AXILINEBalancing shafts, drums, rolls and rollers at the point of operation
A long rotor has its own logic. A conveyor drum, a calender roll, a cardan drive and a machine-tool spindle shaft don't behave like a fan impeller: one correction mass is almost never enough for them. And some call-outs don't end in balancing at all, but in a runout or a caked-on layer of product that we find instead. We arrive with a two-channel analyser, put sensors on both bearing supports, and start by answering the question of whether this is unbalance at all. If it is, we balance in two planes and hand over a report with the before-and-after numbers.
Symptoms: what brings people to call us
Unbalance in a long rotor rarely arrives alone. A conveyor drum starts to knock after spillage builds up on it, an intermediate shaft starts humming after a half-coupling is replaced, a calender roll leaves a streak on the web. Vibration on machines like these is almost always mixed, so our first step is never a trial weight — it's a measurement.
Call us if you recognise your machine in the list below. The sooner we take readings, the cheaper the fix: balancing takes up part of a shift, while straightening a shaft or replacing a wrecked bearing stops the line for days.
- Vibration at the bearing supports has risen above what's normal for this machine, even if it's technically still within limits.
- Humming or runout appeared after cleaning, washing, a lagging replacement or a shaft repair.
- The conveyor belt has started to wave, the drum knocks as it coasts down, the take-up carriage visibly has a life of its own.
- A periodic streak or thickness variation has appeared on the web, film or sheet, spaced at the roll's rotation.
- The cardan drive vibrates at certain speeds and settles down above or below them.
- The machine-tool spindle has started leaving waviness on the workpiece, and surface finish has got worse.
- The support bearings run hot, grease gets forced out, and the support fixings keep working loose.
- Rollers are noisy, the web wanders, and the feed pitch on the packaging line has drifted.
The once-per-turn component is covered in more detail in a separate article on overall vibration, 1x and phase. In short: balancing only brings down the part of the level that sits at the rotation frequency. If a drum has 8 mm/s of overall vibration (the total level from all causes combined) and only 2 mm/s of it is at 1x, weights will remove about a quarter, and the rest has to be tracked down elsewhere in the mechanics.
What equipment in this family we balance
These machines all have one thing in common: an elongated rotor. That's what leads to two correction planes, sensitivity to bow, and the fact that unbalance is often picked up from outside rather than being built in from the start. Below are the subgroups, and what creates unbalance in each of them.
Drive, intermediate and long-run shafts
Drive shafts, intermediate shafts between motor and machine, long transmission shafts. Unbalance is most often introduced by a repair: a new half-coupling, a built-up journal, a replaced section of shaft. We fit the mass on flanges and half-couplings.
Cardan shafts
Drive cardan shafts and power take-off shafts. Besides unbalance, yoke phasing and working angles come into play here: incorrectly assembled yokes produce vibration that no weight will remove.
Spindle shafts and production machine-tool shafts
Spindle shafts, machine-tool drive shafts, shafts of grinding and sharpening heads. Speeds are high, the tolerance is tight, and correction masses are small. We balance them in their own bearings, together with the arbor if it stays fitted during the job.
Drive, conveyor and take-up drums
Drive, bend and tail drums on conveyors, take-up drums on slide rails. Unbalance comes from caked-on spillage, uneven lagging wear, water trapped inside a hollow shell, and factory-fitted weights that have shifted.
Drying and process drums
Dryer drums, granulator, mixer and cooler drums, jacketed process drums. The hot environment produces thermal bow, and condensate and product residue inside change the unbalance from one run to the next.
Print cylinders, roll-mill and calender rolls, papermaking machine rolls
Print cylinders, roll-mill rolls, calender rolls, press and dryer rolls on papermaking machines. The working surface is untouchable, so welding or drilling into the barrel is out of the question — correction goes only through the fitted balancing points.
