Neer's Rotary Transmitting Dynamometer is an inline torque-measuring instrument that reads the elastic angular twist of a calibrated section of rotating shaft to compute transmitted power. The central component is the torsion bar — a precisely sized shaft section whose deflection under load is proportional to torque per the standard torsion equation. It exists because absorption dynamometers waste the power they measure, while a transmitting unit lets the load keep doing useful work. Marine drivetrain trials and pump-set commissioning still use this principle to verify shaft horsepower without disconnecting the driven machine.
Neer's Rotary Transmitting Dynamometer Interactive Calculator
Vary torsion-bar twist, size, material modulus, and speed to see transmitted torque, power, shaft horsepower, and shear stress.
Equation Used
The torsion bar converts the measured angular lag between input and output reference discs into torque. Use theta in radians, G in pascals, d and L in metres, then multiply torque by angular speed to obtain transmitted power.
- Solid circular torsion bar in elastic range.
- Uniform diameter and shear modulus along the calibrated length.
- Twist angle is the measured input-to-output disc lag.
- No correction for temperature, coupling slip, or bending strain.
The Neer's Rotary Transmitting Dynamometer in Action
The instrument sits in the driveline between the prime mover and the load. Power passes straight through it. As torque builds, the calibrated torsion bar inside the unit twists by a small angle θ — typically a fraction of a degree per metre of bar length at rated load. Two reference discs or pointers, one at each end of the bar, rotate together with the shaft but appear angularly displaced relative to each other. Read that displacement and you have torque. Multiply by angular velocity ω and you have transmitted power.
Why build it this way? Because the alternative — a rope brake or Prony brake — converts every watt you measure into heat. On a 200 kW marine auxiliary that means a cooling problem and lost availability. A transmission dynamometer like Neer's lets you instrument a working drivetrain without interrupting it. The torsion bar acts as both the load path and the sensing element, which is elegant but demands tight tolerances. The bar diameter is typically held to ±0.02 mm because the torsion equation T = (π × G × d4 × θ) / (32 × L) carries d to the fourth power. A 1% error in diameter becomes a 4% error in indicated torque.
If the angular readout drifts, the cause is almost always one of three things. Bar material shifting modulus G with temperature — steel loses around 3% modulus per 100°C rise, so an uncooled bar near a hot exhaust reads high. Pointer-to-bar coupling slipping at the pinned end — even 0.05° of slip on a short bar swamps the signal. Or shaft misalignment driving bending strain into the bar, which looks like torsional deflection on a stroboscopic readout. Inline torque sensors of this family fail almost exclusively at the coupling interfaces, not in the bar itself.
Key Components
- Calibrated Torsion Bar: The sensing shaft section, machined from heat-treated alloy steel with a known shear modulus G (typically 79-82 GPa for medium-carbon steel). Diameter is held to ±0.02 mm and surface finish below Ra 0.8 µm to prevent stress raisers. Its angular deflection under torque is the entire measurement signal.
- Reference Discs (Pointer Pair): Two discs rigidly keyed to the shaft, one at each end of the torsion bar's calibrated length. At zero torque their reference marks align. Under load they appear angularly offset by θ. Disc concentricity must hold to 0.05 mm TIR or runout masquerades as twist.
- Stroboscopic or Optical Readout: A flash lamp synchronised to shaft rotation freezes the two pointer marks so the operator reads θ directly off a graduated scale. Modern equivalents use optical encoders at each end and subtract counts. Resolution typically 0.01° on a 500 mm bar.
- End Couplings: Flanged or splined couplings transmit torque from input shaft into the bar and out to the load. These must transmit full rated torque without slip — any micro-slip at the keyway shows up as drift on the readout. Fitted to h6/H7 with shrink-fit or taper-lock preferred over keyways for precision work.
- Support Bearings: Carry the radial load of the bar assembly so bending strain doesn't contaminate the torsional signal. Typically deep-groove ball bearings sized for negligible drag at operating speed. Misalignment beyond 0.1 mm/m introduces measurable bending error.
Industries That Rely on the Neer's Rotary Transmitting Dynamometer
Transmission dynamometers earn their keep wherever you need to know shaft power without dumping that power as heat. Marine, agricultural, mining, and stationary-engine work are the historical strongholds. The Neer pattern specifically — with its angular-deflection readout — appears in older textbooks alongside Bevis-Gibson and Froude designs, and the underlying principle still drives modern strain-gauge torque transducers. Shaft horsepower verification on a working vessel is the flagship use case, but pump-set commissioning, gearbox efficiency mapping, and PTO power audits all use the same idea.
- Marine Engineering: Shaft horsepower trials on tugboat propulsion drivetrains — measuring delivered power between the reduction gearbox output and the propeller shaft on vessels like a Damen ASD 2810 tug during sea trials.
