Chris Hong and Vincent Pearce
Slidedeck: link
Adobe Illustrator files: link
Video of it not really working… : link
brainstorming
While we were looking around the internet to brainstorm different ideas for our final project, we came across a video that illustrated a planetary gear system with floating carriers. I know that’s a mouthful, so here’s what it looks like.
Essentially, this gear system is different from a standard planetary gear system because of the two carriers (blue and green) that are attached to the inner gears (yellow). Because the two carriers rotate about a fixed axis, when the outer gear (pink) rotates, it causes the inner gears to rotate along about the center of the pink gear as well.
We initially tried to brainstorm ways to incorporate this gear system into some sort of a model with planets and rockets. However, we realized very soon that replicating this mechanism will prove itself as a very challenging task alone. In the end, we decided to attempt to make this system ourselves with laser-cut wood and possibly take it further if we are successful.
low-fidelity prototype
For our first prototype, we took advantage of the CAD files that were provided by the creator of the video. As the creator designed this system in Autodesk Inventor, the files were in .ipt format. We opened these files on SolidWorks, then took cross-sections of the parts to import them into Adobe Illustrator. After extracting all the cross-sections into a single .ai file, we cut the file twice so we can double up the parts to provide a bit more structural support. The cut was successful as shown below, but the model was too flimsy to transmit the torque properly. Also, gears cut out of cardboard were definitely not good enough to rotate smoothly.
medium/high fidelity prototyping
For our medium-fidelity prototype, we focused on cutting the gears only out of wood to see how smoothly they interact with each other. We quickly noticed that the gears from the template got stuck quite often, which prompted us to replace the gears with .svg files obtained from geargenerator.com. The new gears themselves spun together much smoother, but they started getting stuck again once they were doubled up and held in place by carriers. We believe there are a couple of reasons contributing to this difficulty, and the first is that the size of carriers is not quite perfect. Since the initial model we started with is a CAD model, it did not take into account problems that may occur with gears in real life. As a result, we had to try 3 or 4 different sizes of carriers to smoothen the interaction of gears.
For our high-fidelity prototype, we attempted a full assembly of the system with the best result taken from medium-fidelity prototyping. We noticed a lot of issues to tackle once we built the full assembly. We will talk more in detail about the issues in the following sections.
post-processing
Before assembling, we spray-painted our parts in different colors. The inner gears were painted yellow, the carriers were painted black, and the remaining parts were painted white. We were rather upset to find the painted surfaces still remaining sticky even after giving it a lot of time to dry because the black and white paint I used were the same ones I used for my box and my box didn’t have that issue at all. We are not sure why this happened, but we suspect that it might’ve been the cold and gloomy weather.
We would’ve liked more time to apply more coats, however, we spent too much time troubleshooting the gears that we could barely wait for the paint to dry perfectly. We actually found on the internet that wax-based polish helps paint dry almost instantly and tried applying it to one of our pieces. However, after leaving it for over 1 hour, we concluded it only makes it stickier.
An alternative solution we tried was instant dry nail polish. Applying this to the painted surfaces got rid of some of the stickiness which we wanted to solve before assembling the parts. Despite our effort, there were still interferences in the gear rotation caused by painted surfaces. Our last resort was to apply lube to all surfaces that may cause friction, and that surprisingly helped a lot.
final product/limitations
In short, our final prototype does not work the way we wanted it to, unfortunately. While the initial video we started working off of seemed like a fascinating mechanical system to us, we now realize that this mechanism will probably only work as smooth in a mass-less, inertia-less, friction-less, gravity-less environment. We feel that the reason it is failing is mostly due to the gears being unable to sping perfectly vertically, and there are a lot of reasons for this. We’ll go one-by-one what is causing our prototype to fail.
- Axis alignment
The 0.5″ dowels we are using as main shafts naturally sag due to the weight of the gear system it is supporting. This causes the carriers which are attached at the end of the dowel to deviate from the intended plane of rotation as well. Since gears are supposed to be parallel to the carrier, the gears also fail to spin vertically.
2. Doubling up parts
All parts in our prototype were cut twice and glued together because building the system out of 1/4″ wood would not be structurally sound. In addition, the gears will have to stay perfectly in place to not slip off. However, there were problems that came with the doubling-up process. Some parts were not perfectly aligned, and some parts were printed in 2 different laser cutters which means the geometry may have been ever so slightly different from each other due to different kerf. This caused the gears to have imperfect teeth.
3. Lack of feasible alignment method for gears
We tried to brainstorm ways to keep the gears, physically. A jig of some sort that would keep the inner gears and outer gear together was something we thought of designing, but we weren’t able to come up with a way to incorporate something that would not interfere with the rotation of the outer gear. Our best attempt at addressing this was waterjet cutting little clamp-like pieces and attaching them at the end of each dowel to fix the carriers and inner gears together. While this definitely helped a lot, it wasn’t able to fully solve the issue.
4. Inner gears causing translation(?)
As shown before, the yellow inner gears have two dowels on two opposite sides that connect to carriers. The dowels are supposed to translate torque from the outer gear to the carriers, but it seems the dowels are causing the carriers to translate rather than rotate along a fixed axis. This caused the carrier and its axle to shift away from their intended positions, piling onto the first problem. We believe this is caused by the excess of friction between all dowels that are supposed to spin smoothly.
cost analysis
Materials:
Cardboard: 1 box, about 6 dollars
Crayons: 1 large crayon, 1.20 dollars
Wood: 1.5 full sheets, 9 dollars
Vinyl: two tiny bees, 0.01 dollars
Paint: maybe a quarter can, 1 dollar
Metal:13 dollars for 864 square inches, 5 square inches used costs 0.20 dollars
Wax: a smidge, 0.01 dollars
Quick-dry nail polish: half a bottle, 2.50 dollars
Grease: a quarter bottle, 1 dollar
Machines:
Laser cutter: 1 hour, 20 dollars
Water jet: 30 minutes, 9.60 dollars
Vinyl cutter: 15 minutes, 5 dollars
Labor:
10 hours each
21 an hour x 2 x 10 = 420 dollars
Total:
$55.52 before labor,
$470 after labor
