The nodal finite element realization of the T-/spl Omega/ method involving a gauged current vector potential, T, is shown to yield erroneous results if applied to 3D eddy current problems with reentrant corners in the conducting region. The reason for the problem is pinpointed and a remedy is suggested.
No takes yet. Share an insight, caveat, or question.
Preis et al. (2000) studied this question.
Synapse has enriched 5 related papers. Consider them for comparative context: