Automatic generation of globally assured collision free orientations for 5-axis ball-end tool-paths

Ben Ezair, Gershon Elber

Research output: Contribution to journalArticlepeer-review


In this paper, we present a method for generating smooth 5-axis tool-paths for computer numerical control (CNC) machining, that are assured to be globally accessible. We do so, for a given ball-end cutter and cutter contact curve, and a general arrangement of model and obstacle freeform surfaces. The proposed method is based on a conservative computation of accessibility for (small) finite surface and orientation elements. A conservative estimate ensures the correctness of the solution, and allows us to define a continuous accessible volume in the tool's configuration space. This global accessibility volume, in turn, allows us to plan an accessible tool-path, and optimize it globally. The conservative approach we use allows accessibility computations to be made for an entire cutter contact curve, eliminating the need to compute accessibility for individual cutter contact points, along with the shortcomings that such a sampling approach would introduce. We present examples using the proposed method, that turns paths specified on the surface (cutter contact curves) into valid 5-axis tool-paths with orientation information. We also present results using a CNC machining simulation software, validating the proposed approach and demonstrating realistic manufacturing processes using the proposed methods.

Original languageEnglish
Pages (from-to)171-181
Number of pages11
JournalComputer-Aided Design
StatePublished - Sep 2018


  • Accessibility
  • Accessible region
  • Interference detection
  • Tool orientation
  • Visibility

All Science Journal Classification (ASJC) codes

  • Computer Science Applications
  • Computer Graphics and Computer-Aided Design
  • Industrial and Manufacturing Engineering


Dive into the research topics of 'Automatic generation of globally assured collision free orientations for 5-axis ball-end tool-paths'. Together they form a unique fingerprint.

Cite this