Prosthetic Ankle Redesign
May 2022 - Jul. 2022
Background
After graduating from the University of Florida, I decided to stay in the area over the summer to continue working in the Human Neuromechanics Lab (HNL) on a prosthesis project. The HNL lab was conducting research on the University of Michigan’s Open-Source Prosthetic Leg, implementing their own controls algorithms on it to use for residual limb research on amputee patients.
Project Assigned
As a research assistant, my task was to redesign the ankle piece on the prosthesis to allow for greater range of motion, specifically more dorsiflexion. The entire ankle mechanism only had 30° total range of motion (ROM) with 11° of dorsiflexion, and my lab mentor wanted to use the prosthesis for walking and stair ascension/descension, which required more dorsiflexion. My task was to determine an appropriate range of motion and redesign the ankle and other components in the mechanism, accordingly, ensuring that I kept the redesigned parts to a minimum to reduce manufacturing costs for the final design. The parts I redesigned would first be 3D printed for prototyping before acquiring a quote for manufacturing; I would assemble the parts into the prosthesis to see if they functioned appropriately and troubleshoot any potential issue that came up.
Research
My first task was to research how the mechanism worked. My lab mentor gave me all the pieces necessary to assembly the ankle portion of the prosthesis and a website that provided a step-by-step on how to do so. Despite having all the pieces and the steps at my disposal, the first attempt at assembling the ankle took me at least two hours. It was relatively easy to understand how the mechanism worked but putting everything together for the first few times was very challenging and meticulous. Below is a video of how the mechanism inside the ankle portion of the prosthesis works, which includes two pulleys and belts. When fully assembled, a motor (part K) would rotate part Q, which in turn would rotate part T. This causes part O to rotate, which leads to the ankle piece (part V) to rotate (see diagram and video below).
After establishing how the mechanism worked, I began to look into different research papers that studied human gait, looking for ankle ROM angles. Upon reading several papers and cross-referencing them, it seemed as though an appropriate angle range for plantar flexion and dorsiflexion for regular walking seemed to be 15-20° and 5-15°, respectively. For stair ascension, plantar flexion and dorsiflexion angles were about 15-25° and 10-20°, respectively. For stair descension, plantar flexion and dorsiflexion angles were about 25-35° and 15-25°, respectively. Therefore, the total range of motion in the ankle for normal walking/gait, stair ascension, and stair descension were found to be around 20-35°, 25-45°, and 40-60°, respectively. Although my goal was simply to increase dorsiflexion as much as possible without sacrificing the 19° of plantar flexion the design currently had, reaching around 40° total ROM seemed suitable for our application.
Redesigning
The previous person working on this project had already created a redesign of the ankle piece, achieving 14° of dorsiflexion. He also created a SOLIDWORKS assembly file with the relevant pieces in the ankle mechanism to simulate plantar flexion and dorsiflexion prior to 3D printing, which I used to guide my simulations. However, since the original file for the ankle piece was a solid-body STL file (in other words, it had an empty Feature Tree), modifications to the part were burdensome to make. Therefore, I had to recreate the ankle piece (part V from the diagram above) from the very beginning on SOLIDWORKS and implement my design changes to that new file.
​
Since the entire ankle mechanism functioned as a four-bar linkage system, my approach to the redesign was to change the distances associated with the two holes on the ankle piece, which can be seen in red below.
%20Dims.png)
Since the SOLIDWORKS assembly file allowed me to simulate ankle rotation via assembly mates, I could change these dimensions on my redesign trial-and-error style and see how different dimensions affected the plantar flexion and dorsiflexion angles of the mechanism. After observing how different dimensions affected the ROM, I also decided to make slight design changes to the top pulley.
​
After achieving necessary dorsiflexion and plantar flexion angles, I sent the files to my lab mentor for 3D printing for prototyping. After assembling the new pieces into the mechanism, we realized that we couldn't achieve the angles SOLIDWORKS was reporting due to a small interference in the mechanism based on the redesign.
​
After modifying the piece and accounting for the interference, we reprinted the parts. However, another issue came up that time. This would happen around 3-5 more times with the redesigns; we would print the parts then notice small changes that needed to be made such as decreasing the possibility of interference, decreasing material in the part, adding fillets, adding height to the ankle piece, and making modifications to other pieces in the mechanism to mention a few.
Results
Eventually, I came up with a redesign that only entailed redesigning the ankle piece and one of the pulleys, ultimately achieving around 24.5° of plantar flexion and 20° of dorsiflexion (44.5° total ROM).
.png)
.png)
After printing the pieces and assembling the whole ankle mechanism, I used an angle locator to determine whether or not I was achieving the angles that SOLIDWORKS was reporting. The angle locator reported around 26° plantar flexion and 19° in dorsiflexion (respectively seen in the images below).
However, after putting the Össur foot piece, which is the part below the ankle piece, into the foot shell, the plantar flexion decreased to 24° and the dorsiflexion increased to around 21°, which was much closer to what SOLIDWORKS predicted (respectively seen in the images below).
Shortcomings & Lessons Learned
One of my shortcomings during this project was not documenting each design change in a clean and consistent manner. I had a notebook where I would jot down necessary changes for the next design iteration, but it would have been beneficial to have cleaner notes detailing my changes not only for potential future modifications but also for the design iteration phase itself. I could have avoided simple mistakes during prototyping by keeping a detailed log of all my changes, preventing myself from printing out parts with features I forgot I had included that no longer served any purpose. I also learned to be patient with the design process because it is a very tedious and meticulous process that requires time.
Overall Experience & Skill Attained
This was a really interesting project that I am very grateful to have been a part of. This project provided plenty of SOLIDWORKS experience, which included part design, assembly simulation, and part drawings, which I had to send out for a manufacturing quote. I also was able to prototype via 3D printing and gain manual assembly experience. With each 3D printed prototype/iteration, I got to use a Dremel to increase the size of the holes in the ankle piece and in a linkage coupler piece (for a different design iteration) for shafts to pass through and an Arbor press to press bearings into certain pieces. Overall, it was an invaluable design and assembly experience in the field of prosthetics. Below is a video of the prosthesis assembled and in motion with a previous iteration of the ankle piece (not the final design).
