Poloidal field coil design for known plasma equilibrium states

The technique for obtaining plasma equilibria with given boundary conditions has long been known and understood. The inverse problem of obtaining a poloidal field (PF) coil system from a given plasma equilibrium has been widely studied, however its solution has remained largely an art form. An investigation, by the writers, of this fundamentally ill-posed inverse problem has resulted in a new understanding of the requirements that solutions must satisfy. A set of interacting computer codes has been written which may be used to successfully design PF coil systems capable of supporting given plasma equilibria. It is shown that for discrete coil systems with a reasonable number of elements the standard minimization of the R M S flux error can lead to undesirable results. Examples are given to show that an additional stability requirement must be imposed on the regularization parameter to obtain correct solutions. For some equilibria, the authors find that the inverse problem admits dual solutions corresponding to two possible magnetic field configurations that fit the constraining relations on the plasma surface equally well. An additional minimization of the absolute value of the limiter flux is required to discriminate between these solutions