Quantification of feature shape complexity for the virtual prototypes and investigation of additive manufacturability

The shape complexity quantification is necessary for the selection of the manufacturing process for any Computer Aided Design (CAD) model or virtual prototype. The primary goal of this study is to quantitatively evaluate the Feature Shape Complexity (FSC) of 3D CAD models by considering the “number of inputs” necessary for their generation. In this approach, the quantification of FSC works on the novel “Digging Approach” concept of CAD-to-Features-to-Sketches. The evaluation involves assessing the complexity of features, which are weighted based on the complexity of the “embedded sketch”. Any feature-based CAD model with feature tree information can be evaluated for shape complexity using this proposed technique. For the investigation of the additive manufacturability of feature-based CAD, an explanatory Multiple Linear Regression (MLR) model is utilised to predict the overall printing time and cost of geometries and examine the impact of design and manufacturing parameters on the material extrusion process (time and cost). The developed build time and cost estimation models have achieved high levels of efficiency, with R square values of 0.963 and 0.99, respectively. Regarding the Root Mean Square Error (RMSE), the build time estimation model yielded a value of 31.65, while the cost estimation model produced an RMSE value of 0.12. It was evident from the MLR models that the weight of the support material and FSC impacted printing time and cost more significantly than other parameters. Hence, process selection must be done based on the shape complexity of CAD with optimum part orientation to reduce the requirement of support material.