GRIPS INDUCED FAILURES IN FATIGUE TESTING Giangiacomo Minak Ph.D., Cristiano Fragassa Ph.D. DIEM, University of Bologna Via Risorgimento 2, 40136 Bologna giangiacomo.minak@mail.ing.unibo.it Abstract. The shape and dimensions of fatigue testing specimens usually are reported in different Standard test methods. Fatigue testing machine or grips manufacturers sometimes suggest specimens ad hoc for particular application. In this work fatigue failures induced by grips on 20MnCrS5 carburized and tempered steel specimens will be shown and compared with failures occurred in the mid section of specimens made of the same material subjected to the same thermo- chemical treatment then shot peened. Introduction. When testing materials very important issues are the load alignment and the repeatability of the failure modes. Standard test methods [1,2,3] and manufacturers [4] suggest shape and dimensions for the specimens so that failures occur in a calibrated zone in which stress can be easily calculated without uncertainties. To achieve this result usually specimens present smooth reductions of the section so that the nominal stress in the grip zone is far lower than the nominal stress in the calibrated zone. In this work will be shown a case study in which more than 80 % of failures occurred in the grip zone, in specimens conforming to grip manufacturer advices, subjected to plain bending. Material and Methods. In order to evaluate the fatigue limit of two groups of 20MnCrS5 carburized and tempered steel specimens, one of them with a subsequent shot peening treatment, plain bending four point tests have been performed on a Rumul Microtron resonant machine. Special Amsler grips (figure 1) have been utilized and the setup procedure recommended by the manufacturer has been followed. The specimens shape is prescribed by the grips manufacturer and it is shown in figure 2. Test were ran at a frequency of approximatively 100 Hz till failure or run-out at 10 7 cycles, with a load ratio R=- 1. A stair case procedure was followed, starting from an estimate given by axial tests results. Failure was detected observing the decrease of the specimen flexural stiffness. The microstructure of the material was fully martensitic with an high number of inclusions, mainly MnS and oxides. Shot peened specimens presented an high number of craters on the surface due to the impact of the peening balls on the hard and fragile surface. In fact the hardness was 62 HRC on the surface and 40 HRC in the bulk material. Figure 1: Four points bending grips Figure 2: Specimens design