Graph-matrix-based automated tolerance analysis and setup planning in computer-aided process planning



Journal Title

Journal ISSN

Volume Title


Texas Tech University


This research focuses on exploring a systematic approach to tolerance analysis and setup planning in Computer-Aided Process Planning (CAPP).

In this research, the geometrical characteristics of various types of dimensional and geometrical tolerances are mathematically represented and compared. Tolerance analysis is based on tolerance zone analysis and aims at mathematical representation. The concept of tolerance weight is defined and applied in tolerance analysis and setup planning. It makes the comparison of different types of dimensional and geometrical tolerances possible. Both dimensional tolerances and geometrical tolerances are analyzed in one-, two-, and three-dimension for tolerance stackup in setup planning.

A graph-matrix-based and tolerance-oriented algorithm is developed for the setup planning of both prismatic parts and rotational parts. The basic concepts on "tolerance graph" are introduced. In a tolerance graph, the faces of a part are represented as vertices, and the tolerance relations among faces are represented as either directed or undirected edges. A tolerance graph is essentially a directed graph by orienting undirected edges into directed edges or representing an undirected edge as two directed edges. The concept of tolerance graph is used consistently in both tolerance analysis and setup planning through the research. Accompanying tolerance graph, tool-approach-direction matrix, machining-feature matrix, and tolerance matrix are defined and applied in representing and computerizing the relations of tool approach directions, machining features, and tolerances in setup planning. The experience principles and rules of setup planning are summarized, explored, and incorporated in the setup-planning algorithm. A setup planning prototype based on the algorithm is implemented in C++ and Microsoft Visual C++®.



Tolerance (Engineering), Production planning -- Data processing