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.
Current between to dissimilar contacts
Posted 6 avr. 2017, 02:12 UTC−4 Low-Frequency Electromagnetics Version 5.2 1 Reply
Please login with a confirmed email address before reporting spam
Hi,
I have made an extremely simple model of two contacts joined together. They are of different material. For one of the contacts I do a parametric sweep of its conductivity.
The expected result does not align with the output from the COMSOL model. At low conductivities, it is expected that the resistance approaches an open circuit and varies linearly.
R = rho*length/Area.
The model suggests that it flattens out. Any ideas?
The model is at the link below:
drive.google.com/drive/folders/0Bw4FmphlWoNeX1ZJUGlWaTd5RFU?usp=sharing
I have made an extremely simple model of two contacts joined together. They are of different material. For one of the contacts I do a parametric sweep of its conductivity.
The expected result does not align with the output from the COMSOL model. At low conductivities, it is expected that the resistance approaches an open circuit and varies linearly.
R = rho*length/Area.
The model suggests that it flattens out. Any ideas?
The model is at the link below:
drive.google.com/drive/folders/0Bw4FmphlWoNeX1ZJUGlWaTd5RFU?usp=sharing
1 Reply Last Post 7 avr. 2017, 09:54 UTC−4
Please login with a confirmed email address before reporting spam
Posted:
8 years ago
Hello Sivaharan,
The analytical solution you expect is
current = A/L * resistivity1/(1+resistivity1/resistivity2)
Your model's results match that expression very well as long as conductivity (=1/resistivity1) is above ~1e3[S/m], but deviates from it at low values of conductivity.
This behavior suggested to me that the reason for this may be that when your two conductivities are too dissimilar the problem becomes ill-conditioned. Your model is set up with an iterative solver, and iterative solvers can be more sensitive to ill-conditioning that direct solvers. This inspired me to switch your model to a direct solver to see if that would help. Sure enough, when I did that, your model's results matched the analytical expression above perfectly for the entire range of values of conductivity.
With an iterative solver, tightening the tolerance also helped.
You can read more on the topic of matrix conditioning and solvers in this blog:
www.comsol.com/blogs/solutions-linear-systems-equations-direct-iterative-solvers/
Best regards,
Jeff
The analytical solution you expect is
current = A/L * resistivity1/(1+resistivity1/resistivity2)
Your model's results match that expression very well as long as conductivity (=1/resistivity1) is above ~1e3[S/m], but deviates from it at low values of conductivity.
This behavior suggested to me that the reason for this may be that when your two conductivities are too dissimilar the problem becomes ill-conditioned. Your model is set up with an iterative solver, and iterative solvers can be more sensitive to ill-conditioning that direct solvers. This inspired me to switch your model to a direct solver to see if that would help. Sure enough, when I did that, your model's results matched the analytical expression above perfectly for the entire range of values of conductivity.
With an iterative solver, tightening the tolerance also helped.
You can read more on the topic of matrix conditioning and solvers in this blog:
www.comsol.com/blogs/solutions-linear-systems-equations-direct-iterative-solvers/
Best regards,
Jeff