- Agricultural Equipment: PTO power verification on tractor drawbar tests at the Nebraska Tractor Test Laboratory, where transmitted-type dynamometry confirmed catalogue power on units like the John Deere 4020.
- Stationary Power: Inline torque measurement on belt-driven line shafts in restored mill installations such as Queen Street Mill, where absorption dynamometry would unload the prime mover from its working line.
- Pump and Compressor Testing: Commissioning verification on large centrifugal pumps — confirming shaft input power on a Sulzer SMD 800 split-case pump matches hydraulic output for efficiency calculation.
- Gearbox Development: Input/output torque mapping on industrial reducers, paired across input and output shafts to compute mesh efficiency on units like a SEW-Eurodrive K series helical-bevel gearbox.
- Wind Turbine Drivetrains: Field-level shaft torque monitoring on legacy wind installations where a permanent transmitting torque sensor sits between rotor hub and gearbox input.
The Formula Behind the Neer's Rotary Transmitting Dynamometer
The governing relationship comes from the torsion of a circular shaft. What matters in practice is how the answer changes across your real operating range. At the low end of typical loading — say 20% of rated torque — the angular deflection is small, often near the resolution limit of the readout, and noise dominates. At rated torque you sit in the design sweet spot where signal-to-noise is best and the bar is well below its yield in shear. Push past 120% rated and you risk permanent set in the bar, which changes its zero and ruins calibration. The formula tells you the torque; multiplying by shaft speed gives transmitted power.
Variables
| Symbol | Meaning | Unit (SI) | Unit (Imperial) |
|---|---|---|---|
| T | Transmitted torque in the shaft | N·m | lbf·ft |
| G | Shear modulus of the torsion bar material | Pa (N/m²) | psi |
| d | Diameter of the calibrated torsion bar | m | in |
| θ | Angular twist between reference discs | rad | rad |
| L | Calibrated length of the torsion bar | m | in |
| ω | Shaft angular velocity | rad/s | rad/s |
| P | Transmitted power | W | hp |
Worked Example: Neer's Rotary Transmitting Dynamometer in a paper-mill refiner drive audit
A pulp-and-paper plant in Kotka is auditing the shaft power delivered to a 600 kW double-disc refiner driven through a long line shaft. Maintenance has fitted a Neer-pattern transmitting dynamometer between the motor coupling and the refiner input. The torsion bar is 50 mm diameter, 800 mm calibrated length, machined from EN24T (G = 80 GPa). The shaft turns at 1480 RPM. At rated load the optical readout shows 0.92° of twist between the reference discs.
Given
- d = 0.050 m
- L = 0.800 m
- G = 80 × 10⁹ Pa
- θ (nominal) = 0.92° = 0.01606 rad
- N = 1480 RPM
Solution
Step 1 — convert shaft speed to angular velocity:
Step 2 — at nominal load, compute torque from the measured 0.92° twist:
Step 3 — multiply by ω for transmitted power at the nominal point:
That sits right on the refiner's 600 kW nameplate — exactly where you want to be on a commissioning trial. Now check the ends of the operating range.
Step 4 — at 30% loading (light pulp, refiner gap open) the readout drops to about 0.28°:
At this end the angular signal is small — only 28 encoder counts on a 0.01° resolution readout — so a 0.02° zero drift is already a 7% error. This is where a poorly-coupled bar lies to you.
Step 5 — at 120% short-term overload during a refiner plate clash, twist climbs to roughly 1.10°:
The bar is still elastic — shear stress at the surface is around 192 MPa against a yield shear of roughly 380 MPa for EN24T — but you are eating into the safety margin. Sustained operation here will fatigue the keyways long before it fatigues the bar.
Result
Nominal transmitted power is 612 kW at 0. 92° of twist and 1480 RPM, confirming the 600 kW refiner is operating at its design point. Across the range, the dynamometer reads 186 kW at light load (where readout noise dominates) up to 731 kW during overload events (where the keyways take a beating before the bar does) — the sweet spot for trustworthy data is 60-100% of rated torque. If your measured power deviates from electrical input power by more than 5%, suspect these in order: (1) temperature drift in G — a hot bar near a steam pipe loses 3% modulus per 100°C and reads high; (2) end-coupling micro-slip at a worn keyway, which produces a slowly creeping zero between cold-start and steady-state; (3) shaft misalignment greater than 0.1 mm/m at the support bearings, injecting bending strain that the optical readout cannot distinguish from torsion.
