In order to extend our understanding to more generalized and practical cyclic loading conditions, severe nonproportional loading has been studied. Tension-torsion, low-cycle fatigue tests were performed on 1Cr-Mo-V steel at room temperature by applying intermittent axial half cycles. Tests with different loading paths but the same amplitudes of maximum shear strain and normal strain across the maximum shear plane were conducted. Different fatigue lives resulted from different loading paths even when amplitudes were held constant. Additional tests showed that the elimination of the compressive half cycles did not change fatigue life which suggest that the compressive half cycles were nondamaging. The principal implication of the work is that successful methods for predicting nonproportional loading must be selected from models which can show loading path sensitivity. Several models with this feature are discussed, and a procedure for predicting endurance from continuous cycling tests is proposed.

No abstract

No abstract

No abstract

The effect of embedded marker particles on global stiffness and strains determined when using these markers as fiducial marks is studied analytically. These effects are important to a recently developed strain measurement method that uses before and after X-ray pictures of the embedded marker particles. Global effect of the markers on the stiffness is studied using micromechanical models available from the literature. The error in the measured strain due to the influence of the markers is assessed by both the finite element method and elasticity theory. It is shown that the global influence is small for marker-volume fractions less than 3%. The induced strain measurement error is small for particles that are five or more marker-diameters apart. A simple formula for the strain measurement error in the case of particles that are five or more marker-diameters apart is derived from elasticity theory.