Abstract:To evaluate the reliability of 3D printed products, the performance of the final printed product,including strength,fatigue,etc.,were taken as direct indicators, then the process parameters that affected these direct indicators were found out, the logical relationship between the three was determined, and it was mapped to the Bayesian Network. Its advantages in polymorphism and uncertainty were used to assess reliability quantitatively. Based on the established Bayesian network, its reliability under different working performance requirements was calculated through forward reasoning, and then the reverse reasoning was used to calculate the influence of process parameters on each performance index under the condition of satisfying reliability conditions. There is a certain reference value for guiding the selection of process parameters and materials, and a more comprehensive and effective method for evaluating the reliability of 3D printing products is also proposed.