Thanks a lot for your clear explanation
Arnoldo
David Meeker <dmeeker@xxxxxxxx> wrote:
ac090960wrote: Hello,The 3.2 version will let you evaluate this problem, but the problem it's solving is not the problem that you probably intended. For on-edge laminations, the total current inside each lamination should add up to zero inside the cross-section of each lamination if you have a tape-wound core, or if you are considering penetrations into a laminated stack. However, this condition isn't enforced in the way that the on-edge problems were implemented in 3.2, so the result that you get is for the analysis of a series of concentric iron rings rather than one long tape in the axi case, or you get the results for a stack of laminations where all the laminations in the stack are shorted together at +/- infinity. This isn't what people are usually intending by doing the on-edge lamination analysis, so I changed the program so that it just doesn't allow the user to run the on-edge model for harmonic problems in the 3.3 version, as well as in the 03Dec02 update of 3.2.
Making some trials with the new release 3.3 (downloaded today) on an
AC axisymmetric geometry an error occurred during FKN execution: (On-
edge lamination not supported in AC analyses).
The strange is the same problem is running using the release 3.2.
There is, however, no problem with the in-plane laminations (e.g. stacked laminations in a transformer or motor core). In the in-plane case, the eddy currents always integrate to zero across the thickness of the lam (unlike the on-edge case), so it is reasonable to represent the combined eddy current/hysteresis effects via an equivalent bulk material with a zero conductivity and a complex-valued permeability (how femm does the in-plane laminations internally).
If you want to analyze problems with on-edge laminations, (e.g.. tape-wound core or penetration of the flux into a laminated planar stack through the sides of the laminations), you'll need to model the individual laminations and impose a separate circuit condition on each one to force the net current in each lamination to add up to zero (I know this is a drag to do). I've attached an example of a tape-wound core that has been done by modeling each lamination in this way.
Dave.-- David Meeker dmeeker@xxxxxxxx http://femm.berlios.de/dmeeker
Your use of Yahoo! Groups is subject to the Yahoo! Terms of Service.
> ATTACHMENT part 2 application/x-zip-compressed name=axi3.zip