Research On Feature Recognition Technology For Mill/Turn Parts

CAPP technology is well known as the bridge connecting the activities in product design and manufacturing and it is also the key to achieve intelligent manufacturing.However,in the actual machining in industry,process plans are still mostly created manually by experienced technical staffs,which is quite far away from the CAPP automation we expected.In most cases,it is a very complicated task to plan the machining process for a part,which is related to the choice of materials,the identification of machining features,the selection of machines and cutting tools,the determination of the sequence of the setup and the order of machining operations.Among these subtasks,recognition of machining features is thought as the most complex one,which is the foundation for machine and cutting tool selection,especially for mill/turn parts.Since there are a large number of intersecting features,including the intersections between turning and turning features,turning and milling features,and milling and milling features,the problem of feature recognition for mill/turn parts is more complicated.To address the above issues,this paper has carried out the related research on the automatic feature recognition for mill/turn parts.It mainly includes the generation of Maximal Turnable State(MTS),the separation and recognition of turning and milling features,the generation of setup plans and turning feature sequencing.In this paper,an inverse MTS generation method based on Attributed Adjacency Graph(AAG)decomposition is proposed for complex mill/turn parts with multiple extreme faces.Here,an extreme face refers to the plane perpendicular to the turning axis of rotation,and its adjacent faces are all located on the same side of the plane.Since the projection positions of the multiple extreme faces along the rotation axis are not monotonically changing and the concave regions at their contours require special treatments,the MTS construction for such parts becomes more difficult.The geometry and topology information of the design part is represented by its attributed adjacency graph.With the geometric reasoning rules,the original attributed adjacency graph is transformed and decomposed into a series of minimum subgraphs.Each minimum subgraph can construct a revolution body that contains all the faces in the subgraph,which is used toconstruct the MTS later.Secondly,a feature classification based on shape descriptor is proposed for the common turning features,and the 2D mesh model of delta volume is constructed.Using the divide and conquer strategy,the 2D mesh model is divided into four separate blocks,each of which is further subdivided into layers.Then,a parallel feature recognition method based on layer tree and recursion scheme is presented,which can effectively resolve the problem in intersecting feature recognition.Since parallel computation is adopted,the efficiency of the feature recognition is higher than the traditional methods.Finally,the relationship between the tolerance constraint and the setup is studied.A tolerance relation diagram is constructed to solve the problem of setup planning for turning operations.Under the constraints of process precisions,the tolerance graph is transformed to generate the setup sequence and the feature operation sequences that satisfy the design requirements.