Division J First Group Design
Key Features:
- Propellers far out to the side
- Increase the balance of the AEV
- Propellers at an angle to propel in the middle of the AEV
- Propellers will be largest size possible
- Increase power caused by propellers
- Propellers far enough on wings to allow this
- Space used effciently
- Very little open space on the middle of the AEV
- AEV is made more lightweight and easier for the propellers to move
- Estimated Weight 2.5 lbs
Evolution of Design
To create a final design for the AEV, each team member created a design and after the screening and scoring lab, two designs were chosen to move forward with. The results from the screening and scoring lab are found in Table 1 and Table 2. Because they received the highest scores of 3.3 and 3.25 respectively, the group design (Design A) and Kenny’s design (Design B) were chosen to proceed with.
Group Design Kenny’s Design
Design A, featured a narrow frame with outstretched wings for the motors a propellers. Design B, had both of the motors in the back with a narrow base towards the front of the AEV. Design A scored well in all categories of the screening and scoring lab and excelled in its stability and reliability. Design B scored well in most categories, excelling in its weight and aerodynamics, but was sub-par in stability. Data was collected for both during AR&D labs and Performance Tests. The first conclusion made from data in AR&D 1 was that to reduce the friction force on the AEV a lightweight AEV would be preferred. Another conclusion was that a pull system will be more effective than a push one. When comparing the two designs in Performance Test 1 Design A, which was a pull method, only used 45.088 joules of energy, while Design B, a push method, used 47.436 joules. Design A also traveled a further distance than Design B. Therefore, it was determined that a push system should be used on the way to the cargo load and a pull should be used on the way back to the starting dock, since more power will be required to transport the cargo load. Design A was used because of the data from Performance Test 1 and modified to operate as a push system when traveling to the cargo load. Another modification to Design A was the location of the battery and the Arduino. To accommodate for the 2 inches rule, the Arduino and arm were moved back and the battery was readjusted for the screws to have holes to go through. The differences made can be seen below:
The next conclusion was that a power brake should be applied to stop the AEV. Using the data collected from AR&D 2 , the power breaking showed to have a shorter and more reliable stopping distance. The power breaking trials had significantly less total stopping distance and coasting distance. This was ideal because it gave the group more control over where the AEV would stop, which is extremely important when trying to stop at certain checkpoints and traveling up and down inclined surfaces. The variability was also important because the AEV needed to be able to stop accurately in small spaces. Despite the energy saved when a coasting brake is used, if the AEV was not able to stop in the correct position, the unused energy would be meaningless. The group determined that the most reliable form of braking should be used because of the limited amount of trials available. During Performance Test 1, this theory was confirmed when the group was unable to have the AEV stop at a consistent location when a coasting break was used, despite a consistent code.
Another modification made was that a push method would be used when the AEV was heading towards the cargo and a pull method would be used when the AEV was on its way back to the starting dock. This modification was made because more power would be required when the AEV would be heading back to the starting dock with the cargo.
Another factor to consider with the AEV was whether to track the AEV using a timer or the reflectance sensors. From the data collect in AR&D lab 1, the reflectance sensors only had an error of 2 marks. During Performance Test 1, the correct value for the amount of marks to move was calculated, but the AEV went much further than intended. A time tracker was then used for the rest of the Performance Test and the AEV distance was able to be more accurately coded. Time was adjusted based on the distance traveled by the AEV. The power break was also timed and used for a short burst.
After further analysis, it was observed that a time based code proved to be very inaccurate in the performance tests. When testing the exact same code, the AEV would end up at different positions. This was because the battery would lose power over time causing the AEV to have less power but the same amount of time. The AEV then ran with the code based on marks. At first, after a few trials, the code was determined to be too inconsistent. During all the runs, the reflectance sensors weren’t able to accurately read the marks, and this would result in the AEV not stopping at all. However, after some brainstorming, a Post-it was taped around the reflectance sensors. This was done to block out interfering light from hitting the sensors. After a few trials, it was observed that the reflectance sensors were accurately reading the marks, and that the AEV was able to complete the runs. Using this data, it was determined that the AEV would use code that would be based on marks.
It was observed that during the final run the AEV was able to complete the task during its first two trial runs with no errors. The average cost of the two runs ended up being $593,417.50. The AEV performed slightly better than the class average, for its first two runs, in both time and energy with an average of 199.675 joules and 53.28 seconds compared to class averages of 212.1717 joules and 55.95 seconds. From this data, it was also seen that the AEV had a good balance of energy and time usage. In the alternate MCR completed in the the third test run, the AEV was able to produce a final cost of $503,169 due to the fact that the AEV was able to carry two loads back instead of one.