AXILINEOn-Site Balancing of Centrifuge Drums and Baskets
A centrifuge basket can shake for three different reasons, and weights only fix one of them. We arrive with a two-channel vibration analyzer, run several tests on the clean rotor, and use phase repeatability to answer whether it's imbalance or unevenly settled product. If it's imbalance, we balance right on the machine, at the standard spots on the rim. We're based in Vila Nova de Gaia near Porto and travel throughout Portugal.
Three sources of vibration on a basket, and weights fix only one
The first source lives in the product: the cake settled unevenly on the screen, dried out on one side, or part of it came away during discharge. The second lives in the tooling: the filter cloth seated with a fold, got wetted unevenly, or the perforation is clogged in spots. The third is actual rotor imbalance: shell erosion, deformation after an impact, a loosened fit on the spindle, or a lost factory weight.
Rotor imbalance is constant and is removed with a correction mass. Product and cloth produce a variable imbalance: it arrives with the load, grows through the spin cycle, and clears after washing. A variable quantity can't be compensated with fixed weights. You'd end up with a machine that shakes harder than before on a clean basket.
| What you observe | Usual source | What to do |
|---|---|---|
| The level rises through the spin cycle and drops after washing | Uneven cake layer on the screen | A loading schedule, the distributor, the feed rate |
| Vibration jumps from batch to batch, with the 1x phase different every time | The product settles differently, or the cloth is shifting | The cloth, the retaining hoop, feed evenness |
| It rose right after the cloth or screen segments were replaced | The set differs in mass, or it seated with a fold | Reposition the cloth, then a trim balance (a quick touch-up) |
| 1x is high on the empty, dry basket, with the phase staying put from run to run | Mechanical rotor imbalance | Balancing in its own bearing housings |
A noticeable 2x component — at twice the rotation frequency — together with axial vibration at the drive points to shaft misalignment. Peaks at frequencies that aren't multiples of the running speed, plus a hot upper bearing, point to the spindle bearings. How to read a spectrum is covered in a separate article of ours.
Sources: ISO 13373-3:2015 · ISO 281:2007
How to separate uneven loading from rotor imbalance
This is the central question on a machine like this, and it's settled by measurement. The procedure below takes half an hour and saves a whole shift.
- 01
Wash the basket and let it dry
A baseline measurement is only meaningful on a clean rotor. A precoat layer, a damp cloth, and cake trapped in a pocket are all variable mass. If a precoat layer is left in place as part of the process, tell us in advance: then we'll work with it, but on a scheduled basis.
- 02
Three runs in a row on the empty basket
We bring it up to operating speed, record the running-speed component's amplitude and phase at each bearing housing, stop, and repeat. Three runs give you repeatability; one run gives you nothing.
- 03
Watch the phase, not the amplitude
If the 1x phase holds within roughly ±10-15° and the amplitude reproduces, the source is rigidly tied to the rotor, and a weight will remove it. If the phase drifts by tens of degrees, the mass is settling differently every time, and weights are useless.
- 04
A run with a load through the full cycle
We load the machine normally and record the level through the phases of the cycle: run-up, spin, wash, braking. The difference from the clean basket is the process's own contribution, in mm/s.
- 05
Decision
If rotor imbalance is greater than the process contribution: we balance. If the process contribution is greater: we remove the baseline component and say plainly that from here it's a question for the loading, the cloth, and the distributor.
We don't start trial runs until we've seen a repeatable phase. The influence coefficient is the machine's response to a trial weight; calculated from a drifting phase, it produces a weight that lands in the wrong place.
Which drums and baskets we take on
Filtering centrifuges
A perforated shell, a screen, or filter cloth. Imbalance comes from perforation clogged in spots, a torn cloth, worn screen segments, or erosion near the discharge zone. The basket is short, and a single correction plane — the section of the rotor where weights are fitted — is usually enough.
Sedimentation drums
A solid wall, where the cake sticks in patches and stays put longer. A baseline measurement only on a washed rotor — otherwise you'd lock in a random buildup pattern with the weights.
