On-Site Rotor Balancing for Agricultural Mulchers and Flail Mowers

A flail mower or agricultural mulcher loses its balance faster than any other implement: one torn-off flail, a pair of new hammers fitted in place of worn ones, mud caked on after a wet field. The drum is long, turns at around two thousand rpm, and every bit of unbalance goes straight into the bearing housings, the belts, the gearbox and the mounting frame. We come out to the farm, measure vibration at both drum bearings driven from the tractor's own PTO (power take-off), and calculate the weight's mass and position. We deal with the flail set first, then fit weights: done in that order, the result lasts longer.

Updated 27 August 2026 · by AXILINE · Vila Nova de Gaia

Symptoms: what the operator feels before any instrument gets involved

The operator recognizes this kind of imbalance before any analyzer does. The mulcher starts hammering into the mounting frame, the tractor's steering wheel develops a fine tremor, and the hum rises exactly with PTO speed. There is almost always an event this started after: replacing part of the flail set, a torn-off flail, working a wet field, hitting a stone. Tell us about that event when you get in touch, and half the diagnosis is done before we even arrive.

• Shaking started right after part of the flail, blade or hammer set was replaced.
• A flail or a paired set tore off, and the hammering appeared within the same shift.
• Vibration travels through the mounting frame and the seat, rising with PTO speed and dropping when it's reduced.
• After working wet ground, the drum caked up with soil; washing helped, but not all the way.
• The belts on the drive pulley run hot, squeal, and jump out of the grooves more often than usual.
• The drum's bearing housings run hot, and grease has to be topped up every week.
• The PTO shaft's universal joints wear out in half a season instead of several.
• Cracks have appeared in the welds on the guards and the drum's end discs, and the bracket bolts have loosened.
• The machine shakes even with the drum turning empty, with no material feeding through, before you've even entered the paddock.

There's a simple test to separate the two. Vibration that's present with the drum turning empty is mechanical, and that's what we reduce. Vibration that only shows up under load and disappears at idle comes from the shredding process and the material feed. Weights won't remove it, and we won't call that balancing.

How a flail rotor is built, and where its imbalance comes from

An agricultural mulcher's rotor is a steel tube drum with brackets welded on in a helical pattern. Hinged working elements hang on pins in the brackets: flails, hammers, cup-shaped or Y-blade knives. The drum runs on two bearing housings at the ends of the housing, is turned by a belt drive off a right-angle gearbox, and the gearbox gets its torque from the tractor's PTO through a driveline shaft.

The hinge changes the whole picture compared with a rigid rotor. At operating speed, centrifugal force swings every flail out to its radius and holds it there with a force of around 11 kN, for a one-kilogram flail sitting at a quarter-metre radius. As long as the pins and bracket eyes are sound, the flails take up the same position run after run, the imbalance is reproducible, the phase (the angular link between vibration and shaft rotation) is stable, and the rotor balances like an ordinary one. The moment a hinge is worn out, the flail sits at an enlarged radius and wanders. Then the phase drifts between runs, and the influence coefficients — the calculated relationship of "fit a weight, get a response" — won't converge.

Matching flails by mass solves most of the problem

Let's be straightforward: on flail machines, the main work is done not by the instrument but by the scale. Imbalance here is almost always built up from mass differences between the working elements, not from a manufacturing inaccuracy in the drum. Get the set in order, and a significant share of the vibration is gone before we ever fit the first weight.

The rule is simple and works without any equipment at all. Working elements are replaced in diametrically opposite pairs, not one at a time. A new pair is matched on a scale, aiming for a difference of a few grams. If wear is already deep, the whole set is replaced and arranged from heaviest to lightest so that opposite positions come out equal. Paired flails on one pin count as a pair, not individually.

Before the season it's worth doing this yourselves, without us coming out. Recheck the positions against the manufacturer's layout, weigh a sample of flails, check the pins for wear, tap the brackets, wash the drum, and run the rotor empty. Often, after a review like this, balancing isn't needed at all. If a properly matched machine still hammers, that's when you call us: it means the problem is in the drum itself, in its fit, or in the drive.

PTO drive: 540 and 1000 rpm, and what the driveline shaft does to the measurement

The rotor is not turned directly by an engine but by a chain: the tractor's PTO, a driveline shaft with universal joints, a right-angle gearbox, a belt drive, the drum. Every link contributes something to the measurement, and this has to be sorted out before the trial weight, not after.

There are two standard PTO speeds: 540 and 1000 rpm. A machine is designed for one of them, and drum speed comes from the gearbox and pulley ratio. On a typical agricultural mulcher, the drum ends up turning at around two thousand rpm; take the exact figure from your machine's manual. Balancing has to be done at whichever mode you actually run in. If you run a 1000-rated shaft at 540 to save fuel, tell us beforehand: drum speed is different, the influence coefficients are different, and a result obtained in one mode will be worse in the other.

The driveline shaft muddies the spectrum in its own right — the picture of vibration broken down by frequency. Universal joints running at an angle turn the driven side unevenly, twice per shaft revolution, and worn crosses and splines add their own components. The driveline's rotational frequency is many times lower than the drum's, so peaks show up in the spectrum that have nothing to do with rotor speed. Weights can't remove them. We always stick the phase mark on the drum itself or its pulley: past the belt drive and the gearbox, the drive side simply turns at a different frequency.

• We hold one fixed PTO mode for the entire job: the baseline run, both trial runs, and the check run.
• The instrument logs drum speed on every run; a shift between runs means the run has to be repeated.
• We set the driveline shaft's operating angle to a minimum, and the implement's mounting height per the machine's manual.
• The universal joint crosses, splines, guards and overrunning clutch are checked before the measurement, not based on its results.
• The phase mark goes on the rotor or its pulley, never on the driveline shaft and never on the PTO stub.
• We warm the tractor up: PTO speed drifts noticeably more on a cold engine.

What we check before we get the weights out

Balancing only reduces the running-speed component of vibration — the part that repeats once per rotor revolution. If most of the level comes from something else, weights will do cosmetic work at best. So the first thing we do is measure overall vibration (the total level of oscillation from all causes at once), the running-speed component, phase, and the spectrum at both bearings — and only from the figures do we decide whether it's worth going further. How to read these numbers is covered in a separate article on overall vibration, the running-speed component, and phase.

• The working-element set against the manufacturer's layout: recounting positions, checking for empty brackets, spot-weighing.
• Pins and bracket eyes: wear, ovaling, and whether the locking elements are in place.
• Brackets and their welds on the drum tube, cracks in the heat-affected zone.
• Cleanliness: the drum cavity, the bracket pockets, the inner guards, the support roller.
• The drum's bearing housings: play, heat, noise, condition of the seals and grease.
• The fit of the pulley and drum on the shaft: a loosened fit gives different vibration from one run to the next.
• Belt drive: tension, groove wear, pulley runout and alignment, and how the tensioner performs.
• Right-angle gearbox: play, noise, oil level, and fastening to the mulcher's frame.
• Driveline shaft and PTO: universal joint crosses, splines, overrunning clutch, mounting angle.
• Frame, mounting stand, support roller and rear roller: loose fasteners feel indistinguishable from imbalance.
• Resonance: we vary the speed and watch how the amplitude and phase behave. A flexible mounting produces resonance zones in places you wouldn't expect.

Worn-out bearings produce vibration on their own, and on a machine that has run half a season with imbalance, they are almost certainly worn out. We assess them by spectrum and temperature and tell you plainly if a replacement is needed first. How imbalance differs from drive misalignment, and why the two get confused, is covered in a separate article.

Sources: ISO 13373-3:2015 · ISO 281:2007

How the work proceeds on the farm

1. 01

   Preparation and clearance to work

   The machine sits on level ground, lowered onto the earth or onto stands, the drum washed and dry, the flail set complete. The tractor is warmed up, the PTO in working order, the driveline shaft with its guards intact. Starts and stops are made only by your operator, on our instruction, and nobody stands in the drum's plane of rotation: a folded-back apron means the throw-off zone is open.

2. 02

   Sensors and phase mark

   We fit two accelerometers on magnets to both drum bearing housings, radially, on pads cleaned down to bare metal. We stick the reflective mark on the end of the drum shaft or on the drive pulley. We keep the measurement points and direction unchanged for the rest of the job.

3. 03

   Baseline run

   We bring the rotor up to its operating PTO speed and let it settle, so the flails swing out to their radius. We record overall vibration, the running-speed component, phase, speed, spectrum and time waveform. This shows how much of the level is imbalance and how much is everything else.

4. 04

   Repeatability check

   We shut down, wait for the drum to fully stop, start again, and compare the phase. This is a mandatory step for a flail machine. If the phase shifts from run to run, we look for worn pins, a loosened fit, or a wandering PTO speed. You cannot balance off a measurement that doesn't repeat.

5. 05

   Trial runs

   With the rotor stopped, we fasten a weighed trial weight in the first correction plane — the rotor cross-section where the weights will go. Run, measure. We move it to the second plane, run again. The readings have to change noticeably, otherwise we increase the weight and repeat. Why a trial weight is needed and how it's sized is explained in a separate article.

6. 06

   Weights and check run

   The software outputs a mass and a location for each plane. We weigh the weight, fasten it, clean up the weld, clear the tools out of the housing, put the apron back, and run a check at the same points and in the same mode. If we haven't reached the agreed value, the software suggests an addition to the weights already fitted.

7. 07

   Report and saved coefficients

   We record the before-and-after figures, the masses, radii and locations of the weights, drum speed, PTO mode, and the measurement points and direction. We keep the influence coefficients on file: after the next flail-set replacement, fine-tuning will take one or two runs instead of a full cycle.

Any hands-on work at the rotor is done only with the tractor shut down, the PTO disengaged, and the drum fully stopped. The drum keeps coasting for a long time after shutdown, and that's deceptive. We do not bypass lockouts, and we do not ask the operator to bypass them either.

Sources: Balanset-1A operation manual

Two planes: where to take them, and what to fasten the weights with on the drum

A flail machine's drum is elongated: the working width is many times the tube diameter. A rotor like this always has a moment imbalance on top of the static one, and one weight can't remove it. Fit a mass in the middle, and one bearing goes quieter while the other gets louder. Hence two planes, one trial run for each, with both bearings checked at once. Why the number of planes is set by rotor geometry is covered in a separate article.

We take the correction planes at the ends of the drum, as far apart as possible: the end discs, the outermost sections of tube between the last rows of brackets, the drive pulley's flange. The further apart the planes, the smaller the mass needed to offset the moment component. Access on an agricultural mulcher is usually easier than on a forestry machine: the rear apron folds back, the side guards are bolted on, and the end disc opens along with the belt-compartment cover.

Drums like this generally have no bolt-on balance positions, so we weld the weight on. The weld is all that holds it, and the load is serious: a 150 g plate at a 300 mm radius, at two thousand rpm, pulls outward with a force of around two kilonewtons. We weld a continuous bead around the perimeter of a weighed low-carbon steel plate, on the end disc or on the tube, well away from the brackets, their welds, and any hardfaced wear areas.

• The weight is weighed, with mass and radius recorded in the report before welding, not from memory afterward.
• The weld is continuous around the perimeter, no quick tack welds, and no undercut on the tube wall.
• We never weld on hardfacing, on the pins, or on the flail brackets.
• Before welding, we clear dry plant debris out of the housing: it's combustible material, and we keep a fire extinguisher on hand.
• If there are ready-made holes in the pulley flange or the end disc, we work with a bolted weight in fixed-position mode and get a position number instead of an angle.
• We remove mass by drilling only on a thick-walled end disc; we don't drill the drum's thin tube.

We agree the target value before starting work: residual imbalance to the G balance quality grades of ISO 21940-11, and the vibration level at the bearings measured against the zones of ISO 20816. We record the applicable part and edition of the standard in the report for the specific machine, because these documents do have formal scope limits, and for mounted agricultural implements those need to be stated separately.

Sources: ISO 21940-11:2016 · ISO 20816-1:2016

When on-site balancing won't help

We say this based on the first measurement, and often even before the visit, from your description and photos.

• The flail set is mismatched or incomplete. We'd be balancing a random state that will disappear with the next pair replacement. Matching the set comes first, weights after.
• Pins and bracket eyes are worn out. The flail radius wanders, the phase doesn't repeat, and the influence coefficients can't be calculated.
• The drum tube is bent, or a bracket has torn out along with a piece of the wall. This is straightening and repair work; weights won't fix geometry.
• The fit of the pulley or the drum on the shaft has loosened. Vibration changes from run to run, and the balancing won't even last out the shift.
• The tractor won't hold a PTO mode, or the speed drifts between runs. There won't be any repeatable measurements.
• Operating speed falls into a resonance in the frame, the mounting stand, or the housing. De-tuning the resonance comes first; we have a separate article on this.
• There's no access to the drum ends: a sealed housing, an apron that won't come off, no inspection hatches. Then the drum has to be removed and balanced off the machine.
• The imbalance comes back after a few shifts in a row. That means the drum is caking up with soil again or losing flails, and the fix is wear and fastening, not weights.

Another common case. If the machine only shakes under load, and the drum is calm turning empty, the source is the shredding process, the material feed, or the drive. We separate these cases honestly with the measurement, and we don't offer balancing in that situation.

What you get, and how to book a visit

The result of the work is a rotor that holds its operating speed without hammering, and a report with figures. The report includes the initial and final vibration at both bearings in mm/s, the running-speed component and phase, spectra, PTO mode and drum speed, the masses, radii and locations of the weights by plane, mechanical findings, and recommendations for next season. A document like this lets you compare the same machine's condition a year later, instead of arguing over impressions.

We are the engineers who design and manufacture the Balanset instruments, and we do the on-site balancing ourselves. We are based in Vila Nova de Gaia, near Porto, and we travel throughout Portugal. Vibration diagnostics with a report cost EUR 300 per unit, balancing adds from EUR 250, the minimum invoice for a visit is EUR 500, and the calculator on the website gives you an exact figure for your machine and address.

Plan the work for the off-season. In winter and early spring the equipment sits on the yard, you can go through the flail set at your leisure before we arrive, and the queue for a visit is shorter. In-season, machine downtime costs more than the balancing itself. Several machines on one farm in a single visit come out cheaper per rotor, so it's worth bundling the mulcher, the mower, and the shredder into one inquiry.

• Tell us the machine type: agricultural mulcher, flail mower, residue shredder, haulm topper.
• State the PTO mode the machine runs on, 540 or 1000 rpm, and the drum's operating speed from the manual.
• Describe the event: a lost flail, a partial set replacement, a stone strike, a bearing repair.
• Tell us whether the working-element set is complete and whether it was matched by mass.
• Send photos of the drum from the ends, the bearing housings, and the belt-drive assembly: from these we can choose the correction planes in advance.
• Have the drum washed before we arrive, a level yard, a tractor in working order with the PTO warmed up, and an operator on hand for the runs.

If your machine doesn't need balancing, or it won't help, we'll tell you before the visit. Sometimes the most useful answer sounds like this: go through the flail set by mass, and call us back if the machine is still hammering afterward.

Sources: Balanset-1A manufacturer specification

Frequently asked questions

Related content