- reminiscent of your analysis of Beagle2's heatshield and entry dynamics, and your prognosis (before the event) of a 10 percent chance of mission success ..

The section in the document relevant to your thermal analysis would be -

RTGs are composed, as all electric power sources, of two units: a thermal source and a thermoelectric converter. For VGNP, the General Purpose Heat Source (GPHS) is the most appropriate and well developed of thermal source technologies.

The GPHS is constructed of Pu238 oxide, formed into encapsolated pellets, and integrated into graphite blocks for support and re-entry protection. These devices are manufactured in 250 W (thermal) units for marriage into thermoelectric generating units. The RTG constructed for Cassini by the Department of Energy and General Electric employs a 4,500 W (thermal) GPHS assembly.

Thermoelectric converters constitute the second portion of the system. Cassini employs silicon-germanium units that convert heat to electricity at an efficiency of 6.8%. These units operate at a nominal hot-side temperature of 1000 deg C and reject to 300 deg C. The Cassini RTG creates approximately 300 W (electric) at 28-30 V.

Reconfiguring these units for the venusian surface will require a development effort to guard the thermoelectrics and GPHS from the atmosphere. RTGs operating in space vent to vacuum the helium nuclei created by the Pu238 decay. This will not be possible in the venusian environment. The retained helium eventually accumolates in the GPHS and provides a high-conduction path for the thermal energy to the surrounding environment, lowering the units efficiency. GE has proposed a solution to this problem for RTG designs supplied to be to the MESUR mission that uses a low-conductivity cover gas for the GPHS.

For Venus, redesigns of the RTG's aluminum cover and the unicouple cold-end attachments will also be required.

I am not qualified to comment on whether this design could reach a stable acceptable equilibrium temperature - interesting point.

"Hot electronics' would solve the whole problem beautifully, for sure - I wonder what progress has been made in that department since Malin posted his document. It's great even to be batting these ideas around ..