Dynamometer

What a Dynamometer Is

A dynamometer is a device for measuring mechanical force, torque, or power. The defining characteristic is that it does not just sense a quantity, it provides a calibrated, traceable measurement against a reference, which is what separates a dynamometer from a simple gauge or an indicator. The instrument family splits into two branches that confuse newcomers because they carry the same name. The rotating power dynamometer simultaneously measures the torque and the rotational speed of a prime mover and reports power in kilowatts or brake horsepower. The static force dynamometer measures a pull or a push along one axis and reports force in newtons, kilograms-force, or tonnes.

The power dynamometer is the older and more famous of the two. Its history runs through the friction brake, where Gaspard de Prony described the Prony brake around 1821 as a band or block clamped on a rotating shaft, with the reaction torque held by a weighted arm. In 1877 William Froude developed the hydraulic or water brake, which dissipates power in fluid shear rather than dry friction and remains in heavy use for large diesel and turbine testing today. The twentieth century added electromagnetic absorbers, the eddy current brake and the powder brake, and finally the AC and DC motoring machines that can both load and drive a shaft. Each generation traded simplicity for controllability and a wider usable map.

The force dynamometer evolved from the spring balance and the proving ring into the strain-gauge instruments used in rigging today. A modern tension dynamometer is essentially a packaged load cell with integral lifting eyes, a display, and an overload-safe structure, rated to be hung in a lifting chain. Mechanical dial models such as the Dillon AP series still serve where no battery or electronics are wanted, while wireless electronic units such as the Crosby Straightpoint Radiolink Plus report load to a handheld or to software over a radio link.

Why keep both under one word? Because both answer the same engineering question in different domains: how much effort is this machine producing or carrying, expressed as a number a buyer can specify, a technician can verify, and an auditor can trust. A power dynamometer answers it for a spinning shaft, a force dynamometer answers it for a straight pull. Throughout this guide, the term is qualified as power dynamometer or force dynamometer whenever the distinction matters, because mixing the two is the single most common selection error.

Four engineering attributes ultimately decide the quality of either family: measurement accuracy against a traceable reference, the usable operating envelope, the energy-handling or overload capacity, and long-term stability between calibrations. These four together govern the total cost of ownership. A cheap instrument with a narrow envelope and poor stability forces frequent recalibration and risks an out-of-tolerance result on the one test that matters, which is far more expensive than the purchase price saved.

Power vs Force Dynamometers

Before any model selection, fix which family the job needs. A rotating prime mover that must be characterised for torque and power needs a power dynamometer. A lift, a tie-down, a cable pull, or a structural pull test that must be quantified along a line needs a force dynamometer. The two share only a name and a heritage. The table below contrasts the families on the dimensions that change which catalogue you open.

Within the power dynamometer branch there are three control architectures. An absorption dynamometer can only take power away, acting as a controllable brake that turns mechanical output into heat. Eddy current, water brake, hysteresis, and powder units are all absorbers. A motoring dynamometer uses an AC or DC machine that can both absorb and drive the shaft, which is required for motored-engine friction sweeps, cold cranking, and transient drive cycles. A universal or four-quadrant dynamometer combines both behaviours, absorbing on one quadrant and motoring on another, and is the standard choice for modern emissions and durability test cells.

Power dynamometers also split by where they couple to the powertrain. An engine dynamometer bolts to the crankshaft or flywheel of an engine on a stand and reads gross output with no driveline in the path. A chassis dynamometer measures at the driven wheels through rollers with the engine still in the vehicle, capturing real transmission and tire losses. A hub dynamometer replaces the wheels and bolts to the hubs, removing tire slip while keeping the rest of the driveline. Because of driveline losses, a chassis dyno typically reads roughly 15 to 20 percent below the equivalent engine dyno on a two-wheel-drive car.

Within the force dynamometer branch the split is mechanical versus electronic. Mechanical dial dynamometers convert load through an elastic element to a geared pointer, need no power, and survive harsh sites, with peak-hold pointers to capture the maximum load. Electronic dynamometers strain-gauge the same elastic element and report digitally, often wirelessly, with logging, peak capture, and multi-point monitoring of a lift. A specialised subfamily, the multi-component cutting-force dynamometer used in machine-tool research, measures force along several axes at once with piezoelectric elements, but it is a laboratory instrument rather than a rigging tool.

Absorber and Drive Technologies

For power dynamometers, the absorber or drive technology determines the usable torque-speed envelope, the response speed, the cooling requirement, and the cost. No single technology is best for all duties: a unit perfect for a 20000 RPM motor is wrong for a slow marine diesel. The table below compares the mainstream technologies on the metrics that drive selection.

