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 microcantilever (Problems connecting multiple physics)
Posted Nov 28, 2010, 7:11 a.m. EST 1 Reply
Please login with a confirmed email address before reporting spam
I'd like to simulate a microcantilever deflecting under electrostatic actuation. The cantilever is of non-uniform width. Only a portion of the area below the cantilever is to act as the electrode. I'd like to calculate the pull-in voltage of the structure. Could you please offer me some directions on how to proceed?
I know that I should use the moving mesh, electrostatics and solid mechanics physics. However,I'm not clear as to what exactly I should do with the modules. Also, how should I model the air around the cantilever? Air has nearly the same permittivity as vacuum, and I'm not modelling squeeze film damping, so do I need to model the air as well?
I added the following:
Solid Mechanics:
Fixed Constraints
Boundary Load (how do I connect the Fes force to this boundary load? What values should I give in the user defined Fx,y,z?)
Prescribed Displacement
Electrostatics:
Electric Potential (value V_in NaN as parameter)
Ground
Force Calculation ( with a force Fes)
Moving Mesh:
Prescribed Deformation (u,v,w)
As of now the system on simulation(varying V_in) is giving no displacement.
I'm not sure how to proceed and any help is welcome.
I know that I should use the moving mesh, electrostatics and solid mechanics physics. However,I'm not clear as to what exactly I should do with the modules. Also, how should I model the air around the cantilever? Air has nearly the same permittivity as vacuum, and I'm not modelling squeeze film damping, so do I need to model the air as well?
I added the following:
Solid Mechanics:
Fixed Constraints
Boundary Load (how do I connect the Fes force to this boundary load? What values should I give in the user defined Fx,y,z?)
Prescribed Displacement
Electrostatics:
Electric Potential (value V_in NaN as parameter)
Ground
Force Calculation ( with a force Fes)
Moving Mesh:
Prescribed Deformation (u,v,w)
As of now the system on simulation(varying V_in) is giving no displacement.
I'm not sure how to proceed and any help is welcome.
1 Reply Last Post Nov 28, 2010, 10:16 a.m. EST
Please login with a confirmed email address before reporting spam
Posted:
1 decade ago
Hi
first if you are in v4 (check latest updates) AND you have small deformations and static or low speed of your air/vacuum section, then you might get around without the moving mesh, taking into account the default now "Spatial frame driven by the deformations" in structural. This means that you need to set up air as some "gelly-solid" with an E around 1 to 1000 Pa, nu=0 and rho=1 (needs some tweeking, just keep E/rho at least 1E6-1E9 times smaller for E/rho than for your structural material). You have a simpler model like that. The electrostatic must be solved or driven by the spatial frame too so check your operators and how you couple.
Start with a simple 2D model (two thin horizontal beams one inside the other (the inner one free on the right side but with commen boundary (fixed to the left), structural material for the inner one + air around (both in structural, external boundaries fixed, inner free), add DC on the external domain, GND on the bottom horizontal external boundary, isolation on the vertical ones, apply Volts on the inner beam boundaries. Calculate the inner rectangle electrical forces and link them to the structurla boundar loads. Note: if the geometry and BC's are too symmetric it might be difficult to solve as the system might then not know towards where to start bending
In any case you must couple electrostatic/dynamic foces to the structural force: domain load or preferred boundary load. Check the units of the calculated electric induced forces. If in N/m^2 (or N/m in 2D) you can normally apply them directly to the boundary loads (and v4 should propose that too for you in the corresponding drop down menu)
Note that you should apply a "distributed" force onto a "distributed" boundary are, and not calculate an average electric force and smear this out regularly over the area. This is often overlooked because of the implicit notation of the COMSOL GUI fields, the formula we apply is defined for a single element placed anywhere along the boundary/domain used in the selection list, the location (x,y,z) and Entity material index "_i" (i.e. solid.rho_i) are not written out and is easily forgotten.
--
Good luck
Ivar
first if you are in v4 (check latest updates) AND you have small deformations and static or low speed of your air/vacuum section, then you might get around without the moving mesh, taking into account the default now "Spatial frame driven by the deformations" in structural. This means that you need to set up air as some "gelly-solid" with an E around 1 to 1000 Pa, nu=0 and rho=1 (needs some tweeking, just keep E/rho at least 1E6-1E9 times smaller for E/rho than for your structural material). You have a simpler model like that. The electrostatic must be solved or driven by the spatial frame too so check your operators and how you couple.
Start with a simple 2D model (two thin horizontal beams one inside the other (the inner one free on the right side but with commen boundary (fixed to the left), structural material for the inner one + air around (both in structural, external boundaries fixed, inner free), add DC on the external domain, GND on the bottom horizontal external boundary, isolation on the vertical ones, apply Volts on the inner beam boundaries. Calculate the inner rectangle electrical forces and link them to the structurla boundar loads. Note: if the geometry and BC's are too symmetric it might be difficult to solve as the system might then not know towards where to start bending
In any case you must couple electrostatic/dynamic foces to the structural force: domain load or preferred boundary load. Check the units of the calculated electric induced forces. If in N/m^2 (or N/m in 2D) you can normally apply them directly to the boundary loads (and v4 should propose that too for you in the corresponding drop down menu)
Note that you should apply a "distributed" force onto a "distributed" boundary are, and not calculate an average electric force and smear this out regularly over the area. This is often overlooked because of the implicit notation of the COMSOL GUI fields, the formula we apply is defined for a single element placed anywhere along the boundary/domain used in the selection list, the location (x,y,z) and Entity material index "_i" (i.e. solid.rho_i) are not written out and is easily forgotten.
--
Good luck
Ivar