A traction recording dynamometer is an inline force-measuring instrument fitted between a prime mover and its load — typically a tractor and a plough, or a locomotive and a test car — that continuously records drawbar pull against distance travelled on a paper drum. Unlike a Prony brake, which absorbs and dissipates shaft power as heat at a fixed point, this device measures useful tractive effort during real work. The trace gives you a force-distance diagram whose enclosed area is the work done in joules. That number, divided by elapsed time, is true drawbar horsepower under field conditions.
Traction Recording Dynamometer Interactive Calculator
Vary drawbar pull, recorded distance, elapsed time, and wheel slip to see work, drawbar power, horsepower, and slip error on a live recording-drum diagram.
Equation Used
The calculator treats the dynamometer chart area as average drawbar force times corrected travel distance. Dividing that work by elapsed time gives drawbar power, then horsepower.
- Average pull represents the area-average value of the force-distance trace.
- Wheel slip creates a linear distance-axis error.
- Drawbar horsepower uses 1 hp = 745.7 W.
- Spring and stylus calibration errors are not included.
Operating Principle of the Traction Recording Dynamometer
The instrument sits in the draft line. A calibrated spring — usually a stack of laminated leaf springs or a heavy helical compression spring — deflects in proportion to the pull. A stylus rides on that deflection and scribes a line across a paper-wrapped drum that rotates in step with ground travel, normally driven by a small toothed wheel running on the soil or by a flexible cable taken off a road wheel. What you get is force on the vertical axis, distance on the horizontal axis, and a continuous record of every variation in pull as the implement crosses hard patches, root mats, and changes in soil moisture.
Why build it this way? Because tractive effort in the field is never steady. A plough averaging 800 kgf will spike to 1,400 kgf when it hits a buried stone and drop to 400 kgf in light loam — and a single peak reading on a dial gauge tells you nothing about the work done. The force-distance diagram captures the integral. The Nebraska Tractor Test laboratory has used recording dynamometers of exactly this type since 1920 for that reason.
Tolerances matter. The spring rate must be calibrated to within ±1% across its working range, and the drum-to-ground ratio has to be locked — if the ground wheel slips even 3% the distance axis stretches and your computed horsepower is wrong by the same margin. Common failure modes are stylus chatter on rough ground (fix: add a viscous damper to the stylus arm), spring hysteresis after overload (replace, do not re-zero), and ground-wheel slip in wet conditions (use a knurled or spiked drive wheel, never a smooth rubber tyre).
Key Components
- Calibrated draft spring: Converts pull into linear deflection. A typical agricultural unit uses a 0–2,000 kgf spring with a working deflection of about 75 mm, calibrated to ±1% across the range. Overload by 25% and you induce permanent set — the zero shifts and every subsequent reading is offset.
- Recording stylus and arm: Mechanically links spring deflection to a vertical scribe line on the paper. The arm runs through a 5:1 to 10:1 magnification linkage so a 75 mm spring travel produces a clean full-scale trace on a 75–150 mm wide chart. A viscous damper kills high-frequency chatter from stones and trash.
- Distance drum: A paper-wrapped drum rotated by a ground wheel or cable drive, normally geared so 1 revolution equals 5 m or 10 m of travel. Slip must stay below 1% — use a spiked wheel in cultivated soil, never a smooth tyre.
- Ground wheel and flexible drive: Drives the drum. A 500 mm spiked wheel running directly on tilled ground gives reliable distance reference. A flexible shaft or roller chain transmits rotation to the drum head without binding through the drawbar's articulation.
- Drawbar eyes and shackles: Rated to at least 2× the spring's full-scale load. On a unit calibrated to 2,000 kgf, the eyes and pins must hold 4,000 kgf without yield — a sudden plough catch can spike well above nominal.
Industries That Rely on the Traction Recording Dynamometer
These instruments were the standard tool for measuring useful work output of any vehicle that pulls a load. Anywhere shaft power isn't the right number — because losses in transmission, wheels, tracks, or coupler slack matter — you measure at the drawbar instead. They still see active use in agricultural tractor certification, locomotive tractive-effort verification, and forestry skidder evaluation. The reason is simple: a Prony brake or water brake tells you what the engine produces, but a recording dynamometer tells you what actually reaches the work.
- Agricultural machinery testing: The University of Nebraska Tractor Test Laboratory uses inline recording dynamometers on its drawbar test track to certify every tractor model sold in Nebraska — a programme running continuously since 1920 under the Nebraska Tractor Test Law.
- Locomotive engineering: British Rail's Rugby Locomotive Testing Station fitted recording dynamometer cars between locomotive and train to measure tractive effort vs distance during 1950s steam and diesel evaluations — the LMS Mobile Test Plant being the best-known example.
- Forestry equipment: John Deere and Tigercat verify drawbar pull on grapple skidders like the 848L using portable inline dynamometers during prototype trials on graded slopes.
- Military vehicle development: The US Army Aberdeen Proving Ground used recording dynamometers to characterise drawbar pull of M4 Sherman and later M60 tanks across soft-soil trafficability tests.