Water brake dynamometers dissipate power in fluid friction inside a casing, where a rotor stirs water against a stator and the reaction torque is reacted by the load cell. Invented by William Froude in 1877, they handle enormous power in a compact package and tolerate brief overloads gracefully, which is why they dominate large diesel and gas-turbine test stands. The trade is a relatively coarse low-speed control and a continuous demand for clean cooling water carried away as warm effluent.

Eddy current dynamometers couple the test shaft to a toothed or disc rotor that spins inside a magnetic field set up by a stator coil. The motion induces eddy currents in the rotor, and the resulting drag is the absorbed torque, controlled smoothly by the coil current. They are the most common absorber on modern chassis dynos and on general engine stands because the load is electrically controllable and fast. The limitation is that absorbed energy becomes rotor heat, so the high-speed end of the map is bounded by cooling, and like all pure absorbers they cannot motor an engine.

Hysteresis dynamometers develop braking torque through magnetic hysteresis in a rotor passing through a field, producing torque that is essentially independent of speed, including down to zero speed. That makes them ideal for testing small motors from no load to locked rotor, with accuracy on the order of plus or minus 0.25 percent of full scale on typical units, but their continuous power is modest, on the order of single-digit kilowatts. Powder brakes use magnetic particles that chain up under a field to transmit torque, giving high torque at low speed but limited high-speed capability because the particles generate heat.

AC and DC motoring dynamometers are electrical machines that can absorb power as a generator and deliver power as a motor, giving full four-quadrant operation. AC machines with regenerative drives are now the default for transient emissions and durability cells because they respond quickly, return absorbed energy to the grid through the drive, and run from zero to full speed in both directions. Published examples include the HORIBA DYNAS3 high-speed AC series rated to roughly 210 kW at up to 20000 RPM, and Mustang Advanced Engineering DGEC eddy current units spanning about 25 kW to 1300 kW for the absorber-only class.

Torque, Speed, and the Power Math

A power dynamometer never measures power directly. It measures two independent quantities, torque and rotational speed, and computes power from them. Understanding that chain is essential because every accuracy claim, every calibration step, and every comparison between machines depends on it. Torque is sensed as a reaction: the absorber tries to spin in its bearings, an arm of known length resists that rotation, and a load cell at the end of the arm reads the force. Torque is then force multiplied by arm length, so a 200 N force on a 0.5 m arm is 100 N·m of torque.

Rotational speed is sensed separately with a shaft encoder, a magnetic or optical pickup on a toothed wheel, or a tachometer. Power is the product of torque and angular velocity. In SI units, power in watts equals torque in newton metres multiplied by angular velocity in radians per second, where angular velocity in radians per second equals RPM multiplied by pi and divided by 30. So 100 N·m at 3000 RPM is 100 multiplied by 3000 multiplied by pi divided by 30, which is about 31.4 kW.

In US customary practice the famous shortcut is power in horsepower equals torque in pound feet multiplied by RPM and divided by 5252. The constant 5252 is simply 33000 divided by 2 pi, carrying the definition of one horsepower as 33000 foot-pounds per minute. A direct consequence is that the torque and power curves on a US dyno chart always intersect at 5252 RPM, because at that speed the formula makes the two numbers equal. That crossing point is a quick sanity check that a plotted chart uses consistent units.

Because power is derived, a dynamometer is only as good as its torque chain and its speed channel. Torque calibration is done by hanging certified dead weights on the known moment arm, which directly verifies the force-times-length product against gravity, the most traceable reference available. Speed is checked against a reference frequency. If either channel drifts, the computed power drifts with it, which is why both are verified at every calibration rather than trusting a single power reading.

One further subtlety separates absorbers from motoring machines on inertia. During a fast acceleration sweep, some of the engine torque goes into accelerating the dyno rotor and shaft, not into the absorber. High-end controllers correct for this rotating inertia so that the reported torque reflects engine output rather than the energy stored in spinning steel. Inertia dynamometers exploit the same physics deliberately, using a calibrated flywheel of known mass moment of inertia and computing torque from the measured angular acceleration. The standards below close the loop on comparability.

These correction standards exist because engine power depends on inlet air density, which changes with temperature, barometric pressure, and humidity. A correction factor is a multiplier applied to measured power to estimate the figure the engine would have produced at the standard reference atmosphere, so that results taken on different days or in different cities can be compared. A power number is therefore meaningless without naming its correction standard. SAE J1349 and ISO 1585 use very similar references, while the older DIN 70020 uses a different temperature and the full standard atmosphere, so the same engine yields slightly different headline figures under each.

Key Specification Parameters

Reading a dynamometer datasheet means looking past the headline number to the envelope and the error budget. For a power dynamometer the parameters that drive selection are peak torque, maximum speed, continuous absorbed power, torque accuracy, speed range, inertia, and cooling capacity. For a force dynamometer the drivers are working load limit, accuracy as a percentage of full scale, resolution, overload and break ratings, and ingress protection. Each is explained below.

