Engine torque is measured using a device called a dynamometer, which applies a controlled braking force to the engine’s output shaft and records how much rotational force the engine produces against that resistance. The dynamometer directly measures torque in pound-feet (lb-ft) or newton-meters (Nm), then uses that torque reading along with engine RPM to calculate horsepower. Horsepower is never measured directly on a dyno. It’s always derived from the torque measurement.
What a Dynamometer Actually Does
A dynamometer works by resisting the engine’s rotation and measuring the force required to do so. At its core, the device is a brake. It absorbs the engine’s power output while a load cell or strain gauge records the force being applied at a known distance from the center of the shaft. That force, multiplied by the distance, gives you torque.
The load cell is the critical measuring component. It contains an elastically deformable material wired into an electrical circuit called a Wheatstone bridge. When force stretches the material, the wire inside changes length, which changes its electrical resistance. That tiny resistance change produces a voltage signal proportional to the applied force. The dynamometer’s electronics convert that signal into a torque reading hundreds of times per second.
Types of Engine Dynamometers
Three main absorber technologies are used in engine dynamometers, each with different strengths depending on what’s being tested.
Water brake dynamometers work by churning water inside a housing. The engine spins internal rotors, and the water resists that motion through momentum and shear forces. The more water flowing through the absorber, the greater the braking force. Water brakes are versatile and relatively affordable, covering everything from low-speed, high-torque diesel applications to high-speed gasoline performance engines.
Eddy current dynamometers use magnetic fields instead of water. Rotating metallic discs spin through a magnetic field, inducing electrical currents (eddy currents) that resist the rotation. Air-cooled versions offer precise control but can only sustain full load for short periods before overheating. Water-cooled versions use the same principle but add coolant flow to handle longer test runs.
AC motoring dynamometers use special AC electric motors that can both absorb power (acting as a brake) and drive the engine (forcing it to rotate). This dual capability makes them ideal for simulating real driving conditions, like deceleration or cold-start testing, where the engine needs to be spun externally.
Engine Dynos vs. Chassis Dynos
An engine dynamometer bolts directly to the engine’s crankshaft or flywheel, measuring output before any power passes through the transmission, driveshaft, or differential. This gives the “crank” torque and horsepower numbers you see in manufacturer specs.
A chassis dynamometer (or “rolling road”) measures torque at the drive wheels. You drive the car onto rollers, and the dyno measures the force the tires apply to those rollers. Because power is lost to friction in the transmission, driveshaft, differential, and wheel bearings, wheel torque is always lower than crank torque. The commonly cited drivetrain loss is roughly 15% for two-wheel-drive vehicles and 25% for all-wheel-drive vehicles, though this varies by vehicle.
How Inertia Chassis Dynos Calculate Torque
Some chassis dynos use an inertia-based method rather than a traditional brake absorber. The car accelerates a set of heavy rollers with a known mass. By measuring how quickly those rollers speed up, the dyno calculates force using the basic physics formula: force equals mass times acceleration. That force is then multiplied by the roller’s radius to give torque at the wheels. The acceleration is typically measured by using the dyno’s eddy current retarders as generators, where the voltage output corresponds to roller speed and its rate of change.
The Torque-to-Horsepower Formula
Since dynamometers measure torque directly, horsepower is always a calculated value. The formula is straightforward: horsepower equals torque (in lb-ft) multiplied by RPM, divided by 5,252. That constant, 5,252, comes from unit conversions built into the relationship between rotational force and power.
This formula creates an interesting quirk on dyno graphs. At exactly 5,252 RPM, dividing RPM by the constant gives 1, meaning horsepower numerically equals torque at that engine speed. Below 5,252 RPM, the torque curve is always higher than the horsepower curve. Above 5,252 RPM, horsepower is always the larger number. So if you see a dyno chart where the two lines cross, they will always cross at 5,252 RPM. As a concrete example, an engine making 500 lb-ft of torque at 7,500 RPM produces about 714 horsepower.
Correction Factors and Standardization
Air density changes with temperature, humidity, and barometric pressure, and those changes directly affect how much power an engine makes. An engine tested on a cool, dry day at sea level will produce more torque than the same engine tested on a hot, humid day at altitude. To make results comparable, the industry applies correction factors that normalize readings to a standard atmosphere.
The current standard is SAE J1349, which normalizes results to 77°F, 29.23 inches of mercury barometric pressure, and 0% humidity. An older standard, SAE J607 (sometimes labeled “STD”), uses 60°F and 29.92 inches of mercury. The older standard’s cooler reference temperature means it produces slightly higher corrected numbers for the same actual engine output, which is one reason comparing dyno results across different shops or eras requires knowing which correction factor was applied.
Shaft-Mounted Torque Sensors
Not all torque measurement requires a full dynamometer. Inline torque sensors can be installed directly on a rotating shaft, measuring the tiny amount of twist that occurs when torque is applied. One approach uses optical interrupter sensors paired with slotted discs mounted at two points along the shaft. As the shaft twists under load, the relative position of the slots shifts slightly. The optical sensors detect this angular displacement with high precision, and the amount of twist, combined with the shaft’s known stiffness, gives a real-time torque reading without any contact-based measurement. These sensors are particularly useful for measuring torque in running vehicles or in components like transmissions and axles.
How Your Car Estimates Torque Internally
Modern vehicles don’t carry dynamometers onboard, but their engine control units continuously estimate torque output in real time. The ECU uses a combination of sensor inputs: how much fuel is being injected, current engine RPM, ignition or injection timing, turbo boost pressure, and the parasitic loads from accessories like the air conditioning compressor and alternator. The ECU also factors in internal engine friction at different speeds and temperatures. From all these inputs, it calculates an estimated torque output, which it then uses to manage transmission shift points, traction control, and stability systems. These estimates are reasonably accurate for control purposes, though they’re not as precise as a dynamometer measurement.
Units of Measurement
Torque is expressed in pound-feet (lb-ft) in the U.S. and in newton-meters (Nm) in most of the rest of the world. Converting between them is simple: 1 Nm equals 0.7376 lb-ft, and 1 lb-ft equals 1.3558 Nm. So an engine rated at 400 Nm produces about 295 lb-ft. Both units describe the same physical quantity, just measured in different systems. If you’re comparing specs across manufacturers from different countries, keeping this conversion handy avoids confusion.