This work deals with the development of a non-contact torque sensor system prototype made from rolled and textured Galfenol, a magnetostrictive alloy of Iron and Gallium. It is already known that this smart material exhibits a linear response to strain in the presence of appropriate biasing magnets. The response of a sensor system built from it follows commercially available strain gages. In this research, the magnetic change in Galfenol due to shear strain experienced by the shaft in torque will be monitored using Hall effect sensors to derive the torque information. Factors affecting the performance of this Wireless Magneto-Elastic Torque Sensor System (WIMETs) such as annealing, rolling process and strain transfer of adhesives shall be explored.
The ability to provide real-time measurements with minimal signal conditioning requirements make a well-designed torque system attractive for applications such as condition based maintenance. Factors such as being non-contact and passive to the shaft, compact and easy to install, accurate, sensitive and cost effective are highly desired for any torque measurement system. A rate-of-change torque sensor that demonstrates both a sensitivity and time resolution high enough to not only recognize failing machinery, but to specifically identify the failing part is also a critical feature. These characteristics have been incorporated in the current design of WIMETs.
A mathematical model for the magneto-elastic coupling along with simulations from COMSOL shall be presented. Results from static tests for various torques and dynamic tests for various torques at different RPMs will be discussed. It will be shown that the WiMET sensor system setup in a clamshell is reliable and exhibits sensitivity of up to 10mV/in-lb. over a wide range of torque (0–150 lb.-in). Its performance will be compared with a commercial torque sensor and results for detection of eccentric loads on the shaft will also be furnished.