Consideration of Fatigue in the Development Process

Fatigue, and not the personal burnout kind -In all the years developing equipment with suppliers, it was always surprising how fatigue analysis was an afterthought, which would disrupt field testing and negatively affect post launch. Shear pins, welds, fasteners, gear reducers, helical springs, press fits, seals, and so on. In one case, a field test with over 300 machines ended up having a mean time between repairs (MTTR) of 50 days for each machine (equates to about 7 reactive repairs per machine annually). At an average cost of about $200 per service call, this represented an annual potential repair and maintenance (R&M) cost of $420,000 for just the test units. If extended further to a launch of 30,000 machines, this R&M cost could balloon to over $43 million annually. It must be noted that about five of the twenty (25%) of the systemic failures were related to component fatigue failure. For a piece of equipment to have this rate of component failure at less than 50,000 cycles when the machine had an expected annual cycle rate of 75,000 was unfathomable. The components making up the 25% of field failures were expected to last the lifetime of the machine, about 7 years, or over 1 million cycles! If the time and effort was taken to conduct a proper fatigue analysis at the right time during the development process, the field test results would not have felt overwhelming as it did. Just in this field test alone, if fatigue analysis had been done, the MTTR could have increased to over 100 days per machine.
The beauty of fatigue analysis is that it does not require actual components or testing (this comes later). This process can be done easily during the conceptual phase. While Finite Element Analysis is a good tool, hand calculations should be overlaid to validate at the very least, sanity. Here are some basic steps:
1.    Determine the worst-case load conditions by categorizing each use case. Use those free body diagrams!
2.    Use your mechanics of materials knowledge to apply those calculated forces to the component geometry to determine the principal stresses.
3.    Calculate the mean and alternating stresses from the time-based load conditions. When doing so, ask, are the stresses reversing from tensile to compression, what is the material, what are the surface characteristics, what is the environment, residual stress, microstructure, size effects, vibrations, and so on?
4.    Apply the most relevant fatigue theory for the application, material, manufacturing method, and loading -these can be selecting the right material stress-strain curve (SN curve), modified Goodman to Gerber diagrams.
5.    Ask, from the fatigue analysis, what is the expected life cycles? If it was determined to be less than adequate, how would the design or manufacturing method be modified to make the requirements be met?
BECS can help!