Discussion Forum

Posted: 1 decade ago Feb 10, 2014, 9:39 p.m. EST

[QUOTE] Hello, i'am also a beginner with COMSOL and my free trial period has just ended. My question is, is COMSOL capable of calculating the current distribution and the magnetic field distribution of a moving conductor in between two fixed conducting planes? For clarity, i attached a picture of what i want to simulate in COMSOL. You see there the upper and lower contact disks with their corresponding contact bolts which will carry the current to/from the contact disks. The conductor will move with a prescribed velocity and has electrical connectivity to the contact disks.. Therefore the current will flow through a continuously moving path. Of interest is the magnetic field distribution while the conductor is moving and how strong the induced eddy current will be in the contact disks or an additional shielding around the contact disks. Therefore i would like to ask some experienced COMSOL users: Can this be done with COMSOL? Kind regards Florian Brabetz [/QUOTE] hi, florian I am not sure whether my understanding is right. Do you mean that the electric current flows frim the upper disk to the down disk through the cylinder conductor, while the conductor keeps sliding electric contact between the disks? If it is the sliding electric contact problem, it will be not easily solved by mf mode directly. My suggestion is that you may try to use the conductivity change to represent the move of the condctor. It means the conductivity in the domain between the two disks is modeled as an expression instead a constant value, and the rotationary movement of the conductor is inclued in the expression. You can refer to the tutorials like "electric impedance sensor" in ACDC model library or the "railgun" model in some earlier version like COMSOL 3.5a. Wish the above will be useful! Good luck! Tang

Posted: 1 decade ago Feb 15, 2014, 1:42 p.m. EST

Hello Lingling Tang, i will keep it in mind since my solution by now is really computationally intensive. The support helped me to setup the moving mesh boundaries correctly. Then the computed current density of the EC interface can be used as an "external current density" in the MF interface. With this approach i get the moving B-field and the induced eddy currents by the moving B-field. I attached a simple model which should clarify how i set up the interfaces. But be careful it took 9 hrs to calculate on my slow machine (core 2 duo 2.93 ghz) and unfortunately the file including the solution was to big to attach. I have to see how this will perform with real geometries and our computing cluster. If it won't run i have to look for another approach, so thanks for your suggestion. Kind regards Florian Brabetz

Attachments:

• moving_mesh_ec_mf.mph