Parametric Accelerated Life Testing (ALT) is proposed as a new reliability technique to improve the design of a new compressor system subject to repeated pressure loading, as a way to detect and minimize fatigue failure due to design faults in mechanical systems. The approach includes: (1) a parametric ALT schedule focused on the BX lifetime of the product, (2) load analysis, (3) a tailored sample of parametric ALTs with design changes, and (4) an assessment of whether the BX lifetime goal is reached by the final design(s) of the product. A newly-designed compressor in a French refrigerator was tested as a test case. To maximize its energy efficiency, the compressor was redesigned. Nevertheless, the refrigerator models rendered undesirable noises and vibrations in the field with this compressor. Consumers have now demanded that their refrigerators be replaced. The outcome was 0.35G in an anechoic chamber when the sound level of the problematic refrigerators was registered for vibration. The noise originated in the reciprocating compressor after closer examination of the refrigerator, where there was friction between the upper compressor shell and the stator frame in the compressor. To test any design defects, the compressor noise needed to be experimentally replicated. A parametric ALT was then performed after the compressor’s mechanical pressure loads were analyzed in a simple vapor-compression cycle. The failure shapes were identical to those of the failed products from the field in the first ALT. The stator frame in the compressor system was reshaped as a corrective measure to increase the minimum distance between the upper shell of the compressor and the stator frame. There were no issues during the second ALT. The lifetime of the newly built compressor was ensured to have a B1 life of 10 years with a failure rate of 0.1 percent / year following parameter ALTs with corrective action plans.
Author (s) Details
Dr. Seong-woo Woo
Reliability Association of Korea, 146-102 Sunyoo-ro, Yeongdeungpo-gu, 07255 Seoul, Korea.
Dennis L. O’Neal
School of Engineering and Computer Science, Baylor University, Waco, TX 76798-7356, USA.
View Book :- https://stm.bookpi.org/AAER-V2/issue/view/31