WHP vs Crank HP: Reading Dyno Numbers

What WHP and crank HP mean

Crank horsepower, sometimes written as flywheel horsepower, is the power an engine produces measured right at the crankshaft, before it passes through the transmission, driveshaft, differential, and wheels. This is the figure manufacturers usually advertise on a spec sheet.

Wheel horsepower, or WHP, is what a chassis dyno measures: power delivered at the wheels, after it has already passed through the entire drivetrain. See WHP in the glossary for how Moddo's own numbers use this term.

The two numbers are never equal on the same car, because the drivetrain itself absorbs some power turning its own gears, shafts, and bearings before any of it reaches the wheels. Neither number is more real than the other; they simply describe power at two different points in the same mechanical path.

What a dyno actually does

A chassis dyno holds the car's drive wheels on a set of rollers and measures power while the engine accelerates or pushes against those rollers, without the car actually moving down a road. This lets a shop measure power repeatedly, in a controlled space, without the variables a real road introduces, like wind, traffic, and changing grade.

The car is strapped down, the transmission is put in a gear (commonly the gear closest to a 1:1 ratio, since that generally minimizes drivetrain losses from gear reduction), and the engine is run through its rev range at wide-open throttle while the dyno's sensors record the result. The output is a graph, not just one number: horsepower and torque plotted against engine speed or road speed for the length of the pull.

A dyno session is where a tune is normally verified, since it gives a controlled, repeatable place to load the engine while watching air-fuel ratio, knock, and boost the whole time. See What Does an ECU Tune Do? for how those channels relate to what the tune is actually doing during a pull.

A dyno pull is a snapshot of wide-open-throttle performance, not a complete picture of how a car drives. It says little about part-throttle response, how the car behaves at low RPM in traffic, or how it handles, which is why a dyno sheet is one tool among several a tuner or owner uses to judge a car, not the whole judgment on its own.

Why the gap between WHP and crank HP exists

Every mechanical connection between the engine and the wheels loses some power to friction and to spinning the mass of the drivetrain itself. A manual transmission and a single differential typically lose less than a setup with more rotating assemblies, like an all-wheel-drive car with a transfer case and two differentials.

This loss is commonly discussed as a range, often cited loosely as somewhere in the area of 10 to 20 percent of crank power for a typical rear- or front-wheel-drive car, with all-wheel-drive cars often losing more. This is a general range for orientation, not a number to calculate a specific car's crank output from, since actual drivetrain loss varies by vehicle, gear oil temperature, and even how hard the car is driven on the dyno.

Automatic transmissions with a torque converter add another variable, since the converter itself can slip under load in a way a locked manual clutch does not, which is part of why loss percentages are given as a range rather than a fixed number even within one drivetrain layout.

Why this matters when comparing cars or claims

A WHP number by itself does not say whether a car's engine is strong or its drivetrain is efficient; it is the product of both. Two cars with the same crank output can post different WHP numbers purely because one has a lighter-loss drivetrain, and two cars with the same WHP can have meaningfully different crank output for the same reason in reverse.

This is why a WHP figure quoted casually, without knowing the drivetrain layout, the dyno type, and the correction factor behind it, tells you less than it sounds like it does. Reading a dyno sheet correctly means reading all of that context, not just the peak number on the page.

Drivetrain layout and where the losses come from

The path power takes from the crankshaft to the wheels differs by drivetrain layout, and each stage in that path is a place some power is lost to friction or to spinning mass. A front-wheel-drive or rear-wheel-drive car generally has the engine, a transmission, and one differential in the path; an all-wheel-drive car adds a transfer case and a second differential, each with its own gears, bearings, and seals turning under load.

More rotating parts generally means more loss, which is the main reason all-wheel-drive cars are often discussed as losing a larger share of crank power than two-wheel-drive cars. It is a reasonable generalization, not a fixed rule; the actual loss on a specific car depends on its particular gear ratios, bearing design, and fluid, not just how many driven wheels it has.

None of this makes an all-wheel-drive car's WHP number less valid; it is still an accurate measurement of power at the wheels for that car. It simply means a WHP comparison between a two-wheel-drive car and an all-wheel-drive car says less about the two engines than it looks like it does, since the drivetrains behind them are not doing the same amount of work.

