Essential Tremor is a debilitating disorder that in the US alone is estimated to affect up to ten million people.
Unfortunately current treatments (i.e. drug therapy and surgical procedures), are limited in effectiveness and often pose a
risk of adverse side-effects. In response to this problem, this paper describes an active cancellation device based on a
hand-held Shape Memory Alloy (SMA) actuated stabilization platform. The assistive device is designed to hold and
stabilize various objects (e.g. eating utensils, tools, pointing implements, etc.) by sensing the user's tremor and moving
the object in an opposite direction using SMA actuators configured in biologically inspired antagonistic pairs. To aid in
the design, performance prediction and control of the device, a device model is described that accounts for the device
kinematics, SMA thermo-mechanics, and the heat transfer resulting from electrical heating and convective cooling. The
system of differential equations in this device model coupled with the controller gain can be utilized to design the
operation given a frequency range and power requirement. To demonstrate this, a prototype was built and
experimentally tested under external disturbances in the range of 1-5 Hz, resulting in amplitude reduction of up to 80%.
The extent of cancellation measured for both single-frequencies and actual human tremor disturbances demonstrate the
promise of this approach as a broadly used assistive device for the multitudes afflicted by tremor.
A commonly noted disadvantage of shape memory alloys is their frequency response which is limited by how fast
the material can be cooled. This paper presents a feasibility study of using vertically aligned carbon nanotubes (CNT) as
microscopic cooling fins to improve convective heat transfer. Using DC plasma enhanced chemical vapor deposition
(PECVD), aligned CNT's were successfully grown directly on ½ of the surface of a 0.38 mm diameter SMA wire,
achieving desirable thermal contact. Cooling speeds were measured with a thermal imaging camera, and the effective
convective coefficient was extracted from the temperature profiles using a basic cooling model of the wire. From this
model, the effective convective coefficient was estimated to have increased by 24% (from 50 W/m<sup>2</sup>K for untreated SMA
wire to 62 W/m<sup>2</sup>K for the nanotube treated wire), indicating that the deposition of CNT's indeed increased performance.
By extrapolating these results to full wire coverage, up to a 46% improvement in frequency response with zero weight or
volumetric penalties is predicted. Further improvements in cooling performance are likely to occur with higher CNT
densities and longer nanotube lengths, allowing further developments of this technology to benefit many future
applications utilizing high-speed miniature/micro-scale SMA actuators.
In urban combat environments where it is common to have unsupported firing positions, wobble significantly
decreases shooting accuracy reducing mission effectiveness and soldier survivability. The SMASH (SMA Stabilizing
Handgrip) has been developed to cancel wobble using antagonistic SMA actuators which reduce weight and size relative
to conventional actuation, but lead to interesting control challenges. This paper presents the specification and design of
the SMA actuation system for the SMASH platform along with experimental validation of the actuation and cancellation
authority on the benchtop and on an M16 platform. Analytical dynamic weapon models and shooter experiments were
conducted to define actuation frequency and amplitude specifications. The SMASH, designed to meet these, was
experimentally characterized from the bounding quasi-static case up to the 3 Hz range, successfully generating the ±2
mm amplitude requirement. To effectively cancel wobble it is critical to produce the proper output functional shape
which is difficult for SMA due to inherent nonlinearities, hysteresis, etc. Three distinct electrical heating input functions
(square, ramp, and preheat) were investigated to shape the actuator output to produce smooth sinusoidal motion. The
effect of each of these functions on the cancellation response of the SMASH applied to the M16 platform was
experimentally studied across the wobble range (1-3 Hz) demonstrating significant cancellation, between 50-97%
depending on the smoothing function and frequency. These results demonstrate the feasibility of a hand-held wobble
cancellation device providing an important foundation for future work in overall system optimization and the
development of physically based feed-forward signals for closed-loop control.
Shape memory alloy (SMA) wires are used increasingly in place of traditional actuators because of their compactness, high work density, low cost, ruggedness, high force generation, and relatively large strains. One well known issue with SMA wires is degradation in performance as actuation cycles accumulate, with significant reductions observed as soon as only tens or hundreds of cycles; thus, manufacturers typically recommend very conservative limits on the operation regime. This paper introduces an alternative approach of cycling or "shaking down" SMA wires under controlled conditions prior to installation. This enables the designer to design to the stable post-shakedown specification of the wire to produce actuators with repeatable larger forcing capabilities. This paper presents a preliminary experimental study which explores the functional dependence of shakedown performance on loading and strain history. A methodology is developed by which an SMA wire can be thermally cycled under electrical heating and the performance characterized with a double-exponential empirical model fit which captures the steady state performance of the wire and the rate at which shakedown occurs. Several sets of experiments are conducted to explore the functional dependence of the shakedown performance varying the load applied (29 to 78N), the allowed strain (4 to 7%), and the form of the loading function (linear spring vs. constant). These experimental studies expose important shakedown parameters affecting SMA actuator performance and provide a first step towards creating detailed SMA wire shakedown protocols tailored to the application that will enable the design of higher performance, stable SMA actuators.
Due to physiologically induced body tremors, there is a need for active stabilization in many hand-held devices such as
surgical tools, optical equipment (cameras), manufacturing tools, and small arms weapons. While active stabilization has
been achieved with electromagnetic and piezoceramics actuators for cameras and surgical equipment, the hostile
environment along with larger loads introduced by manufacturing and battlefield environments make these approaches
unsuitable. Shape Memory Alloy (SMA) actuators are capable of alleviating these limitations with their large
force/stroke generation, smaller size, lower weight, and increased ruggedness. This paper presents the actuator design
and quasi-static characterization of a SMA Stabilizing Handgrip (SMASH). SMASH is an antagonistically SMA
actuated two degree-of-freedom stabilizer for disturbances in the elevation and azimuth directions. The design of the
SMASH for a given application is challenging because of the difficulty in accurately modeling systems loads such as
friction and unknown shakedown SMA material behavior (which is dependent upon the system loads). Thus, an iterative
empirical design process is introduced that provides a method to estimate system loads, a SMA shakedown procedure
using the system loads to reduce material creep, and a final selection and prediction for the full SMASH system
performance. As means to demonstrate this process, a SMASH was designed, built and experimentally characterized for
the extreme case study of small arms stabilization for a US Army M16 rifle. This study successfully demonstrated the
new SMASH technology along with the unique design procedure that can be applied to small arms along with a variety
of other hand-held devices.
Due to stresses encountered in combat, it is known that soldier marksmanship noticeably decreases regardless of prior training. Active stabilization systems in small arms have potential to address this problem to increase soldier survivability and mission effectiveness. The key to success is proper actuator design, but this is highly dependent on proper specification which is challenging due to the human/weapon interaction. This paper presents a generic analytical dynamic model which is capable of defining the necessary actuation specifications for a wide range of small arms platforms. The model is unique because it captures the human interface--shoulder and arm--that introduces the jitter disturbance in addition to the geometry, inertial properties and active stabilization stiffness of the small arms platform. Because no data to date is available for actual shooter-induced disturbance in field conditions, a method is given using the model to back-solve from measured shooting range variability data the disturbance amplitude information relative to the input source (arm or shoulder). As examples of the applicability of the model to various small arms systems, two different weapon systems were investigated: the M24 sniper weapon and the M16 assault rifle. In both cases, model based simulations provided valuable insight into impact on the actuation specifications (force, displacement, phase, frequency) due to the interplay of the human-weapon-active stabilization interface including the effect of shooter-disturbance frequency, disturbance location (shoulder vs. arm), and system parameters (stiffness, barrel rotation).