Criterion for failure under reversed transverse shear superimposed upon static tension

Abstract

A stress testing machine was designed and built to study the combined effect of mean static stress and reversed shear stress on the fatigue life of a material. Aluminum alloy 7075-T6 was tested under different combinations of reversed shear and static tension. Test specimens were of the double shear model, with minimum diameters of 0.2±0.003 inch. For application of the reversed shear load, specimens were tested at three levels of loading taken as percentages of the static rupture shear load. The static rupture shear load was determined by statically loading the specimen until it ruptured. Shear loads were chosen as 20%, 30% and 40% of this rupture load. For combined loadings, 10%, 20% and 30% of the tensile yield strength at 0.2% offset were chosen and applied in combination with different levels of the shear load. All tests were conducted at one loading frequency, 1000±10 cycles per minute. The effect of loading condition was measured by recording the number of cycles to failure (rupture in shear area). Superposition of mean static stress (tension) resulted in considerable reduction in permissible amplitude of the reversed shear stress. The empirical equations representing the form of the relationship between stresses and cycles to failure, were obtained by utilizing a non-linear regression analysis. The failure relationship between reversed shear stress and static tension was found to be linear. Under such conditions failure may be predicted by a general criterion which expresses the octahedral shear stress as a linear function of the sum of normal stresses. The values for constants of this criterion were calculated from data obtained in this study. It was observed that if an average value at 30% yield strength in static tension is used for the linear constant of the criterion, the 1 resulting error will be less than 10%. The effect of loading frequency was studied for one combination of reversed shear stress and static tension. Tests were conducted at approximately 1000, 700, 500, and 400 cycles per minute. An increase of 60% in the loading frequency resulted in approximately 40% increase in the resistance to failure of the test specimens.