A generalized analytical methodology for generative process planning

Ramesh Srinivasan, Purdue University

Abstract

A generative process planning methodology for machining operations is presented in this thesis. A study of the literature revealed that the following are lacking: (i) methods to include the geometric model of the raw workpiece in the selection and planning of machining operations, (ii) geometric models of shape generation processes in machining centers, and (iii) concise representation of and algorithms for the recognition of three dimensional machined features (set of connected or disconnected surfaces generated by machining processes). Analytical models and tools have been developed to address these problems. The above list is far from being exhaustive. Important issues which have not been investigated in this thesis pertain to the topics of tolerancing and multitechnology process planning. A scheme to decompose two and three dimensional polyhedra in a unique manner and an algorithm to perform the same have been presented. Such decomposition of the excess material to be removed (geometric subtraction of the finished part from the raw workpiece) is performed; the decomposition data is used in the determination of precedence relationships among machining operations based on material removal and reconstruction of machined features for feature recognition. The decomposition scheme can also be employed to obtain a constructive representation of any polyhedron bounded by planar and quadric surfaces in two and three dimensions from its boundary representation. In order to model the shape generation processes in a machining center, a kinematic chain representation has been given to the machining centers. Operation planning at the machining center is performed using inverse kinematics. Machined features are represented concisely using attributed tree grammars and feature recognition is performed using a parsing algorithm. Testing the approachability of a machined feature and the noninterference of the tool with the finished part during machining are performed using point-set-approximations of the cutter. These models and procedures have been implemented on an experimental basis on a VAX 11/780 computer and the programs have been coded on FORTRAN77.

Degree

Ph.D.

Advisors

Liu, Purdue University.

Subject Area

Industrial engineering

Off-Campus Purdue Users:
To access this dissertation, please log in to our
proxy server
.

Share

COinS