Why turbo sizing starts with mass flow, not with boost
Boost pressure on its own tells you almost nothing. Fifteen pounds through a 2.0 litre four at 7,000 rpm and fifteen pounds through a 5.7 litre V8 at 6,500 rpm are completely different demands on a compressor, and the same turbo cannot serve both. What a compressor is actually rated on is mass flow — pounds of air per minute — against pressure ratio, and every compressor map is drawn on those two axes.
So the sizing question is never “which turbo makes 15 psi?” It is “how many pounds per minute does my engine want at the pressure ratio I need, and does that point fall inside a decent efficiency island on this map, away from surge on the left and choke on the right?” This calculator produces the coordinates. The map tells you the rest.
Mass flow also connects directly to power, which is why it is the right currency. Burning a pound of air releases a fairly predictable amount of energy regardless of how the air got there, so a gasoline engine makes roughly 9.5 to 10.5 horsepower per pound per minute of air. That coefficient is what lets a compressor map be read as a horsepower ceiling.
From displacement to pounds per minute
The calculation has three stages: volume, density, and then mass.
Volume. A four-stroke engine draws one swept volume per cylinder every two revolutions, so the volumetric flow rate at manifold conditions is displacement × rpm × VE, divided by 3,456. That constant is 1,728 cubic inches per cubic foot multiplied by the two revolutions per cycle. Volumetric efficiency accounts for the fact that the cylinder never fills perfectly at manifold density — it depends on cam timing, port flow and induction tuning, and it is the least certain input on this page.
Density. Apply the ideal gas law in the engineering units the industry uses: ρ = 144P / (53.35T), with pressure in psia, temperature in degrees Rankine (°F + 459.67), and 53.35 the specific gas constant for air in ft·lbf per lbm per °R. The 144 converts pounds per square inch into pounds per square foot. Check it at standard conditions: 144 × 14.7 ÷ (53.35 × 519.67) = 0.07635 lb/cu ft, which is the familiar sea-level figure.
Notice that density goes as P/T. That is the whole reason intercooling matters: pressure and temperature pull in opposite directions, and a compressor that raises pressure also raises temperature, so some of the boost you paid for is given straight back as heat unless you remove it.
Mass. Multiply volumetric flow by density and you have pounds per minute. Divide the absolute manifold pressure by ambient pressure and you have the pressure ratio. Those two numbers are the point you plot on the map.
Discharge temperature. An ideal compressor would raise temperature by a factor of PR(γ−1)/γ, which for air is PR0.2857. A real one is less efficient, and all of the shortfall appears as extra heat, so T2 = T1[1 + (PR0.2857 − 1)/η]. This is the temperature entering the intercooler, and the gap between it and your manifold temperature is what the intercooler has to remove.
Worked example: 350 cu in at 6,500 rpm and 15 psi
A 350 cubic inch V8 at 6,500 rpm with 95% volumetric efficiency, running 15 psi of boost with the charge intercooled to 120 °F, at sea level.
- Volumetric flow. 350 × 6,500 × 0.95 ÷ 3,456 = 2,161,250 ÷ 3,456 = 625.36 CFM.
- Absolute manifold pressure. 14.7 + 15 = 29.7 psia.
- Absolute charge temperature. 120 + 459.67 = 579.67 °R.
- Charge density. 144 × 29.7 ÷ (53.35 × 579.67) = 4,276.8 ÷ 30,924.4 = 0.13829 lb/cu ft.
- Mass airflow. 625.36 × 0.13829 = 86.48 lb/min, which is 0.6538 kg/s.
- Pressure ratio. 29.7 ÷ 14.7 = 2.020.
- Supported power. 86.48 × 10 = 865 hp.
So the map coordinates are 86.48 lb/min at a pressure ratio of 2.02. Now check the heat. From a 90 °F compressor inlet at 72% isentropic efficiency: 2.0200.2857 = 1.2226, so the temperature rise factor is (1.2226 − 1) ÷ 0.72 = 0.3091, and the discharge temperature is 549.67 × 1.3091 = 719.6 °R = 259.9 °F. The intercooler has to shed 140 °F to deliver the 120 °F charge the density calculation assumed.
What happens if it does not? Suppose the intercooler only manages 180 °F at the manifold. Density becomes 144 × 29.7 ÷ (53.35 × 639.67) = 0.12532 lb/cu ft, mass flow falls to 78.37 lb/min, and supported power drops to 784 hp. Twenty-one degrees of extra manifold temperature per hundred is worth roughly 9.4% of the power — 60 °R on 639.67 °R — before any detonation consideration at all.
Taking your two numbers to a compressor map
Plot mass flow on the horizontal axis and pressure ratio on the vertical, and look at where the point lands.
Too far left and you are near or beyond the surge line, where flow through the compressor becomes unstable and reverses periodically. Surge is audible, damaging, and usually means the compressor is too large for the engine at that operating point — the classic case of a big turbo on a small engine at low rpm.
