Coinsiding nodes on Electromagnetic Egde Based Harmonic Analysis??

[JonLueke]

I was wondering if anyone knew how to deal with the following problem.

I am modeling a simple rectangular prism (400x50x2) microns, and want to study the current density and skin effect of the volume when exited by AC voltage bias. I am using the SOLID117 element, with Keyopt(1)=1, which is the classical formation of the element, including the AZ and VOLT dofs. I applied a voltage difference over the length (400 microns) of the model, as well as assigning parallel magnetic flux boundary conditions to the appropreate faces of the prism. I attempted a harmonic solution @ 40 Ghz, using a ramped boundary condition. Ansys returns an error saying:

P3dx2s: 1st and 2nd nodes coincide.

and

Edg117: Can not compute ESYS

I have searched the documentation and the web in order to find out what these errors actually mean, but have not found anything. If anyone had any ideas, or a listing of Ansys error messages and what they mean, it would be greatly appreciated.

Jonathan Lueke
Researcher
University of Alberta

 

[ifegermany]

you did specify a non-zero resistivity I assume, as
this is a pre-requisite for calculating currents.

SOLID117 minimum dimension is refered in Element Reference as 1e-6 length units. So really, you should be save with MKS, having a actual lenght of 400e-6 units. But if you want
to change see:

Coupled-Field Guide
Chapter 1. Coupled-Field Analyses
1.3. System of Units
Table 1.4

IF you change from MKS, make sure you issue/update EMUNIT command as well, as this does define free space permeability value in corresponding units.



Thing with loading in the VOLT DOF is, this is in units
of [Volt x sec time-integrated potential] - see Element Reference.

Have a look at:

Coupled Field Guide
Chap. 7.2. Electromagnetic-Circuit Simulation

to see if you actually can model the desired skin effect
and current density distribution using SOLID117, or instead
you might be able to uttilize PLANE53 (2D) or SOLID97(3D)
in conjunction with CIRCU124.