Method: Kamlet-Jacobs (1968)
D = 1.01·√φ·(1 + 1.30ρ0) and P = 15.58·ρ02·φ, where φ = N·√M·√Q is the entire chemistry-dependent content of the method and ρ0 is the actual loading density (not theoretical maximum, unless the charge is void-free). N, M and Q come from an assumed decomposition hierarchy applied to the pooled elemental formula of whatever you specify: all nitrogen to N2, then hydrogen to H2O, remaining oxygen to CO2, and any leftover carbon as solid graphite. Against experiment for neat CHNO explosives the method runs about 1% average absolute error (Politzer & Murray, Propellants Explos. Pyrotech. 2019); individual compounds can deviate more, and the self-test below shows exactly where.
Strict CHNO only. The decomposition hierarchy has no branch for metals, chlorine, or fluorine: a metal fuel burns after the detonation front rather than in it; a fluoropolymer forms HF preferentially over H2O, removing hydrogen from the term the method is built on. This tool refuses rather than extrapolating. Of the 34 published PBX/PBXN/LX formulations on the Thermal Analysis page, 20 contain aluminium, ammonium perchlorate, Viton, or Kel-F and cannot be evaluated by this method at all.
Of the remaining 14, none can be fully evaluated either, and that is a real finding rather than a gap in this tool. Every one of them contains a polymeric binder or plasticizer (Estane, HTPB, CAB, BDNPA/F) with no citable condensed-phase heat of formation. One commonly repeated HTPB value traces to a single survey's own group-additivity estimate, not a measurement, so it is excluded rather than laundered in. This tool computes the 10 neat energetics and oxidizers below, and mixtures among them, and lets you add your own component with your own sourced data. It will not guess a binder's contribution on your behalf.
1. Compute
2. Result
3. The trade-off: performance is not free
HMX at its own theoretical maximum density is the ceiling nothing in the formulation can exceed. Every formulated explosive sits below it, not because the chemistry got worse but because binder buys handling safety, mechanical integrity, and castability that a neat crystal does not have.
| Material | Predicted D | Measured D (published) | What the gap buys |
|---|---|---|---|
| Neat HMX (TMD 1.902) | 9.12 km/s (this tool) | — | The ceiling: 100% energetic filler, no binder, not a real handleable article |
| PBX 9011 (90% HMX / 10% Estane) | not computable: no citable Estane ΔHf | ~8.8 km/s† | Castable, machinable, far less sensitive to impact |
| PBX 9502 (95% TATB / 5% Kel-F 800) | not computable: contains fluorine | ~7.7 km/s† | Among the most insensitive HE formulations fielded, at a real velocity cost vs. an HMX system |
† Published measured values, LASL Explosive Property Data (Gibbs & Popolato, 1980) and the LLNL Explosives Handbook (Dobratz & Crawford, UCRL-52997). These are measured numbers for the formulated article, shown beside a Kamlet-Jacobs PREDICTION for the neat filler only. The two are not computed by the same method, and the comparison is illustrative of the trade-off rather than a validation of the formulated value.
4. Does the implementation actually work?
Acceptance criteria are fixed before the result: reproduce accepted experimental detonation velocity and pressure for HMX, RDX, TNT and PETN to within the method's own documented scatter, reproduce Kamlet's own published RDX/TNT 77/23 mixture value, and correctly refuse a non-CHNO input and a missing-heat-of-formation input rather than silently computing something.
Related tools
See how binder choice changes what a DSC, TGA or DMA trace looks like, and load any of the 34 published PBX/PBXN/LX formulations, on Thermal Analysis. Turn multi-rate TGA or DSC runs into activation energy with an explicit multi-step test on the Decomposition Kinetics tool. Every term here is defined in the Glossary.