THE QUEST FOR RELIABILITY STANDARDS

摘要

As electronic assemblies become more complex each Original Equipment Manufacturer (OEM) struggles with the question as to whether the product will work in the intended environment, for the length of time expected by the user. Many methods have been developed to simulate the end-use characteristics of equipment in the field, but a clear understanding is required to differentiate between the methods used to assess quality and those used to validate reliability. The considerations and relationships between the customer and supplier are also now a major factor, especially since most OEMs outsource not only the fabrication, but also the design of the product. The industry trade associations have written many standards which define methods for both quality conformance and reliability testing. Sometimes these methods are misapplied; however there should be no misunderstanding about the fact that poor quality can never achieve the intended reliability of the product. Quality conformance is meeting the requirements of the customer. Since each end-use varies classes or levels of complexity and performance were developed in most of the performance standards in order to simplify the procedures for determining whether the finished goods meet the customer requirements. There is no doubt that the requirements for a hand-held commercial product are different than those for an aerospace application. Also the degree of validation of both quality and reliability differs, as does the cost of producing the item. Some of the existing reliability methods have become suspect in simulating the end-use, so some OEMs are requiring new methods in order to accept delivery of those products indicated in their fabrication documentation package. Reference to classes or levels of quality are not a guarantee that the product will perform according to its reliability assessment. An example is one where an OEM has developed a product and the prototype version has been tested to validate the reliability of the unit in its intended use. The production version is then outsourced to a new fabricator and it is unclear as to whether the new supplier's process is robust enough to achieve the same reliability as the original prototype. The dilemma facing the industry is that the supply chain is hard pressed to keep up with all the new 'home-grown' methods and new stress techniques being developed. The concepts badly need industry consensus on robustness and reliability assessment methods linked to the end-use environment. This paper explores the various alternatives.