Search Results

Ring gear bolts in differentials are often modified in size to accommodate the additional clamp load that is required due to an increase in torque from a vehicle's powertrain. Depending on a given program several constraints need to be considered. These include cost, validation time, reliability / durability and timing for implementation. In this paper, a Finite Element Analysis (FEA) procedure for analyzing stresses in ring gear bolts within a rear differential assembly is outlined and the computational results are then compared to quasi-static bench test results that were developed to measure bending and tension loads in the ring gear bolts during loading and unloading of the axle pinion. A dynamometer test is then developed to duplicate the failure mode and provide a comparison of the design changes proposed and the expected improvement in durability.

An essential part of the SHM validation effort is to check the presence and adequacy of the methods required to validate the correct functionality of each SHM task, which can be targeted at detecting structural faults. The ultimate proof of the correct functionality is validation evidence, e.g. crack detection evidence, observed during the operation of the aircraft. However, the occurrences of structural faults such as cracks are infrequent, and hence, years of flight tests might be required to collect validation evidence; small numbers of flights would be only sufficient to prove the system's “fitness for flight” and would be insufficient to prove “fitness for purpose”. Validation evidence can be collected during laboratory tests by inducing faults in structural specimens and examining the SHM detection capability.

Quality has been the illustrious word of the 80s and 90s as we speak organizations are chasing quality problems through the engineering teams and into production. Taskforces of workers in white coats are being sent on to the production line to furiously check components, monitoring process capability in an attempt to improve product quality. Unfortunately it's only after several years of production that the first “real” data gets back to the engineering teams, when it is often too late to remove the causes of these quality problems. The organization is left kicking itself over the same old catch twenty-two situations, “If only the team knew this process data before they decided to engineer it like that!”