Some all-wheel-drive dynos measure all four wheels at once for this reason, using rollers under both axles rather than just the driven axle a two-wheel-drive car would use. A car tested on the wrong type of dyno for its drivetrain, or with power sent to wheels the dyno is not measuring, produces a reading that understates its actual output, which is a setup issue rather than anything to do with the car's real power.

Why dyno numbers are not comparable across dynos

Different dyno designs measure power differently, even on the identical car on the same day. An inertia dyno, common in shops, calculates power from how quickly it accelerates a known roller mass, which reads differently from a load-bearing dyno that applies a controlled resistance and measures torque against it directly.

Neither type is more "correct" than the other; they are different measurement methods that tend to produce different absolute numbers for the same car. This is the single most common reason two dyno sheets for what should be the same car disagree, and it has nothing to do with the car itself.

Even within the same dyno brand, individual units can read slightly differently from each other due to calibration, roller condition, and setup, which is a smaller effect than the difference between dyno types but still enough to matter when comparing numbers closely.

None of this makes any of the dynos wrong. Each is doing what it is designed to do, measuring power by its own method, consistently, on its own rollers. The mistake is not in what any single dyno reports; it is in treating numbers from two different dynos as if they came from the same measurement.

Why before-and-after comparisons need the same dyno

Because different dyno types and even different dynos of the same type can read a car differently, the only way to reliably measure a change from a part or tune is to run the car on the exact same dyno, in a similar state of tune-in and similar weather, before and after the change. Comparing a before number from one shop's dyno to an after number from a different shop's dyno mixes the change being measured with the difference between the dynos themselves.

This also means a WHP number from one dyno sheet should not be compared directly to a WHP number from another car's dyno sheet at a different shop as if they were measured the same way, even if both cars are the same model.

• Use the same physical dyno for the before and after pull, not just the same brand of dyno.
• Keep the transmission in the same gear for both pulls.
• Note the correction factor used (SAE or STD) and confirm both pulls used the same one.
• Run the pull under similar weather and engine temperature conditions where possible.

Reading beyond the peak number

A dyno operator or shop typically reports a single peak whp and a single peak torque figure as headline numbers, but the sheet itself is a full curve. Reading only the peak throws away most of the information the pull actually captured.

A car that is being evaluated for how it will actually feel and perform, in a straight-line pull or on a track, benefits from looking at the curve's shape as much as its highest point. Two cars with an identical peak WHP number can produce very different results in the real world if one holds strong torque across a wide RPM range and the other spikes narrowly.

This matters most in situations where the engine spends time away from its power peak: exiting a slow corner, pulling out of traffic, or accelerating from partial throttle rather than a dead stop at full throttle. A car with a wide, flat torque curve stays closer to its strongest output across more of these situations than one with a narrow spike, even if the spike itself reaches a higher peak number.

What the dyno operator controls, and why it matters

A dyno reading is not purely a property of the car; how the pull is set up also affects the result. Tie-down tension, tire pressure on the rollers, whether a cooling fan is placed in front of the car to simulate airflow, and how the operator ramps through the rev range can all shift the number somewhat, which is part of why the same car can read slightly differently between visits even at the same shop.

A consistent, experienced operator reduces this variability but cannot eliminate it entirely, since some of it comes from the car itself: tire temperature building through a pull, or a fuel tank that is nearly empty versus full, changing the car's weight slightly on an inertia-based dyno. None of this means dyno numbers are unreliable; it means small differences between two close numbers are often not meaningful, while a large, consistent difference across repeated pulls usually is.

• Tie-down tension and wheel alignment on the rollers
• Tire pressure and tire temperature during the pull
• Whether a cooling fan is used and how it is positioned
• The ramp rate: how quickly the pull moves through the rev range
• Ambient temperature, humidity, and barometric pressure on the day

Correction factors: SAE and STD

Air density affects how much power an engine makes, since colder, denser air lets an engine produce more power for the same throttle input than hot, thin air does. Dynos apply a correction factor to normalize a run to a standard set of atmospheric conditions so that results from different days and temperatures can be compared.

The two correction standards commonly seen are SAE (the current automotive industry standard) and STD, an older standard with more generous reference conditions that usually reads higher than SAE from the identical raw run. An uncorrected number is a third thing again: the raw reading on that day, in that weather. A dyno sheet should state which correction was applied; comparing an SAE-corrected number to an STD-corrected number treats two different calculations as if they were the same measurement.

