The research aims to develop a hybrid model to estimate the surface roughness of 3D printed parts, considering the significant process parameters and measurement uncertainties inherent in additive manufacturing (AM). The methodology involves a Taguchi L-27 orthogonal array design of experiments (DOE) to identify significant factors. This is followed by a deep dive analysis of the roughness profile and the modelling of the measurement process uncertainties. The ANOVA tests reveal that build orientation and layer height are critical contributors to surface roughness. However, it was evident from the experimental analysis that the layer height and build orientation shares a complex relation on surface roughness. The two novelties of this paper are: one, the development of a hybrid model of surface roughness by mathematically quantifying the staircase (step and stack) and corner geometry effects of the part geometry. Two, strengthening the model’s surface roughness prediction accuracy by incorporating the stylus flight and probe reachability measurement uncertainty functions. The final mathematical function was validated with multiple experimental data as well as literature data. The results show that the proposed model is robust to estimate the surface finish accurately at build orientations in the full range of 0–90 degree. The statistical analysis shows that the proposed model is capable of accurately predicting about 90 % variations in the data at different layer heights and build orientations. The proposed model is useful to accurately predict the surface finish of 3D printed parts, a priori, during the part design or before printing, thereby improving the surface quality of the AM parts.
Read full abstract