7.3 - Design Process - Bottle Opener
Tell us how you approached your problem and settled on your final design. This includes brainstorming, early iterations that differ from your final design, and any iterations that made it off the drawing board. We want to see the way you thought through this problem.
Our design process began with the challenge of removing a crowned bottle cap using an automated system. After brainstorming several concepts, we settled on a crank-rocker mechanism. into the limited-range prying motion (rocker) required for the task.
Insert brainstorming sketches here
We used the Grashof equation to calculate the specific link lengths that would ensure the mechanism functioned as a crank-rocker. We then modeled the assembly in CAD to visualize the path of motion. The structure of the digital prototype consists of one sheet metal ground link as well as a custom surfaced connector from ground to the output link. The sheet metal link was chosen for speed of design and ability to hold it upright as a handle and the custom connector was designed so as not to interfere with the bottle cap itself (as seen in the side view). The path of motion itself is limited since a crowned bottle cap only requires around 2mm of displacement to be removed, but to account for error we added extra range of motion. The output link itself does not exceed going above parallel as that motion is typically only used for securing the hook underneath the bottle cap. For our final design we plan to simply slide the bottle cap into the groove where the hook will be.
The next iteration we did was a low-fidelity prototype made with cardboard, popsicle sticks, and wooden skewers. The purpose of this was to ensure we could validate the kinematic behavior that we sought in our CAD. Physical testing ensured our system would be viable and exposed possible improvements such as increasing motion profile and exposing the need to ensure the geometry of the bottle opener properly interfaced with the bottle.
Following this, we detailed a full assembly in OnShape for our physical prototype.
The design remained similar to the digital prototype but changes were made to account for manufacturing methods/tolerancing as well as forcing being applied during manual actuation. Firstly, a base plate was added for ease of testing and without having to account for manual actuation tipping over an upright structure. The second major change was 3D printing a conical input link as opposed to laser cutting a flat disk. A flat disc ran the risk of wobbling due to bending forces applied when the user grabs the handle (as we saw with our first build assignment). Swapping over to a conical structure allows for some of the downward force being transferred into compressive forces (instead of pure bending). Having a greater surface area of contact between the shaft and the input link also made 3D printing more advantageous than laser cutting as it is the point force is being applied. One other notable change was thickening the output link. Previously in the digital prototype it was around 15mm thick but in the physical prototype we doubled it to around 30mm in order to anticipate force transmission that it will be experiencing in the final design. Lastly, we made sure all of the lasercut bearing holes were a interference fit with a diameter of 21.75mm to account for kerf and 3D printed shaft holes were left at 8mm because thermal expansion and nozzle diameter caused it to be a interference fit as well.
On the whole the manufacturing and assembly was a success and we were even able to recycle parts from Build Assignment 1 which helped cut down on volume of parts needed. One thing we noticed when assembling was that while the 3D printed connector is good for maintaining link 1 length and avoiding the bottle cap, it leaves a small surface area of contact between the connector and the shaft. This plus the geometry of the offset ball bearing made it very susceptible to bending which resultingly can cause wobble and slipping. The other point of concern to address in the future would be to watch out for slip when the mechanism is mounted upright, especially at the joint between the output link and coupler.
To address further interference, we modified the coupler and ground link as seen in the figures below. The following attached are CAD models of the digital prototype incorporating electrical components, custom bottle opener link, and different iterations of ground links.
The difference between the the above figures lies in the modification in the coupler and ground link. In the figure to the left, we discovered that a redundant link would interfere with the motor as it rotates a full 360 degrees.
To avoid that, we opted for a single link coupler that wouldn’t have interference issues. We also modified the ground link to be more sturdy and added a slot for the planned path. We also added limit switches to signal the motor to reset after each use.
Initially, we planned for the output link to be sandwiching a 3D printed spacer. However, after testing the link on a bottle on hand, we realized that it is too thick to maintain a good contact with the bottle cap. Realizing that, we substituted the spacer with nuts to decrease the thickness.
The video below shows the motion of one of our members successfully prying open a bottle on the second try, proving that decreasing the thickness would allow for a better contact with the bottle cap. However, a problem that we noticed was that the tip of the output link (the part that interfaces with the crown of the bottle cap) was not thin enough to maintain a good contact and the pivot was not centered on the bottle cap when compared to a commercialized bottle opener.
To resolve this, we have decided to mill down the unwanted parts of the output link to create a sharper prying edge.
After that, we also designed a standalone frame with a dedicated bottle slot to reduce play and human error, ensuring the bottle is perfectly aligned with the mechanism every time.
The above figures show the stand that we designed specifically for our bottle. To ensure reliable execution and protect the mechanism, we integrated two limit switches. These define our specific motion range, once the arm hits the top switch after opening the bottle, it sends a signal to the Arduino Uno to reverse the motor direction and return the linkage to its home position.
This is the final design. It contains a 3D printed part that allowed for the pivot to be at the tip of the bottle opener and updated the input link to be curved instead of straight. The base is also redesigned to hold the bottle better.