A Systematic Approach to Preparing Drive Files for Squeak and Rattle Evaluations of Subsystems or Components 2007-01-2269
Many decisions need to be made when test track data is used to set up Squeak & Rattle evaluations of subsystems or components. These decisions are judgment-based so different people with different backgrounds and experience levels will make different decisions - few of which can be called right or wrong - but they are different which causes problems.
Squeak & Rattle evaluation has become more scientific in recent years as subjective evaluation has been replaced by quantitative methods like N10 Loudness and shakers have become quiet. It is the authors' contention that the variations caused by different judgment calls can no longer be tolerated. Therefore a methodical process was developed which assures that different people will get the same results from the same set of test track data.
Factors which were included in this process include (but are not limited to):
Determining if random vibration replication is appropriate for a particular test track
How to obtain a reasonable worst case random vibration level
How to obtain an appropriate PSD shape
Frequency spacing and data sampling rates
What portion of the test track data to use
How and when to combine various test track surfaces into one composite test
How to decide and, if appropriate for S&R, implement a non-Gaussian distribution (via Kurtosis).