- Animal traction research: FAO and CEEMAT field studies in West Africa used compact spring-and-drum dynamometers to record sustained pull from oxen and donkeys hitched to ard ploughs, producing the force-distance traces that justified yoke and harness redesigns.
- Heavy haulage: Mammoet and Sarens log drawbar force during heavy modular trailer pulls to verify tractor unit selection on transformer and refinery-vessel moves.
The Formula Behind the Traction Recording Dynamometer
The instrument records force vs distance. The work done is the area under the trace, and drawbar horsepower is that work divided by elapsed time. What changes across the operating range is which term dominates the noise. At the low end of typical agricultural pull — say 400 kgf in light loam — the force trace is smooth and the integral is dominated by mean pull, so a simple average works within 2%. At nominal 800 kgf in cultivated clay, peaks and troughs roughly balance and the planimeter integration matches mean-pull estimates within 5%. At the high end, 1,400+ kgf in compacted or stony ground, peaks dominate and you must integrate the actual trace — averaging will under-report work by 10% or more.
Variables
| Symbol | Meaning | Unit (SI) | Unit (Imperial) |
|---|---|---|---|
| Pdb | Drawbar power (useful power delivered to the load) | W | hp |
| F | Instantaneous drawbar pull recorded by the spring | N | lbf |
| ds | Differential distance travelled, from the drum trace | m | ft |
| t | Elapsed time over the recorded run | s | s |
| Favg | Mean drawbar pull from planimeter integration of the trace | N | lbf |
| v | Mean travel speed during the run | m/s | ft/s |
Worked Example: Traction Recording Dynamometer in a sugarcane-harvester drawbar pull test
A sugarcane equipment maker in Piracicaba, São Paulo, is testing a new 3-row cane billet harvester behind a Valtra BH210 tractor on flat, tilled red latosol. They fit an inline recording dynamometer with a 0–3,000 kgf spring, a 500 mm spiked ground wheel geared so 1 drum revolution equals 10 m of travel, and run a 100 m straight-line pass. The planimeter reads a mean drawbar pull of 1,800 kgf over the run. Time to traverse the 100 m is 60 s. They want drawbar horsepower at nominal load, plus an estimate of how the figure shifts at the light-cane and heavy-cane extremes the harvester will see in production fields.
Given
- Favg = 1,800 kgf
- distance = 100 m
- t = 60 s
- spring full-scale = 3,000 kgf
- drum ratio = 1 rev per 10 m —
Solution
Step 1 — convert mean pull from kgf to newtons. 1 kgf = 9.81 N, so:
Step 2 — compute mean travel speed from distance and time:
Step 3 — drawbar power at nominal load:
That's the useful power reaching the harvester at the drawbar. The Valtra BH210 is rated 210 hp at the engine, so the tractor is loafing at roughly 19% of rated engine power — typical for cane harvesting on flat ground where most of the engine load goes to the PTO-driven cutter and chopper rather than the drawbar.
Step 4 — at the low end of the operating range, light first-ratoon cane in dry soil, mean pull drops to about 1,000 kgf:
The trace at this load is smooth and almost flat — peaks within ±10% of mean. You can read the integral by eye and the planimeter is barely needed.
Step 5 — at the high end, fourth-ratoon heavy cane on damp soil with root mat, mean pull climbs to 2,600 kgf and the trace becomes spiky:
At this load the trace shows ±25% excursions around the mean, and the spring approaches 87% of full-scale. The planimeter integration is essential — eyeballing the average will under-read by 8–12% because the eye weights the trough regions too heavily.
Result
Nominal drawbar power is 29. 4 kW, or about 39.5 hp, at 1,800 kgf mean pull and 1.67 m/s travel. That's a steady, comfortable working load — the operator feels no surging, the tractor holds gear without lugging, and fuel burn is in the 18–22 L/h band. The range from 22 hp on light first-ratoon to 57 hp on heavy fourth-ratoon shows where the drivetrain has to live: the sweet spot sits near 40 hp, and the harvester should be geared so that point falls in the tractor's peak-torque RPM band. If your measured drawbar horsepower comes in 15% lower than predicted, check three things in order: ground-wheel slip on the dynamometer's distance drum (a smooth tyre on damp soil can slip 5–8%, stretching the distance axis and depressing the integrated power), spring hysteresis after a previous overload (any shock pull above 3,000 kgf permanently sets the leaves and shifts zero by 30–80 kgf), and stylus-arm friction (a dry pivot bushing adds drag that flattens peaks and clips 3–5% off the integral).
