A top-down neutron-transport model of a bounded circular fuel pile. Each nucleus in the lattice is pre-assigned fissile (U-235) or non-fissile according to the enrichment fraction. A neutron that reaches a fissile nucleus fissions it with probability 0.85 if the neutron is thermal (slowed) or 0.55 if fast — moderation sets the odds a freshly-born neutron starts thermal — and a non-fissile nucleus mostly just absorbs it (2% stray fission). Each real fission releases a stochastically-rounded ν neutrons (so a fractional value like 2.43 still averages out correctly over many fissions), which fly off in random directions.
Two loss channels compete against that multiplication: a neutron that reaches the edge of the pile leaks out and is gone — raising fuel density shortens the average flight between nuclei and cuts this loss, which is the real critical-mass/geometry effect (a small or sparse pile is subcritical no matter how enriched it is) — and a neutron that flies into an inserted control rod is absorbed and removed outright, the classic reactivity-control mechanism. The live multiplication factor k is not set directly by any single slider; it emerges from the race between fission multiplication and these two absorption/leakage channels, tracked generation-over-generation by the fission-rate readout.
- Fuel density — nuclei packed into the fixed-radius pile. Sparser packing lets more neutrons escape before hitting anything (subcritical geometry).
- Enrichment — fraction of nuclei that are fissile U-235 versus inert absorber.
- Moderation — chance a new neutron is born thermal, which raises its own fission probability.
- ν (neutrons/fission) — new neutrons released per real fission event.
- Control-rod insertion — five vertical rods lowered from the top of the pile; any neutron inside the inserted depth is progressively absorbed and removed, the same way real rods throttle a reactor toward subcritical.