# Screw compressor and turbocharger balancing: what's possible on site

> A screw airend and a turbocharger share one unwelcome trait: their rotors can't be balanced where they operate. We tell you this before the visit, not after. Meanwhile, the vibration on machines like these is most often produced not by the screws or the cartridge at all, but by the driver, the coupling, the bearings, and pulsations. That part we can work with on site, and below we go through exactly how.

**In short:** Honestly: screw rotors and turbocharger rotors are not balanced on site, and we're no exception. The airend is sealed inside a casing with clearances in the hundredths of a millimetre, and intervening in the profiled rotors isn't permitted. A turbocharger rotor is balanced only on a specialised high-speed test bench. On site, we do something else: we measure the vibration, separate the causes using the spectrum (the breakdown of vibration by frequency) and phase (tying the vibration to the shaft's angle of rotation), and balance what's accessible — the drive motor's rotor, the half-coupling, the belt-drive pulley, the cooling fan. Going by our measurement results, that's exactly where most of the vibration on screw units actually sits. If the cause is inside the airend or the cartridge, you get a written report and go to a specialised repair shop with numbers, not guesses.

Source: https://axiline.pt/en/equipment/screw-compressor-turbocharger-balancing/  
Publisher: AXILINE · Vila Nova de Gaia, Portugal · +351 931 831 229 · axilinegeral@gmail.com

## How an airend is built: why the rotors are out of reach

Inside an airend, two profiled rotors turn together — the male and the female. The profiles roll against each other, and the clearances between the screws, the casing, and the end covers run to hundredths and tenths of a millimetre. Thrust bearings hold each rotor's axial position, and even slight wear in them changes the end clearance faster than any imbalance could damage anything. In oil-free machines, the rotors never even touch each other: a separate set of timing gears drives them in sync.

The practical conclusion is simple. Correction planes — the rotor cross-sections where corrective mass can be added or removed — aren't accessible on the screws without fully stripping down the airend. There's nowhere to put weights: screw rotors have no standard balancing land, and metal can't be removed from the profile, since that changes the contact line and the clearances. Stripping and rebuilding an airend calls for special tooling and axial-clearance adjustment, so it's done as a specialised repair, not on the compressor frame in a workshop.

There's good news too. Screw rotors are rigid, short, and factory-balanced on a balancing machine. Inside an oil-flooded airend, they have almost no way to pick up in-service imbalance: no product build-up, no blade erosion, no weld repairs. So when a screw compressor starts shaking, imbalance in the screws themselves turns out to be the cause least often of all.

> How a rigid rotor differs from a flexible one, and why that determines where balancing happens, is covered in a separate article on rigid and flexible rotors. Here, the short version is enough: the screws are rigid, but out of reach.

Sources: [ISO 21940-11:2016](https://www.iso.org/standard/54074.html)

## Where a screw unit's vibration actually comes from

A screw unit isn't just the airend. It's also the drive motor, the coupling or belt drive, the cooling fan, the oil cooler, the frame on its vibration isolators, and the discharge piping. Every element signs its name in the spectrum at its own frequency, and we read the spectrum before we ever reach for weights.

| What the measurement shows | Likely cause | What we do |
| --- | --- | --- |
| The motor shaft's 1x dominates (vibration at its rotation frequency), phase is stable from run to run | Imbalance in the motor rotor, half-coupling, or pulley | We balance on site, in the unit's own bearing supports |
| Second harmonic (a peak at twice the running frequency) and elevated axial vibration near the coupling | Shaft misalignment between the motor and the airend, often after replacing the flexible insert | Shaft alignment, then a verification measurement |
| High-frequency peaks not related to running speed as multiples, the envelope (the signal used to detect bearing defects) is rising | Bearing wear in the airend or the motor | We don't balance. A report and bearing replacement; for the airend, that's a specialised repair |
| A peak at the number of lobes on the male rotor times running speed, harmonics and sidebands rising | Compression pulsations; if it grows from measurement to measurement, profile wear and increased clearances | We check the operating regime and the piping; if the profile is worn, only an airend repair will do |
| Peaks below running frequency (subharmonics) and slapping on the belt drive | Stretched belts, pulley runout, a loose fit | Repair the drive; we balance the pulley after the repair |
| Low-frequency rocking of the whole frame | Sagging vibration isolators, loose fasteners, cover resonance | We replace the isolators, torque the fasteners, then measure again |

> On a typical airend, the lobe-passing frequency falls somewhere around the fourth to sixth running-speed harmonic, and it's always present in the spectrum. What's a warning sign isn't the peak itself, but its growth from measurement to measurement.

