Search Results

For a vehicle or repairable system, incidents (conditions) are neither necessarily independent nor identically distributed. Therefore, traditional statistical distributions like Weibull, Normal, etc, are no longer valid to estimate reliability. The Non-homogeneous Poisson process (NHPP) model can be used to predict reliability and warranty of the field product. It can also measure the reliability improvement during the development cycle. The NHPP model is discussed in this paper. In applying a NHHP model to reliability data on a repairable system, one may have few or no failures. This paper presents the I/100 and reliability derivations when the parameter β in the ROCOF function is assumed to have a known value.

Most products are designed to operate for a long period of time, and in such case, life testing is a relatively lengthy procedure. Lengthy tests tend to be expensive and the results become available too late to be of much use. To reduce the experimental cost significantly and provide an efficient tool to assess the life distribution for highly reliable product, a step-stress accelerated test (SSAT) was developed. An example of a rear suspension aft lateral link is used to validate the SSAT method.

With the increasing emphasis on Systems Engineering, there is a need to ensure that Electrical/Electronic (E/E) Systems Endurance Testing of vehicles, in a real world environment, prior to Production Launch, is performed in a manner and at a technological level that is commensurate with the high level of electronics and computers in contemporary vehicles. Additionally, validating the design and performance of individual standalone electronic systems and modules “on the bench” does not guarantee that all the permutations and combinations of real-world hardware, software, and driving conditions are taken into account. Traditional Proving Ground (PG) vehicle testing focuses mainly on powertrain durability testing, with only a simple checklist being used by the PG drivers as a reminder to cycle some of the electrical components such as the power window switches, turn signals, etc.

The changes in reliability of the Electrical/Electronic Systems of a vehicle-line during its early design and development engineering processes have been studied. A computerized vehicle failure tracking system was used to provide results from several stages of early development vehicle testing at the proving grounds. The data were analyzed using a software program that assumes that failures in a repairable system, such as a car, occur as a nonhomogeneous Poisson process. Results suggest that, under normal circumstances, a significant and quantitative improvement in reliability is achievable as the system or component design progresses through the early design and development processes. This also provides a means of predicting future system(s) reliability when the system(s) is in production.

This paper presents a method of accounting for sample variability and sample size in establishing the acceptable bogey levels. The technique makes use of the statistical tolerance theory which accounts for the variability of the sample mean and standard deviation by determining a K-factor adjusted for sample size. The result is a tolerance that is reasonably assumed to cover a specified fraction of the population of parts. The technique, although not as simple as a fixed bogey, does discriminate between designs with different levels of energy management robustness.