there are many examples of most cases, (also harmonic /frequency sweeps) but these are often with 2-3 physics so they are not trivial. Often its usefull to start simpler, by just extracting one physics at the time. It's less "sexy" but easier to learn with than a complex multiphysics study.
The default frequency name is "freq" in 3.5, but to be able to distinguish the variable names, in 3.5 we add "_smsld"
to the variable name to explicitely define it internal to the "smsld" = structural physics 3D application mode.
In the newer v4, the names are inverted and "smsld" becomes "solid" and its a prefix à la java: "solid.freq"
--
Good luck
Ivar
OK I have to get back to 3.5a convention then. 1) frequency sweep is a harmonic solution type, so you are on the right way. 2) Indeed Comsol does not "tell explicitely" what it's sweeping, it's as many things with COMSOL it's implicit. They explain it once in the doc and expect us to understand, I have still often doubt's allthough I have been using COMSOL for a few years now.
The main word is read the doc, and as its rather dense, each word counts, expect to read the doc several times, then pass over , in detail the exercices from the doc (and the model in the model library).
Then in the doc, and in the 3.5 there are many examples of most cases, (also harmonic /frequency sweeps) but these are often with 2-3 physics so they are not trivial. Often its usefull to start simpler, by just extracting one physics at the time. It's less "sexy" but easier to learn with than a complex multiphysics study.
The default frequency name is "freq" in 3.5, but to be able to distinguish the variable names, in 3.5 we add "_smsld"
to the variable name to explicitely define it internal to the "smsld" = structural physics 3D application mode.
In the newer v4, the names are inverted and "smsld" becomes "solid" and its a prefix à la java: "solid.freq"
--
Good luck
Ivar