Pressure Lab Cycling does not hide its maths. This page lists every source the pressure model is built on and every coefficient the engine actually uses — the numbers below are read straight out of the running code, so they can never drift from what the calculator does.
This page is maintained by BAM, the publisher of Pressure Lab Cycling. The model is our own derivation, calibrated against openly published cycling research. None of the researchers, magazines or manufacturers referenced below are affiliated with this app, and nothing here is an endorsement, a certification or a safety standard.
We describe the principles in our own words and fit our own coefficients to them; no tables or text from these publications are reproduced here.
The idea that a tire should sink a roughly fixed share of its own height under load — the “15% tire drop” convention popularised by Frank Berto — is what ties rider weight and casing width to a target pressure. We use the principle, not the published tables: our load and width exponents are fitted so pressure rises sub-linearly with load and falls steeply with width.
Bicycle Quarterly / Rene Herse CyclesIndependent roller and field testing shows that rolling losses stop falling and start rising once pressure passes a surface-dependent “breakpoint”, because vibration losses in the rider and bike overtake casing losses. That is the basis for the surface factors below, and for the mild rise in optimal pressure with speed.
Rolling resistance & impedance backgroundFront/rear static weight splits per bike type, and the relative rolling losses of race, training and reinforced casings with latex, butyl or tubeless setups, are widely published figures. We apply the weight split as a real wheel load rather than as a pressure fudge factor.
Tire testing methodology overviewEach wheel is solved on its own, from the load that wheel actually carries:
P = 5748 × L^0.3 / w^1.608
× (622 / D)^0.908
× surface factor
× (1 + 0.0026 × (speed − 17.5))
× casing coefficientL is the wheel load in kg, w the measured casing width in mm and D the rim diameter in mm. The result is in psi; bar is a straight unit conversion afterwards.
| Constant | Value | What it does |
|---|---|---|
| BASE_COEFFICIENT | 5748 | Overall scale of the model. Fitted so a reference road setup lands on a target tire drop rather than chosen to match any other calculator. |
| LOAD_EXPONENT | 0.3 | Pressure rises sub-linearly with wheel load: doubling the load does not double the required pressure, because the contact patch grows too. |
| WIDTH_EXPONENT | 1.608 | Pressure falls steeply with casing width, because air volume scales roughly with the square of width. |
| DIAMETER_EXPONENT | 0.908 | Scales pressure relative to a 622 mm rim. A smaller wheel spreads the same load over a shorter contact patch and needs slightly more pressure. |
| SPEED_SLOPE | 0.26 % / mph | Optimal pressure creeps up with speed — about 0.26% per mph — because suspension losses matter relatively less as aerodynamic and rolling loads grow. |
| SPEED_REFERENCE_MPH | 17.5 mph | The pace at which the speed term is neutral. Slower riding lowers the recommendation, faster riding raises it. |
| DROP_STIFFNESS | 5.5 | Contact-stiffness constant in the tire-drop model, calibrated so a 70 kg rider on a 28 mm casing at about 80 psi lands on a 15% drop. |
| DROP_WARN | 55 % | Casing compression, as a share of casing height, where pinch-flat risk starts to be flagged. |
| DROP_CRITICAL | 70 % | Casing compression where the tire is effectively bottoming out on the rim and the risk is flagged as extreme. |
Surface factor is the loss-optimal pressure relative to fresh tarmac (= 1.00). Impact is the peak-to-static load ratio used for the pinch-flat check — rougher ground means bigger single hits.
| Surface | Surface factor | Impact |
|---|---|---|
| Velodrome — indoor boards | 1.340 | 1.8× |
| Velodrome — outdoor concrete | 1.120 | 2.0× |
| Fresh tarmac | 1.000 | 2.4× |
| Aged tarmac, light cracking | 0.947 | 2.7× |
| Hardpack dirt / smooth gravel | 0.907 | 3.0× |
| Rough tarmac / chipseal | 0.869 | 3.2× |
| Loose gravel, small stones | 0.823 | 3.5× |
| Cobbles / pavé | 0.774 | 3.9× |
| Chunky gravel, washboard | 0.730 | 4.2× |
| Rocky, technical off-road | 0.668 | 4.6× |
A supple race casing is run at the reference pressure; stiffer and more heavily reinforced casings sit slightly lower because their own hysteresis losses are already higher.
| Casing & setup | Coefficient |
|---|---|
| Race casing — tubeless or latex | 1.00 |
| Training casing — tubeless or latex | 0.97 |
| Training casing — butyl tube | 0.94 |
| Reinforced / endurance casing | 0.91 |
Static share of total system weight carried by each wheel, applied as a real load before the pressure is solved.
| Bike type | Front | Rear |
|---|---|---|
| Balanced 50/50 — TT, tri, track | 50.0 % | 50.0 % |
| Road bike — 48/52 | 48.0 % | 52.0 % |
| Gravel bike — 47/53 | 47.0 % | 53.0 % |
| Mountain bike — 46.5/53.5 | 46.5 % | 53.5 % |
Tire drop is estimated from the wheel load, the pressure and the casing width, then mapped through a saturating curve so the modelled compression can never exceed the casing height — real casings stiffen progressively as they squash.
For the pinch check the static load on the more heavily loaded wheel is multiplied by the surface impact factor. If the resulting compression passes the warn threshold the result is flagged as increased risk; past the critical threshold it is flagged as extreme.
The recommended width is the narrowest casing, never narrower than the one you selected and capped at 65 mm, that clears the warn threshold at its own recommended pressure.
Deliberately outside the model:
Treat every result as a starting point to test and adjust from. Your rim and tire manufacturers' limits always win: if a recommendation sits above any stated maximum, lower the pressure or change the setup.