USE OF INTRINSIC ELECTRICAL RESISTANCE CHANGES IN SHAPE MEMORY ALLOYS AS ROBUST ACTUATOR STATE AND FAULT DETECTION SENSORS

摘要

Shape memory alloy (SMA) used as electrically controlled on-demand actuators provide engineers new opportunities to create lighter automated components and devices in vehicles due to their compact size, silent operation, and inherently low mass. Outstanding and critical issues are cost-effective and robust control and protection of the SMA actuator element within the device to achieve long lasting service. SMA responds autonomously to external conditions such as temperature and stress and exhibit many property changes during excitation, but many current devices only use SMA as compact actuators; not making use of their intrinsic sensing capabilities. Inherent SMA property changes during use can provide significant utility for improved optimal control strategies. The motivation for this work is to create a robust control method for electrically controlled SMA actuation to simplify device implementation and improve reliability by using intrinsic material property changes. The current work demonstrates the use of electrical resistance feedback in an integrated controller to allow reduction of parasitic mass, cost, and complexity in 2-position devices. Using signal processing and algorithm logic states, we create virtual sensors that successfully identify start of the actuation, end of actuation, reset, and stress overload events. Using electrical resistance to sense the start of actuation allows successful/repeatable performance over a wide range of environmental conditions. Sensing the end of actuation and reset readiness prevents overheating and allows for shorter actuation cycle times, respectively without additional position and state sensors. While many previous efforts have examined the use of resistance in control schemes, one critical need not addressed in previous controllers is the ability to detect stress overload of the SMA during excitation. To protect against unintentional blocked deployment, many current devices include bulky mechanical overload protection systems that prevent stress spikes and SMA damage accumulation. Using resistance feedback, we demonstrate the detection of stress overloads thus extending device lifetime without the need for external mechanisms. The time derivative of the electrical resistance, logic state of the controller, and detection and use of peak/valley widths and thresholds define control events. These events become software based sensors that can augment or replace dedicated external sensors. Software based sensors were successfully employed to control an SMA wire actuator under various environmental temperatures and stress conditions. The control algorithm is not affected to changes in electrical contact resistance, material degradation and other noise sources yielding a powerful method for simple control of two position SMA devices without the need of external sensors.