Test Results
Spring 2020
The results for each of the tests is given below.
​
Test #1:
The system weight was 5.39-oz, which was 25% below the maximum allowable value of 7.2-oz. From the 3D-model, it was predicted that the weight would be 5.35-oz (Req. 1).
Test #2:
The system could be easily mounted & removed without the use of hand tools (Reqs. 3 & 4). Additionally, the average time to mount the system was 14-sec, which was 30% less than the prediction of 20-sec. The average time to dismount the system was 7-sec, which was also 30% less than the prediction of 10-sec.
Test #3:
Both the smallest (3x3x3") and the largest (8x6x3") package sizes could fit easily in the payload system (Req. 9). On average, the 8x6x3" package was able to loaded into the system in 7.7-sec, which is 23% less than the prediction of 10-sec.
Test #4:
Over 10 runs, the system was able to release packages while landed and take off with a 90% success rate, which is above the minimum threshold of 80%. The prediction was 100% success. (Req. 6).
Test #5:
Over 10 runs, the system was able to take off and release packages in flight with a 90% success rate, which is above the minimum threshold of 80%. The prediction was 100% success (Req. 7).
Test #6:
The system was able to take off, land, and release packages of up to 1.25-lb with a 100% success rate, which is above the minimum threshold of 80%. The drone could carry up to 1.75-lb and 2.25-lb, but with success rates of 62.5% and 60%, respectively. (Req. 8).
Test #7:
While loaded with a 1.25-lb, 8x6x3," package, the drone could travel up to 4 miles on a single battery charge, which was much farther than the predicted 1 mile. While loaded with a 3.75-oz, 8x6x3," package, the drone could travel up to 6 miles (Req. 8).
Test #8:
Over 50 runs, the average force to release the trigger mechanism was 1.42-oz with a success rate of 100%, which is above the 80% minimum threshold. The average number of cycles to release was 3, with 82% of them being greater than 1. This is just above the minimum threshold of 80% (Reqs. 2 & 5).
Test #9:
These were the speed runs at 35-mph. Over 14 runs, the success rate of the drone of retaining 3x3x3" packages weighing 3.75-oz was 71%, which is just below the minimum threshold of 75%. Over 15 runs, the success rate of retaining 8x6x3" packages weighing 1.25-lb was 53%, far below the 75%. threshold. The average success rate for all runs was 62% (Reqs. 8, 10, & 11).
Test #10:
Over 100 runs, the drone's camera exerted an average of 1.2-oz, which is 20% greater than the prediction of 1-oz (Req. 2).
Test #11:
At a drone tilt angle of 42-deg, the average number of cycles to release was 4. At 35-deg, the average was 3. At 30-deg, the average was 4, although, the system could not be released a camera angle of 0-deg. At 25-deg, the average was 4, although, the system could not be released at all at camera angles of 7.5 & 0-deg. At 20-deg, the system could no be released at all.
It was initially predicted that the system could be released inadvertently at tilt angles greater than 25-deg (Reqs. 8 & 11).
Did It Succeed?
How well did the project meet the original requirements (See 'Analysis')?
​
SUCCESS
The results of Test 1 far surpassed the​​ requirement (by 25%). This was very representative because it was a simple weight measurement.
SUCCESS
Test 8 best surpasses this requirement, as the mechanism could be released every time. This test is representative because there were 50 runs, which is a large sample size.
Although not focused on this requirement, Tests 7 & 9 were also representative because they were actual flight tests. The average release rate was almost 100%, far above the minimum threshold of 80%.
Test 10 is not as relevant, because the force values measured by the electronic scale are highly inaccurate and inconsistent. The actual force value exerted by the camera and the value required to trip the mechanism aren't that important; as long as the mechanism can be released consistently.
SUCCESS
Test 2 easily proves this requirement follows its exact steps. Additionally, the inherent design does not require any modifications to the drone itself.
SUCCESS
Test 2 easily proves this requirement. Additionally, the inherent design does not require tools because it uses thumbscrews. Surpassing the time predictions, although not requirements, is also a marker for success.
SUCCESS
Test 8 surpasses the minimum threshold of 80% by only 2%, with the average number of release cycles being 3.
This test is not very representative because number of cycles ranged from 1 to 5 and occasionally spiking up to 15. The inconsistency is due to the use of friction to increase the number of required cycles. The original design called for a multistage ratchet, which would have been more consistent.
SUCCESS
Test 4 surpasses the minimum threshold of 80% by 10%, with 9 out of 10 runs being successful.
This test could be more representative with more trials (50 vs. 10).
SUCCESS
Test 5 surpasses the minimum threshold of 80% by 10%, with 9 out of 10 runs being successful.
This test could be more representative with more trials (50 vs. 10).
SUCCESS​
Test 6​​ far exceeds this requirement, as even up to 2.25 was able to be carried (albeit, not with 100% success)
Test 7 is very practical proof of this, as the drone and system were able to carry 1.25-lb over 18-mi of total flight distance without any issue. Surpassing the distance predictions, although not requirements, is also a marker for success.
Test 9 is also practical proof of this, as neither the drone nor system had any issues with carrying the max load at 35-mph, as far as power and structural strength go.
Test 11 is not as representative because it is a simulated tilt test; however, the system was able to sustain structural loading acquired during high-speed, jolting maneuvers.
All of the tests targeting this requirement prove that it was represented well.
SUCCESS​
Test 3 easily satisfies this requirement, proving that the required range of packages does fit in the system. Surpassing the time predictions, although not requirements, is also a marker for success.
SUCCESS​
Test 9 best satisfies this requirement, proving that the loading incurred by drag forces on large packages will not surpass the system's structural strength
The test is representative because it follows the exact steps in the requirement.
FAILURE​
Test 9 best represents this requirement, because it follows its exact conditions. With the 3x3x3" package, the success rate (71%) is marginally below the minimum threshold of 75%. With the 8x6x3" package, the success rate (53%) is far below. At such high speeds, the vibrations cause the very light trigger mechanism to release prematurely.
Test 11 also points towards failure at high speed, because at high angles of tilt (at high speeds), the system is prone to be released prematurely. This test is not as representative because it was simulated and not an actual flight test.
The results of this requirement are not extremely bad, as Test 7 proves that the package retention rate is nearly 100% at speeds of 30-mph. Reducing the maximum speed by 5-mph is no issue, as a pilot would likely not be flying 35-mph due to efficiency and safety concerns.
Conclusion
In summary, this project was largely successful, satisfying 10 out of 11 of the original requirements. There is still much work to be done on the trigger mechanism, but the original concept was still proven: that a purely mechanical package-release system could be designed and built. Some aspects of the project far surpassed the initial predictions, such as being able to carry 1.25-lb of payload for 4-miles.
​
All in all, this project's development was very beneficial for the principal engineer to gain knowledge and experience. With some future iterations and improvements, the design should be reliable enough to be made into a consumer product.