Design and reliability of a MEMS thermal rotary actuator.

A new rotary MEMS actuator has been developed and tested at Sandia National Laboratories that utilizes a linear thermal actuator as the drive mechanism. This actuator was designed to be a low-voltage, high-force alternative to the existing electrostatic torsional ratcheting actuator (TRA) [1]. The new actuator, called the Thermal Rotary Actuator (ThRA), is conceptually much simpler than the TRA and consists of a gear on a hub that is turned by a linear thermal actuator [2] positioned outside of the gear. As seen in Figure 1, the gear is turned through a ratcheting pawl, with anti-reverse pawls positioned around the gear for unidirectional motion (see Figure 1). A primary consideration in the design of the ThRA was the device reliability and in particular, the required one-to-one relationship between the ratcheting output motion and the electrical input signal. The electrostatic TRA design has been shown to both over-drive and under-drive relative to the number of input pulses [3]. Two different ThRA designs were cycle tested to measure the skip rate. This was done in an automated test setup by using pattern matching to measure the angle of rotation of the output gear after a defined number of actuation pulses. By measuringmore » this gear angle over time, the number of skips can be determined. Figure 2 shows a picture of the ThRA during testing, with the pattern-matching features highlighted. In the first design tested, it was found that creep in the thermal actuator limited the number of skip-free cycles, as the rest position of the actuator would creep forward enough to prevent the counter-rotation pawls from fully engaging (Figure 3). Even with this limitation, devices were measured with up to 100 million cycles with no skipping. A design modification was made to reduce the operating temperature of the thermal actuator which has been shown in a previous study [2] to reduce the creep rate. In addition, changes were made to the drive ratchet design and actuation direction to increase the available output force. This new design was tested and shown to operate in one case out to greater than 360 million cycles without any skipping, after which the test was stopped without failure. The output force was also measured as a function of input voltage (Figure 4), and shown to be higher than the previous design. The maximum force shown in the figure is a limit of the gauge used, not the actuator itself. Continued work for this design will focus on understanding the actuator performance while driving a load, as all current tests were performed with no load on the output gear.« less