15 Characterization and Prognosis of Multirotor Failures 171 Fig. 15.22 PSD comparison for asymmetric damage on one arm of the quad-rotor 0 1 2 3 4 5 100 105 Asymmetric Tip Damage Normalized Frequency Intensity [(m/s2)2] Arm 1 (Damaged) Arm 2 Arm 3 Arm 4 Motor Frequency Fig. 15.23 PSD comparison for symmetric damage on one arm of the quad-rotor 0 1 2 3 4 5 100 105 Structural Damage (Loose Motor Mount) Normalized Frequency Intensity [(m/s2)2] Arm 1 (Damaged) Arm 2 Arm 3 Arm 4 Motor Frequency 15.5.5 Additional Observations Beyond the quantitative analysis, certain trends that were not reflected in the data are important to note. These observations were recorded periodically throughout the testing. 15.5.5.1 Battery Battery overheating was a common failure mode observed, which was recorded to avoid misclassification of battery failure as hardware malfunction. The five-cell lithium polymer batteries used on the test platforms were not rated high enough to simultaneously control the four high performance motors and often overheated, occasionally causing catastrophic failure in the system. The high current sourced to the motors induced the melting of several power connectors, which opened the electrical circuit, cutting off the power supply to the vehicle and causing the aircraft to crash instantly. Such high battery temperatures were indicated by deformation of the plastic coating around the battery and corresponding melting odor. Temperatures of the battery exceeded 160 ıC during testing. In future research, it is advisable to include a constant temperature measurement on the battery in addition to the voltage readings. 15.5.5.2 ESC In one case, an ESC on the quad-rotor vehicle burnt out due to its PWM signal wire becoming unplugged during normal operation. This change presumably caused an unregulated current to continuously enter the ESC for the duration of the test. The thermocouple embedded in this particular ESC failed, likely due to the extreme internal temperature and melting of the ESC casing. A thermal image taken after the test showed the ESC at 105 ıC, as shown in Figs. 15.24 and 15.25. Figure 15.26 shows the burnt ESC after it had cooled down and been removed from the system. This test case highlights the need for an improved monitoring system for electronic speed controllers, with either a more sophisticated temperature sensor or improved placement of the thermocouple within the ESC.
RkJQdWJsaXNoZXIy MTMzNzEzMQ==