NVH & Specialized Testers

BTI engineers and manufactures manual, semi-automatic, and fully automatic industrial NVH (noise vibration and harshness)

Most end-of-line rejections in a modern driveline are not failures. They are sounds. The part works, it meets every dimension, and a customer would return it anyway — which means the acceptance criterion is a measurement of noise and vibration, not of geometry.

Balance Technology Inc. engineers and manufactures NVH test stands and specialized testers for most part types and sizes, at any production rate and any level of automation.

Why an Overall Number Is Not Enough

The instinct is to set a vibration limit and pass anything below it. That catches gross faults and misses the ones customers complain about, because human hearing responds to tones, not to total energy. A unit can sit comfortably under an overall limit and still produce a whine that is instantly objectionable.

The measurement that matters is order analysis: resolving vibration by its relationship to shaft speed rather than by absolute frequency. Each mechanism then lands in a predictable place.

• First order — unbalance. Rises with the square of speed.
• Tooth-count order — gear mesh. Tonal, and noticed at low amplitude.
• Bearing orders — non-integer, characteristic of the bearing geometry, which is what makes bearing faults identifiable at all.
• Electromagnetic orders — slot and pole passing in electric machines.

That separation is the whole value of the test. A stand that reports one number tells you a unit is bad. A stand that resolves orders tells you why, which is the difference between sorting parts and fixing a process.

Speed Sweeps, Not Spot Checks

Resonances only appear when something excites them. A unit tested at one speed can pass and then sing at a speed the test never visited, because a structural or mounting resonance sat just outside the test point.

Sweeping through the operating range — and often in both directions, since some behaviors are load- and direction-dependent — produces a map of order amplitude against speed rather than a single value. That map is also what makes the test diagnostic: a resonance shows as a peak that stays at a frequency while the orders move through it.

Electrification Changed the Requirement

A combustion engine generates a broad, loud, irregular signature that masks a great deal of driveline noise. Remove it and tones that were inaudible for decades become warranty conversations.

Two things follow for the test stand. Acceptance limits have tightened, in some cases by a wide margin, on components whose design did not change. And the stand itself has to be quieter than the part — its own drive, bearings, mounting and structure must not contribute at the orders being measured, which is a harder engineering requirement than it sounds.

The Stand Is Part of the Measurement

This is the point most often underestimated. Mounting stiffness changes what the unit does: tooling more rigid than the vehicle produces readings the vehicle will never reproduce, and tooling too compliant hides real faults. The drive introduces its own orders. Sensor location determines which path you are listening to, and structure-borne and airborne measurements answer different questions.

Repeatability has to be established before limits mean anything. If the same unit measured twice gives results that differ by a meaningful fraction of the tolerance, the limit is sorting noise rather than parts.

Specialized Testers

Beyond NVH, many characteristics are only meaningful under operation, and BTI builds dedicated stands for them — torque-to-turn, backlash, end play, functional performance, leak, high-speed spinning and destructive testing among them. These are engineered to the part rather than selected from a catalog, because the fixture and the loading condition are usually the hard part of the problem.

Setting Limits Honestly

NVH limits are frequently inherited rather than derived, and the inherited number is often the amplitude at which somebody once complained. That is a legitimate starting point, but it should be replaced as data accumulates.

The defensible method is to build a population, correlate the measurement against subjective assessment or field return data, and set the limit where the two separate. A limit set tighter than the measurement system can repeat will reject good parts and teach the line to ignore the stand.

Configurations

Layouts range from single manual stations for development and validation through to fully automatic end-of-line cells with load, test, sort and data collection, and vary to accommodate any production rate.

NVH measurement also combines readily with balancing and gaging on one machine — see combination equipment — which is common where a unit needs characterizing several ways before it ships.

Specifying a Stand

What determines the machine: the part and how it mounts; the speed range and whether load must be applied; which orders and frequency range matter; whether structure-borne, airborne or both; the acceptance criteria and whether they already exist; cycle time; and what has to be recorded and for how long.

Related Equipment and Applications

BTI also engineers resonant frequency measurement systems, balancing machines and spin test systems. NVH testing is commonly specified for electric vehicle drivelines, axles and RDU components and transmission components.