You Don't Save Money by Doing Less Testing – You Save Money by Doing More of the Right Testing!

Abstract

Like so many aspects of life, we are looking for value-for-money. But we need to consider the value in terms of both short and long-term gains. Although certification standards require verification that requirements have been met, we need to recognize that verification is also there to bring value to a project and to the business as a whole. However, prioritizing the value to the project over the value to the business can result in sub-optimization and an overall higher cost to the business. This paper examines a specific case, the prediction of the fatigue lives of critical parts in gas turbine engines, to illustrate the more general case of performing tests to calibrate models that then have general applicability across multiple projects, rather than focusing testing on the needs of a specific project. In some circumstances, testing may not even be the best approach to take; if some level of error escape into service is acceptable (unlike the life prediction example given in this paper) then more focus on requirements validation and design review may provide a more cost-effective approach. This is where the linkage in a systems engineering model between requirements, functions, failure modes and effects analysis, verification test cases, and available calibrated models can help with identifying opportunities and risks.