Search Results

There is a continual growth of test and validation in high reliability product applications such as automotive, military and avionics. Principally this is driven by the increased use and complexity of electronic systems deployed in vehicles, in addition to end user reliability expectations. Higher reliability expectations consequently driving increased test durations. Furthermore product development cycles continue to reduce, resulting in less available time to perform accelerated life tests. The challenge for automotive electronic suppliers is performing life tests in a shorter period of time whilst reducing the overall associated costs of validation testing. In this paper, the application of prognostic and health monitoring techniques are examined and a novel approach to the validation and testing of automotive electronics proposed which it is suggested may be more cost effective and efficient than traditional testing.

Heavy duty diesel engine development has always faced high customer durability requirements, short development timelines and increasingly stringent emissions legislations. However, more frequently heavy duty engines are being used in multiple vehicle platforms across the globe with increasingly stringent quality demands in emerging markets. In order to meet engine life requirements, Delphi Diesel Systems has adapted accepted validation procedures to evaluate their system performance for the global market. In addition to durability and structural testing Delphi Diesel Systems has introduced specialized tests to validate their product at extremes of environmental conditions and fuel properties and has increased OEM collaboration. This paper details some of the adjustments made to the validation test suite to meet the specific challenges of the Heavy Duty market.

Close coupled catalysts and new ceramic catalyst substrates have significantly improved the light-off performance of automotive converters required to meet stringent emission requirements. The hotter environment of these catalytic converters and the lower structural strength of the ceramic substrates require the rethinking of converter designs. The development of new package requirements to accommodate the change in environment and new substrates are discussed. A historical perspective on converter durability is presented as reference. Development of durability test protocols is essential to verifying product durability performance to these new environments. Data collection and documentation of testing templates are shown to demonstrate the effectiveness of tests that represent real world environments. Design improvements to address failure modes are discussed along with durability improvement results.

This paper discusses the effect of the field stress variance on the value of demonstrated reliability in the automotive testing. In many cases the acceleration factor for a reliability demonstration test is calculated based on a high percentile automotive stress level, typically corresponding to severe user or environmental conditions. In those cases the actual field (‘true’) reliability for the population will be higher than that demonstrated by a validation test. This paper presents an analytical approach to estimating ‘true’ field reliability based on the acceleration model and stress variable distribution over the vehicle population. The method is illustrated by an example of automotive electronics reliability demonstration testing.

The paper presents a model-based approach to testing embedded automotive software systems in a real-time. Model-based testing approach relates to a process of creating test artifacts using various kinds of models. Real-time testing involves the use of a real-time environment to implement test application. Engineers shall use real-time testing techniques to achieve greater reliability and/or determinism in a test system. The paper contains an instruction how to achieve these objectives by proper definition, implementation, execution, and evaluation of test cases. The test cases are defined and implemented in a modeling environment. The execution and evaluation of test results is made in a real-time machine. The paper is concluded with results obtained from the initial deployment of the approach on a large scale in production stream projects.