Abstract
Typical brake validation includes the fatigue testing of major structural components such as operating shafts, carriers, bridges and housings. These components are regularly redesigned to suit different Customer applications and installations with the result of a heavy demand on test laboratory resources and rig capacity.
Further to this the drive to reduce weight and cost of the final product has driven the aggressive optimisation of these components with the move from a test philosophy of limited testing with high safety factor to one with a statistical approach with accurately predicted component lives and confidence values. The former approach usually involves the testing of a relatively low quantity of products with the aim to complete a predetermined load collective. As a result these collectives, which are often derived from known usage data, will contain safety factors that take into account the low number of parts tested. The statistical approach involves the testing of a much higher number of samples to failure to determine a B10 life for each component at different load levels with known confidence. For any one component and failure mode this will need a minimum of 42 test results to failure.
This paper describes how strain measurements are recorded in the areas of high stress and known fracture locations can be used to determine the most suitable test regime to provide results with statistical confidence while minimising validation resources, costs and lead times.