Issue 48

K. Kimakh et alii, Frattura ed Integrità Strutturale, 48 (2019) 429-441; DOI: 10.3221/IGF-ESIS.48.41 438 These observed profiles (Fig. 8) illustrate geometric irregularities that occur periodically at constant steps. However, their depth varies for the different roughnesses. In fact, these irregularities represent the mark of the cutting tool, they are considered as surface defects which generate stress concentrations. They are likely to decrease the crack initiation and consequently decrease the fatigue lifetime. From these observations, we have therefore measured the height of the profile that represent the total roughness Ry. The values recorded are very close to the values measured by the surface roughness tester. The table below shows the data determined from the profile observations (a, 2b and the equivalent defect size √area). batches 2b (µm) a /2b area (µm) Ra Ry Ra Ry L1 212 0.01 0.10 0.10 0.01 L2 136.88 0.01 0.12 0.12 0.01 L3 64.32 0.01 0.12 0.12 0.01 L4 28.91 0.01 0.14 0.14 0.01 Table 7 : Equivalent defect size for the different surface roughnesses. Result and discussion From the measurement of the geometrical parameters and the evaluation of the roughness equivalent defect size, the fatigue limit were predicted for each surface roughness. It can be estimated as a function of Ra or Ry. Table 8: Comparisons of Predicted and Experimental Fatigue Limit Values for Different Roughness Studied The results obtained for the fatigue limit prediction revel the effect of surface roughness. In fact the fatigue limit increase with the decrease of the surface roughness in the two cases where a = Ra or a = Ry. On one hand the difference between the fatigue limit obtained experimentally and that predicted is reduced when we consider a = Ra. This result joins the results obtained by Itoga [16]. However Y. Choi [8] found that the predicted values are rather close to those of the experiment when we consider a = Ry. On the other hand, the difference obtained in our case between the predicted and the experimental fatigue limit is smaller than values obtained by Itoga and Choi [16, 8] which are important. For the AISI 1045 carbon steel, the Murakami model [18] allowed us to estimate the fatigue limit of the test specimens without going through the fatigue tests which are very costly, just the determination of equivalent defect size and a simple measurement of the hardness will be sufficient. This result will allow us to improve the fatigue performance of our mechanical parts while acting on surface roughness. The ideal is to be able to analytically determine the cutting conditions generating a part having a good fatigue lifetime. To do this, we could use our model expressing the roughness according to the cutting parameters that we had developed in a Batches Ra (µm) σ w (predicted) [MPa] σ w (exp)[MPa] Gap Function of Ra Function of Ry Ra Ry L1 3.125 150 112 180 30 68 L2 2.263 159 118 183 24 65 L3 1.242 176 133 196 20 63 L4 0.5 204 149 202.5 1.5 53.5