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 The information provided may be out of date.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.
DC current model: initial value issue?
Posted Mar 1, 2012, 11:19 a.m. EST Low-Frequency Electromagnetics Version 4.2 6 Replies
Please login with a confirmed email address before reporting spam
Hi,All,
I am building a DC current model using COMSOL 4.2. it is 3D structure, consisting of 2 parts: at bottom, a "L" shape semiconductor mesa ; on top of it, there are two metal electrodes: one is circle shape and the other is square. I am driving DC current from the "square" metal electrode to the "circle" electrode. I set the top surface of the "square" electrode as "normal current density" to account for the incoming current , while I set the top surface of the "circle" electrode as "ground" to let the incoming current to go out. plus, I set a "pair contact impedance" to account for the electrical contact resistance at the interface between the metal electrodes and the semiconductor, for each individual interface. inside the "pair contact impedance", I am using "surface impedance" option instead of the "thin layer" to describe the interface impedance. I want to know the electrical potential of the whole system. this model solves and gave a 3-D results of the electrical potential. However, I don't believe the results because if I keep everything else untouched and just vary the "initial value", for example from 0.1v to 0.2V, the solution(electrical potential of the system) changes closely with the "initial value" . my instinct says that the physics in my model may be missing one boundary condition. could you give some hint what the problem may be? I also attached the model file too.
thanks,
I am building a DC current model using COMSOL 4.2. it is 3D structure, consisting of 2 parts: at bottom, a "L" shape semiconductor mesa ; on top of it, there are two metal electrodes: one is circle shape and the other is square. I am driving DC current from the "square" metal electrode to the "circle" electrode. I set the top surface of the "square" electrode as "normal current density" to account for the incoming current , while I set the top surface of the "circle" electrode as "ground" to let the incoming current to go out. plus, I set a "pair contact impedance" to account for the electrical contact resistance at the interface between the metal electrodes and the semiconductor, for each individual interface. inside the "pair contact impedance", I am using "surface impedance" option instead of the "thin layer" to describe the interface impedance. I want to know the electrical potential of the whole system. this model solves and gave a 3-D results of the electrical potential. However, I don't believe the results because if I keep everything else untouched and just vary the "initial value", for example from 0.1v to 0.2V, the solution(electrical potential of the system) changes closely with the "initial value" . my instinct says that the physics in my model may be missing one boundary condition. could you give some hint what the problem may be? I also attached the model file too.
thanks,
Attachments:
6 Replies Last Post Mar 5, 2012, 3:16 p.m. EST
Please login with a confirmed email address before reporting spam
Posted:
1 decade ago
Hi
not that I think it should change the results (but perhaps) now in V4 you do not need to define "pairs" and use assembly mode for geometry in contact with "thin" physics, as you have the standard "Contact Impedance" that overrides the default Union continuity, they behave the same from my understanding
In anycase, if you use the Assembly mode, I would use "create imprint"s to limit the contact pair region to the two overlapping boundaries and not include the full L shape area
--
Good luck
Ivar
not that I think it should change the results (but perhaps) now in V4 you do not need to define "pairs" and use assembly mode for geometry in contact with "thin" physics, as you have the standard "Contact Impedance" that overrides the default Union continuity, they behave the same from my understanding
In anycase, if you use the Assembly mode, I would use "create imprint"s to limit the contact pair region to the two overlapping boundaries and not include the full L shape area
--
Good luck
Ivar
Please login with a confirmed email address before reporting spam
Posted:
1 decade ago
Hi, Ivar,
thanks for the reply.
I followed your advice: on the "finalize" node, I now chose "form union" over "form assembly" ( the reason I chose "form assembly" earlier is because someone did for me since I am new to COMSOL). then those "identity pairs" despair,and the "pair contact impedance" is useless too. so I use "contact impedance" instead now. and with another change in my model, here I use "external potential" on the top surface of the "square" electrode to account for the incoming current, where i used to use "normal current density". with above modification, the model seems work well. thanks.
but if I use "normal current density" to account for the incoming current, my results looks funny. I am curious which boundary condition should be used for the external current in common DC current structure. for example, if you want current to flow in one electrode and flow out from another electrode.
the modified version is attached too.
regards,
thanks for the reply.
