The objectives of the second performance test are as follow:
- All performance test 1 objectives, and
- Reach the loading zone at the end of the monorail without recoiling
- Attach to the caboose and wait for 5 seconds
- Exit the loading zone by reversing
The team accomplished the above tasks using the final AEV design. The code used can be found in the Performance Test Code section.
The following plots show the power usage vs. time (s) and power usage vs. distance figures, respectively, of 3 successful performance test 2 runs. Note that they are the exact same as the performance test 1 plots, which also accomplished performance test 2 objectives.
Figure 1: Power vs. Time plot
Figure 2: Power vs. Distance plot
Initially, the team was still using the smaller-pitch propellers for the motors on the AEV. After attaching to the caboose, however, the team realized that the AEV could not pull the caboose back, even at 100% power. Because of this power insufficiency, the team swapped out the smaller-pitch propellers for larger-pitch ones. This switch accounts for the decrease in motor speed command values in the code for performance test 2.
After the minor change design and further fine-tuning of the code, the team was able to successfully accomplish the performance test 2 objectives. The servo was once again critical in stopping the AEV in the correct location.
The AEV motors were even more inconsistent during the performance test 2 phase. At several points during this testing period, the motors would suddenly accelerate a very high power (presumably 100%) and remain in that state until one team member had to physically catch the AEV and turn off the motors. The team tried to isolate for the cause of the phantom power spikes, but were unable to find a clear explanation. Analysis of the code resulted in the removal of celerate() commands, which were believed to be a culprit (but not the only one) in the power spikes. These power spikes were so frequent and burdensome that the team did not have time to provide new sets of data for the successful runs, nor perform energy analysis on each run.