A destructive test used to measure the stress to cause failure by fatigue in a material, part, structure, or weldment after applying a fixed number of cycles of load. Generally, the stress to cause failure is plotted against the number of load cycles on a logarithmic scale. In fatigue testing, it is important to decide on and document the repetitive loading cycle, including base (minimum) load and peak (maximum) load, and
frequency of loading. Loading is usually expressed as a ratio, R = maximum stress/minimum stress, considering compressive stresses as positive (+), and tensile
stresses as negative (-), so that load reversals between tension and compression result in a negative (-) value of R (stress ratio).
Testing for the fatigue strength of a material is so laborious that many materials have not been tested at all, so data is simply not available. In some cases the material has been tested by a user, and the resulting data is often treated as proprietary and is not available in general references or in the open literature. For a hard steel, a test of 2 x lo6cycles duration is necessary to establish a definite fatigue strength. For soft steel, a
test of lo7cycles duration is necessary, while for aluminum and magnesium and many other non-ferrous metals and alloys, 5 x lo8 cycles may be necessary, since these materials exhibit an endurance limit, or stress below which the material could sustain an infinite number of loading cycles without failing.
There are many types of fatigue testing machines. Most commonly used are those which use a rotating beam or rotating cantilever. These rotating tests give a completely reversed stress in which the maximum unit of tensile and compressive stress in the surface of the specimen is equal. The speed of rotation varies with this machine from 2000 rpm to 12 000 rpm.
Fatigue test specimens can be of almost any size, depending on the amount of available material, although certain standard sizes (as opposed to non-standard, and especially, sub-size specimens) are preferred (e.g., by ASTM). Cross-sectional shape is generally, but not necessarily, round. Regardless of size and shape, to give the maximum and most repeatable test results, the surface of fatigue test specimens must be carefully prepared and finished so that they are free of holes, notches, abrupt changes of cross section, machine (kerf) marks and scratches, and even residual stresses from processing (unless these are expected to be used in service in the actual item). The slightest corrosion or flaw will greatly reduce the fatigue limit of a part in service.