Guide, conveying and nip rollers
Guide rollers, roller-table conveying rollers, nip rollers. Small ones are often cheaper to replace; on larger, rubber-covered rollers balancing pays off. We first check the runout of the covering, the fit on the shaft, and the bearings.
Textile-machine and packaging-line shafts
Textile machine shafts, pulling and winding shafts, packaging and labelling line shafts. Unbalance is introduced by wound-on fibre, adhesive and film. The masses are small, but the requirements for smooth running are high.
If your machine isn't named in the list, that's not a refusal. Three things decide it: whether the rotor can safely be run up to its working speed, whether sensors can reach the bearing supports, and whether there's somewhere to fit a weight.
What we check before balancing on rotors like these
On an elongated rotor, we reach for a dial indicator before we reach for weights. The reason jobs often fail is geometric: the shaft has runout or bow, and balancing doesn't cure either one. A weight corrects how mass is distributed, not the shape. If a roll barrel runs with runout, the streak on the web will stay whatever the residual vibration.
We measure runout at several cross-sections along the length, on the journals and on the working surface, while turning the shaft slowly by hand. We look separately at thermal bow: a dryer drum or a hot roll has a different shape cold than it does running, so we base the decision on the warmed-up machine at its working condition.
Then comes the product. A build-up on a drum or roll is unbalance that will come back, so balancing over a dirty surface makes no sense. We ask you to wash the rotor down and work with a clean surface — otherwise you'll be calling us again in a week.
- Shaft runout and bow, checked with a dial indicator at several cross-sections, turning slowly.
- Condition of the working surface: build-up, uneven lagging wear, erosion, signs of previous truing.
- Cavities inside the drum: water left over from washing, condensate, product residue.
- The fit on the shaft: key, interference fit, a loosened hub assembly, signs of the hub having turned on the shaft.
- Bearing-support fixings, soft foot, the condition of the frame and the take-up slide rails.
- Drive alignment. Shaft misalignment calls for shaft alignment, not balancing.
- Bearings. High-frequency peaks and a rising crest factor (the ratio of the peak level in the signal to its average level) point to a defect that balancing must not be allowed to mask.
- The drive: belts and tension, pulley runout, gearing, the kinematics of the cardan drive.
- Rotation speed. Below 600 rpm, the once-per-turn component falls below the usual 10–1000 Hz measurement band, so we widen the measured range.
Sources: ISO 13373-3:2015 · ISO 281:2007
One plane, two planes, or flexible-rotor mode
The rotor's shape decides it. The length-to-diameter ratio on shafts and drums is usually well above half, so one mass physically cannot remove the vibration at both supports at once. It will lower the level at the near support and raise it at the far one. The rule of thumb is simple: anything longer than a disc, we balance in two planes, which is exactly why we bring two sensors.
A second effect shows up on very long rotors. The shaft stops being rigid: at running speed it bends on its own, and its bow changes with rotation speed. Rotors like this are called flexible, and a correction found at one speed works worse at another. In that case we measure at several running conditions, track how the phase behaves, and fit the correction to the working speed. If the working speed sits close to a critical speed (the speed at which the shaft goes into resonance), the honest answer is often that you need a balancing rig with run-up, or a change of operating regime.
| Rotor | Shape and speed | Number of planes |
|---|---|---|
| Roller, short drum, pulley | L/D up to 0.5 | Often one is enough |
| Conveyor drum, take-up drum | L/D from 1 to 5, usually 30–200 rpm | Two, on the end discs |
| Drive and intermediate shaft | L/D above 5, up to 1500 rpm | Two, on flanges and half-couplings |
| Cardan shaft | L/D above 10, up to 3000 rpm | Two, at the ends of the shaft tube |
| Calender roll, papermaking machine roll, long-run shaft | L/D 8 and above, working speed close to critical | Two, flexible-rotor mode, readings at several speeds |
| Machine-tool spindle shaft | L/D from 3 to 8, above 3000 rpm | Two, small masses, tight tolerance |
The L/D rule itself is covered in the article on choosing the number of planes; here we only give the conclusion for this family of machines. On site, the cost of the question is measured in runs: one plane means two runs, two planes mean three runs plus the stops needed to move the trial weight.
