Key takeaways
Power is torque multiplied by rpm, so the two curves are two views of the same measurement. The peak number is a single point; how a car feels and accelerates depends on the whole curve, particularly the area beneath it across the rpm range you actually use. Always check the axes, the smoothing and the correction standard before trusting a chart.
The two curves are one measurement#
A dyno chart is a compact summary of an engine's behaviour, but it is easy to read the wrong thing from it. Headline numbers get quoted, curves get compared by eye, and the details that decide whether a comparison is fair are often not shown. This article covers how to read a chart properly. The dyno measures torque and engine or wheel speed. Power is calculated from them: power equals torque multiplied by rotational speed. In imperial units, horsepower equals torque in lb-ft multiplied by rpm, divided by 5,252. In metric units, power in kW equals torque in newton metres multiplied by rpm, divided by about 9,549. The 5,252 figure is a useful sanity check. When a chart plots horsepower and lb-ft on the same numeric axis, the two lines must cross at 5,252 rpm. If they cross somewhere else, the traces are plotted in different units or on different scales, or the data has been altered. It also follows that peak power always occurs at a higher rpm than peak torque, because power keeps rising until torque falls faster than rpm rises.
Why the area under the curve matters#
A car accelerates broadly in proportion to the torque delivered to the wheels, which is engine torque multiplied by the overall gear ratio. Because you can change gear, what matters is the torque and power available across the whole band of rpm you use, not just at the top. A broad curve, with strong output across a wide range, lets you stay near the strong part of the curve between gear changes. Imagine two engines with the same peak power. One reaches it with a narrow, sharp peak and falls away on both sides. The other produces nearly the same figure across a wide range. On a chart the peaks match, but the second engine has far more area under its curve, so it accelerates harder for more of the time and is easier to drive. This is also why gearing matters: shorter gearing keeps the engine in the strong part of its curve for more of a given speed range. For conversion, 1 lb-ft is about 1.356 newton metres. The average power between gear changes tells you more about real performance than the maximum.
Spool, plateau and drivability#
On a turbocharged engine the early part of the torque curve shows spool: the rpm at which the turbocharger builds enough boost for torque to climb. A turbo that needs a lot of exhaust flow before it builds boost may show very little torque at low rpm, then a steep rise. That steep rise can look impressive on a chart but can be awkward to drive, because power arrives abruptly. After the climb, a well-controlled boost curve produces a plateau where torque stays roughly level before easing off at high rpm. A flat, predictable plateau is generally easier to use than a peaky curve. Torque falling steeply right after the peak is also worth noticing: it shows where the engine runs out of breath, and where changing up makes sense.
Tip
How to read a chart quickly
Look first at the rpm range where torque stays close to its peak value. The wider that band, and the lower it starts, the more usable the engine is. Then look at where power peaks and how much of the curve sits close to that figure.
Ways a chart can mislead#
None of these require bad intent. Most happen through default software settings or through comparing charts made under different conditions. The table lists the main ones, and the checklist after it shows what must match before a before-and-after overlay can be trusted.
| Issue | What happens | What to check |
|---|---|---|
| Smoothing | Heavy smoothing rounds off real dips and spikes, which can hide a boost drop or misfire | Whether the trace is raw or smoothed, and by how much |
| Axis scaling | A truncated or stretched vertical axis makes small differences look large | Whether the axis starts at zero and uses the same scale on both charts |
| Mixed corrections | Corrected and uncorrected runs, or runs using different correction standards, are overlaid | The correction standard and conditions listed for each run |
| Mixed measuring points | Crank, hub and roller figures are compared as if equivalent | Where each figure was measured |
| Different gears or sweep rates | Changes the result and the shape of the curve | Gear and test method for each run |
| Peak-only reporting | A single best run is quoted without the shape of the curve | Whether several runs are shown, and the spread between them |
- The same dyno, on the same day or with the same stated correction standard.
- The same gear, sweep rate and start and end rpm.
- The same smoothing settings and the same axis scales on both curves.
- The same measuring point (crank, hub or roller) for both runs.
- More than one run behind each curve, so you can see the spread.
What this means in practice#
Read the chart as a shape, not a number. Ask how wide the useful band is, how smooth and predictable the build-up is, and whether the comparison is like for like. Be wary of any single peak figure shown without its conditions. For background on how the numbers are produced in the first place, see the earlier article on how hub dynos measure power. The dyno test cell page covers the testing setup.
