Predicting fatigue life of metal LPBF components by combining a large fatigue database for different sample conditions with novel simulation strategies

Recent developments and future trends in fatigue life assessment of additively manufactured metals with particular emphasis on machine learning modeling

“…Wits et al 144 also used linear regression algorithm to study the influence of initial defect size on fatigue life of SLM AlSi10Mg specimens and achieved considerable results. Elangeswaran et al 143 developed fatigue life prediction models for SLM metal structures with varying manufacturing directions, heat treatment, and surface treatment parameters using linear regression and Gaussian process regression algorithms. They compared the results of two models and found that the performance of linear regression models was poor, as indicated in Figure 5B.…”

“…Comparison of predicted results of different algorithms: (A) linear regression, support vector machine, and kernel ridge regression; 142 (B) Gauss process and linear regression 143 . Figure 5A is reproduced from Luo et al 142 with permission from Elsevier.…”

“…Comparison of predicted results of different algorithms: (A) linear regression, support vector machine, and kernel ridge regression; 142 (B) Gauss process and linear regression. 143 Figure 5A is reproduced from Luo et al 142 with permission from Elsevier. Figure 5B is reproduced from Elangeswaran et al 143 alloy samples by LMD titanium alloy was carried out by verifying the correctness of the method.…”

“…143 Figure 5A is reproduced from Luo et al 142 with permission from Elsevier. Figure 5B is reproduced from Elangeswaran et al 143 alloy samples by LMD titanium alloy was carried out by verifying the correctness of the method. Zhan and Li 150 also adopted the above method to establish fatigue life prediction models for various AM metals, which achieved widely applicable performance, as plotted in Figure 7C.…”

Failure Modes in Dual Layer Thickness Laser Powder Bed Fusion Components Using a Novel Post-Mortem Reconstruction Technique

“…This method enables a finite element model to interpolate between test stages and observe the full progression of subsurface pores or cracks under given strain conditions and could be improved further with additional mid-test data. The information is likely to be more useful under cyclic loading conditions to aid in modelling fatigue response, since the role of internal defects is amplified and current state of the art is lacking in pore behaviour information [35]. With enough preliminary data, fracture surfaces can also be reverse engineered to approximate subsurface information around the failure location prior to mechanical testing, without the need for XCT data, and is significantly more time and cost effective.…”