Suspended (batch) centrifuges
A basket on a vertical spindle, with the machine hanging on three tie rods fitted with springs and dampers. It has its own causes: a shifted reinforcing ring, a residual precoat layer, a worn scraper knife, or a sagged damper on one of the tie rods.
Food-industry lines
Sugar, starch, salt, dairy, and juice production. A stainless basket, CIP washing, and a ban on foreign parts in the product zone. We choose how to fasten the masses to fit the procedure, not the other way around.
Textile spin-dry centrifuges
Loading is more often manual, and the laundry tends to bunch up rather than settle evenly. The share of variable imbalance here is especially large. Balancing only helps if there's a repeatable baseline component underneath that noise.
After screen segments are replaced, the shell is straightened, or the discharge zone is hard-faced, the previous factory weights are no longer valid. The drive is a separate system: pulley imbalance can't be removed with weights on the basket. We don't balance the sealed bowls of disc-stack separators on site.
Three-point suspension, a flexible shaft, and critical speed
A suspended centrifuge is made compliant on purpose. The basket sits on a relatively flexible shaft, and the machine hangs on three tie rods fitted with dampers. Above the first critical speed — the speed at which the rotor passes through resonance — it self-centers about the axis of its own masses, and part of the unbalance damps itself out. That's why the machine genuinely wobbles on run-up and settles down at operating speed.
Measuring and balancing has to be done on the stable plateau of the operating speed, outside the resonance zone. The peak on run-up can't be removed with weights, and trying to remove it will spoil the operating condition. If the machine has started lingering in the resonance zone longer than usual, look for the cause in the drive, the tie rods, and the dampers. A rotor like this formally counts as flexible, and it's more correct to assess it against the applicable part of ISO 21940 for flexible rotors.
- We place sensors on the spindle bearing housings, not on the cover and not on the suspension frame. A hanging casing moves as a single body and would show the suspension's motion, not the force at the bearing.
- The primary point is the upper spindle bearing, measured radially. We take a second reading at the lower bearing if there's access. We keep the direction the same from run to run.
- A magnet won't hold on austenitic stainless steel. We mount the sensor on an adhesive pad or with a clamp, after cleaning the spot.
- We stick the marker on the spindle above the upper bearing, or on the drive pulley, so the laser phase sensor has a clear line of sight. The cover is opened only on the stopped, locked-out machine.
- The speed needs to hold on the plateau for at least a minute. If the cycle doesn't allow a hold, we ask for the service mode to be engaged.
A casing on dampers produces higher vibration velocities than the same machine on a rigid base, so a zone table can't be picked at random. We choose the applicable part and edition of ISO 20816 for the specific machine and record the points, frequency band, operating mode, and mount type.
Sources: ISO 20816-1:2016 · ISO 21940-12:2016
Where we fit weights: the rim, standard spots, and mass limits
A perforated basket is a thin shell, loaded by the product's centrifugal pressure. It works under hoop tension, and every perforation hole is already a stress concentrator. We don't drill new holes in the shell, and we don't weld anything onto it.
- The top rim or the reinforcing ring. The most common standard spot: threaded holes or a groove for a bolted weight. Fixed-position mode is handy here, where the instrument outputs a position number and a mass instead of an angle.
- The bottom ring or the basket hub. A second correction plane, for when the basket is deep and one plane isn't enough.
- A standard balancing ring with movable weights, if the design has one.
- Removing metal instead of adding mass, only in a zone the manufacturer permits and only on substantial elements.
- Welding is ruled out. A weld changes the structure of the stainless steel and creates a corrosion zone and a gap where product accumulates.
- We take the mass limit from the manual and work out the load in advance. It's the rim that should carry it, not the shell.
- If the mass exceeds the allowed limit: we split the correction across two planes, or decline and explain why.
- Actual values on a short basket are modest — tens of grams at a radius of a few hundred millimeters. A calculation calling for hundreds of grams almost always means you're compensating for product or a bent shell with the weight.
- We lock every weight in place: a lock nut, a lock washer, or wire locking.
We use the same radius for the correction mass as for the trial weight. A weight that comes loose here isn't just a lost balance — it's a projectile.
Sanitary requirements and the product zone
- Not a single foreign part appears in the product zone. If the only available mounting spot is inside the basket, the weight doesn't go there.
