S. Bedi

BASc (ITT, Kanpur), MASc (University of British Columbia), PhD (University of Victoria, BC)

Professor

Fax: (519) 888 6197  
Tel: (519) 888 4567,  ext. 2178 
E-mail: sbedi@mecheng1.uwaterloo.ca  


Laboratories

Research Interests

Geometric Modelling, Computer-Aided Design,  NC Machining and Automated Polishing

My research interests are in developing improved methods for machining complex curved surfaces, which are found in tools, dies and molds.  Towards this end, two methods of machining called the Principle Axis Method (PAM) and the Multi-Point Method (MPM) have been developed with Prof. F. Ismail.  These methods use the flexibility of a 5-axis machine to improve the efficiency of the finish machining phase by more than 50%.

In PAM, the shape of the surface, characterized by its principle curvature, is matched to the shape of the radiused end milling cutter.  This method has been successfully used in machining numerous complex surfaces.  The current work in this area is focused on developing gouge detection and avoidance mechanisms, efficient tool path planning algorithms, scallop height control methods and in extending the PAM to automated polishing.

In MPM, the tool is inclined so as to machine the surface at two points simultaneously with a flat end-milling cutter or with a radiused end-mill.  This method has been successfully implemented and tried experimentally.  The current research in this area is focused on developing robust algorithms for MPM, tool path planning strategies, surface finish estimation, machined surface evaluation and automated polishing.

The methods described above are off-line methods, but depend greatly on the location of the work piece on the machine and the geometry of the machine.  Many efficiencies can be gained by moving the tool path planning directly to the controller.  The design of such a controller is currently in progress.  In future, the work will be undertaken to allow a solid modeller and an NC machine to share their work space and allow virtual projection of a part onto the work volume of a machine.  This will allow the development of an NC machine which does not require a tool path and is capable of machining a virtual part by mere selection of control buttons  (This is my dream).

Selected Publications

[UW Home][Engineering Home Page][Mechanical Home Page]


Last updated: September 19, 2014