Presented in this paper is a procedure to identify machining features of powertrain components. Machining feature recognition is one of the most important steps for machining process planning. In the case of powertrain components, the first step is to compare a machined model (finished part model) and the corresponding rough part model to identify the volume which should be removed from the rough part model. In regard to the comparison, the most intuitive idea is to use a 3D BOOLEAN operation. Although this approach looks fine, it might not take advantage of the inherent attributes of powertrain component machining. This paper focuses on two important attributes of powertrain machining: (1) a machined model and the corresponding rough part model are very similar and have many identical faces and (2) a rough part model always contains the machined model. Based on these two attributes, we develop an efficient procedure for identifying powertrain machining features. Since the proposed procedure employs well-known 2D geometric algorithms instead of 3D BOOLEAN operations, it is very efficient and robust.
Highly interacting machining feature recognition via small sample learning