Sources: ISO 21940-12:2016 · ISO 21940-11:2016
How the work goes on site
- 01
Measurement before work starts
We fit two accelerometers to the bearing supports, close to the bearings, on cleaned-off mounting spots. We stick a reflective marker for the optical tachometer onto the trunnion, the drum end, the half-coupling or the pulley. We record the overall level, the once-per-turn component, the phase (the angle showing where in the turn the vibration peaks), the speed and the spectrum (how the vibration breaks down by frequency) at the working condition.
- 02
Decision: balance or repair
We work out the share of 1x in the overall level and look at the spectrum. At the same time we check runout, fixings and signs of misalignment. If something other than unbalance dominates, we say so before any trial runs.
- 03
Choosing the correction planes
We work out where a mass can physically be fitted: the drum's end discs, a flange, a half-coupling, the roll's fitted balancing holes. We agree with you on the fixing method and on what must not be done to that surface.
- 04
Trial weight in the first plane
We fit a weighed mass at a known radius and run the machine. We count the response as usable if it's at least 20–30% by amplitude or by angle. If the response is weak, we increase the mass and repeat, rather than work out a correction from noise.
- 05
Trial weight in the second plane
We move the weight and run the machine once more. The software gets the influence coefficients for both planes — how a weight in each one changes the vibration at the supports — and calculates the correction mass and angle for each.
- 06
Correction and check run
We fit the weights at the positions given, remove the trial mass and run a check. If there's a small amount left to reach the target, we add small trim masses, usually in a single extra pass.
- 07
Measurement after the work and the report
We repeat the measurement at the same points, in the same direction, at the same running condition. We record the before-and-after numbers, the weights fitted, their radii and positions. We save the influence coefficients so the next trim balance can be done without trial runs.
We work with the Balanset-1A: two vibration channels, an optical tachometer reading a reflective marker, a laptop with the software, a fixed-position mode and a drilling calculation. For drums and rolls, fixed-position mode is especially handy: instead of an angle and a protractor, you get the number of a fitted hole and a mass.
Sources: Balanset-1A operation manual
Where we fit weights: each group has its own rules
It isn't physics that decides the correction method, it's the construction and what interference is allowed. Welding a plate to the end disc of a conveyor drum is fine. Doing the same to a print cylinder or a calender roll ruins an expensive part and gives you grounds for an argument with the manufacturer. So we agree the method with you before any trial runs, not after.
| Rotor group | Where we fit the mass | What we don't do |
|---|---|---|
| Drive, intermediate, long-run shafts | Bolted weights on flanges and half-couplings, clamps, removing metal from thickened sections | We don't weld near fitted journals and seals |
| Cardan shafts | Plates on the shaft tube near its ends, clamps | We don't touch splines, universal joints or factory weights unless necessary |
| Conveyor drums, take-up drums | Welded or bolted weights on end discs and hubs, drilling into the disc | We don't weld into the lagging or the drum barrel |
| Drying and process drums | Masses on end flanges and ribs, outside the heated zone and away from the jacket | We don't drill a jacketed shell or weld on the seams of pressurised components |
| Print cylinders, calender rolls, papermaking machine rolls, roll-mill rolls | Only the fitted balancing holes, grooves, rings and end recesses | No welding and no drilling of the working surface, ever |
| Guide, conveying, nip rollers | End caps and trunnions, screw-mounted weights | We don't weld onto a rubber-covered or polished surface |
| Textile and packaging-line shafts, machine-tool shafts | Fitted rings and holes, small screw-in weights | We don't break the coating on the working surface |
If a precision roll has no fitted balancing points, on-site balancing is off the table. A shaft like that gets removed and balanced in a rig on its trunnions, or handed back to the manufacturer. We'll say so straight away, not after three trial runs.
