Torsional fatigue tests were conducted for both shot-peened and non-peened spring steel, SUP7, with different Vickers hardness by changing tempering temperature. As a result, the torsional fatigue strength of shot peened specimen exceeded that of non-peened ones for the whole hardness level. Furthermore, higher Vickers hardness leads to higher torsional fatigue strength. Also, for the shot peened specimen with the hardness of 620HV, there was no decrease in fatigue limit, as seen in the non-peened specimen. In addition, the results of the measured residual stress for pre-fatigue and post-fatigue tests indicated that there was decay of residual stress after 107cycles fatigue test. The quantity of decay can be predicted by the residual stress before fatigue test and applied stress and yield strength of the specimen. Finally, we proposed the estimating equation for torsional fatigue limit of shot peened specimen. According to our estimation formula for torsional fatigue limit, we can predict accurately the torsional fatigue strength of shot -peened specimen with a hardness of up to 550HV by using the slope of the approximation curve in the modified Goodman diagram and values of residual stress for both specimens after 107 cycles fatigue test.
