Design and Validation of a Partial-Assist Knee Orthosis with Compact, Backdrivable Actuation

Abstract

This paper presents the mechatronic design and initial validation of a partial-assist knee orthosis for individuals with musculoskeletal disorders, e.g., knee osteoarthritis and lower back pain. This orthosis utilizes a quasi-direct drive actuator with a low-ratio transmission (7:1) to greatly reduce the reflected inertia for high backdrivability. To provide meaningful assistance, a custom Brushless DC (BLDC) motor is designed with encapsulated windings to improve the motor's thermal environment and thus its continuous torque output. The 2.69 kg orthosis is constructed from all custom-made components with a high package factor for lighter weight and a more compact size. The combination of compactness, backdrivability, and torque output enables the orthosis to provide partial assistance without obstructing the natural movement of the user. Several benchtop tests verify the actuator's capabilities, and a human subject experiment demonstrates reduced quadriceps muscle activation when assisted during a repetitive lifting and lowering task.

Fig. 1.

Photo and CAD model of the powered knee orthosis. The shank attachment, which incorporates a double hinge as well as other linear and angular adjustments, allows for a wide range of leg curvatures to experience a comfortable fit. The CAD model is built and rendered using Fusion 360.

Fig. 2.

The presented custom-designed outer rotor BLDC motor. The green parts are the N35 permanent magnets (PMs), the yellow part is the winding, and the transparent part is the encapsulation material. The blue parts are the stator core and the rotor yoke.

Fig. 3.

The presented actuator with a high package factor design. A single-stage planetary transmission is nested inside of the stator. The housing supports both the stator and ring gear. The high package factor design can reduce the total weight and size of the actuator.

Fig. 4.

The rotor magnetic field design. There is no current in the winding for this simulation. By using a high performance magnetic material (Hiperco 50) to build the stator, the magnetic strength of the stator can reach 2.4 T.

Fig. 5.

A comparison between a stator with (right) and without (left) encapsulation technology. Left: a stator without encapsulation, where the white part is the isolation layer made of nylon material. Right: the designed stator with encapsulation, which is the black material.

Fig. 6.

Schematic of the electrical system for the powered knee orthosis. The C2000 controller receives IMU and force sensitive resistor (FSR) feedback. A quadrature incremental encoder provides position information through signal channels A, B, and Z. By implementing a high level controller, a torque command is sent to the Elmo driver for controlling the motor torque.

Fig. 7.

The actuator test platform includes a magnetic powder brake (left), a FUTEK torque sensor (center), and two misalignment couplings. The actuator (right) is mounted to the testbed frame by its housing, with an output shaft connecting the transmission to the misalignment coupling. The brake is similarly attached on the opposing side of the load cell.

Fig. 8.

Output torque measured over a range of current inputs (5–35 A) to identify the actuator torque constant and verify peak torque. Torque constant (0.59 Nm/A) and offset (−0.21 Nm) computed by fitting the data (red) with linear regression (blue).

Fig. 9.

Results from the continuous torque step response test. Torque values shown are from torque sensor readings (blue), and command current reference (black, dashed). The output torque measured by the FUTEK torque sensor was low-pass filtered (Butterworth, third order, 100 Hz cutoff) for presentation in this figure, but was left unfiltered for the rise time calculation.

Fig. 10.

The temperature curve over 30 minutes of continuous motor operation at 18 A. Maximum stator temperature was recorded at five minute intervals (red), and fit with a two term exponential curve (blue).

Fig. 11.

Thermal image after 30 min of continuous operation. The triangular marker indicates the software-identified point of maximum temperature.

Fig. 12.

Dynamic backdrive torque test result. The test platform seen in Fig. 7 was used for this test, in which the left misalignment coupling was manually deflected in a sinusoidal pattern at approximate frequencies of 1 Hz and 2 Hz. The backdrive torque measured by the FUTEK sensor (solid black, left axis) is plotted against the deflection angle measured by the U.S. Digital encoder (dashed blue, right axis).

Fig. 13.

EMG comparisons between bare, passive, and active modes for RF. The black dashed (bare), green dotted (passive) and blue solid (active) lines represent the time-normalized ensemble averages across the 15 repetitions. The red line represents the time-normalized ensemble average of the commanded torque during active mode.