- The weight and fastener material matches the rotor's material for corrosion resistance: stainless steel, no zinc, no lead, no galvanized hardware.
- No adhesives, no tapes, no magnetic weights. They won't survive CIP washing, temperature, and washing chemicals.
- No gaps or stagnant pockets: the weight sits flush, with no burrs on the edges.
- We count tools and hardware before and after the work, and remove the reflective marker from the product zone.
- We agree the fastening method with your quality team in writing, before the visit.
If your procedure doesn't allow any of the available fastening methods, we say so in advance. In that case the visit turns into vibration diagnostics: you get a report with numbers instead of weights you'd have to remove later.
How the visit proceeds
- 01
Conversation before the visit
You send us the model, speed, basket and product type, photos of the rotor and rim, and the manual's page on balancing and mass limits.
- 02
Inspection on the stopped machine
Anchor and suspension tie-rod tightness, damper condition, the basket's fit on the spindle, rim runout, the condition of the cloth and perforation, and the clearance to the scraper knife. Some call-outs end right here.
- 03
Sensors and the marker
Two accelerometers on the spindle bearing housings, and the laser phase sensor on the reflective marker. We clean the mounting spots and fix the sensor with an adhesive pad or a clamp.
- 04
Baseline measurement on the clean basket
Several runs in a row: overall vibration (the total level across all frequencies) and the running-speed component with phase, speed, the FFT spectrum (the breakdown of vibration by frequency), and the time waveform for each bearing.
- 05
Trial weight
A weighed mass at an agreed standard spot at a known radius. We consider a 20-30% change in 1x amplitude or 20-30° in phase to be valid.
- 06
Correction and check run
The software outputs a mass and an angle, or a fixed-position number. We fit the weight, lock it, and repeat the measurement at the same points under the same conditions.
- 07
A cycle with product, and the report
We run through a working cycle and show how much the process itself adds. We save the influence coefficients: the next trim balance after a cloth change will take one run instead of three.
The seven steps fit into a single visit if the mechanics are sound and there's access. A bearing defect, a sagged damper, or a loosened fit changes the plan: repair first, then a repeat measurement.
When balancing won't help, or isn't needed
- Vibration appears only with a load and disappears on the clean basket. That's the load, not the rotor.
- The 1x phase doesn't repeat from run to run: the influence coefficients would come out random.
- The basket's fit on the spindle has loosened, or the reinforcing ring has shifted.
- The shell is deformed, there's a crack at a perforation hole, or through-wall erosion. A rotor like that is taken out of service: weights would mask a dangerous condition.
- The manufacturer prohibits fitting masses, or only permits it at their own service center.
- The machine gives no speed plateau: the cycle is built on run-up and braking with no hold.
- The dominant vibration comes from the spindle bearings, shaft misalignment at the drive, or worn suspension dampers.
- The sanitary procedure doesn't allow any of the available methods for fastening masses.
In any of these cases you get measured numbers, spectra, and an assessment of the probable cause. We'd rather say so before the visit, based on the model and photos.
Sources: ISO 13373-3:2015 · ISO 21940-12:2016
What you get, the 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 our website works out the exact amount for your machine, the number of rotors, and the distance.
The engineers who come out are the same ones who design and manufacture the Balanset instruments and do the on-site balancing themselves. We work with the Balanset-1A: two accelerometers, a laser phase sensor reading a marker, a two-channel USB module, and software on a laptop. Balancing in one and two planes by the influence-coefficient method, fixed positions and drilling calculations, trim balancing, tolerance to grades G, and an archive with reports. A Balanset-1A OEM version without the case is available, for building into rigs and machines.
- A report with before-and-after numbers: overall vibration and the running-speed component in mm/s RMS (root-mean-square value) at each bearing, phase, speed.
- FFT spectra and time waveforms, taken at the same points under the same conditions.
- The process's contribution shown separately: the difference between the clean basket and the loaded one, through the phases of the cycle.
- The mass, radius, and angle or position number of every weight, plus the fastening and locking method.
- A condition assessment by zone under the applicable part of ISO 20816, stating the edition, points, frequency band, and mount type.
