WHP to HP Calculator

Two horsepowers: what the crank makes and what the wheels see

Advertised horsepower is an engine number. It is measured at the crankshaft on an engine dyno, with the gearbox, driveshaft and differentials out of the loop, and it is the figure printed on the window sticker. The EPA collects those rated horsepower figures for its light duty vehicle trends report, which is why a 300 hp sedan means 300 hp before the drivetrain.

Wheel horsepower, or whp, is what survives the trip. A chassis dyno straps the car down, spins the rollers through the driven wheels and reads the power that made it past every bearing, gear mesh and clutch plate. That reading is always lower, and the gap is the drivetrain loss.

Both numbers use the same unit. Mechanical horsepower is 550 foot pounds per second, which works out to 745.7 watts, per the NIST SI conversion tables. So a 353 hp crank rating is about 263 kW, and the 300 whp version is about 224 kW.

The formula, and why it divides instead of adds

The drivetrain keeps a fraction of the crank power and turns the rest into heat. Call the share it keeps k, where k = 1 minus the loss. Then whp = crank hp x k, and the direction people ask for most is the reverse: crank hp = whp / k.

The division matters. On a rear wheel drive car at a 15 percent loss, 300 whp means 300 / 0.85 = 352.9 hp at the crank, not 300 plus 45 = 345. Adding 15 percent of the wheel figure understates the crank number, because the loss is a share of the bigger crank value, not of the smaller wheel value. The bigger the loss, the wider the two answers drift: at a 25 percent loss, dividing gives 300 / 0.75 = 400 hp while adding gives 375 hp, a 25 hp gap.

The reverse direction is the one to sanity check before a build. A 400 hp engine in an all wheel drive chassis at a 22 percent loss puts down 400 x 0.78 = 312 whp, so a dyno sheet reading 312 is exactly on plan, and one reading 360 means the loss estimate was optimistic.

Drivetrain loss by layout: FWD, RWD, AWD and 4WD

The loss depends on how many parts sit between the crank and the contact patch. A front wheel drive car runs a compact transaxle right behind the engine, so the power path is short. A rear wheel drive car adds a long driveshaft and a separate rear differential. An all wheel drive car adds a transfer case or center differential and a second driven axle on top of that.

Manuals generally beat automatics on loss, and modern lockup torque converters and 8 to 10 speed gearboxes have pulled automatic losses down from where 1980s three speeds sat. Tires matter too: a chassis dyno reading includes tire flex on the rollers, which can be worth a couple of percent on a soft sidewall.

Use the middle of the range as the estimate and the edges as the honest band. If a seller claims a loss under 8 percent on a street car, or over 30 percent without a good story, treat both as a reason to ask questions.

Worked example: a 300 whp rear wheel drive car

A coupe straps onto a chassis dyno and the operator reads 300 whp at the wheels, corrected to standard conditions. The car is rear wheel drive with an automatic transmission, so the working loss is 15 percent and the share kept is 0.85.

Crank hp = 300 / 0.85 = 352.9, so the engine is making about 353 hp. The drivetrain is eating 352.9 - 300 = 52.9 hp, which leaves as heat in the transmission and differential, enough to warm a case you would not rest a hand on after a few pulls. In other units, 300 whp is 224 kW at the wheels and 353 hp is about 358 PS on the metric scale used by many overseas spec sheets.

The band matters as much as the point estimate. A 15 percent assumption with a realistic 14 to 18 percent spread gives a crank range of 300 / 0.82 = 366 down to 300 / 0.86 = 349. Quote the number as 353 hp, or about 349 to 366, and it will survive scrutiny far better than 353 alone.

Same wheels, different drivetrains, different crank claims

Drivetrain loss explains why two cars with identical dyno sheets can wear different badges. Take 300 whp as fixed and move the loss around.

The AWD car needs 44 hp more at the crank than the FWD car to match the same wheel number. That is the price of the transfer case and second axle, and it is why AWD versions of the same model are often rated a touch higher on engine power or a touch slower on paper than their grip at the drag strip suggests.

The reverse comparison bites buyers of modified AWD cars. A 500 crank hp AWD sedan at a 22 percent loss promises 500 x 0.78 = 390 whp. If the dyno day returns 340 whp, the shortfall of 50 whp is a real gap of about 64 crank hp worth hunting down in boost, fuel or tune, not a quirk of the drivetrain.

What dyno numbers hide: corrections, altitude and dyno type

A chassis dyno does not read absolute truth; it reads a number corrected to a standard day. The common US standard, SAE J1349, references 77 F, 29.235 inHg and zero humidity, and the SAE J1349 standard caps valid corrections near 3 percent before the run stops being comparable. A hot day in Denver can need a double digit correction, and outside that band the correction stops meaning much.

Dyno type moves numbers too. Inertia dynos, the roller kind common in tune shops, spin a known mass up and compute power from acceleration. Load bearing dynos hold the car at a steady speed and measure torque directly, and the two can disagree by a few percent on the same car. Operators also pick correction standards, SAE or the older STD curve, and the same run quoted both ways can differ by 4 to 5 percent before any drivetrain math starts.

None of this makes the number useless. It makes the number a comparison tool, most meaningful when the only thing that changes between runs is the car.

Where the percentage method breaks down

Treating drivetrain loss as a fixed percent is an approximation, and it leans optimistic for powerful cars. A real drivetrain has two kinds of loss. Some parts scale with how much power flows through them, but a base load of churning fluid, bearing drag and gear mesh is closer to a fixed number of horsepower. A transmission might cost a fixed 15 hp plus a few percent of the load, not 15 percent at every power level.

Consequence one: the fixed share shrinks as a percentage as power rises, so a 200 hp car and a 1,000 hp car with the same gearbox do not both lose 15 percent. The percentage method quietly overstates the crank number on high output builds. Consequence two: the same 300 whp through three different gear ratios gives three slightly different answers, because gear mesh losses change with the pair in mesh.

For a clean measurement of the loss itself, coast down testing on a load dyno measures how fast the drivetrain slows the car with the engine off throttle, and that lets a shop subtract measured loss instead of assuming a percent. Most owners will never need that level. What everyone should do is keep the loss inside the sane 8 to 30 percent band, use the same layout and loss on every comparison, and remember that a crank figure derived from a wheel figure inherits every correction and quirk the dyno run carried.