Since modular engines are on the way, i havent really played about with them much but when i did, its when the 3x3 parts were still pretty new, and temps would exeed 2500, if they STILL get that hot with engine overheating off, that might open up the first door to a melt down/blow out/overheating mechanic, im also planning on releasing a logic overhaul update on the savannah here in the near future mainly the pumps and power systems are my focus at the moment, such as making the pumps have an actual purpose other then to just be on to allow the reactor to start and run, there is no overheating mechanic, in reality with a real rbmk reactor, loss of feedwater even for a brief time could cause the reactor to run away because of the positive void coefficient, the breakdown of the positive void coefficient (PVC) i applies to atomic physics itself, basically a nuclear fission chain reaction CAN NOT self sustain(go critical) without a moderator, a moderator is a way, mainly an element used inside the reactor core to slow down the neutrons, as of now there are 2 currant ways of sustaining a nuclear chain reaction with 2 moderator types, #1 a reflector and #2 a refractor, a reflector is just as it sounds, it reflects the neutrons, a common reflector used is graphite, RBMK specific, which has 2 moderators, the graphite and water, water is a refractor, meaning it slows down the neutrons with water drag, it also does not release extra heat energy like the reflection process does, unlike modern western reactors, which only use water as their moderator, say the water all evaporates away, the chain reaction would actually stop instead of run away, now for thr RBMK core, thats the complete opposite, as the water inside the channels convert to steam(a void) the neutrons flying through those channels will no longer be refracted by the water inside those channels, itd be just the graphite surrounding those channels thatd allow the reaction to accelerate in that area, usually the automatic systems would detect that and lower the control rods in the channels needed to try and compensate for that, but the core being so large, having so many channels, and those channels being their own pressure/coolant vessel there would often be hot spot issues with the core, hopefully i will soon be able to simulate all of this, along with the meltdown/blowout/overheat mechanics as well as the "runaway reaction", just need an explosion that can send the lid off of the reactor and make it land sideways like the real one did and that will be one of the hardest parts is to simulate the cap jumping along with the massive 2 explosions, 1 being the upper biological shield flying off the reactor, and 2 the hydrogen explosion from the molten graphite and uranium fragments