Issue 48

M. Laredj et alii, Frattura ed Integrità Strutturale, 48 (2019) 193-207; DOI: 10.3221/IGF-ESIS.48.21 197 Due to the symmetry with respect to the X-Z, Y-Z planes, only a quarter of the specimen and sleeve are considered. This allows a much greater mesh refinement around the hole and reduces the required computational resources. A highly refined FE model is required to ensure convergence of the given solution at the area of high located deformation around the hole. For this reason, a non-uniform mesh distribution was used. Smaller elements were taken near the entrance and exit surfaces and longer elements away from the hole. The workpiece was discretized using linear brick elements (C3D8R) with reduced integration. This type of element is adapted in the case of elasto-plastic calculations and also is well conjugated with contact elements. In the case of a thickness equal to 3 mm, this latter is divided into 10 elements. The model mesh is illustrated in figure (2-a) consisting of 17749 elements and 23043 nodes. Surface to surface contacts were used to represent the interaction between the different elements (hole surface – sleeve and sleeve – mandrel). Figures (2-b-c) illustrate the geometry and the boundary conditions of the 3D FEA contact model of the mandrel and the handle. Figure (3) shows the residual stress field with a 3.36% cold expansion degree for the 6005T6 aluminum alloy with a 5 mm thickness. The deformed zone generates compressive residual stresses at the edge of the hole. We find that the residual circumferential stresses are not uniform through the thickness. Hence, Figures (4) shows that the level of compressive residual stresses on the entrance face is lower than that at mid-thickness. The stress varies between a maximum value of approximately 320 MPa around the hole in the exit face and with minimum value of 140 MPa at the entrance face. The difference in the residual stresses in the two specimen faces is explained by the level of retained expansion combined to the material volume carried by the mandrel movement during the cold expansion process. This result is in agreement with the works of Liu [23] and Elajrami [24]. These results lead to conclude that the entrance face is less resistant to fatigue than the exit face. Therefore, we chose the study of the residual stresses in the entrance face. Figure (5) illustrates an example of typical tangential residual stress distribution around an expanded hole. The important factors of the residual results profile (outputs simulations results) are summarized in Table (2). Figure 3 : Distribution of circumferential residual stresses on the three faces 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 -400 -300 -200 -100 0 100 entry face mid-thickness exit face Distance from hole edge(mm) Residual tangential stress (MPa) Figure 4 : Tangential residual stress distribution at three different planes of the plate