Development of an Algorithm to Automatically Detect and Distinguish Squeak and Rattle Noises University of Cincinnati

Format:

Conference/Event

Author/Creator:

Lee, Lee, author.

Contributor:

Kim, Jay

Kim, Kichang

Conference Name:

SAE 2015 Noise and Vibration Conference and Exhibition (2015-06-22 : Grand Rapids, Michigan, United States)

Language:

English

Physical Description:

1 online resource

Place of Publication:

Warrendale, PA SAE International 2015

Summary:

AbstractSqueak and rattle (S&R) noises are undesirable noises caused by friction-induced vibration or impact between surfaces. While several computer programs have been developed to automatically detect and rate S&R events over the years, no reported work has been found that can detect squeak and rattle noises and distinguish them. Because the causes of squeak noises and rattle noises are different, knowing if it is a squeak noise or rattle noise will be very helpful for automotive engineers to choose an appropriate measure to solve the problem. The authors have developed a new algorithm to differentiate squeak noises and rattle noises, and added it to the S&R detection algorithm they had developed previously. The new algorithm utilizes a combination of sound quality metrics, specifically sharpness, roughness, and fluctuation strength. A three-dimensional space defined by the maximum values of sharpness, roughness, and fluctuation strength of the noise are used to differentiate squeak and rattle noises. The developed algorithm has been applied to 86 recorded squeak and rattle noises and the results have shown that the correct type of noise was successfully identified nearly 100% of the time. Also discussed are possible performance improvement and best application of the developed S&R differentiation algorithm

Notes:

Vendor supplied data

Publisher Number:

2015-01-2258

Access Restriction:

Restricted for use by site license