Testing
Spring 2020
There are 11 different tests designed to acquire data to see if the project fulfills the project requirements (see Analysis). Note that Tests #4, 5, 6, 7 & 9 were flight tests performed outdoors. Contrarily, Tests # 1, 2, 3, 8, 10, & 11 were not flight tests and were performed indoors.
Test #1:
Find the weight of the system using a digital scale (Req. 1).
Test #2:
Mount & remove the system without any use of hand tools (Reqs. 3 & 4).
Test #3:
Determine if packages will fit in the grappling system, which is tested by placing both the smallest (3x3x3") and largest (8x6x3") package sizes into the system (Req. 9).
Test #4:
Determine if the system will release packages when landed by activating the release of the system (Req. 6).
Test #5:
Determine if the system will release packages when in flight by activating the release of the system (Req. 7).
Test #6:
Determine if the system will be able to carry the maximum payload weight of 1.25-lb (Req. 8).
Test #7:
Determine the maximum distance that the drone can travel while with the maximum payload (Req. 8).
Test #8:
Determine the amount of force necessary to activate the trigger mechanism and the number of cycles required (Reqs. 2 & 5).
Test #9:
Determine if the drone can retain the payload when traveling at a maximum speed of 35-mph (Reqs. 8, 10, & 11).
Test #10:
Determine the amount of force exerted by the drone camera’s rotation (Req. 2).
Test #11:
Determine the drone tilt angle at which the mechanism will be inadvertently released, and the number of cycles required (Reqs. 8 & 11).
Testing Data
To see the data tables for each of the tests, including photo and video documentation, click on the following links:
Test #1: Click here to see the data tables
Test #2: Click here to see the data tables
Test #3: Click here to see the data tables
Test #4: Click here to see the data tables
Test #5: Click here to see the data tables
Test #6: Click here to see the data tables
Test #7: Click here to see the data graphs
Click here to see the data tables
Test #8: Click here to see the data tables
Test #9: Click here to see the data tables
Test #10: Click here to see the data tables
Test #11: Click here to see the data tables
Testing Highlight
Drone Flight Instability (Tests 4, 5, & 6)
Intro:
One of the most concerning problems that arose was that of the drone's flight stability. As one might expect, attaching a heavy foreign object to a drone causes it to behave differently that initially designed. Over the course of testing, two unique stability issues were discovered and addressed:
Continuous upward flight (Tests 4 & 5): After takeoff, the drone would continue to fly upward, presumably indefinitely. This could only be stopped by the pilot manually forcing the drone to land, although this was often difficult. This issue occurred whenever the drone payload storage and release system (DPSRS) was attached to the drone, regardless of whether a package was loaded or not.
Forced oscillations (Test 6): At payload weights of 1.25-lb and above, the drone began to experience sideways rolling oscillations, swinging back and forth like a pendulum. These oscillations became stronger and stronger until the violent motions caused the package release mechanism to be tripped prematurely.
Requirements:
The requirements relating to these issues were 7, 8, 9, 10, & 11. All of these pertain to flight in general. Req. 8 is focused on having the structural strength to be able to carry packages up to 1.25-lb. The first issue was problematic for all flight situations, but the second issue only arose at high payload weights.
Drone Flight Instability
Summary
In Test 4, the drone consistently rose to heights of 10 to 20 feet before finally hovering. This was concerning, because the drone's default hover height was 4 feet. Even more concerning was the fact that the drone sometimes rose to heights of 50 feet before the pilot finally forced the drone to land. In theory, the drone could have continued to fly upward until it ran out battery or lost signal, which could be disastrous. It was finally hypothesized that the payload system attached to drone was interfering with the drone's intelligent sensors. The DJI Phantom 4 Pro utilizes visual and infrared sensors to actively detect and avoid obstacles. Unfortunately, since the payload was clamped to the bottom of the drone, it interfered with the downward facing sensors. To avoid this perceived obstacle, the drone’s “smart” nature attempted to fly upward, away from obstacle. However, since it was still attached to the drone, the drone kept flying continually upward, trying to avoid it. The solution to this problem was to disable all obstacle sensors, after which the drone hovered at its default of 4 feet.
In Test 5, the same continuous upward flight problem manifested again. In an attempt to solve this, the drone's GPS and GLONASS navigation systems were disabled. This did not appear make any appreciable difference. However, it was then discovered that not all of the drone's obstacle avoidance sensors had actually been disabled; there were some hidden in the settings. After making these changes, the drone finally hovered consistently at 4 feet.
The purpose of Test 6 was to determine the drone's maximum payload. At payloads of 1.25-lb and above, the drone began to experience oscillations, which were initially induced by motions of the drone in flight. However, these oscillations began to grow stronger and stronger, getting so violent until the vibrations inadvertently tripped the trigger mechanism and released the package, allowing it to settle back to equilibrium. These oscillations were clearly driven by some force that caused them to increase. Again, the drone’s intelligent nature was the primary culprit. The DJI Phantom 4 Pro utilizes gyroscopes to monitor the drone’s tilt angle and correspondingly control the propeller speeds in an attempt to remain stable. However, the addition of a large weight produced the effect of a pendulum. When the drone attempted to stabilize, the heavy mass below it caused it to swing past that point, tilting it out of level in the other direction. Subsequently, the drone attempted to stabilize back in the other direction, with the extra mass again causing it to swing too far. With each cycle, the drone overcompensated more and more, causing the swinging motion to increase dramatically over a short period of time. This was resolved by switching the drone into fully manual flight mode, shutting off the gyroscopes and all navigation systems. Although oscillations still occurred, they eventually became damped and were not a problem.
An unfortunate side effect of disabling the obstacle avoidance sensors, gyroscopes, and GPS navigation system is that the drone is much more difficult and dangerous to fly. However, this was necessary, as these problems would have been "show stoppers" for the project if not properly addressed.
