Multiple cutting tool selection in Automated Process Planning & CNC code generation

Open Access
Author:
Jayanthi, Bhanu Kishore
Graduate Program:
Industrial Engineering
Degree:
Master of Science
Document Type:
Master Thesis
Date of Defense:
December 01, 2014
Committee Members:
  • Christopher Saldana, Thesis Advisor
Keywords:
  • Compute aided process planning
  • multiple cutting tool selection
  • accessible area
  • decomposable area
Abstract:
Cutting tool selection is one of the important stages in Computer Aided Process Planning (CAPP) that integrates Computer Aided Design (CAD) and Computer Aided Manufacturing (CAM). Cutting tool selection is a decision making stage in process planning that provides input to CAM for CNC (Computer Numerically Controlled) code generation. Most of the state-of-art CAM systems still require a human process planning engineer to manually look after catalogues and manuals for selecting appropriate tools for machining a pocket geometry. This consumes a significant amount of time in process planning and hence, cutting tool selection can be considered a bottleneck stage. Also, in order to machine pockets, a process planner has to select more than two tools for optimizing the cutting time and to reduce the tool wear. We consider multiple cutting tool selection problem in this thesis. Since the cutting time is dependent upon several factors like step-over (radial depth of cut), step-down (axial depth of cut), tool geometry, feature geometry, feed rate, cost, etc, multiple cutting tool selection becomes highly constrained and large scaled problem. Using one small tool to machine a feature can eliminate un-machined and gouged areas inside the feature geometry. However, the material removal rate for a small tool is very low and hence consumes a lot of time and incur high production costs. Large tools are known to have higher machine removal rates despite maintaining un-machined areas. Hence, a trade-off or optimization solution is necessary between large and small tools from tools database in order to optimize the fitness function (cost or time taken to manufacture) while machining 2.5-D pocket geometries. In this thesis, feasible tools from the tools library are determined by generating accessible areas of the pocket feature geometries which are inputs from a process planning system. A new approach to open pocket accessible area generation is discussed. The variation of optimal tools sequence by using open pocket approach in place of normal accessible area generation approach is discussed in the results. Decomposable areas are generated which compute the areas traversed by a tool provided another large tool has already machined the pocket feature. Then optimal tool sequences are determined by using optimization algorithms. Two different optimization routines are explained in this thesis. The first is using Dijkstra’s algorithm which generates exact optimal tool sequences but take a lot of computation time. The second is using genetic algorithm which generates near to optimal solution, however, the computation time is less compared to Dijkstra’s approach. The variation of cutting tool sequences and computation times with these optimization algorithms are presented in the test cases. Also, the variation of tool sequence with different feed rates that impact the cutting tool life has been discussed in the results.