For the midterm project I decided to make a mechanical device which converts water bottles into plastic twine. I had seem something similar on YouTube and thought I would have a go at it. I carried out this project in three phases: low fidelity, medium fidelity, and high fidelity. During each of these phases, I worked on improving the main components of the project.
(Low -> medium -> high)
The main components of the project are the mounts, the blade holder, handle, gears, flywheel, and spool. Each component saw numerous revisions.
Mounts:
l.
The mounts saw very little change throughout this project. I used the same shape during low fidelity and high fidelity. The biggest improvement seen is in the structural category.
Low Fidelity:
During this phase, the mount was simply a shape glued onto a flat piece of cardboard. There are no cutouts or braces to keep the mount upright. The mount was too flimsy and needed something to keep it stable.
Medium Fidelity
In medium fidelity, I added mounting slots so the mount would slip into the base. This vastly improved stability, though there was still need for more improvement.
High Fidelity
For my final prototype, I added an additional layer to the heaviest half to increase rigidity. This worked. Rigidity is now satisfactory. Along with more layers, I added more mounting holes and made them snugger. Along the sides of the mount, I used some scrap material
Blade Holder:
This part had no significant changes, more so refinements. Everything was based off he same initial shape.
Low Fidelity:
The first prototype was a box encasing a notched post. I quickly realized that there was no part holding the blade secure. At the time, I believed the instabilities were the result of the cardboard, and not the design.
Medium Fidelity:
The medium fidelity prototype saw an addition of mounting pegs to keep the blade holder secure as well as notches to keep the blade secure. This iteration made me realize the force of the bottle being cut could tip over the post.
High Fidelity:
High fidelity saw the addition of a beefier post to prevent tipping and a redesign of the box. I added a large adjustable screw so twine thickness could be adjusted.
If I were to redesign this, I would cut redo the box tightening mechanism and add a platform for cut water bottles to rest on. These two improvements would add to ease of use and improve the user experience.
Handle:
The handle saw a common trend towards increased strength. The low fidelity was cobbled together while the high fidelity is one of the most refined components in the project.
Low Fidelity:
This low fidelity handle is a cobbled together proof of concept. It is weak and could barely stay together due to the cardboard giving out.
Medium Fidelity:
Moving to medium fidelity, I redid the cylinder part of the handle by adding notches. Unfortunately, I believed wood thickness to be ¼”, so nothing properly fit together. This weakness, combined with a weak structure, meant that I had a lot of things to change before the final prototype.
High Fidelity:
High fidelity is where I remade all components of the handle. I adjusted the bar kerf so it fit snugly. I lengthened the handle and adjusted the all dimensions. I wanted the handle to spin on a dowel and that is what I achieved.
I am completely satisfied with how this part turned out. It is well dimensioned, sturdy, and I feel the amount of work I put in reflects the handle’s function.
Gears:
Low Fidelity:
I used an online generator to generate these gears. They meshed well when they were cardboard so I knew they’d work when they were acrylic.
Medium Fidelity:
I chose acrylic due to the rigidity and lower friction compared to wood. In medium fidelity I had issues with kerf affecting the fit of the gears on the shafts and the width of the gears was two narrows.
High Fidelity:
I doubled the gears and made their fit tight on the shafts. When epoxying, I added a lot of washers to increase the surface area that the epoxy could bond to.
The only complaints I have about the gear is there is still some slop. The gears don’t make perfect contact (horizontally or vertically) but I blame that on rushed construction.
Flywheel:
Low Fidelity:
The low fidelity flywheel was solely for testing fit. It had no function because of the low weight. Fit worked fine I knew the future cuts would work fine.
Medium Fidelity:
This flywheel was constructed of two layers of acrylic and proved the flywheel design was viable. The two layers of acrylic were heavy enough to keep momentum going. The fit onto the shaft was loose, so I cut a piece of acrylic with various sized holes at .5”, .492”, and .486”. The .492” worked best.
High Fidelity:
The intention was for this flywheel to be constructed from a single thick piece of steel but after looking for material, I learned there was no thick steel of the proper dimensions. Acrylic was heavy enough, so I plasma cut thin steel to place between two acrylic sheets. The weight is great and I am satisfied with how things turned out.
A recurring issue with all parts on the shafts is crookedness. The flywheel is crooked mainly due to rushed construction. Doing this again, I would be more careful with assembly.
Spool:
The spool changed the most of all the parts. It began as identical to the handle but I realized that with a larger spool, it would move faster and potentially have more leverage. The more I think about it, the more I realize that the force would be the same but my misconception led to an interesting solution.
Low Fidelity:
The low fidelity spool was the same as the handle. After looking into the function, I believed that this design would be weak and unable to cope with the stresses of pulling the twine through the blade.
Medium Fidelity:
The medium fidelity prototype was the same size, but had added holes for fit. Similar to the handle, this spool was designed with improper measurements for kerf and material thickness. This spool fit but did not fit well and would not work for the final prototype.
High Fidelity:
This spool was a complete redesign. Initially, the side panels would have been constructed from waterjet cut parts but the waterjet cutter wouldn’t cut. I attempted an acrylic cut but mis-dimensioned a hole, this resulted in the part shattering while trying to fit everything. Attempt #3 is the final attempt built from laser cut wood. This spool has shortcomings, such as there being no part that actively holds the twine, it must be tied to a post.
The spool is the part that can use the most work. There are many components that were rushed and the spool is slightly crooked. Though it work, there is room for improvement.
The final prototype cut bottles into twine:
There has been no testing with anything larger than a 500ml bottle, but I’m hoping the device can cut a 2L.. This project taught me a lot about prototyping, mainly about structure and dimensioning. Parts such as the handle, mount, and spool were not inherently bad at first, they were simply too weak. I didn’t understand the types of stresses these parts would undergo, so my solutions were poorly designed. Dimension-wise, many parts had to be cut and re-cut because I was hesitant to check my measurements and kerf. Near the end of building, I would double check every number before cutting which allowed me to find many mistakes.
I feel much more confident with the tools used than I did when I began the project. I’m looking forward to continuing to improve my illustrator and prototyping skills.
Notes on construction:
Kerf: .008″
Settings for cutting wood: 100p/4s/10f
Settings for cutting acrylic: 100p/2s/100f
Cost (for final)
3 sheets of .2 inch wood @$20 each: $60
12″x 12″ stainless steel: $10
2x bearing: $10
Access to tools: $20
Misc parts (dowels, epoxy, wood glue): $30
Labor (13 hours @ $15 an hour):$195
Total: $325
