Systems like the LFTR have passive high temperature thermal control based on thermal expansion of a near-critical mass density. As the temperature increases, thermal expansion produces a rapid drop in power production thereby stabilizing the reactor core.
Systems like the E-Cat HT are solid state and, in any event, are not dependent on critical mass density, but another approach to utilization of thermal expansion might work: Thermal Convection To make thermal convection work, passive (free) convective forces must be large enough to move enough thermal capacity past the power source and must be in a regime where the rate of cooling exceeds the power production at the target temperature. The 3 variables one has to play with to reach the target temperature are material thermal properties, power density of the E-Cat and g forces. Of these three, only g forces and power density are amenable to continuous alteration via centrifugation and reactor fabrication respectively. In my ultracentrifugal rocket engine patent, the g-forces are so enormous that enormous fluid flow, hence enormous thermal capacity flow enables relatively small heat exchange surfaces to cool the engine. A material that might be worthwhile analyzing in this regard is NaCl (sodium chloride) with a melting point near the high end of the E-Cat HT, and a heat capacity comparable to that of H2O. It is problematic to run molten NaCl in an ultracentrifuge due to material strength limits as they detemper at high temperature. On the other hand, power density might be reduced to the point that the heat capacity flow rate, even under only 1-g, might be sufficient. Clearly some arithmetic needs to be done here.