When it won't work on site, and what to do instead
Turning down a balancing job is a result too, and it saves you a shift. Below are the cases that come up most often in this family of machines.
- The shaft has runout or bow. The fix is straightening, machining true, or replacement. Balancing comes after the geometry is restored, not instead of it.
- The drive is misaligned. First soft foot and shaft alignment, then a repeat measurement, and only after that a decision on balancing.
- The working surface is untouchable and there are no fitted balancing points. The rotor gets removed and balanced in a rig.
- The rotor is flexible and the working speed is close to critical. On site the result would be unstable — you need a run-up rig, or a change of operating regime.
- Product keeps building up. Balancing would buy you a week, so cleaning and fixing the cause of the build-up come first.
- There's water or condensate inside the hollow drum. Readings drift from run to run — drainage comes first.
- The rotor can't be brought to a stable running speed, or there's no access to both supports and the correction planes.
- The bearings are worn out. Balancing would mask the level without restoring the bearing life.
A full rundown of situations where balancing doesn't help is covered in a separate article. On a visit like that, you get diagnostic results and repair recommendations instead — cheaper than balancing a machine that doesn't need weight at all.
What you get
The main result is numbers you can put in front of someone. We hand over before-and-after readings at the same points, at the same running condition, in the same frequency band. Without that, the comparison means nothing, because the vibration level changes with load and temperature.
- A report with overall vibration and the once-per-turn component for each support, before and after the work.
- The correction masses fitted, their radii, angles or fixed-position numbers.
- The spectrum and a condition assessment: bearings, fixings, signs of misalignment, resonance.
- A finding on geometry: runout, bow, the condition of the working surface and the fit on the shaft.
- An assessment against the applicable part of ISO 20816 as a reference point, and, where needed, a residual unbalance calculation against a balance quality grade.
- The influence coefficients saved for this machine, so the next trim balance goes ahead without trial runs.
- Recommendations: what to do before the next shutdown, and what to watch for in vibration monitoring.
We use the ISO 20816 zones and the G balance quality grades (the standard limits on allowable residual unbalance) as a reference, and we state the applicable part and edition of the standard, because there are exceptions there for power, speed and bearing type. For contractual acceptance, we fix the target in advance: measurement points, frequency band, running condition and target level.
Sources: ISO 20816-1:2016 · ISO 21940-11:2016
Price and how to book
vibration diagnostics with a report costs 300 EUR per unit, balancing adds from 250 EUR, and the minimum invoice per visit is 500 EUR. The calculator on the website gives you the exact figure: it accounts for the number of rotors, travel and the scope of work. Diagnostics without balancing is worthwhile too, because you get an answer on exactly what to repair before you order parts.
We're the engineers who design and manufacture the Balanset instruments, and we balance with them ourselves on site visits. We're based in Vila Nova de Gaia, near Porto, and we travel throughout Portugal.
- The type of machine and rotor: drum, roll, intermediate shaft, cardan, roller, spindle.
- The speed and drive type: direct, belt, through a gearbox, power take-off.
- The approximate length and diameter of the rotor, so we can estimate the number of planes in advance.
- Photos of the supports, the rotor ends, and the places where a weight could be fitted.
- The measured vibration level, if a reading has already been taken, and the measurement points.
- Constraints: what must not be touched, site access requirements, the shutdown window.
Preparation cuts the visit down by almost half: access to both bearing supports, cleaned-off spots for the sensors, a clean rotor surface, the ability to reach working speed, and to stop as many times as needed. More on this in the article on preparing equipment for balancing.
Sources: Balanset-1A manufacturer specification
Frequently asked questions
Can a conveyor drum be balanced without removing it or taking the belt apart?
