For our midterm, Vanessa and I decided to design a custom-made wooden candy dispenser. Halloween is just around the corner, so we took this opportunity to put our engineering skills to test!
Check-in 1: 2D Drawings
As soon as Dr. Wettergreen gave the instructions and criterias for the midterm, millions of ideas just flew through our mind. We were tasked with creating a device with a gear mechanism using the tools/machines that we had operated up until now. In our brainstorming session, what started as a pencil dispenser eventually turned into a candy dispenser, and both Vanessa and I immediately set to work. As we began to draft a very general sketch of our idea on a paint software, we also discussed our idea with Dr. Wettergreen who showed us the gumball machine in the OEDK. As he took the machine apart, we were amazed at the inner network and framework for the dispenser. One of our favorite aspects was seeing the bevel gear mechanism that translated the rotation of the handle to the movement of one gumball towards the exit upon inserting a coin. Based on this, we created a few 2D drawings of what our model could look like.
2-D sketch of candy dispenser and its components.
At this point, Dr. Wettergreen cautioned us that the mechanism may be a little complicated so we would need to brainstorm ways to make it a little simpler given the tools and resources we had. We decided to only use one rotating gear at the bottom of the box with six slots so that the candy could fall into a slot, rotate, and exit through a hole at the bottom of the box.
Check-in 2: Low/Medium fidelity prototype
The next step was to construct a low-fidelity prototype. To do this, Vanessa and I headed to the computer lab and got started on upgrading our design from a general sketch in Paint to a 3-D model in AutoCAD. Having no CAD experience, I was amazed to see how useful the software was as Vanessa translated our idea into a more detailed model. She made it look so easy and fun to use:) (I’ll definitely be using this software in future projects).
Model of our Wooden Candy Dispenser in AutoCAD
We first started with the main component of the device which was the box that contained the candy and worked our way down. Since our goal for our low-fidelity prototype was to test our idea, we focused on designing the major components for our candy dispenser and set dimensions on the shorter side. This included the gear system and candy storage space. Since we were able to find a wooden dowel that had the appropriate dimensions to serve as a handle, we decided to neglect this part in the CAD model. One downside of AutoCAD was that the modeling was performed in 3-D and it did not have a feature that captured the 2-D dimensions of the model as Solidworks did. This caused us to manually go back and select each side of our model and create the 2-D dimensions so that we could open it in Adobe Illustrator later for laser cutting. After completing our model, we went downstairs to get cardboard and then proceeded to laser cut our prototype. We couldn’t wait to put the pieces together to see the outcome!
From here, we used hot glue to put the cardboard pieces and the wooden dowel together. Seeing how quick this was made me think that high fidelity prototype wouldn’t take that long, but boy were we wrong! To replicate the candy going through the dispenser, we used beads which were smaller than desired but did the job quite perfectly to demonstrate the movement.
Through the low-fidelity prototype, it was determined that there was a need to insert a funnel within the dispenser that would allow one piece of candy to exit the machine and direct all the candy inside the machine towards the exit. The problem observed was that the candy would spread and become scattered into the corners of the machine, and we would have to shake the machine to get them to come to the center where the hole was. With the help of Dr. Wettergreen, we brainstormed our model for the funnel design on the whiteboard. In this way, the low-fidelity prototype helped to provide a tangible method to not just see whether our device functioned but also highlighted issues or features that would need to be added to make the actual device be user-centric as well.
Check-in 3: Medium/High fidelity prototype
After seeing that our design and gear mechanism worked in the low-fidelity, we headed back to the computer lab to update our model on AutoCAD. We increased the scale of our design, and added unique features such as a rotating handle at the top, open flaps to open/close the dispenser, and ramp style exit for the candy to slide out and be contained within the space made.
We laser cut the pieces and assembled our box using wood glue and hot glue on some parts. When we did this, we realized there were a few components that needed to be edited so that our device would function better.
One change that needed to happen was that we needed to make the gears thicker by adding additional layers of wood so that a piece of candy would fit (we chose to use Hershey’s Kisses). Additionally, the rotating mechanism we had was not as smooth as we would have liked. This was due to the friction of the wood of the gear with the base of the box. To address this, we originally tried to sand the gear, but this caused the wood to chip, so we planned on using a bearing for a much smoother rotation.
On the hinges of the doors at the top of the box, we realized that we did not align the pieces appropriately, so the doors did not close properly. We planned to use tape before screwing the pieces together to make sure that they were in the right positions first.
Lastly, we decided to add a funnel mechanism so that the candy does not get stuck at the base of the box. This would be made out of layers of wood where the opening at the bottom gets larger at each layer, effectively making a funnel.
Implementing an inner funnel system to direct candy towards the exit (made with 16 layers of wood!)
Final Functioning Model
Before laser cutting our final design, we edited our CAD model and transferred it over to Adobe Illustrator. The file is shown below.
Device Parts Ready to be Laser Cut (Part 1)
Device Parts Ready to be Laser Cut (Part 2)
Because of the number of pieces that were being cut, this took longer than we were expecting, about an hour in total compared to the 20 minutes it originally took for our medium fidelity prototype.
After the pieces were cut, we decided to spray paint them so that they would be evenly coated when we put them together. We chose to use a combination of blue and gray spray paint. We chose to do one coat of paint, sand the pieces, and then apply a clear coat.
