Search Results

Variable amplitude torsion, bending, and combined torsion and bending fatigue tests were performed on an axle shaft. The moment inputs used were taken from the respective history channels of a cable log skidder vehicle axle. Testing results indicated that combined variable amplitude loading lives were shorter than the lives of specimens subjected to bending or torsion alone. Calculations using strain rosette readings indicated that principle strains were most active around specific angles but also occurred with lesser magnitudes through a wider angular range. Over the course of a biaxial test, cyclic creep narrowly limited the angles and magnitudes of the principal strains. This limitation was not observed in the calculated principal stress behavior. Simple life predictions made on the measured strain gage histories were non-conservative in most cases.

In design for durability, it is generally believed that a compressive mean stress is beneficial and a tensile mean stress is detrimental. Quantitatively the effect of mean stress on fatigue life however is still inconclusive and may very well depend on both the material used and the loading conditions. Over the years, many models have been proposed to help predict mean stress effects. For example, in the long life region, Goodman's formula is widely used, while the Smith-Watson-Topper damage parameter seems the most popular for use in computerized local strain based fatigue tools. In this paper, several frequently cited mean stress correction methods together with the most recent crack closure based method are compared in various ways. Particular emphasis is given to the effect of yield level mean stresses, which has been traditionally neglected but is of practical importance to the ground vehicle industry.