High vibration on a coating-line blower: one in-situ balancing pass, six points under 3 mm/s

At-a-Glance

The Setup

A 10 HP centrifugal blower, rated 15,500 CMH at 60 mmWC, on a surface-coating line. Not a glamorous duty — a mid-size, medium-pressure air mover that sits behind a booth or an oven and simply has to run.

Which is why it matters. On a coating line the fan is not an accessory to the process; it is part of the process — airflow across a booth governs overspray capture, airflow through a cure oven governs the temperature profile the coating actually sees. So when that fan shakes, there is no luxury of "we'll look at it next shutdown": a vibrating rotor puts bearing life, shaft life and finish quality on the same clock. The requirement on site was to restore continuous running without pulling the rotor out and sending it back to a works balancing machine.

The Complication

The complaint arrived as a bald symptom: vibration level. No cause, no direction, no history.

That is the point at which most fans get blamed for something they didn't do. The obvious suspect is the fan — "you sent us a bad rotor". But a machine that ran well and then vibrates in the field is rarely telling you about what was despatched. It is telling you about everything that has happened since: the truck, the crane sling, the grouting, the alignment, the foundation — and, on a coating line above all, the process itself. A coating-line impeller lives in an aerosol of paint, powder or resin. It deposits, and it deposits unevenly. That condition does not exist on day one; it grows in service. Hence the trap: the fan was fine when it left, is not fine now, and neither fact contradicts the other.

The Diagnosis & Fix

Our method on a field vibration call: measure before concluding, and rule out in ascending order of cost.

1. Characterise, don't guess. Six-point survey first — DE and NDE, three axes each. Unbalance has a signature: high radial, low axial. Misalignment pushes the axial reading up. Soft foot or a foundation problem shows as vertical-to-horizontal asymmetry at one bearing. None of that is separable from a single number shouted down a phone line.
2. Rule out the cheap causes first. Mounting, hold-down bolts, soft foot, coupling condition, ducting strain on the casing. An hour here saves a rotor removal.
3. Balance in place if the signature supports it. In-situ balancing — on the installed machine, in its own bearings, on its own foundation — corrects the rotor as the plant actually runs it, including accumulated process deposit. A works balance on a bare rotor cannot do that.

The machine responded to step 3. A single in-situ balancing pass brought it home — itself the strongest evidence that rotor unbalance, not alignment or foundation, was the dominant contributor. Post-balance readings, as measured:

Six points, all under 3 mm/s. For context, ISO 14694 sets fan acceptance at 4.5 mm/s rigidly mounted, BV-3 (alarm 7.1, shutdown 9.0) and 6.3 mm/s flexibly mounted (alarm 11.8, shutdown 12.5). At a peak of 2.9 mm/s this blower sits in the unrestricted long-term operation band on either basis. Balance quality is assessed to ISO 21940.

The chain below is the generic pattern this class of failure follows — reusable, not a forensic finding on this rotor.

The Result

• What we proved: after one in-situ balancing operation, all six measured points read below 3 mm/s, peak 2.9 mm/s — inside ISO 14694 acceptance for both rigid BV-3 and flexible mounting, with margin.
• What we changed: the rotor was corrected on site, in its own bearings. No removal, no transport, no works balancing slot. The plant kept its fan.
• What we fed back into our own system: log the as-found vibration alongside the after readings; record balancing planes, correction masses and residual unbalance, not just the mm/s; state the measurement condition on every report; add an impeller build-up check to the PM checklist for coating-line fans.
• What the record does not confirm: the as-found figures were never captured, so we cannot say "we took it from X to Y" — only that it finished under 3 mm/s. Readings were taken in a lifting condition, so an at-duty confirmation is not separately evidenced. And the file carries no formal customer sign-off or closure rating. We will not invent one. What we stand behind is the corrective action performed and the numbers measured after it.

That last bullet is deliberate. A service story that always ends in applause is a marketing story, not an engineering one.

The Takeaway + Call to Action

The transferable lesson is about where you point the finger. A fan that ran well and then starts to vibrate is almost never a fan that was built wrong — it is a fan changed by its transport, its installation, or its process. Diagnose in that order, and survey six points before you take a decision. On any fan handling paint, powder, resin, tar or sticky dust, treat the impeller as a deposit-and-wear item: a scheduled clean and a periodic vibration check cost far less than a bearing, a shaft, or a shift of scrapped finish.

If a fan on your plant is shaking, running a hot bearing, or simply not doing what its nameplate says — we will come and measure it. We service centrifugal fans and blowers of any make, not only our own: in-situ balancing, vibration surveys, alignment, bearing and rotor work, and performance troubleshooting.

— Jitamitra Electro Engineering · Technical Services

Engineered for Every Application.