Fixed physical constants used: ethanol stoich AFR 9.01, methanol stoich AFR 6.47, ethanol density 0.789 g/cm³, methanol density 0.794 g/cm³, ethanol molar mass 46.07 g/mol, methanol molar mass 32.04 g/mol. Gasoline's effective molar mass isn't a fixed physical constant (gasoline is a mixture, not one molecule) — literature cites a "typical" range of roughly 100–120 g/mol depending on composition, which is why it's left editable. RON blending method: Anderson, Kramer, Mueller & Wallington, "Octane Numbers of Ethanol- and Methanol-Gasoline Blends Estimated from Molar Concentrations," Energy & Fuels 24 (2010); corroborated by Pearson, Turner et al., "Iso-stoichiometric fuel blends," IMechE Proceedings (2015). AFR method: standard stoichiometric mass-balance across fuel components. Ethanol-sensor model: fit to a user-supplied 21-point measured sweep on a real Continental flex-fuel sensor (99 RON E7 base, methanol added 0→400ml into 200ml base fuel, quadratic regression R²=0.9998). The methanol-only component is isolated from that sweep and the actual ethanol% is added back in (scaled by the sensitivity coefficient) to generalize beyond the single tested dilution line. The whole curve is then rescaled by a factor derived live from the "Methanol full-scale %" field so the signal hits exactly 100% (5V / 150Hz) at that methanol level with the E7 base ethanol included — the raw regression's own crossing point was closer to ~65%, so this trades a little mid-range fit accuracy (results run a few points low around 40–65% methanol) for an exact, adjustable saturation point matching the sensor's real electrical limit. Treat results as most reliable near the E7 base/tested range and more approximate far outside it. Ethanol-equivalent output is scaled linearly to a 0–5V analog signal (E0→E100); the methanol-scaled output uses the same underlying signal on a 50–150 Hz / 0–68% methanol scale. Adjust the fields above if your specific sensor/interface differs. Fuel cost is a simple volume-weighted average of the three £/L prices you enter. The consumption multiplier is the ratio of base gasoline AFR to blend AFR (gasAFR ÷ AFR_blend, both already computed above) — a richer blend needs more fuel mass per unit air to hold stoichiometric, so at equal airflow the fuel volume burned scales with that same ratio (e.g. E85 ≈1.49×, M50 ≈1.41× with the default constants). It doesn't separately correct for the blend's own density, and it doesn't credit any efficiency gain a properly optimized engine can get from methanol/ethanol's charge cooling and extra knock margin, so treat it as a fuel-volume estimate, not a range estimate. Cost per mile converts your entered imperial MPG to miles/litre (MPG ÷ 4.54), then divides the consumption-corrected cost per litre (blended £/L × consumption multiplier) by that figure — the gasoline reference column uses the same MPG with no multiplier applied, so the comparison isolates the effect of the blend itself rather than any change in driving style.