Performance reliability analysis of Ti6Al4V manufactured by selective laser melting

Objective This paper takes Ti6Al4V manufactured by selective laser melting as an example to analyze the influence of the main parameters on the material performances. Methods Elman neural network (ENN) is used to obtain three-level and three-factor all-factor design data based on a small amount of experimental data so as to save time and cost. Multiple regression analysis (MRA) is used to fit the data of three-level and three-factor full-factor design so as to obtain the polynomiasl of each performance with respect to parameters. The limit state equations corresponding to each performance are obtained by the polynomials. Second moment method (SM) is used to calculate the reliability of Ti6Al4V manufactured by SLM. The results are compared with those obtained by Monte Carlo method (MCM). Through the calculation of reliability sensitivity, the influence degree of each parameter on the reliability of each performance is analyzed. Results ENN has a good predictive ability for the three properties of Ti6Al4V specimen, such as shear strength, hardness and density. The fitting equations of each performance obtained by MRA all have good prediction accuracy. The reliability index, failure probability and reliability are calculated by the Monte Carlo method and the second moment method. The calculation results of the two methods are very close. Under the given parameters and performance values, the reliability of obtaining qualified products is 0.999675. Conclusion Laser power is positively correlated with reliability, while scanning speed and scanning spacing are negatively correlated with reliability. The order of influence of each variable on reliability from large to small is scanning speed, laser power and scanning spacing.