MIT SEMINAR SERIES IN MANUFACTURING AND
PRODUCTIVITY
Place: Room 33-116 Time: 12:00 P.M. Tuesday, November 14th, 2005
Timken Chair in Design and Development, Clemson University
For most of modern industrial history, the vast majority of precision machines have been designed using a stacked-axis serial kinematic layout, often in a Cartesian arrangement. This arrangement leveraged centuries of accumulated skill and knowledge of how to fabricate straight lines, planes, and right angles with great precision. Over the past two decades, advances in computational power and efficiency have enabled designers to relax the requirements for precision construction in such machines, and replace it with software-based error compensation methods; thus creating very high precision machines without the requirement for extraordinary precision in construction and assembly. Essentially these software-based methods acknowledge the fact that the actual kinematic model of the “as-built” machine does not exactly conform to the simple Cartesian model the designer proposed.
More recently, designers have begun to explore the the use of parallel kinematic layouts (PKMs) where the complexity of the kinematic model of the machine may preclude a meaningful intuitive understanding of the relationship between the control inputs and the resulting outputs of the system. Early enthusiasm for PKMs led to many unsubstantiated claims of superior stiffness and accuracy because all actuators connected directly between the motion platform and ground. However, it was discovered that these claims were very difficult to demonstrate in prototype machines.
This presentation will review several examples of precision metrology devices developed by the author, and their successes and shortcomings. We will also point to some research challenges needing solution.