In most cases, yes. We work on the drum as installed, in its own bearing housings: sensors on the bearing housings, a marker for the tachometer on the trunnion or half-coupling, weight on the end discs or hubs. Removal is needed when there's no access to the ends, the lagging is unevenly worn, or the shaft has runout.
Why is a long shaft almost always balanced in two planes?
Because on an elongated rotor, unbalance is spread along the length and creates not just a force but a couple. One mass balances out the force, but the couple stays: the level at the near support falls while the far one rises. Two planes give two independent masses and remove both components. That's why we bring a two-channel instrument to machines like these.
The shaft has runout. Will balancing help?
No, and it's better to establish that before any trial runs. Runout and bow are geometry, while balancing works with mass distribution. A weight might lower the readings slightly at one speed, but it won't remove the streak on the web, the knocking, or the uneven wear. The right order is: straighten or true up the part first, then balance on the restored geometry.
Weights can't be welded onto a print cylinder or a calender roll. What do you do then?
We work only through the fitted balancing points: screw holes, rings, grooves, end recesses. If none exist, on-site balancing isn't carried out, and we say so before any work starts. Then the shaft is removed and balanced on a rig on its trunnions, where correction is done by a method the construction allows.
The cardan shaft is vibrating. Is that unbalance?
Not necessarily. A cardan drive has a kinematic irregularity that produces a second harmonic, and it has yoke phasing: if the yokes were assembled turned relative to each other, vibration shows up without any unbalance at all. We first check the assembly, the working angles, the spline joint and the universal joints. Balancing is worthwhile once the once-per-turn component genuinely dominates.
The drum only turns at 60 rpm. Is there any point balancing it?
There's a point to it, but the approach is different. At low speed, vibration velocity is small, and the usual 10–1000 Hz band simply cuts off the once-per-turn component, since 60 rpm is only 1 Hz. We widen the lower end of the range and look at more than just velocity. On drums like this, the vibration is often not unbalance at all but the belt joint, build-up, or worn-out bearings, and that comes out in the measurement.
Related content
On-Site Balancing of Driveshafts and Propeller Shafts
Yes, we balance driveshafts and propeller shafts on site, assembled on the machine, without removal. Four conditions apply: the shaft is assembled to its alignment marks, there's no play in the U-joints or splines, the shaft reaches stable running speed, and the running-speed component 1x — vibration at the rotational frequency, which is what imbalance produces — dominates the vibration. We correct in two planes, at the ends of the tube near the yokes. If the vibration sits at the second harmonic from the working angles, or if there's play in the joints, we'll say so before any trial runs: what's needed there isn't a weight but correct assembly and repair.
On-Site Balancing of Combine Threshing and Chopping Drums
Yes, we balance threshing drums, forage-harvester chopping drums, and straw choppers on-site, with no removal from the combine. Conditions: a complete, matched set of rasp bars or knives, the drum washed clean of crop residue and dried out, an unbent shaft, and access to the drum ends through the hoods, inspection hatches, or a lowered concave. The drive from the combine's engine holds a steady speed, which only helps the balancing. If the set of working elements is mismatched, or the shaft is knocking after a foreign object went through, we'll say so after the first measurement, and we won't use weights to compensate for something that needs matching or straightening instead.
Shaft Misalignment: Types, Signs, and How to Check It
Shaft misalignment is a mismatch between the axes of two shafts joined by a coupling, in the machine's running condition. It comes in parallel (axes offset), angular (axes tilted), and combined form, which in practice covers nearly every real case. With an instrument you'll see a noticeable 2x component (vibration at twice the running speed), elevated axial vibration, a phase shift of around 180° across the coupling in the axial direction, and a level that drifts as the machine warms up. Misalignment is confirmed not by vibration but by direct measurement of the geometry — with dial indicators, the reverse-indicator method, or a laser system — and only after you've ruled out soft foot.
Describe your equipment and the problem
We'll answer your questions, clarify the details, and let you know what's needed for an estimate and a visit.