Peak torque and maximum speed define the corners of the usable map, but they cannot be reached at the same time. An absorber holds full torque only up to a corner speed, then power-limits as speed rises because heat rejection caps it. The honest way to read a datasheet is to overlay the device-under-test torque-speed curve on the dyno envelope and confirm the test curve sits inside it everywhere, not merely under the single peak-power figure.

Continuous versus intermittent power matters because absorbers turn mechanical input into heat. The continuous rating is what the cooling system can carry away indefinitely; the intermittent rating is a short-duration ceiling. A water brake quotes both against a stated cooling-water flow and temperature rise, and an eddy current unit quotes both against air or water cooling. Running above the continuous rating without the matching cooling overheats the absorber and shifts calibration.

Torque accuracy and speed accuracy combine into power accuracy because power is their product. Torque accuracy is usually quoted as a percentage of full scale or of reading, traceable to dead-weight calibration on the moment arm. For hysteresis motor-test dynamometers, accuracy on the order of plus or minus 0.25 percent of full scale is typical. Always confirm whether a quoted accuracy is on torque alone or on computed power, and whether it holds across the full speed range or only at a reference point.

For force dynamometers, the parameters shift to rigging language:

• Working load limit (WLL): the maximum load the instrument is rated to carry in service, the figure that must exceed the actual lift with margin.
• Accuracy: typically quoted as a percentage of full scale; quality wireless units such as the Crosby Straightpoint Radiolink Plus claim plus or minus 0.1 percent of full scale.
• Capacity range: packaged tension dynamometers span roughly 0.5 tonne to 500 tonnes and beyond, with mechanical dial units commonly to about 25 tonnes.
• Proof and break load: the proof load verifies the structure without damage; the minimum break load sets the safety factor over WLL.
• Ingress protection and certification: IP67 sealing for site use, and ATEX or IECEx (for example Ex ia IIC T4 Ga) for hazardous-area lifts.

Inertia and response round out the power-dynamometer spec. Low rotor inertia lets the absorber follow a fast transient sweep without lagging the engine, and controllers correct residual inertia torque during acceleration. Sample rate matters on the force side too: a rigging dynamometer that updates at a few hertz suits static weighing, while dynamic load monitoring on a moving lift needs the higher sampling some electronic units offer, up to a couple hundred hertz on capable models.

Selection Decision Factors

To turn the preceding chapters into a chosen model, follow the decision sequence below. Most selection mistakes come not from a single wrong number but from deciding the detail before the family. Settle the family and the architecture first, then the envelope, then the interfaces. These eight steps double as a fixed RFQ template.

1. Confirm the family: rotating power measurement needs a power dynamometer; a tension, compression, or rigging measurement needs a force dynamometer. This single choice opens an entirely different catalogue, so never skip it.
2. For power, pick the architecture: absorption only for steady-state power verification; motoring or universal four-quadrant for friction sweeps, cold cranking, and transient drive cycles. Then choose engine, chassis, or hub coupling.
3. Map the envelope: overlay the device-under-test torque-speed curve on candidate absorber envelopes and confirm it fits at peak torque, corner speed, and continuous power simultaneously, with margin for overspeed and overtorque.
4. Match the absorber technology: water brake for very high power, eddy current for general controllable absorption, hysteresis or powder for small low-speed motors, AC motoring for transient and regenerative cells, per Chapter 3.
5. Size cooling and utilities: confirm cooling-water flow and temperature rise for a water brake, or air and water capacity for an eddy current or AC unit, against the continuous absorbed-power rating. An undersized cooling loop caps real-world capacity below the headline.
6. For force, set WLL and safety factor: choose a working load limit so the normal load sits between 30 and 80 percent of full scale, and verify proof and break ratings against the lift, with ASME B30.26 compliance for below-the-hook rigging hardware.
7. Define accuracy and standards: state required torque or force accuracy as percent of full scale or reading, name the correction standard for power (SAE J1349, ISO 1585, or DIN 70020), and require traceable calibration certificates.
8. Specify interfaces and environment: control and data interface (analog, CAN, fieldbus, wireless), display and logging, ingress protection, and hazardous-area certification (ATEX or IECEx) where the site demands it.

One dimension is routinely overlooked: serviceability and calibration support. A dynamometer must be recalibrated at least annually and after any overload, transducer change, or relocation, and a force dynamometer must be proof-tested per the applicable rigging standard. Before buying, confirm local calibration capability, dead-weight or reference-machine access, spare-part lead time, and controller firmware support, because these determine downtime five and ten years into service. Established suppliers across the families include AVL, HORIBA, SuperFlow, Mustang Advanced Engineering, and Dynapack for power dynamometers, Magtrol for motor-test dynamometers, and Crosby Straightpoint and Dillon for force and rigging dynamometers.