Developing Robust Vibration Excitation and Control Methods for Evaluating Rattle Noise in Automotive Components

The authors participated in a task force that was required to develop a repeatable, dependable, and reliable test procedure to compare, rate, and evaluate the severity of rattles. The assemblies involved in the study are designed and manufactured by different companies and are tested by different people on test equipment and instrumentation from different suppliers. The challenges therefore, were considerable and involved both the vibration inputs and responses as well as the acoustic responses. At the beginning of this activity, it was observed that different test labs using the same Ford vibration specs were obtaining different sounds from the same test item! Clearly, this was unacceptable and the test methods had to be improved and standardized.

This paper focuses on vibration related to rattle testing. The particular assemblies used in this study were seat belt retractors. However, the lessons learned can be applied to vibration testing of any automotive component that rattles (e.g., HVAC modules, instrument panels, ashtrays, radios, etc).

Challenges in producing appropriate vibration conditions in a test lab included:

1. 1
   minimizing the sound levels generated by the shaker itself (without test item) while creating the vibration
2. 2
   selecting the type of vibration to use: sine tones vs. random vibration vs. time histories
3. 3
   out-of-band energy caused by rattles in the test item which “shake” the shaker
4. 4
   minimizing or eliminating the resonances of the excitation system
5. 5
   undesirable, off-axis motion of the exciter that results from offset payloads
6. 6
   minimzing instrumentation problems caused by measuring relatively low vibration levels (ground loops, noise floor, etc).

Firm specs were established on all these parameters, and qualification tests for the test system(s) were developed. The end result was a specification that contained enough detail and guidance to ensure that good tests could be run by different test labs. This produces more consistent results that will be directly comparable across different models from different seat belt retractor suppliers.