Date of Award

Spring 2012

Project Type


Program or Major

Mechanical Engineering

Degree Name

Master of Science

First Advisor

Barry Fussell


Central to creating a smart machining system is the challenge of collecting detailed information about the milling process at the tool tip. This work discusses the design, static calibration, dynamic characterization, and implementation of a low-cost wireless sensor for end-milling. Our novel strain-based sensor, called the Smart Tool, is shown to perform well in a laboratory setting with accuracy and dynamic behavior comparable to that of the Kistler 3-axis force dynamometer. The Smart Tool is capable of measuring static loads with a total measurement uncertainty of less than 3 percent full scale, but has a natural frequency of approximately 630 Hz. For this reason, signal conditioning of the strain signal is required when vibrations are large.

Several techniques in signal processing are investigated to show that the sensor is useful for force estimation, chatter prediction, force model calibration, and dynamic parameter identification. The presented techniques include a discussion of the Kalman filter and Weiner filter for signal enhancement, Linear Predictive Coding for system identification, model-based filtering for force estimation, and sub-optimal linear filters for removing forced vibrations.