Correction factors also move by a larger amount on days far from the reference conditions, meaning the gap between a car's raw and corrected number is not fixed; it depends on how hot, humid, or high-altitude the day of the pull actually was. A correction factor close to 1.0 means the day was close to the reference conditions; a factor further from 1.0 in either direction means more of the reported number is the correction doing work rather than the engine.

Torque curves, area under the curve, and the horsepower formula

A single peak horsepower or peak torque number describes only one point on the run, not how the engine performed across the whole rev range. A dyno sheet's real value is the full curve: where torque peaks, how it holds or falls off before redline, and how horsepower builds as a result.

Horsepower and torque are related by a fixed formula: horsepower = torque (in pound-feet) x engine speed (in RPM), divided by 5252. Because 5252 is the constant that formula uses, a torque curve in pound-feet and a horsepower curve plotted on that same numeric scale always cross at exactly 5252 RPM, the one point where the two units happen to read the same number; it is an artifact of that unit combination and the shared scale, not a physical event in the engine. Separately, horsepower keeps climbing as RPM rises even where torque is roughly flat, since power is torque multiplied by RPM; it peaks and starts to fall once torque is falling faster than RPM is rising, and that trade-off plays out across the whole curve, not just at the 5252 RPM crossing point.

A broader torque curve, meaning strong torque across a wider RPM band rather than a narrow spike, generally makes a car feel stronger and often performs better in real driving and acceleration than a car with a slightly higher peak number but a narrower curve. This is sometimes summarized as the area under the curve: the total power delivered across the pull, not just its highest point. Two cars with the same peak WHP number can have meaningfully different area under the curve, and therefore feel and perform differently.

Drivetrain loss is not a fixed percentage

Because drivetrain loss varies with vehicle layout, gear oil viscosity and temperature, tire pressure on the rollers, and even how the car is driven on the dyno, there is no single formula that converts a WHP number to a crank number reliably. A loss estimate is only ever a general range, useful for orientation, not for back-calculating a precise crank figure from a WHP reading.

Loss also is not constant across the rev range on most cars; it tends to be a larger percentage of a small torque figure at low RPM than of a large one at high RPM, since some drivetrain losses (like bearing drag and seal friction) are closer to fixed values rather than scaling with power. This is one more reason a single percentage figure oversimplifies what is actually a curve of its own.

This is also why estimating a car's crank output backward from a WHP number, by adding an assumed loss percentage, is treated as a rough estimate at best by most tuners rather than a real figure. It can be useful for a ballpark comparison to a manufacturer's claimed crank rating, but it should not be presented as a measured number, since nothing about that arithmetic was actually measured on the car.

The only way to measure crank power directly is with an engine dyno, which holds the engine on a stand, outside the car, and measures it before the drivetrain is even attached. This is common in engine development and racing but rare for a street car, since it means removing the engine, which is why chassis dyno WHP figures are what most owners and shops actually work with day to day.

Reading a dyno sheet with a tune in mind

When comparing dyno runs to evaluate a tune, the useful checks go beyond the peak number: whether the correction factor and dyno were the same as the baseline run, whether the AFR and knock activity logged during the pull stayed in a reasonable range across the whole curve, and whether boost held its target instead of tapering off. See Datalogging: What to Log and Why for the channels that matter alongside the dyno's own output trace.

A dyno operator repeating a pull two or three times back to back and confirming the results are consistent is a stronger basis for trusting a number than a single pull, since a one-off pull can be affected by a momentary condition, like a brief cooling fan cycle or a slightly different starting engine temperature, that will not repeat reliably.

Altitude, weather and why some days need more correction

Barometric pressure drops with altitude, which means a car dyno-tested at a high-elevation shop is making less actual power on that day than the identical car at sea level, purely because there is less oxygen in a given volume of air. The correction factor exists specifically to account for this, scaling a high-altitude or hot-weather run up to what it would have read under the reference conditions.

This means a raw, uncorrected number from a high-altitude dyno session looks lower than the same car would read at sea level, and the corrected number is what makes the two comparable. It also means the size of the correction itself carries information: a run needing a large correction was further from the reference day than one needing almost none, which is part of why some tuners note the correction factor's value, not just which standard was used, when judging how much to trust a given number.