I followed your advice: on the "finalize" node, I now chose "form union" over "form assembly" ( the reason I chose "form assembly" earlier is because someone did for me since I am new to COMSOL). then those "identity pairs" despair,and the "pair contact impedance" is useless too. so I use "contact impedance" instead now. and with another change in my model, here I use "external potential" on the top surface of the "square" electrode to account for the incoming current, where i used to use "normal current density". with above modification, the model seems work well. thanks.
but if I use "normal current density" to account for the incoming current, my results looks funny. I am curious which boundary condition should be used for the external current in common DC current structure. for example, if you want current to flow in one electrode and flow out from another electrode.
the modified version is attached too.
regards,
Attachments:
Please login with a confirmed email address before reporting spam
Posted:
1 decade ago
Hi
I would propose to use a "terminal" in current or voltage mode, this lumped BC sets an average value over the boundary, but gives you directly in the "Results - Derived Values - Global Evaluations" sub nodes the impedances currents and voltages, so you do not need to bother about surface integrations
--
Good luck
Ivar
I would propose to use a "terminal" in current or voltage mode, this lumped BC sets an average value over the boundary, but gives you directly in the "Results - Derived Values - Global Evaluations" sub nodes the impedances currents and voltages, so you do not need to bother about surface integrations
--
Good luck
Ivar
Please login with a confirmed email address before reporting spam
Posted:
1 decade ago
Ivar,
I tired to use "terminal" to account for the external current, it works very well. thanks.
actually I once considered to use it, but I find out in COMSOL help file: under the entry "Boundary Conditions for the Electrostatics Interface", "terminal" is listed as one of "Interior Boundaries", also as one of "Exterior Boundaries". I don't understand the difference between the 2 situations, so I decide to drop it. but glad it works out at the end.
regards,
I tired to use "terminal" to account for the external current, it works very well. thanks.
actually I once considered to use it, but I find out in COMSOL help file: under the entry "Boundary Conditions for the Electrostatics Interface", "terminal" is listed as one of "Interior Boundaries", also as one of "Exterior Boundaries". I don't understand the difference between the 2 situations, so I decide to drop it. but glad it works out at the end.
regards,
Please login with a confirmed email address before reporting spam
Posted:
1 decade ago
Hi
for me in ES or EC there is no difference for internal or external Terminal BCs.
However in MF or MEF there are some issues with the consisitency of terminal and physics. For the magnetic field one need to have a current loop (a full one) as one cannot have a current source appearing in 3D space like that, and dissapearing somewhere else in the middle of our model, that is not "physical" we need to lead the current back to form a loop. In 2D there is the workaround to say the current loops out of the plane, and we solve for only some of the full vector field
--
Good luck
Ivar
for me in ES or EC there is no difference for internal or external Terminal BCs.
However in MF or MEF there are some issues with the consisitency of terminal and physics. For the magnetic field one need to have a current loop (a full one) as one cannot have a current source appearing in 3D space like that, and dissapearing somewhere else in the middle of our model, that is not "physical" we need to lead the current back to form a loop. In 2D there is the workaround to say the current loops out of the plane, and we solve for only some of the full vector field
--
Good luck
Ivar
Please login with a confirmed email address before reporting spam
Posted:
1 decade ago
Hi, Ivar,
thanks for the explanation. back to this model, I had another simple question: what if I wan to get average value over a specific area?
in my case, I am calculating the electrical potential over my 3D structure. On any boundary surface, I could use "results ---->derived value----->average--->surface average" to calculate the electrical potential over the whole surface which is selected boundary. but what if I am only interested in part of one boundary surface. how can I do it? for example , if I only want to calculate the average over a specific area of 1x2um on my "L" shape mesa.
regards,
thanks for the explanation. back to this model, I had another simple question: what if I wan to get average value over a specific area?
in my case, I am calculating the electrical potential over my 3D structure. On any boundary surface, I could use "results ---->derived value----->average--->surface average" to calculate the electrical potential over the whole surface which is selected boundary. but what if I am only interested in part of one boundary surface. how can I do it? for example , if I only want to calculate the average over a specific area of 1x2um on my "L" shape mesa.
regards,