2012 Autumn Project
This Easter we were on the lake testing the IMU based control system, the results were encouraging but the controller was no where near as robust as the system with four wands, tested during winter 2010. As autumn came, we gave it an other go. The objective of Autumn Project 2012 was to be able to stand on the surfboard and to acquire some valid data regarding the tow force. The Mettler SS1321 load cell had been destroyed by water during the previous project, its successor was an ip67 rated Sensortronics 60001A100LBS-1000. Sadly the GS407 GPS unit failed, this for no apparent reason. Interestingly, this GPS module is no longer available anywhere, seems we were not the only ones experiencing problems. The rest of the hardware setup remained unchanged.
To eliminate the ratched lashes and guy wires, the blocks were attached with four M8 threaded bars to the board; the struts were fixed with a second screw to the block, eliminating the rotational degree of freedom. These two simple modifications made assembly much more enjoyable. The LQR controller was completely redesigned. Furthermore, the pilot was considered in the inertia tensor again. Since the board would be flown by different people (with different weight) and it was very likely that we would stand in the wrong place the whole time, an integrator was also added. An integrator corrects steady state errors and can therefore compensate modeling errors . Already from the first take-off the new controller proved to be valid, after some minor tuning, standing on the board became almost too easy.
Because the craft is not stabilized on the yaw axis, the issues encountered during Easter persisted. Telemetry provided valuable insight, from the imgae below it can be seen that the flaps operate with much reduced travel compared to Easter. Our simulation overestimates drag, drag was stable at around 200N, the simulation predicted 275N. The plots are from a low speed flight, notice the positive pitch of 5 degrees and the constant positive flap deflections to increment camber.
After the second day of testing, the structure started to show signs of fatigue. Several areas had delaminations or cracks. This was to be expected: the foils had been designed in summer 2009 to be used with guy wires, and to last for a couple of days, not four projects. Here is some video footage of the project: