# On-Site Belt-Drive Pulley Balancing: Geometry and Belts First, Weights After > You changed the belts on a fan, and vibration went up. Or a pulley started humming after the grooves were re-machined, after it was refitted onto a new bushing, or after a bearing change. We come out and balance the pulley right on the shaft, in its own bearing supports, but first we check runout, fit and tension: on a belt drive, half of "imbalance" cases turn out to be geometry. We're based in Vila Nova de Gaia, near Porto, and travel throughout Portugal. **In short:** Yes, we balance a belt-drive pulley on site, without removing it from the shaft. Conditions: the fit on the key, taper bushing or clamping sleeve is tight and doesn't turn, the rim's radial and face runout are within the manufacturer's tolerance, and the belts are a matched set with tension set correctly. A thin pulley is closed out with a single correction plane (weights fitted at one rotor cross-section) in two to three runs; a wide multi-groove pulley, or a shaft assembled with a pulley and coupling half, sometimes needs two. If the rim runs out or the grooves are worn shiny at the bottom, weights won't help: refitting or replacing the pulley comes first, and we'll say so before starting work. Source: https://axiline.pt/en/equipment/on-site-belt-drive-pulley-balancing/ Publisher: AXILINE · Vila Nova de Gaia, Portugal · +351 931 831 229 · axilinegeral@gmail.com ## Symptoms: how a pulley specifically gives itself away A pulley gives itself away by being tied to an event. Vibration doesn't build up gradually but shows up after a specific operation: belts were changed, the pulley was refitted, the grooves were re-machined, or it was removed and put back on during a bearing change. If the noise appeared the same day, look for the cause in whatever hands touched. The second sign: the vibration lives at the bearings of the shaft the pulley sits on, and changes noticeably when the belts come off. On a belt drive there are three suspects at once: the pulley, the belts, and the fit. Telling them apart takes measurement, not listening. - [x] Noise appeared after a belt change, a pulley refit, or groove re-machining - [x] Vibration at the drive bearings drops sharply if the belts are removed and the motor is run on its own - [x] Belts run hot, squeal on start-up, rubber crumbles off the sides, black dust collects under the drive - [x] One belt in a multi-groove set sags noticeably more than the others - [x] The pulley rim wobbles when the shaft is turned slowly - [x] Rust-coloured dust sifts out from under the hub or taper bushing — a sign of fretting, corrosion from micro-movement in a loosened fit - [x] You replace the drive-side bearings more often than the ones at the free end of the shaft - [x] Vibration along the line between the pulleys is noticeably higher than across it > The last point is worth remembering. Pure imbalance gives similar levels at the bearing in the horizontal and vertical directions. Strong directionality along the drive line points to pulley eccentricity (its centre offset from the rotation axis) or tension pulsation, and weights aren't the main fix here. ## Pulley construction, and where its imbalance comes from A pulley is most often a cast-iron casting: a grooved rim, a hub, and a disc or spokes between them. Part of the imbalance is built in at the casting stage: uneven rim wall thickness, an offset core, blowholes. Machining finishes the grooves and the bore, but the rim's inner surface often stays as-cast, and the non-uniformity lives there for the pulley's whole service life. The second group of causes shows up at installation. Every type of fit adds its own mass shift relative to the rotation axis. ### Key and clearance fit Clearance in a keyed joint shifts the pulley toward one side of the shaft. The key itself is also a mass, and the pulley doesn't always have compensation built in for it. After every reinstallation, the assembly's imbalance is different. ### Taper bushing A split bushing is tightened with two or three screws. Uneven tightening seats the pulley eccentrically; under-tightened screws let it spin under load. The split itself and the screws shift the centre of mass: on a new bushing that's accounted for, on a worn one it no longer is. ### Clamping sleeve A self-centring sleeve holds without a key, by friction, and centres better. But the screws need to be tightened around the circle in several passes to a specified torque. One over-tightened side is eccentricity all over again. ### Groove wear Grooves wear unevenly around the circumference and between grooves. The belt sits deeper over part of a revolution, and tension pulsates at the rotational frequency. On the instrument it looks like imbalance, but it's fixed by replacing the pulley. ### Repair re-machining Turning removes metal relative to the axis the pulley was clamped on in the chuck. If the lathe setup had runout, the new geometry is eccentric to the shaft. After any re-machining we check the assembly with a dial indicator and balance it again. ### Pulley assembled with the shaft and coupling On a fan or transmission shaft, the pulley sits alongside the coupling half and the rotor. We balance the assembly as a whole: the parts' imbalances add up vectorially, direction included, and parts balanced separately still produce a new residual once assembled. ## A sequence you can't reorder Balancing compensates for mass asymmetry. Runout, a loosened fit, and a faulty drive are different physical causes with vibration at the same frequency. A weight can partly mask them at one operating point, but the compensation falls apart the moment tension or load changes. So the sequence is strict. - Geometry. A dial indicator on the rim and face, a full turn of the shaft. We check radial and face runout against the manufacturer's tolerance, usually tenths of a millimetre depending on diameter - Fit. Tightening torque of the bushing or sleeve screws, play at the key, signs of fretting. A pulley that moves gets refitted first - Belts. Worn belts, or ones of mismatched length, only get replaced as a set with a single batch marking - Tension. Set according to the belt manufacturer's data: by deflection under a specified force, or by the frequency method based on span vibration - Relative position of the pulleys. We line them up in one plane with a straightedge, a string, or a laser tool. Where there's a coupling to a rotor, we separately assess shaft alignment - Only now, balancing. Everything above changes either the vibration itself or the system's stiffness, and with it the influence coefficients — the relationship "weight fitted, vibration changed by this much" that the instrument uses to calculate the correction > Doing it in the reverse order doesn't save time, it doubles the work. Balance first, then change the belts: tension and stiffness change, and part of the result is gone. Balance first, then refit the pulley: the angular position of the masses is different, and the imbalance is new. Any work on the drive after balancing means a confirmation reading, and often a repeat correction too. How to distinguish imbalance from misalignment by phase is covered in a separate article. ## How pulley balancing proceeds on site 1. **Dial indicator before electronics** — The machine is stopped and locked out. We measure rim and face runout, check the fit, the groove profile with a gauge, and the tension. This is also where we decide where the weight will physically go and whether a drill will be needed. 2. **Supports and the mark on the correct shaft** — We mount two accelerometers (vibration sensors) on the bearing supports of the shaft carrying the pulley, radially, on bare metal near the bearing. We stick the reflective tape onto this same shaft. Tape on the motor shaft would give the phase of the wrong rotor: on a belt drive the shafts turn at different speeds. 3. **Baseline measurement** — Running speed, steady state. We record overall vibration in mm/s RMS (root-mean-square value), and the amplitude and phase of the running-speed component 1x — vibration exactly at the shaft's rotational frequency; the phase shows its angular position relative to the mark. We also record the spectrum (vibration broken down by frequency) and the time waveform. We check speed against the transmission ratio from the pulley diameters. If 1x doesn't dominate, we stop and explain why. 4. **One run per plane** — We number the spokes or mark the rim into fixed positions. A weighed trial weight is attached at position 1 with a magnet or a clamp. The pulley is light, the amplitude-and-phase response is usually strong, and a valid run comes out on the first try. 5. **Drilling or a bolt** — The software gives mass and position, and for metal removal it calculates the hole diameter and depth. We drill the rim or fit a bolt, and remove the trial weight. A confirmation run at the same speed and the same tension. 6. **Figures before and after** — A small trim correction if needed. Then the report: readings at the bearings before and after, spectra, weight masses and positions, and the results of the geometry checks. We save the influence coefficients in the archive, so a repeat balancing job on this assembly will go ahead without trial runs. > Typical duration on an accessible pulley: two to four hours with four to six machine stops. The drive guard and lockout procedures eat up the most time, not the calculation. ## One plane or two: the thin disc and its exceptions An ordinary V-belt pulley is a thin disc: the width is noticeably less than the diameter, and the length-to-diameter ratio is far from 0.5. The mass is concentrated in one plane, static imbalance dominates, and a single correction plane is enough. How the number of planes is chosen in the general case is covered in our article on single- and two-plane balancing. Two situations break the rule. First: a wide multi-groove pulley with six to eight grooves. It gains axial extent, and with it a couple imbalance — a pair of imbalances at different ends that rocks the shaft. Second: we're not balancing the pulley but the whole shaft, with a pulley at one end and a coupling half or impeller at the other. Then we spread the correction planes along the length of the shaft: one on the pulley rim, the other on the coupling-half bolts or the impeller disc. A two-channel measurement settles the question with data, not guesswork. Similar 1x phases at the two bearings indicate static imbalance and a single plane. Opposite phase indicates couple imbalance and two planes. We make the decision after the baseline run, not from a catalogue. ## Where to place the weight: why drilling the rim beats welding On a pulley it's almost always more convenient to remove metal than to add it. The reason is the material. Grey cast iron doesn't take well to welding: it needs preheating, special electrodes, and slow cooling, otherwise the weld zone cracks. Field-welding a weight onto a cast-iron pulley is a gamble, and it's one we don't take. Drilling solves the problem cleanly. The instrument converts the correction mass into a hole diameter and depth for a known material. We drill in the thickened part of the rim or in the disc, on the side where the hole doesn't touch the grooves and doesn't collect dirt. - First choice: factory balance holes, where the manufacturer provided them - Drilling the rim from the inside or the face side, clear of the belt tracks. Several shallow holes are better than one deep one - A bolt with washers and a lock nut in a hole in a spoke or the disc: fast, reversible, and the mass is easy to adjust during trim correction - A steel welded pulley: welding on a plate is acceptable, with a runout check after it cools - Magnetic weights only as trial weights - we remove them before the confirmation run - We don't touch: the bottom and walls of the grooves, the thin edge of the rim, or the area near the taper bushing's split > The grams removed by drilling don't affect the rim's strength, as long as you don't exceed the depth calculated by the software. Weight-attachment methods on different rotors are covered in detail in a separate article. ## The belt and tension: what masquerades as imbalance The drive adds its own lines to the spectrum, and the main one is the belt-pass frequency. It's calculated from the kinematics and always lies below the rotational frequency of both shafts. A defective spot on the belt passes both pulleys within one revolution, so its second harmonic often stands out in the spectrum. A vibrometer will show a high overall level that's easy to blame on imbalance. The spectrum sorts everything out in a single measurement: a detailed breakdown with formulas is in our article on belt-drive vibration. Tension acts more subtly. It isn't vibration but a steady radial force that loads the bearings of both shafts and sets the system's stiffness. For ball bearings, life falls off roughly as the cube of the equivalent load, and an ISO 281 calculation shows this directly once you plug in the actual force from an over-tightened belt. For balancing, what matters is something else: influence coefficients taken at one tension are no longer accurate at another. So tension is set before balancing and isn't touched afterward. Sources: [ISO 13373-3:2015](https://www.iso.org/standard/40840.html) · [ISO 281:2007](https://www.iso.org/standard/38102.html) ## When on-site pulley balancing won't help We close out some requests without fitting any weights, and we say so within the first hour on site. ### The rim runs out beyond tolerance Runout above the manufacturer's tolerance means an eccentric fit or deformation. A weight will remove the vibration at one operating point, but the belt will keep pulsating in tension every revolution and eating away at the grooves. Refitting or replacement comes first. ### The grooves are worn through A shiny groove bottom means the V-belt has sunk in and is riding on its bottom instead of its sides. The drive is faulty, and readings on it won't repeat. The pulley needs replacing, and we balance the new one. ### The bushing or hub spins The 1x phase wanders from run to run, the influence coefficients don't converge, and the instrument sees it right away. There's no point balancing a pulley that moves - the assembly needs refitting. ### The belt slips The driven shaft's speed drifts with the load, the spectrum lines blur, and the phase can't be trusted. Tension or a new set of belts comes first, then a measurement. ### A crack in a spoke or the rim A cast-iron pulley with a crack isn't balanced, it's taken out of service. A weight would mask the one external symptom and leave a part spinning that could fly apart. ### An acceptance record with residual unbalance is required Acceptance with a measured residual unbalance in g·mm/kg by grade G is done on a lathe. On site we calculate the tolerance against the applicable part of ISO 21940 and confirm the result with the residual vibration: for operational purposes that's usually enough. Sources: [ISO 21940-11:2016](https://www.iso.org/standard/54074.html) ## What you get, and how to order The result of the work is numbers, not impressions. You get a report with before-and-after readings for each bearing support: speed, overall vibration in mm/s RMS, 1x amplitude and phase, spectra, weight masses and positions, and data on runout, fit and tension. We assess residual vibration against the zones of the applicable part and edition of ISO 20816, and calculate the residual-unbalance tolerance by grade G. On price: vibration diagnostics with a report 300 EUR per unit, balancing from 250 EUR, minimum invoice for a visit 500 EUR. The calculator works out the exact amount, depending on the number of rotors and correction planes, access, and how far the site is. We're based in Vila Nova de Gaia, near Porto, and travel throughout Portugal. The engineers who design and manufacture Balanset instruments come out and do the balancing themselves. The Balanset-1A can also be bought, so you can balance pulleys with your own staff: two accelerometers, a laser phase sensor, a two-channel USB module, and software with fixed positions, drilling calculations, trim balancing, and a report archive. - [x] The ability to stop the machine four to six times for a few minutes each - [x] A removable or openable belt-drive guard - [x] Photos of the pulley, hub, and machine nameplate with the request: they show the fit and the correction planes - [x] Belt markings and the manufacturer's tension data, if available - [x] A person on the operations side responsible for start-up and lockout Sources: [ISO 20816-1:2016](https://www.iso.org/standard/63180.html) · [ISO 21940-11:2016](https://www.iso.org/standard/54074.html) · [Balanset-1A manufacturer specification](https://vibromera.eu/product/balanset-1/) · [Balanset-1A operation manual](https://vibromera.eu/balanset-1a-operation-manual/) ## Frequently asked questions **We changed the belts, and vibration didn't go away. Is that pulley imbalance?** Not necessarily. First check rim runout and the fit with a dial indicator: vibration after a belt change often means the pulley was refitted with the bushing skewed, or the tension changed. If the geometry is within tolerance and the running-speed component of the shaft carrying the pulley dominates the spectrum, then yes, we balance. **Can the motor pulley be balanced with the belts removed?** The driving pulley, yes: the motor turns it fine without the drive, and such a run is useful diagnostically, since vibration without the belts is the pulley's own imbalance together with the motor's rotor. The driven pulley doesn't turn without belts, so we balance it under running conditions with tension set. **Why doesn't balancing hold on a pulley with a taper bushing?** Most often the bushing has loosened or was tightened askew. The pulley shifts gradually under load, the angular position of the masses changes, and the compensated imbalance comes back. Signs: rust-coloured fretting dust near the hub and a 1x phase that drifts from run to run. Refitting with the correct screw-tightening torque comes first, balancing after. **Where should the reflective tape go on a belt drive?** Only on the shaft whose pulley you're balancing. The motor shaft and the driven-machine shaft turn at different speeds, and tape on the wrong shaft would give the instrument the phase of the wrong rotor. After sticking it on, we check the measured speed against the calculation from the pulley diameters. **One belt in the set is more worn than the others. Can just that one be replaced?** No. A new belt is shorter and stiffer than the stretched ones, it will take on most of the load and quickly stretch out itself, while tension across the grooves keeps pulsating and ruins the reading. Replace as a set with a single batch marking, set the tension, and only then balance. **Can a correction weight be welded onto a cast-iron pulley?** It's not worth it. Grey cast iron cracks in the heat-affected zone under field welding, and instead of balancing you'd end up with a defective part. On a cast-iron pulley we work by removing metal: we drill the rim or disc to a diameter and depth calculated by the software, or fit a bolt in a hole in a spoke. Welding is acceptable only on steel welded pulleys.