A.T. said:
I don't get this 65% reduction. The total released energy is still the same. So the energy that didn't go into the shock-wave, must go somewhere else, like radiation. But it cannot just disappear.
.
Total released energy from detonation in the vacuum of space would typically be less than the total released energy from detonation well within the atmosphere, mainly due to the pressure and mass of the atmosphete acting as a tamper effectively delaying disassembly allowing additional fissions and fusions to occur.
Additionally, surrounding atmosphere will act to some very limited degree as a neutron reflector, sending some otherwise escaped neutrons back through for another pass at the fissile material in the "spark plug", casing and disassembling originally supercritical mass to increase fission yields, as well as with lithium to produce additional tritium and He3 production to increase fusion yield.
There would also be the effect of differing heat transfer rates immediately following detonation. Heat transfer being dominated by radiation in a very short period immediately following detonation and the lack of atmosphere forming an opaque barrier to many wavelengths at very high temperatures, the disassembling device will experience greater heat transfer briefly after detonation causing temperatures to fall more quickly shortly after detonation making conditions for fusion less favorable that those for a detonation in atmosphere.
.
As to a 'shockwave'developing, what occurs in the vacuum of space is going to be very different for what occurs in the atmosphere. Not having the resistance of the atmosphere, the fastest particles will attain higher speeds. Without atmospheric resistance, outward bound material does not slow much until it has something to interact with. Without atmospheric resistance, there will be no coalescence into a leading edge shock front. The fastest particles will continue to move further ahead of the slower particles. At sufficient distance particles from the blast will be effectively segregated into a smooth gradient of speed with the particles with greatest speed arriving first and successively slower particles arriving there after.
I too am suspicious of the '65%' reduction in damage because it would need to treat all the various types of damage as fungible across various targets and this is simply not the case. Damage from various effects is heavily dependent on the type of target, so altering the relative strength of such effects would require specification of target type to begin to compare relative damage.