Sources: [ISO 13373-3:2015](https://www.iso.org/standard/40840.html) · [ISO 281:2007](https://www.iso.org/standard/38102.html)

## What we actually balance on site on a screw compressor

In every case listed below, the same method applies — the influence coefficient method: a trial weight, a measurement of the response, calculation of the correction mass and angle, fitting, a verification run, and a trim pass if needed (a final precise adjustment with a small weight). The Balanset-1A calculates one and two planes, checks the residual imbalance against the chosen G balance quality grade (the residual-imbalance tolerance), and saves the machine's influence coefficients for future visits.

### The drive motor's rotor

The most common find. Dirt on the cooling fan under its cover, a lost balancing weight, a rotor after remounting. We balance it in its own bearing supports at operating speed; correction planes are the fan and the half-coupling. Weights are mounted with bolts into standard holes or with clips on the blades — welding on aluminium or plastic is out of the question. There's a separate page dedicated to motors after rewinding.

### The half-coupling and the assembled coupling

After replacing the flexible insert or repositioning the half-coupling, vibration often increases: the part ended up in a different angular position. We check the alignment first, then balance the coupling as an assembled unit, usually in a single plane. Weights are bolt-on, into the flange's standard holes.

### The belt-drive pulley

Some screw machines are belt-driven. Before balancing, we check the pulley's runout with a dial indicator, along with the condition of the grooves and the belt tension, otherwise the weights would be correcting for a geometric defect. We balance the pulley right on the shaft: one correction plane for a narrow pulley, two for a wide multi-groove one.

### The cooling fan and the oil-cooler fan

The cooler's axial fan picks up dirt and loses balance just like any other impeller. We balance it on site through the standard access point. We've covered the specifics of fan units in detail on the fan balancing page.

Sources: [ISO 21940-11:2016](https://www.iso.org/standard/54074.html) · [Balanset-1A manufacturer specification](https://vibromera.eu/product/balanset-1/)

## Turbochargers: very high speeds, bench work only

A turbocharger's rotor is a cartridge: the turbine wheel, the shaft, and the compressor wheel inside a bearing housing with floating journal bushings. Operating speed runs from thirty to over two hundred thousand rpm depending on size. The rotor passes through critical speeds (the rotation rates at which the shaft starts to bend) and behaves as flexible. So balancing is only possible on specialised equipment: first the wheels and shaft separately at low speed, then the assembled cartridge at operating speed, on a test bench with its own oil supply unit. Correction there is done by removing metal from the nut and the wheel faces, not with weights.

A portable instrument on the engine can't handle a job like that, and we don't take it on. What we do on site: measure housing vibration, take a spectrum, assess the shaft's axial and radial play, and look for signs of the wheels rubbing against the housing, coking, and oil in the flow path. Then comes the report: cartridge for replacement, or rotor to a specialised turbocharger repair shop.

Process turbocompressors and centrifugal blowers behind a speed increaser (a step-up gearbox) are a different story. The high-speed shaft there is also out of reach, but the unit's low-speed side is open: the motor rotor, the intermediate shaft, the couplings. On site, we separate the shaft frequencies from the speed increaser's gear-mesh frequency, balance the low-speed side, and give a report on the high-speed side. For what we do with the rest of the compressor fleet, see the compressor balancing section page.

Sources: [ISO 21940-12:2016](https://www.iso.org/standard/50429.html) · [ISO 13373-3:2015](https://www.iso.org/standard/40840.html)

## How the visit goes

1. **Request and photos** — Send us the compressor model, photos of the unit with the cover removed, the drive arrangement (coupling or belt), and a description of the symptom: when it started, after what work, and at which regime it's worse. From this information, we'll tell you honestly, before the visit, whether there's anything on your machine to balance, or whether it needs repair right away.
2. **Measurement at operating temperature** — We mount two accelerometers (vibration sensors) on the motor's bearing supports and on the airend casing, magnetically, on cleaned mounting spots, and we record the measurement directions in the report. We aim the laser phase sensor at the reflective mark on whichever shaft is accessible. We record overall vibration (the total level across all frequencies), the running-speed 1x, phase, speed, an FFT spectrum, and the time waveform, both at idle and under load: compression pulsations depend on load, while mechanical causes barely do.
3. **Separating the causes** — We compare 1x with the overall level, check the second harmonic near the coupling, the bearing envelope, the lobe-passing frequency, the fasteners, and the vibration isolators. We assess the levels against the applicable part and edition of ISO 20816, adjusted for the mounting type, but the trend relative to your machine's usual level matters more than the absolute table.
4. **Balancing the accessible planes** — The influence coefficient method: a trial weight, a measurement of the response in amplitude and phase, calculation of the correction mass, fitting, a verification run, and a trim pass if needed. One or two correction planes depending on the component. We check the residual imbalance against the G-grade tolerance.
5. **Report and conclusions** — You receive a report with before-and-after numbers for each point, spectra, and the recorded measurement conditions. If the cause is inside the airend or the turbocharger cartridge, the report will say exactly that, backed up by the spectrum, so the conversation with the repair shop can be specific and to the point.

