Finite element analysis of the temperature distribution within a Conduction-Cooled, MgB2-based MRI superconducting coil segment

• Published in: Cryogenics (Guildford) Vol. 127; p. 103563
• Main Authors: Zhang, Danlu, Sumption, Mike D., Majoros, Milan, Collings, Edward W., Panik, Dean, Rindfleisch, Matt, Doll, David, Tomsic, Mike J.
• Format: Journal Article
• Language: English
• Published: Elsevier Ltd 01-10-2022
• Subjects: Conduction-cooling; FEA; Heat transfer; MgB2; Simulation; Superconducting magnet; Superconducting magnet; Conduction-cooling; FEA; Simulation; MgB2; Heat transfer
• ISSN: 0011-2275; 1879-2235
• DOI: 10.1016/j.cryogenics.2022.103563

•• Temperature-dependent thermal conductivity values for a magnet winding were calculated.•• A 3D FEA model was built for thermal evaluations of a superconducting coil.•• Our 3D model showed a ΔTmax of 5.1 K and an Ic margin of 12.75 A.•• New current designs reduced ΔTmax to as low as 3 K.
Superconducting magnets used for Magnetic Resonance Imaging (MRI) scanners need to keep temperature gradients minimized in order to retain thermal and operating current margin. We have used 3D finite element analysis (FEA) simulation in COMSOL Multiphysics software that includes both conductive heat transfer and radiative heating to calculate the temperature distribution both along the winding direction and across the cross-section of an MRI segment coil at its equilibrium operating temperature. We have also modelled the evolution of the thermal properties during cool-down from ambient temperature. The heat capacity and thermal conductivity of the magnet winding were computed for use within this simulation. The heat capacity as a function of temperature was calculated using a rule of mixtures. This procedure was also used for the thermal conductivity along the direction of the wire. However, the thermal conductivity within the composite cross section (x- and y-directions) was computed using a 2D FEA model. Based on this, a time-dependent, 3D coil model was built to calculate the coil temperature throughout the winding during cool-down in our test cryostat system. The model included a heat leak component to the coil current contacts via conduction through the current leads as well as a radiative component from the surfaces of the cryostat. A key result was that a maximum coil ΔTmax = 5.1 K (=maximum temperature within the winding -minimum temperature in the winding) was seen and a coil Ic margin of 12.75 A was predicted at steady state, with our first current lead design. A second set of more optimized current leads significantly lowered the ΔTmax within the coil at the steady state. The coil Ic margin has been analyzed for different current lead designs.