Prior to our second test flight, we decided to add around 600g of dummy weight in order to simulate the OBC payload and release mechanism. Our other hardware is likely to be around the same weight as the final hardware; although we’ll be using additional radios, the electronics will be considerably smaller than the current ODROID-X2 with its 6 full-size USB ports, Ethernet, HDMI and audio out, and enormous heatsink.
Although we paid more attention to the CoG location than we had prior to our first flight, we made the opposite mistake in placing the dummy payload too far forward. As a result, after launching the X8 was unable to gain altitude and landed hard about 20m out from the launch location.
The propeller was still spinning at the time of impact, and one of the blades shattered on contact with the ground; since we had no spare propellers with us at the time, we decided to return the following day after we’d replaced the broken prop.
On the 27th, we returned to the field and having re-positioned the dummy payload and enabled data logging, we ran two successful flights of 10–15 minutes. The weather was cloudy, with wind speeds of < 3m/s. The main objectives of these flights were:
With about 25 minutes of in-flight data collected, the above objectives were achieved. Unfortunately, due to a few moments of significant vibration, one of the I/O board USB cables became intermittent and the second I/O board failed to collect data for 60% of the first flight and 80% of the second. The impact of this was limited, since both boards have identical sets of sensors, but it did prevent full comparison of sensor data between boards.
After flying, we noticed that the plywood motor mount had cracked, and one of the elevon control horns had snapped off during landing. As a result, we decided to strengthen the motor mount using an aluminium sheet on the inside of the fuselage, bolted to the repaired plywood mount on the outside. We also decided to strenghten the elevons by covering them in carbonfibre cloth and epoxy.