Too far right and you are approaching choke, where the compressor wheel simply cannot pass more air and efficiency collapses. The compressor is too small; boost will fall away at the top of the rev range no matter what the wastegate does.
Inside a high efficiency island is what you want, and it is worth checking more than one point. Run this calculator at peak torque rpm and at peak power rpm, plot both, and check that the whole line between them stays in reasonable territory. A turbo that is perfect at 6,500 rpm and surging at 3,000 rpm is a bad match for a road car.
Efficiency matters directly, not just as a label. Every point of isentropic efficiency you lose appears as charge heat, which reduces density, which reduces the mass flow the same boost delivers. That is why the discharge temperature output on this page is worth reading alongside the flow figure.
Once the air side is settled, size the fuel side to match. Convert the airflow into fuel demand at your target lambda with the air fuel ratio and lambda calculator, then check the injectors have headroom using the injector duty cycle calculator. And confirm the cylinder head can pass the air the compressor is delivering, because a restrictive head caps volumetric efficiency and quietly makes the whole calculation optimistic — the cylinder head airflow calculator sets that ceiling.
Mass airflow and supported power against boost
| Boost (psi) | Absolute (psia) | Pressure ratio | Charge density (lb/cu ft) | Mass airflow (lb/min) | Supported hp at 10 hp/lb·min |
|---|---|---|---|---|---|
| 0 | 14.7 | 1.000 | 0.06845 | 42.81 | 428 |
| 5 | 19.7 | 1.340 | 0.09173 | 57.36 | 574 |
| 10 | 24.7 | 1.680 | 0.11501 | 71.92 | 719 |
| 15 | 29.7 | 2.020 | 0.13829 | 86.48 | 865 |
| 20 | 34.7 | 2.361 | 0.16158 | 101.04 | 1,010 |
| 25 | 39.7 | 2.701 | 0.18486 | 115.60 | 1,156 |
Charge temperature is deliberately held at 120 °F down the whole table so the effect of pressure alone is visible. A real installation gets hotter as pressure ratio rises, so the upper rows are only reachable if the intercooler grows with the boost.
Volumetric efficiency under boost means the same thing it always did
VE is measured against manifold conditions, not against atmospheric conditions, so a turbocharged engine does not have a VE of 200% because it runs two bar. It has whatever VE its cam, ports and exhaust give it — typically 85 to 100% for a well-developed engine near peak torque — and the boost appears in the calculation through density, not through VE. Entering an inflated VE to represent boost double-counts the pressure and roughly doubles the airflow answer. If you have measured VE from a mass airflow sensor and manifold pressure log, use that; otherwise 90 to 95% is a reasonable assumption for a modern four-valve engine and 80 to 90% for an older two-valve one.
Assumptions and where they break
- The horsepower coefficient is empirical. Around 10 hp per lb/min holds for gasoline at a sensible power mixture and reasonable ignition timing. It rises with alcohol fuels and falls with poor combustion phasing, and it says nothing about whether the engine can survive that power.
- Air is treated as an ideal gas. At the pressures and temperatures involved this is accurate to a fraction of a percent, and it is not the source of any meaningful error here.
- Humidity is ignored. Water vapour displaces air and reduces density slightly; the density altitude calculator quantifies it properly if you need to.
- One operating point is not a turbo match. Run peak torque and peak power separately and check both against the map. Transient response, turbine housing sizing and manifold design are decided by considerations this calculation does not touch.
- Pressure drop through the intercooler and piping is not modelled. The compressor has to produce more pressure than the manifold sees, so its real pressure ratio is higher than the one calculated here — commonly by one to three psi on a well-built system.
Working backwards from a horsepower target
Most people arrive with a power number rather than a boost number, and the calculation runs just as well in reverse. Divide the target horsepower by the coefficient to get the mass flow you need: 800 hp at 10 hp per lb/min needs 80 lb/min. Then adjust boost on this page until the mass flow output reaches it, and read off the pressure ratio.
Doing it that way exposes the levers clearly. More displacement, more rpm or better volumetric efficiency all raise mass flow at the same pressure ratio, which keeps the compressor in a happier place and puts less heat into the charge. Boost is the lever you reach for last, not first, because it is the one that costs the most in charge temperature and in cylinder pressure.
It also shows why intercooling is worth as much as boost. The worked example loses 8.11 lb/min — 9.4% of its airflow — when the manifold temperature rises from 120 to 180 °F at unchanged boost. Recovering that with pressure would take about 3.1 psi more boost — 29.7 × 639.67/579.67 = 32.8 psia, along with the extra heat and cylinder pressure that comes with it. Fixing the intercooler is the cheaper and safer route every time.
Finally, remember the engine has to survive what you are planning. Dynamic compression ratio and boost interact directly, so check the geometry with the dynamic compression ratio calculator before settling on a boost target, and confirm the whole combination against what a naturally aspirated version of the same engine would make with the boost horsepower gain calculator.