edfas.org ELECTRONIC DEVICE FAILURE ANALYSIS | VOLUME 26 NO. 3 32 RELIABILITY DESIGN OPTIMIZATION OF WIRE BONDS In this section, the objective is to minimize the failure probability or maximize the reliability of wire bonds under the structural volume constraints. To optimize the reliability the physical dimensions c, d, e, r, ALPHA, and the mean value of Y_PAD are the design variables (Table 5). Figure 7 shows the new design and the new stress distribution after the application of the reliability-based optimization procedure. The stress field is more homogeneous than the distribution in the initial configuration. The integration of the reliability analysis in the optimization process can control the cost with the reliability level and minimize the structural material in the uncritical regions of the structure. CONCLUSION In this study, a probabilistic approach combined with a nonlinear thermomechanical finite-element analysis is used to predict the reliability of the bonding wire of an electronic power module. The finite-element (FE) model developed in this paper is used to investigate the reliability of the wire bonding by considering the thermomechanical coupling effects within a power chip and its environment. The 2D FE model takes into account the bonding wires and the material nonlinearity properties of plastic behavior of the wire bonds. Fig. 6 Approximate explicit functional between input and output response. Table 5 RBDO optimization results: initial point and optimal solution Description Initial point Optimal point Mean value of Y_PAD, μm 25 12.25 c, μm 250 75.87 d, μm 250 293.14 ALPHA 30° 36° e, μm 1375 1221.34 r, μm 1250 650.10 Von Mises stress, MPa 308 273 Volume, μm2 53,497.5 43,602.6 Reliability index 1.176 2.733 Failure probability, % 11.979 0.982 Reliability level, % 88.02 99.01
RkJQdWJsaXNoZXIy MTYyMzk3NQ==