Creep Effects in the Toroidal Field Coils of Fire and Other Burning Plasma Tokamaks

All three burning plasma experiments discussed at Snowmass during the summer of 2002, use preloaded structures to resist some component of the operating loads. For the resistive pulsed reactors, it is the preloads which introduce the most noticeable creep responses because these loads are applied for much longer than the operating loads. If the preloads are maintained during shut-down and maintenance periods, then the structure experiences the preload stresses at room temperature. OFHC copper has significant creep behavior, predominantly at high stress and high temp, but copper experiences finite creep even at cryogenic temperatures. The Beryllium copper used in the FIRE inner leg has better creep properties than OFHC copper. The purpose of these analyses is to characterize the influence of creep on the magnets of the Fusion Ignition Research Reactor(FIRE) and compare it with the creep response of the other proposed burning plasma experiments. The concern is that the desirable features provided by coil preloads will be lost over the lives of the experiments. Structural finite element models of FIRE and IGNITOR are used with creep equations derived from NIST[6] data to explore the structural sensitivity of the machines to creep. For both FIRE and IGNITOR, copper coil material, creep has been found to have a minimal effect on magnet performance. IGNITOR's generally lower stresses (with respect to FIRE’s BeCu TF stresses) and the use of active as well as passive preload systems helps reduce creep to acceptable levels. FIRE's structure is more sensitive to creep due to the free standing wedged TF coil, but the BeCu used in FIRE's inner TF legs has a much lower creep behavior than ETP or OFHC copper. This reduces creep to acceptable levels. For FIRE, however, there is some creep in the horizontal legs which relaxes some of the support of the inner leg. Recommendations are presented to support the OFHC copper horizontal legs more effectively. More work is needed to address the multiple load-unload cycling effects on creep.