Discussion Forum

Posted: 1 decade ago 2 mar 2011, 15:09 GMT-5

Hi first I would ask how do you intend to integrate a distance between two surfaces, in the sens how would you write out the mathematical formula to obtain what you need ? Once you have that I assume it will be rather simpel to find the right integration operator -- Good luck Ivar

Posted: 1 decade ago 2 mar 2011, 15:18 GMT-5

Hi, well I first defined a variable for the surface of plate 1: temp1=epsilon0/(1m+w) with w describing the deformation of surface 1 on z-axis and 1m=the original distance between the two plates. Then I define an integration-operator called intop1 for surface 1 and a global function: mod1.intop1(temp). That works and I get the capacitance with respect to the deformation of surface 1. Now I have to replace the value of (1m) in the formula for temp1 with the deformation in z-axis of surface of plate 2. And I don't know how to get these two things coupled. Do you know what I mean? Chris

Posted: 1 decade ago 2 mar 2011, 16:30 GMT-5

Perhaps I can simplify the problem a little bit: I just don't know how to calculate the difference in z-axis (same x und y coordiantes) of two points, each one located on the surface of 2 parallel plates... Sorry but sometimes I think too difficult :)

Posted: 1 decade ago 2 mar 2011, 17:33 GMT-5

Hi do we agree that w is in fact a function of w(x,y,z) ? so if you look at w on a boundary, that I assume is a plane (in the material frame) perpendicular to Z then w(x,y,Z1), hence temp1(x,y,Z1) is also a function of (x,y,Z1) all for (x,y) belonging to the top boundary/capacitor plate 1 (hence for a given Z1. The same applies for the w(x,y,Z2) on the other capacitor at a given height Z2, but this w value cannot be accessed on plate 1 because these are at two different "z" values. Then if you want the NORMAL distance between the plates and both are alinged with their normal along "z" then you need to map w(x,y,Z2) onto the corresponding point (x,y,Z1) this is obtained by a projection coupling operator from plate 2 to plate 1. But now are you sure that its enough to ONLY consider the normal distance for the capacity calculation ? -- Good luck Ivar

Posted: 1 decade ago 2 mar 2011, 17:44 GMT-5

Hi Ivar, well, what else to do? I integrate the distances between the two deformed electrodes over the wohle surface of the electrodes. I think that's all. Am I wrong?

Posted: 1 decade ago 3 mar 2011, 07:56 GMT-5

Hi probably not but to be sure compare your integration with a hand calculation, and with the induced energy or via port/terminal method, just to see how they all behave -- Good luck Ivar

Posted: 1 decade ago 7 ott 2011, 16:00 GMT-4

[QUOTE] Hi do we agree that w is in fact a function of w(x,y,z) ? so if you look at w on a boundary, that I assume is a plane (in the material frame) perpendicular to Z then w(x,y,Z1), hence temp1(x,y,Z1) is also a function of (x,y,Z1) all for (x,y) belonging to the top boundary/capacitor plate 1 (hence for a given Z1. The same applies for the w(x,y,Z2) on the other capacitor at a given height Z2, but this w value cannot be accessed on plate 1 because these are at two different "z" values. Then if you want the NORMAL distance between the plates and both are alinged with their normal along "z" then you need to map w(x,y,Z2) onto the corresponding point (x,y,Z1) this is obtained by a projection coupling operator from plate 2 to plate 1. But now are you sure that its enough to ONLY consider the normal distance for the capacity calculation ? -- Good luck Ivar [/QUOTE] Hi Ivar I am working on the same type of problem. "Then if you want the NORMAL distance between the plates and both are alinged with their normal along "z" then you need to map w(x,y,Z2) onto the corresponding point (x,y,Z1) this is obtained by a projection coupling operator from plate 2 to plate 1." Can you show me one example how I can map w(x,y,Z2) onto the corresponding point (x,y,Z1) by a projection operator thanks Mustafa

Posted: 1 decade ago 11 ott 2011, 10:31 GMT-4

Hi try this model, this is a typical example how I test out the COMSOL features, simple models simple cases easy to verify (actually I havent verified this one, pls do it, I hope its correct ;) -- Good luck Ivar

Attachments:

• BndSim1.mph