Very interesting design. I can see the performance benefit in having the fuel cores also contain the coolant, but not having the coolant in a separate component block makes the puzzle a little too easy, as being able to pump enough coolant and shutting the reactor down when out of coolant is enough.
The best designs would always be a block of cores surrounded by reflectors that way, as it is the most efficient, not even requiring control rods, when the heat exchanging setup is good enough. I.e. it would be Factorio reactor, but with much simpler piping.
Requiring a separate coolant component block however would make sure that each core would need to touch at least one of those (or a “heat wire”/component heat exchanger), with potentially a limited cooling rate, which would make heat management a thing again.
Even when not limiting the cooling, it would allow for more reactor optimization possibilities as the coolant component block could have several “side effects” depending on the coolant, like water acting as a moderator.
Also, will breeder reactors exist? Then it would probably be good to decide between fast and moderated neutrons, which would open up a lot of possibilities, not just for reactor components (graphite moderator block) and therefore diversifying the reactor layouts, but also for fuels (emitting fast or moderated or both, different effects when getting hit by them, etc).
EDIT:
Maybe my concerns are invalid, because I’m still thinking within the IC2 model of heat management and fuel efficiency. Maybe the puzzle isn’t the fuel efficiency, but controlling neutron economy, the reactivity.
In IC2 reactors, the reactivity is always exactly 1, meaning each tick the reactor will generate the same number of neutrons and thus produce the same amount of energy/heat.
If GT6 reactors however had non-1 reactivity levels, controlling the reactivity with control rods would make a lot of sense, to keep reactors from going critical or turning off. It would also be very elegant to implement:
Each reactor core would need to check the neutron count of the neighboring blocks each ticks, add them together and multiply that by a quarter (only a forth of the neutrons from the neighboring block hit this one, because sides) to get the neutrons that react with the core.
- Multiply this number of neutrons with the “HU per reaction” factor of the fuel to get the heat the core produces.
- Multiply this number of neutrons with the “Durability loss per reaction” factor of the fuel to get the durability loss on the fuel.
- Multiply this number of neutrons with the reactivity of the fuel to get the new neutron count of the block.
Neutron-economy successfully simulated with an algorithm of O(n). Obviously requiring some starting neutrons, which could be generated on random block ticks or just by adding a +10 (or and other fuel specific value) to the result of the “new neutron count” equation. Would also work flawlessly with moderated neutrons.
A neutron economy based system would be very interesting and be less focused on the reactor layout and more on the redstone/computer behind controlling the control rods and thus retroactively the reactivity. However one potential problem I see with this system is that keeping the neutron economy very low would be the incentivised option, which is the most boring option. It’s the option with the least risk and the same fuel efficiency, the only downside is a lower HU production, which some would also count as a pro.
I therefore think it’s reasonable to incentivise the danger of a higher neutron count by giving it better fuel efficiency. Maybe it could even be made fuel specific for some fuels (not all), having an optimal temperature for optimal fuel efficiency.
This system would work even better with moderated neutrons, as converting normal neutrons into moderated ones could be an important additional balancing factor, because of the different multipliers of the fuel for moderated neutrons.