- A residual-unbalance tolerance calculation against grade G under the applicable part of ISO 21940, if your team needs it.
- This machine's saved influence coefficients, for future trim corrections.
We don't promise a final vibration figure up front: it depends on the condition of the rotor, the bearings, and the suspension, and on the manufacturer's restrictions. What we do promise is measured before-and-after numbers, and a straight answer on whether the standard is achievable without a repair.
Sources: ISO 20816-1:2016 · ISO 21940-11:2016 · Balanset-1A operation manual · Balanset-1A manufacturer specification
Frequently asked questions
How can you tell it's the product shaking, not the drum?
By phase repeatability on the clean rotor. Wash the basket, let it dry, and run three starts in a row at operating speed. If the running-speed component's amplitude and phase repeat, the source is rigidly tied to the rotor — that's mechanical imbalance. If the phase lands somewhere new every time, the cause is the loading, the cloth, or residual cake.
Can the basket be drilled and a weight bolted on?
Not on the perforated shell. It works under hoop tension, every perforation hole is already a stress concentrator, and a new one would become a crack initiation site. Weights go into the standard spots: threaded holes or a groove on the top rim, the reinforcing ring, the bottom ring. If none of these exist, we agree a solution with the manufacturer, or we stop at diagnostics.
The machine started shaking after the filter cloth was replaced. Is that imbalance?
Check the cloth itself first. A fold, misalignment, an unevenly seated retaining hoop, or a mass difference from the previous set all produce exactly this picture. Reposition the cloth and repeat the run on the clean, dry basket. If the vibration stays and the phase is stable, it's imbalance, and a trim correction using the saved influence coefficients will take a single run.
The centrifuge wobbles on run-up but is calm at operating speed. Is that normal?
For a suspended machine, that's normal behavior. The rotor sits on a flexible shaft and runs above the first critical speed, so on run-up it passes through a resonance zone where the amplitude rises and the phase swings quickly. What matters is that the zone is passed through quickly. Balancing needs to be done on the operating-speed plateau, and the run-up peak should not be removed with weights.
We're a food producer with CIP washing. Can weights even be fitted?
Yes, but in a specific way and outside the product zone. We choose the weight and fastener material to match the rotor's material for corrosion resistance, with no zinc or lead. We don't use adhesive, taped, or magnetic weights: they won't survive the washing and the temperature. The weight is fitted flush, with no gaps, and locked in place. If none of these options work, we do vibration diagnostics only.
How many correction planes does a basket need?
Usually one. A filtering centrifuge's basket is short, with a length more often less than half its diameter, and it behaves like a disc: a single mass on the top rim is enough. We use two planes when the basket is deep, when a high level remains at the second bearing after correction, and when a second standard mounting point exists.
Related content
Decanter Centrifuge Balancing: What Can Really Be Done On Site
Honest answer: not always. We balance the decanter bowl on site if the manufacturer permits fitting weights, standard balancing positions exist on the end hubs, and the measurement confirms that the bowl's own 1x running-speed component dominates — that is, vibration at its rotation frequency. The scroll can't be balanced on site as a matter of principle: its correction planes — the spots where balancing weights go — are hidden inside the bowl, and reaching them means a full teardown of the rotor. So some visits to decanters end not with weights but with a measurement that separates out the causes: cake, worn flighting, bearings, the gearbox, or imbalance. You get a report with numbers that lets the conversation with the service shop stay concrete. We usually understand your case before the visit even happens, from the model, photos, and vibration trends.
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.
Building Your Own Balancing Machine: Supports, Bed, Drive, and Measurement System
Yes, you can build a rig on your own, and the soft-bearing (above-resonance) scheme is the most accessible way to do it. You need four things: a rigid, heavy bed, supports with a known suspension natural frequency (the frequency at which the support's moving part oscillates on its own) well below the running speed, a drive with stable speed, and a two-channel measurement system with a phase-angle sensor. The electronics get solved by buying a ready-made measurement core; everything else has to be designed and verified by you. The key point that separates a working rig from an expensive piece of hardware: acceptance testing by geometry, by dynamics, and against a reference rotor with a known trial unbalance.
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.