Jabil Strategic Capability Development – Evaluating Acoustic Emissions on Gears

Fri Mar 25 23:00:00 EDT 2016

Introduction and Objective:

Edward J. Collins, our dedicated principal test engineer has led the project to evaluate acoustic emissions as an analysis technique for reducing noise (dba) within precision electric motor and gear box assemblies used in medical device applications. Generally, noise related failures are disassembled and various mechanical and electro-mechanical parts are replaced until the final assembly passes the noise specification.

The project is looking at how to characterize each component of the motor/gear box assembly using acoustic emission so that primary contributors to increased noise is fully characterized with design, material, and supplier corrective actions being implemented where possible to enable higher yields during manufacturing. Edward has described some of the procedures, tools and analysis techniques used when evaluating the acoustic emissions of a gear driven product in a whitepaper.

Process Flow:

For the evaluation of the acoustic emissions of a motor and gear driven product there is a process that will facilitate the diagnostic effort. The following process illustrates our recommendation for the diagnosis of these types of product.

  • Step 1 :  Choose acoustic emissions equipment & software.
  • Step 2 :  Select accelerometer & microphone mounting method.
  • Step 3 :  Setup tachometer input for accurate RPM measurement.
  • Step 4 :  Calculate gear ratio.
  • Step 5 :  Calculate the fundamental order frequencies.
  • Step 6 :  Select the proper FFT (Fast Fourier Transform) sampling rate.
  • Step 7 :  Select linear averaging for the collection of the FFT data (Pulse Acoustic Software).
  • Step 8 :  Perform FFT analysis.
  • Step 9 :  Normalize the FFT data samples if necessary.
  • Step 10:  Failure discrimination and diagnosis.


X-section of actual motor/gear box assembly under evaluation


While this evaluation project focuses on the use of acoustic emissions measurements as a diagnostic tool, gear noise reduction is best addressed in the design phase. Establishing optimized tooth counts, ratios, pitch, pressure angle, helix angle, tooth proportions, and modifications early in the design cycle can provide a robust and cost-effective solution to noise reduction. Listed below are some examples of noise inducing attributes and components.

Feature Description
Armature Shaft The shaft clearances in the motor bearings if excessive, will allow the armature shaft to bounce back and forth in the clearances. The frequency of these bounces will be random but tend to be one to five times the shaft rotating frequency in small motors.
Excessive Tooth Wear Listen for a rumbling noise. Compare both sides of the tooth profile. Increased noise floor.
Frictional Forces These are of a lesser consequence than the factors that affect motion transmission. What is notable is that they change direction at the operating pitch diameter.
Housing Housing design and material yield resonant frequencies that often promote noise propagation. The mounting of the gear shafts and their arrangement and stiffness in the housing also play a role in noise transmission.
Interference Any form of interference would be a major noise contributor. This interference is usually a result of the tip on one gear touching the root area of another.
Lead Errors While not as great a noise factor in the meshing of the gears, heavy contact on the end of the tooth will present a noise issue. If the lead error changes direction multiple times per revolution, a disturbing rattle or shuttling noise could result.
Motor The motor itself inputs a torque ripple that is a function of its electrical and magnetic cogging attributes.


Summary / Benefits:

Acoustic emissions testing using sound and vibration measurements can be accurate and repeatable on gear assemblies. The vibration emissions from gear box samples varied only 5 dB when tested in a lab environment with minimal control of the ambient acoustic conditions. The calculated gear ratios and fundamental frequencies matched the measured FFT spectra with a high degree of accuracy. The microphone based measurements are more likely to require noise isolation in the form of an anechoic chamber and can be just as repeatable in that environment. Gear defects were very detectable, and by enhancing encoder signals could become even more detectable certainly down to the cog level. For more details, please contact Jasmine Ooi for facilitating technical discussions/interest relating to acoustic emissions and specific applications with appropriate Jabil ETS acoustic emissions application developers/expertise.


1. White paper by Edward J. Collins, Principal Test Engineer, Jabil ETS. 2. Brüel & Kjær 3. Richard R. Kuhr (ABA-PGT, Inc).