Robert Poulsen’s Wiki Report
| Date: November 29, 2006 | Palm Concept |
 |  |
| Author : Robert Poulsen | Team : Robot Hand Mechanics |
Rob’s Home Page
Problem Definition
- Our team’s objective was to design and build a robotic finger in order to demonstrate the usability of Shape Memory Alloy (SMA). Also, we were given the long term goal of designing the device such that it could be used as a model for a prosthetic device.
- Originally, my part in this project was to redesign the previous design team’s transmission. Also, I was to design hardware to combine multiple fingers into a single hand as well as utilize the redesigned transmission efficiently. The previous team’s transmission design can be referenced here. For reference, the home page of last years Team Robotic Hand 2005–2006 can be found below along with a picture of their transmission.
- The primary problems with the previous design were high friction, bulky size, lack of precision, and the inability to be implemented multiple times effectively to build an entire hand. As our design process progressed, Evan Coomb’s and my designs began to overlap. Once the decision was made to go to SMA springs, and the design was found to meet the customer requirements (discussed below), the bulk of my design portions shifted to implementing multiple fingers into a hand assembly. Therefore, reference Evan’s Wiki Report for further information on the current transmission design process.
- Because the finger and the transmission were chosen as the critical function prototypes (CFPs), the team focused on progressing these designs to the necessary level. This determined that though I have a rough model of the palm, its final design is yet to be determined until the transmission and finger designs are fully developed. Therefore, the bulk of my time this semester was spent aiding in the refinement of the transmission and the finger.
Back To Top
Requirements and Specifications
- The customer requirements for our team were defined as:
- Demonstrate a Possible Use of SMA Wires
- Anthropomorphic (human-like) Finger Design
- Reduce Friction
- Have Predictable Design
- For further details on each of these customer requirements, please refer to our project proposal.
- Specifically, I am required to work between the designs of the other three members of the team. My device, from here on called the “palm”, will transfer the force and motion generated by each transmission assembly to each cam within each joint of every finger. The choice was made to use Kevlar thread as the effective tendons between the transmissions and the fingers. Therefore, the palm will implement a design that routes these threads with minimal friction as well as accepts the threads at a specified angle and deploys them at the necessary angle to utilize the cams properly.
- In addition, we are to meet the requirements and specifications that were determined at the beginning of this project based on last year’s design. The main constraints that we must meet are to maintain an actuator force of 10 Newtons (from last year’s design), not exceed the physical size of the previous SMA wire array, and meet the above customer needs. Also, we are required to use SMA’s in the project, which seems obvious, however multiple other methods could be implemented for actuation.
Back To Top
Concept Generation and Sketches
- Below you will find the idea generation documents, drawings and models of my initial concept generation.
- Because the design of each uniting component is not complete enough to a point were attachment points can be designed, the overall goal for this semester was to develop just the concept of the palm to ensure that the concept is valid and can be implemented next semester. The initial design selection was to be made based on old transmission designs utilizing SMA wire, some of which are included above. When the decision was made to switch to SMA springs, the palm’s primary focus changed from meshing with the transmission to just routing the tendons. Therefore, external pulleys were first considered as the primary design.
Back To Top
Initial Downselect & Design Refinement
- The following design matrix was used to down select to a few ideas: Selection Matrix.
- Below you will find the ideas that were down selected based on benchmarking and a design selection matrix.
- In order to maintain an anthropomorphic design, a palm device which routes the tendons internally was chosen based on the selection matrix above. Again, the matrix above also incorporated the possibility of part of the transmission being included with the palm. An internal routing device was chosen, with an external transmission. Also, throughout the report, because the palm will not be finalized until the spring semester, I have included different ideas and contributions that I have made to the other aspects of the project. Please reference the other team member’s pages for this information. One concept that was researched was the opportunity to have the finger extension springs or similar device built into each individual joint. This was to minimize the size of the transmission and eliminate further friction problems. This concept was modeled roughly above, and was included in the design matrix. However, after researching the issue further, strength and size restrictions eliminated this as an effective alternative.
- This was the beginning of the cam idea, which is now being implemented as discussed. For further information on the cam, reference Bryan Adam’s report.
Back To Top
Final Selection, Supporting Calculations or Data, and Assembly Drawing
- In trying to design the finger, the reverse biasing cams, and the transmission, a force analysis was necessary to determine the strength we would need out of each joint and corresponding subsystem. These calculations were completed and can be found below.
- It was determined that our fingers were going to need more strength, based on these calculations and so more benchmarking was necessary. This time the benchmarking was done against the human hand instead of the previous year’s design team’s research. The human finger can apply a much greater force(40–60 N vs 10 N) than our current design. In order to simply future calculations, the torque equations above were set up.
- With all of these factors in mind, the palm design was again halted in order to accomplish the more critical design of increasing the applicable force of our transmission. Also, based on the information learned from our CFP’s, the finger will be modified (please reference Tyson Skinner’s report for more information on this). Therefore, the palm design will be pushed back and will be replaced by mechanical systems design of custom SMA springs to develop the necessary force for our design. The palm design was left at the design shown below.
- Having met the force requirement to match last year’s design (10 Newtons/joint) we have begun further spring design to attempt to match the actual average maximum force of the human finger, which is approximately 40–60 Newtons per joint. These springs are not available as COTS (Commercial Off the Shelf) parts, so Ni-Ti wire will be purchased and custom springs will be fabricated. This will be discussed in greater detail as the spring semester progresses.
Back To Top
Conclusions
- Much was learned from the design process that was implemented throughout the fall semester. The design at this point has met the basic customer requirements that we were given. However, much work is left in order to meet the goals that we set as a team for this project.
Back To Top
Page last modified on November 30, 2006, at 03:28 PM