Spray Painting Wooden Components
While the first coat of paint was drying, we went to the plasma cutter to cut out our functional metal pieces. We decided that we would couple our bearing with custom-made bearing/washers as part of our internal rotating gear to minimize the friction between the gear and the wood.
Before Angle Grinding the Plasma Cut Metals
After Angle Grinding Plasma Cut Metals
Attaching plasma cut washer into gear (on both sides) for smooth rotation
When the painting of the pieces was complete, we began to assemble the device using wood glue and super glue for the metal portions (attaching the washer and bearing to the wood).
Attaching the bearing to make the gear mechanism smoother
Documentation
Here you can find a video of our functional prototype, as well as a presentation showing the various stages of the prototyping process.
Click to see how our Wooden Candy Dispenser Works!
Setbacks are learning opportunities in disguise
One of the many things that we learned from this design project was the need to have flexibility. Translating ideas into something tangible is not always a smooth process. Along the way, contingencies have to be incorporated into the design to account for material quality and property, user inputs, machine error, and human error (such as in miscalculating dimension size, etc.). Looking back at our low-fidelity prototype and comparing it to our final product, it’s fascinating to see the difference in outcome. There were so many factors that we had not considered earlier that the design process helped us to recognize and incorporate into our final model. Creating the low-fidelity prototype really underscored the relevance of creating quick and easy artifact to test our idea. For us, it helped to put us in the user’s shoes and interface with the device, expanding our perspective and understanding of the design process. As we moved forwards towards creating a higher fidelity prototype, we were able to take into consideration varying aspects of interacting with the device and ensuring that it met its purpose which we might not have considered if we neglected this critical step of low-fidelity prototyping. In retrospect, we are definitely happy with the way our wooden candy dispenser turned out! Despite the countless challenges we encountered, there were so many learning moments and techniques that we gained that we will continue to practice and apply in future projects.
Voila, finally done 🙂
Raw material cost
| Material | Ticket Price | Spent |
| Plywood | $29.88 (4×8 feet) | $16.80 (3 sheets of 2×3 feet) |
| Wooden dowel
https://www.amazon.com/Wooden-Dowel-Crafting-Macrame-Craftiff/dp/B08R77NWJR/ |
$9.85 (15 pack) | $0.66 (1) |
| Masking tape | $22.49 (60 yards) | $4.50 (12 yards) |
| Wood glue
https://www.amazon.com/Elmers-E7010-Carpenters-Interior-Ounces/dp/B0045PTHH8/ |
$3.69 | $3.69 (overestimate) |
| Super glue
https://www.amazon.com/Super-Glue-SGH24J-4-0-07oz-0-28oz/dp/B002JPVSYQ/ |
$5.42 (4 pack) | $1.36 (1) |
| Hinges with screws
https://www.amazon.com/Zorveiio-Nickel-Hinges-Stainless-Folding/dp/B097QRR485/ |
$10.29 (10 hinges) | $2.06 (2 hinges and screws) |
| Metal bearing
https://www.amazon.com/2-Pack-1621-ZZ-Groove-Radial-Bearing/dp/B08TYTBZ21/ |
$12.35 (2 pack) | $6.18 |
| Steel | $15.93 (864 in2) | $1.38 (~75 in2) |
| Spray paint
https://www.amazon.com/Rust-Oleum-7727830-Stops-Spray-12-Ounce/dp/B000SABRKE/ |
$6.98/can | $6.17 (overestimate of half a can for each color) |
| Total | $42.80 | |
Labor cost
Assuming minimum wage in Texas of $7.25
| Task | Time | Spent |
| Brainstorming | 2 hour | $14.50 |
| CAD modeling | 2 hours | $14.50 |
| Laser Cutting | 2 hours | $14.50 |
| Plasma Cutting | 1 hour | $7.25 |
| Angle grinder | 0.5 hours | $3.63 |
| Assembling | 8 hours | $58.00 |
| Post-processing | 3 hours | $21.75 |
| Total | $134.13
$268.26 (for 2 people) |
|
Machine time/machine hour cost
Machine hour cost can be calculated by dividing the cost of the machine by the total hours it should function, and multiplying that result by the time that the machine was in use.
The EPILOG Fusion laser cutter at the OEDK costs an average of $32,500, and is expected to last for somewhere around fifteen years. For the case of fifteen years of use, and since we used the laser cutter for 2 hours, the total machine time cost was $0.49.
$32,500/(131,490 hours) = ($0.247/hour)×2hours = $0.49
The plasma cutter at the OEDK costs about of $10,728, and is expected to last for about 20 years. For this case of 20 years, and since we used the plasma cutter for 1 hour, the total machine time cost was $0.06.
$10,728/(175,200 hours) = ($0.061/hour)×1hour = $0.061
An average angle grinder costs about $150, and is expected to last between two and five years. For the extreme case of only two years of use, and since I used the angle grinder for about 30 minutes, the total machine time cost was $0.004.
$150/(17,520 hours) = ($0.0086/hour)×0.5hours = $0.004
A Black+Decker drill costs about $85 and should last for about 3 years. We used the drill for about 1 hour, so the machine time cost was $0.003.
$85/(26280 hours) = ($0.0032/hour)×1hour = $0.0032
Project total: $311.62 (for low, medium, and high fidelity prototypes)