When to Use a Traction Recording Dynamometer and When Not To
Drawbar pull isn't the only way to measure power, and the recording dynamometer isn't always the right tool. Here's how it stacks up against the two common alternatives — the Prony brake for shaft power, and a modern strain-gauge load cell with digital data logging for inline force measurement.
| Property | Traction Recording Dynamometer | Prony Brake | Strain-Gauge Load Cell + Logger |
|---|---|---|---|
| What it measures | Useful drawbar power under real load | Engine shaft power dissipated as heat | Inline force only — needs separate distance/time measurement |
| Accuracy (typical) | ±2–3% of full scale | ±1–2% of full scale | ±0.1–0.5% of full scale |
| Sample rate / bandwidth | Mechanical, ~5–10 Hz limited by stylus damping | Steady-state only - no dynamic capture | 1 kHz+ digital, full dynamic capture |
| Power range | 10–500 hp typical drawbar applications | 1–1,000 hp at the shaft | Any range with appropriate load cell — 1 N to 1 MN |
| Output format | Force-distance diagram on paper | Single dial reading or torque arm scale | Digital time-series, any post-processing |
| Field portability | Self-contained, no power needed | Stationary test bed only | Needs DC power, signal conditioning, datalogger |
| Capital cost | $2,000–8,000 restored or new mechanical unit | $3,000–15,000 for a calibrated brake | $1,500–10,000 depending on cell + DAQ |
| Useful for | Tractor and locomotive certification, real-world work measurement | Engine bench characterisation | Modern R&D, fatigue spectra, control development |
Frequently Asked Questions About Traction Recording Dynamometer
Size up. Running a leaf-spring or helical-spring dynamometer above 80% of full-scale routinely is asking for permanent set within a season. The spring rate stays linear up to about 85% — beyond that, individual leaves yield slightly on peak loads and the calibration walks. You'll see it as a slow zero drift over weeks, then a sudden offset after a particularly bad shock pull.
Rule of thumb: pick a spring whose full-scale is 1.5× your expected mean pull. For an 1,800 kgf mean, fit a 3,000 kgf spring. That gives headroom for the peak transients without compromising resolution at the mean.
Because drawbar power is always less than PTO power, and PTO power is always less than engine power. The losses stack: engine to gearbox (≈3%), gearbox to final drive (≈5%), final drive to wheel (≈2%), and then the big one — wheel to ground. On hard pavement you lose 10–15% to rolling resistance and tyre flex; on tilled soil you lose 25–40% to slip and soil deformation.
A John Deere 8R rated 410 PTO hp will deliver around 320–340 hp at the drawbar on concrete and 240–280 hp on cultivated clay. If your number is below 60% of PTO rating on hard ground, look at tyre slip — over 8% slip on a hard surface means the ballast is wrong or the tyres are over-inflated.
Load cell, every time, for R&D. The mechanical recording dynamometer is brilliant for certification work where you need a single defensible number traceable to a paper trace — that's why Nebraska still accepts them. But for development, the 5–10 Hz mechanical bandwidth hides everything interesting: implement entry transients, stone-strike spikes, slip-stick on root mats. A 1 kHz strain-gauge log captures all of it.
Use the recording dynamometer when you need a robust, power-free, paper-trail-defensible record. Use the load cell when you need to understand why the pull looks the way it does.
That's almost always a tyre-tread or wheel-runout artefact in the distance drive, not real load variation. A 500 mm spiked ground wheel has a circumference of 1.57 m — if a single spike is bent or one section of the rim is out of round, the drum advances unevenly once per revolution and the trace shows a periodic pattern at that spacing.
Diagnostic check: lift the dynamometer off the ground, rotate the ground wheel by hand through ten full turns, and watch the drum. If the drum advance per quarter-turn varies by more than 5%, the wheel needs trueing or the spikes need straightening. The other suspect is a kinked flexible drive shaft — replace it before you trust another run.
You add or subtract the gravity component of the towed implement's weight. On a 3% upgrade, an implement weighing 2,000 kgf adds 60 kgf to the recorded drawbar pull that has nothing to do with cutting or tillage work — it's just lifting mass. On a downgrade you subtract.
The correction is Fcorrected = Fmeasured − (Wimplement × sin θ), where θ is the grade angle. For certification work, the Nebraska protocol requires grade below 1% on the test track for exactly this reason. If you're testing on real fields, log grade with a digital inclinometer at 1 Hz and apply the correction in post.
Twenty percent is too big to be a single small calibration error — it's a structural mistake. The two most common are: (1) wrong drum ratio assumed in the calculation. If the dynamometer was rebuilt and someone swapped the ground-wheel sprocket, the documented '1 rev per 10 m' may now be 1 rev per 8 m, and every distance you read is 25% wrong. (2) Spring calibration certificate is stale — leaf springs lose 3–5% rate per decade from creep, so a 1980s unit using its original cert reads light by that margin.
Pull the spring out and dead-weight calibrate it before trusting another run. Hang known weights at 25%, 50%, 75% and 100% of full scale, measure deflection, and recompute the rate. If it's drifted more than 2% from the cert, recalibrate or replace.
You can, but only for the steady-state pull-back force, and only if your spring is rated for compression as well as tension. Most agricultural units are tension-only — they have shackle eyes at both ends and the spring stack only resists pull. Reverse the load and the trace goes off-scale on the opposite side or the stylus disengages.
For EV regen work, a bidirectional strain-gauge load cell is a far better fit — regen events happen in tenths of a second and the mechanical instrument's 5–10 Hz bandwidth misses the peak entirely. Use the recording dynamometer for sustained towing tests where you want a paper record of average regen-assisted descent force, not for transient capture.
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.