Sources: [ISO 20816-1:2016](https://www.iso.org/standard/63180.html)

## When we say straight away: this needs a specialised repair

- [x] A metallic howl or grinding noise from the airend, a changed coast-down sound
- [x] Metal shavings or aluminium dust in the oil and on the magnetic drain plug
- [x] Axial or radial play in the turbocharger shaft, signs of the wheels rubbing against the housing
- [x] Coking on the compressor wheel, oil in the flow path, blue smoke from the exhaust
- [x] Rising discharge temperature and motor current together with vibration: looks like profile wear and increased clearances
- [x] High-frequency crackling and envelope peaks at the airend's bearing frequencies
- [x] The compressor has been through a hydraulic shock, condensate ingestion, or a foreign object

> Balancing a machine with symptoms like these would mean wasting your shift and masking a degradation that will end in the airend seizing. We won't sign up for that: measurement and a report first, then repair, and only after that, balancing whatever's accessible.

Sources: [ISO 13373-3:2015](https://www.iso.org/standard/40840.html)

## If the airend does get stripped down: balancing during repair and the Balanset-1A OEM

During an airend overhaul, once the journals are restored and the bearings replaced, the rotors are balanced on a balancing machine. That's the repair shop's job, and there's no point expecting it from a field crew. The same applies to turbocharger cartridges: they're finished off on high-speed test benches with metal removal.

If you repair compressors or turbochargers yourselves, the measuring section of a machine or bench like that can be built around the Balanset-1A OEM. It's the same two-channel measuring module with accelerometers and a laser phase sensor, but without the case, for building into equipment. Calculation in one and two planes, fixed weight-mounting positions, G-grade tolerances, archiving and reports. We can help with selection and integration.

Sources: [Balanset-1A manufacturer specification](https://vibromera.eu/product/balanset-1/) · [Balanset-1A operation manual](https://vibromera.eu/balanset-1a-operation-manual/)

## Price, timing, and how to order

We are the engineers who design and manufacture Balanset instruments, and we do the on-site balancing ourselves. Base in Vila Nova de Gaia near Porto, we travel across all of Portugal.

For the visit to fit into a single day, prepare access in advance: remove or loosen the coupling and fan cover, arrange for a series of short starts and stops, and give the operator a heads-up. We take the measurement on a warmed-up machine under its usual load.

- Balancing an accessible rotor (motor, half-coupling, pulley, fan): packages from 550 EUR per unit (diagnostics 300 EUR + balancing from 250 EUR)
- Minimum invoice per visit 500 EUR, including cases where the measurement shows there's nothing to balance: you still get diagnostics and a written report
- The calculator on the site works out an exact figure for your machine and site

Sources: [Balanset-1A manufacturer specification](https://vibromera.eu/product/balanset-1/)

## Frequently asked questions

**Can an airend be balanced without taking the compressor apart?**

No, and that has nothing to do with the instrument. The correction planes on screw rotors are sealed inside a casing with clearances in the hundredths of a millimetre, there's nowhere to put weights, and metal can't be removed from the profile. The screws are balanced on a machine when manufactured and during an airend overhaul. On site, we find and fix the external causes of vibration, and that's most of them.

**The compressor started shaking after the coupling was replaced. Is this imbalance?**

Possibly, but we check the alignment first. After replacing the flexible insert, two scenarios are typical: shaft misalignment, with a rise in the second harmonic and axial vibration, or the half-coupling reassembled in a different angular position, with a rise in the 1x running-speed component. The first is fixed by shaft alignment, the second by balancing the coupling on site. A measurement tells these two cases apart in a single run.

**Do you balance turbochargers?**

Not on site. The cartridge rotor runs at tens and hundreds of thousands of rpm and is balanced only on a specialised bench with metal removal. On site, we measure vibration, check shaft play and the condition of the wheels, and give a report: cartridge for replacement, or off to a specialised turbocharger repair shop.

**What if you come out and it turns out there's nothing to balance?**

You still get a result: measurements at each point, spectra, cause separation, and a written report you can take to a repair shop or a supplier. A visit like that is closed out at the minimum invoice of 500 EUR. To reduce the risk of a wasted visit, we look at your photos and symptom description in advance.

**Does production need to be stopped for the work?**

Not entirely, no. The measurement is taken on the machine running under its normal regime. Short stops are needed: to fit the sensors and the reflective mark, and to fit the trial and correction weights. That's usually a series of three to six runs of a few minutes each. The longest part is waiting for the machine to warm up to a stable regime.

**What standards do you use to assess compressor vibration?**

Against ISO 20816, using the part and edition applicable to your machine, accounting for power and bearing type. The absolute table is secondary here: the trend relative to the level at which your machine ran smoothly, and the structure of the spectrum, tell you more. For balancing, we check the residual-imbalance tolerance against G grades under ISO 21940.