A blower that tested clean and still shook: noise and vibration traced to the site support, not the rotor

At-a-Glance

The Setup

The customer builds paint-shop and pretreatment equipment — coating booths, pretreatment tunnels, curing ovens. The fan was a 30 HP centrifugal blower rated 34,000 CMH at 120 mmWC, handling air up to 40 °C.

On that kind of line, airflow is not a detail; it is the process. Booth airflow governs overspray capture and film quality; pretreatment and oven airflow govern how the panel dries and cures. A blower that is unwell — or is merely believed to be unwell — puts the whole line under a question mark, and a question mark becomes a stoppage. So when the site reported the machine as noisy and shaking, it was urgent.

The Complication

The complaint was specific: noise and vibration on the running blower, with photographs, a nameplate image and two short videos attached. No duty shortfall. No bearing failure. No process-air complaint. The machine was moving the air; it was simply misbehaving mechanically — or appeared to be.

The obvious suspect in any noisy, shaking fan is the rotating assembly: balance, bearings, alignment, a balance weight shifted in transit. That is where most engineers look first — and where the customer's imagination goes immediately.

But the obvious suspect had an alibi. The same machine had been dynamically tested before dispatch, and its vibration report was already in the customer's hands: V = 2.7, H = 1.0, A = 2.0 mm/s. For a rigidly-mounted fan, ISO 14694 category BV-3 puts acceptance at 4.5 mm/s, alarm at 7.1, shutdown at 9.0. Those readings do not sit near the accept line — they sit comfortably under it, on all three axes.

A machine that tests clean at works and shakes in the field is telling you that whatever changed between those two states, it is not the rotor.

The Diagnosis & Fix

We worked outward from the rotor rather than inward from the noise.

1. Balance and rotor condition — ruled out. The pre-dispatch dynamic test (balance per ISO 21940; performance tested to IS 4894 / ISO 5801 / AMCA 210 method) put the machine well inside BV-3 acceptance on all three axes. Nothing in the complaint — no bearing heat, no bearing noise, no duty loss — contradicted that record.
2. Duty and process — ruled out. No airflow or pressure shortfall was reported. A fan fighting a duty it cannot meet usually announces itself in the ammeter first, and it had not.
3. The interface between machine and building — examined. Our Technical Services engineer opened a direct line to the site and asked the question that separated the two states: what is this machine standing on?

The answer was a 50 mm angle support. A 30 HP centrifugal blower is not a light dynamic load, and a light angle section under it is a low-stiffness base. A soft base need not be broken to cause trouble; it only has to flex and amplify near the running frequency. The rotor stays inside tolerance, and the structure turns a compliant machine into a noisy one.

Our recommendation: replace the 50 mm angle with a 100 mm channel support beneath the blower, sized to carry the dynamic load and stiffen the base out of the amplifying range. No spare part was consumed. The fix was structural, and it sat on the site side of the flange.

The Result

• We proved the machine was not the fault. V = 2.7 / H = 1.0 / A = 2.0 mm/s against a BV-3 accept line of 4.5 is a dated measurement, not an opinion.
• We identified the cause: an under-sized site support carrying a 30 HP dynamic load on a 50 mm angle.
• We issued the corrective action: a 100 mm channel in place of the angle, with the reasoning stated plainly to the site.
• The customer accepted the recommendation and undertook to strengthen the structure.
• What our record does not confirm: we hold no post-modification vibration reading and no written closure. We will not tell you the shaking stopped when our file does not say so. The corrective action was delivered; the confirmation loop is open.
• What we changed: for field-mounted heavy blowers we now issue a minimum support-stiffness specification (channel, not light angle) with the GA package, and press for a post-installation vibration check — because a works test report proves the machine, not the mounting.

The Takeaway + Call to Action

A fan is only half of a vibration system. The other half is whatever it is bolted to — and that half is usually built by someone else, on a busy site, in a hurry. If a machine left the works inside ISO 14694 BV-3 and shakes at site, the question is not "what is wrong with the fan" but "what has changed" — and the base, the frame and the duct restraints all sit ahead of the impeller on that list. Get the works test report out before you get the spanner out.

We service and diagnose industrial fans and blowers of any make, not only our own. If a machine is noisy, hot, shaking, drawing more current than it should, or simply not doing the duty it was bought for — bring it to us. We will look at the whole system, tell you what we find, and put the reasoning in writing.

— Jitamitra Electro Engineering · Technical Services

Engineered for Every Application.