Peter T. While1, Michael Poole2, Larry K. Forbes1, Stuart Crozier2
1School of Mathematics and Physics, University of Tasmania, Hobart, Tasmania, Australia; 2School of Information Technology and Electrical Engineering, University of Queensland, Brisbane, Queensland, Australia
A method is presented for the theoretical design of gradient coils with minimum maximum temperature. A previously reported analytical model is used for predicting the spatial temperature distribution and sequential quadratic programming is used to minimize peak temperature directly. The method is applied to both symmetric and asymmetric gradient coils and considerable reductions in maximum temperature (up to 50%) are achieved at no cost to coil performance. The new winding patterns are characterized by a fish-eye effect to smooth the hot spot regions. The final coil solutions are dependent on the assumed thermal material properties rather than direct coil spreading.