Overview
Prosthetic Hand
Our team was tasked with designing a modular prosthetic hand capable of actuating under human power.
Design Process and Methods
Initial Design
Group Structure
Group Structure
Building a prosthetic hand was one of a dozen options for our senior design project. This project resonated with the team as our 5 mechanical engineers wanted a challenge that tested our mechanical thinking and design. After product and market research we decided to go with a human powered modular prosthetic to decrease price and make it a viable solution for the largest audience.
Group Structure
Group Structure
Group Structure
After market research was accomplished, we began researching design options. Five major sub-assemblies formed and each member volunteered to head a sub assembly. I had done the research on the actuating sub assembly, the Whipple Tree, and took that on as my area of resposibilty.
Iterations
Group Structure
Whipple Design
All team members got to work designing their sub-assemblies in SolidWorks. I built an assembly file and put all the sub-assemblies together. We worked as team to iterate thru fitting issues and mechanical issues and 3d printed our first prototype. This gave us a better idea of tolerance and fit issues, forces applied to sub-assemblies and parts, and the strength and range of mobility of all parts and sub-systems. All members then completed redesigns of their parts using this information.
Whipple Design
Manufacturing and Assembly
Whipple Design
Each team member was charged with the design and optimization of a sub system. Since I had researched and presented Whipple Tree's as an actuating and force distribution system, I headed up the design and construction of the system. I used nTop and SolidWorks FEA to design a Whipple housing that would withstand the force requirements of actuating under a 20lb load. Next I created a pivot that would rotate around a bearing in the housing to allow for unequal force distribution. Then, added steel cable to attach the Whipple mechanisms to each other. This allowed for different force distributions to individual fingers, giving us a compliant grip that could grip non uniform objects of all sizes.
Manufacturing and Assembly
Manufacturing and Assembly
Manufacturing and Assembly
With over 100 individual parts machining became a bottleneck. We decided as a team to 3d print parts that saw less force and machine force bearing parts out of Aluminum 6061. I experimented with alternative manufacturing techniques for parts of the Whipple tree mechanism. These techniques included electroplating 3D printed pieces to add strength, embedding metal into 3D prints, extrusion orientation and alternative support techniques and advanced material 3D printing. After Manufactuing 120 custom parts and procuring the needed pins, springs, heat inserts, and cabling, we began assembly and testing.
Final Product
Manufacturing and Assembly
Manufacturing and Assembly
Our final product included machined aluminum parts, PLA 3d printed parts, PETG 3d printed parts, PETG-CF 3D printed parts, and purchased springs, pins, cable, screws, and heat inserts. The final product was able to be affixed to a partial palm or wrist amputation. Flexion and extension of the wrist activated the Whipple Tree to close the fingers with a compliant grip. The final version was able to lift objects over 20lbs and could safely grip different shapes and sizes of objects, both major goals of our initial design. The project was showcased at the Senior Design Capstone Expo with a highly positive reception.
Implementation Details
Design & Analysis
Prototyping & Testing
Prototyping & Testing
Our team used all the tools available to us to design and analyze our modular prosthetic hand. Design was done on paper and built in SolidWorks. FEA analysis was done in SolidWorks and nTOP. Physical testing was done by putting real world forces on our parts. Fatigue testing was done physically and modeled in Matlab.
Prototyping & Testing
Prototyping & Testing
Prototyping & Testing
Our first prototype was 3D printed entirely with PLA. This allowed us to identify parts that wouldn't stand up to real world stresses for machining. It allowed us to examine the tolerance needed for pin and bearing press fits, and the motion of the parts as they worked together.
Deliverables and Future Work
A functioning prosthetic hand
Our team delivered a functioning, modular prosthetic hand actuated solely by human movement capable of lifting up to 20lb objects of different shapes and sizes. We delivered system and force models of each sub-assembly; we delivered drawings of each sub-assembly and the completed assembly. We delivered FEA analysis of all components, a cost of materials, and a projected cost to build.
There is still work to-do to make this hand practical for use and a viable option for low-cost manufacturing and sale. The splay function would need to be integrated into the hand which would require a redesign of the finger to palm connections. A redundant lock under load mechanism would need to be created for safety. Additional fatigue and strength testing would need to be done for alternative materials.
To lower the cost of manufacturing a hybrid injection molding and metal casting operation would make more sense for any large scale production.
Reflection and Lessons Learned
Lessons were learned on an individual level and as a team.
Individual Lessons Learned
One of the most important lessons I learned was to back up all FEA analysis with real world stress and strain testing. In fact, physical testing should be done before FEA analysis to gather accurate information about the forces that a part will experience. The simulation tools are only as good as the data that we give them. The overall experience also gave me valuable information and experience of working with a group on a singular project. Communication and flexibility are necessary, sometime a great idea needs to be sacrificed to meet the needs of another part of the team, and how ideas are presented and communicated are often just as important as the idea itself.
For the technical issues, I would try to work on assembling prototype designs earlier in the process. As an example, the initial Whipple Tree to finger interface was not effective for closing the grip of the fingers. Since the Whipple Tree connected all the subassemblies, I was in charge of a last-minute troubleshooting. I identified the main issues being an unaccounted-for friction of the steel cables over metal and plastic parts of the palm, and incorrect tolerances of parts in the finger. A simplification of the Whipple Tree and adding PTFe guide tubing for the cables corrected the friction issues. A few small modifications of the finger assembly allowed for smoother grip operation. Combined, these modifications allowed for a fully functional grip actuation.
Group Lessons Learned
Many of the lessons were encountered in the group setting. Ideas that were presented, even agreed upon, sometimes existed as different priorities and at different levels of understanding to individual group members.
One example was the splay mechanism of the hand. This was to be incorporated as part of the Whipple Sub-assembly and was agreed upon by the group as a priority target. I designed a working system, but we realized early on that the finger and palm subassemblies were incompatible with the movements. Instead of working on a redesign of these systems, we decided as a group that we would finish our current designs and try to fix the problem at a later design stage. This decision sealed the fate of the splay controls, as the fundamental flaw in the finger palm connection was reinforced and the fingers were unable to move with the degrees of freedom necessary for a proper splay function. The lesson, regular check in about the way different systems will interact is crucial. Additionally, we need to work on design problems when they come up or remove a feature completely.