Neer's Rotary Transmitting Dynamometer vs Alternatives
The choice between a Neer-pattern transmitting dynamometer, an absorption dynamometer like a rope brake or water brake, and a modern strain-gauge torque transducer comes down to whether you want the load to keep working, how accurate the result must be, and how much you can spend. Here is how they line up on the dimensions that actually matter when you are scoping a test rig.
| Property | Neer's Rotary Transmitting Dynamometer | Rope Brake (Absorption) | Strain-Gauge Torque Transducer |
|---|---|---|---|
| Accuracy at rated load | ±2-3% typical | ±3-5% with careful cooling | ±0.1-0.5% |
| Power dissipated as heat | Near zero — load keeps working | 100% of measured power | Near zero |
| Operating speed range | Up to ~3000 RPM, limited by readout | Below 1000 RPM (heat limit) | Up to 20,000+ RPM with telemetry |
| Capital cost (relative) | Medium — mostly precision machining | Low — rope, pulley, scale | High — sensor + telemetry + DAQ |
| Setup complexity | Requires inline integration and shaft alignment | Bench-mounted, simple | Inline, plus electronics calibration |
| Continuous-duty rating | Continuous — no thermal limit | Minutes to hours, water-cooled | Continuous |
| Best application fit | Field shaft-power audits, marine trials | Lab teaching, low-speed engine tests | R&D drivetrain mapping, production test cells |
Frequently Asked Questions About Neer's Rotary Transmitting Dynamometer
Steel torsion bars lose roughly 3% of their shear modulus G for every 100°C rise. Your formula assumes a fixed G, so as the bar heats, the same applied torque produces a larger angular twist θ — and the calculation reads that as more torque. A 4% high reading after warm-up suggests roughly a 130°C rise at the bar.
Two fixes: either temperature-compensate by logging bar temperature with a thermocouple and correcting G in software, or shroud the bar from radiant sources like exhaust manifolds and steam lines. On marine drivetrain installations the bar is often jacketed for exactly this reason.
Work backwards from your readout resolution. Pick a target full-load twist θ that gives you good signal — typically 0.5° to 1.5° across the calibrated length. Below 0.5° and encoder noise eats your low-load accuracy; above 2° and you are running too close to yield in shear.
Rearrange T = (π × G × d4 × θ) / (32 × L) to solve for d, then round up to the nearest standard stock size. Then check the surface shear stress τ = 16T / (π × d3) stays below 30-40% of the material's yield shear for fatigue life. A common mistake is sizing d too large to feel safe — that kills your signal at part-load.
For a one-off field commissioning where ±2-3% is acceptable and budget is tight, the transmitting unit wins — no telemetry, no slip rings, no excitation electronics to fail in a wet pump room. For ongoing R&D work or production test cells where you want ±0.5% and digital data logging, a strain-gauge transducer like an HBM T40B is worth the cost.
The deciding question is usually environment. Strain gauges hate condensation, EMI from VFDs, and shock loading. A mechanical Neer-pattern unit shrugs all three off.
If you have ruled out thermal effects on G, the next suspect is the support bearings. As the bearings warm and their internal clearance changes, they can shift the bar axially or radially by tens of microns. That movement re-seats the reference discs against any backlash in their pinned attachment, producing an apparent zero shift.
Diagnose by logging the readout for 30 minutes at no-load after a cold start. A monotonic drift that asymptotes around the time the bearings reach steady temperature is the giveaway. Fix it by replacing keyed disc attachments with shrink-fit or interference-fit hubs.
Yes, and this is a classic use, but the error budget is brutal. If each dynamometer is ±2.5%, the worst-case efficiency error is roughly ±5% — and you are usually trying to resolve a 2-3% efficiency loss. Random errors partially cancel, so in practice you might see ±2% on η, but that is still on the same order as the loss you want to measure.
For meaningful gearbox mapping, calibrate both units against the same reference torque source on the same day at the same temperature. Better yet, use matched strain-gauge transducers if the budget allows — the systematic error cancels much more cleanly.
Three things, in order. First, the readout — a stroboscopic optical readout becomes hard to resolve above about 3000 RPM because the flash duty cycle blurs the pointer marks. Optical encoders push this to maybe 6000 RPM with high-count gratings. Second, the bar's first torsional natural frequency — if shaft speed approaches it, the bar resonates and the angular signal becomes meaningless. Third, balance — the rotating disc/coupling assembly must be balanced to G2.5 or better, or vibration overwhelms the deflection signal.
For most industrial work below 1800 RPM none of these bite. Above 3000 RPM you should be looking at telemetry-based strain gauge transducers instead.
References & Further Reading
- Wikipedia contributors. Dynamometer. Wikipedia
Building or designing a mechanism like this?
Explore the precision-engineered motion control hardware used by mechanical engineers, makers, and product designers.
