nx_thermal_process.nx

nx_thermal_process.nx -- FOOD-SCIENCE SUITE / THERMAL PROCESS rung. The heat side of food safety: how much lethality a time-temperature hold actually delivers, whether a proposed process meets the legal pasteurization schedule, and -- the decision that governs everything -- whether a food needs a botulinum cook at all. This completes a triangle the ecosystem already had two corners of. nx_ferment_safety refuses unsafe FERMENTS on the pH 4.6 botulinum line. nx_chem_solution owns that line as chemistry (CS_HIGH_ACID_PH_MILLI). Neither could say anything about HEAT, so a mix that is pasteurised rather than acidified had no safety path -- and ice cream mix is exactly that case, since 21 CFR 135.110 requires the mix to be pasteurised. ACID AND HEAT ARE ALTERNATIVE BARRIERS, NOT A CHECKLIST. That is the whole logic of tp_barrier_required, and it is why this imports nx_chem_solution instead of restating 4.6: pH < 4.6 -> the ACID barrier holds; C. botulinum cannot grow; pasteurisation suffices. pH >= 4.6, shelf-stable -> nothing but heat stands between the product and botulism -> full 12D cook. pH >= 4.6, kept cold -> pasteurise AND hold the cold chain; refrigeration is the second barrier. Ice cream mix is the third case. Treating it as the second would be a pointless retort; treating it as the first would be dangerous. WHAT IS ASSERTED VS WHAT IS DERIVED -- read this before trusting a number. The only organism constant here is C. botulinum proteolytic spores, D(121.1 C) = 0.21 min, z = 10 C. That pair is canonical, universally published, and is the entire basis of the 12D cook and of F0. I did NOT bake in D/z values for the pasteurisation organisms, because I could not verify them to the precision a safety calculation deserves. Instead the z-value for pasteurisation is DERIVED FROM THE LAW ITSELF. The PMO gives two schedules it considers equivalent -- 145 F/30 min and 161 F/15 s -- and two equivalent schedules uniquely determine a z: z = (T2 - T1) / log10(t1/t2) tp_implied_z_milli computes it and gets 4.279 C (exact arithmetic on the pair gives 4.275, so the integer path is good to 0.1%). That sits just BELOW the commonly quoted 4.3-5.0 C band for the milk-pasteurisation design organism -- close agreement, but stated precisely rather than rounded up