Noise and vibration from a blower inside a vacuum package: the shaft told the story

At-a-Glance

The Setup

The customer is an equipment OEM. They build a vacuum-department package — a complete skid their own end-user buys and runs inside a process plant. Our blower is one component inside that skid: a 3 HP direct-driven centrifugal fan rated 1100 CMH at 300 mmWC on a 60 Hz supply.

Being a component inside someone else's package is what makes these calls awkward. The OEM's brand is on the skid, and the blower is the loudest rotating thing in the enclosure — so the blower gets blamed. The stakes were never just a bearing. They were our customer's package, in front of their customer.

The Complication

"Noise and vibration" is the least diagnostic complaint in our trade — the symptom of practically every rotating-equipment fault there is. The obvious suspect, the one most people reach for first, is imbalance: a fan that shakes must be out of balance, therefore the impeller is bad, therefore it is a manufacturing problem. That reflex is wrong more often than it is right, and it was wrong here.

The second complication was procedural. The end-user had already dismantled the motor before the complaint reached us, so there was no running machine to take a reading from — no vibration velocity in mm/s, no sound level, no record of the operating point when the noise began. Severity could not be graded against the ISO 14694 acceptance bands (rigid mount 4.5 accept / 7.1 alarm / 9.0 shutdown mm/s; flexible 6.3 / 11.8 / 12.5). We say so plainly: a case study quoting a vibration number it never measured is lying to you. What we had instead was a set of strip-down photographs. So we read the photographs.

The Diagnosis & Fix

The method on a bare "noise and vibration" complaint is to work outward from the rotating assembly, ruling out in order of how cheap the evidence is.

1. Imbalance of the impeller. The default assumption — and the only one that would have pointed back at our own works. But imbalance leaves a signature: it does not scratch a shaft, and it does not put rust anywhere. Balance is controlled to ISO 21940 at build. Nothing in the images supported it.

2. The bearing. Cheap to check on a photograph, because a bearing in distress writes on the shaft. It had. The images showed scratches on the shaft, and rust. That is not an assembly-defect signature. That is a bearing that has corroded and gripped, dragging on a shaft still turning inside it.

3. The impeller-to-hub locking. A parallel, independent vibration path: a taper-lock hub held by grub screws. If a screw backs off, the impeller vibrates however good the bearing is. We asked for photographs of the locking arrangement to judge grub-screw condition — and were explicit that this branch stayed open, not cleared.

The recorded root cause was the bearing: it had jammed due to rust — a coherent single source of both reported symptoms on a direct-driven rotor.

Corrective action specified: replace the bearing. Interim action on the open second branch: tighten the taper-lock grub screws. Preventive action, now generalised into our failure-mode library: for direct-driven fans in wet, wash-down or moisture-laden duties, specify effective bearing moisture exclusion — sealing, drain path, correct grease, correct orientation — and instruct the end-user to keep water off the bearing housing.

The Result

• What we proved: a bearing seized by rust from in-service moisture ingress, evidenced in the scratching and corrosion on the shaft. A lifecycle failure in operation, not a works defect.
• What we ruled out: impeller imbalance as the driver. The reflex answer was the wrong answer.
• What we left open: the taper-lock grub-screw condition, as a parallel vibration path. We flagged it, asked for the photographs, and did not close it out.
• What we changed: bearing moisture exclusion is now a specification question for any wet duty, and a taper-lock torque-and-check step belongs in commissioning.
• What the record does not confirm: no post-repair re-check, no confirmation the bearing was actually changed, no customer sign-off. The replacement was specified, not witnessed by us. We will not claim a closure the file does not carry.

The Takeaway + Call to Action

A fan that suddenly gets loud is usually telling you about its bearings, not its balance. Balance faults are born at build and show up on day one; bearing faults are earned in service and show up in year three. If a machine ran quietly for a season and then started shouting, look at what the environment did to it — water, dust, heat, a missing seal — before you look at what the factory did to it. And read the shaft: on a rotating assembly, it is the only component that records its own history.

We service centrifugal fans and blowers of any make — strip-down diagnosis, bearing and shaft assessment, balancing to ISO 21940, vibration assessment against ISO 14694, and performance testing to IS 4894 / ISO 5801 / AMCA 210 method. If a fan in your plant has started making a noise it did not use to make, send us the photographs before you send us the fan.

— Jitamitra Electro Engineering · Technical Services

Engineered for Every Application.