The construction up until that point (minus the raw airfoil surfaces themselves, which were made on a CNC hot-wire foam cutter) had been conducted on the floors and porches of various team members’ apartments using only a dremel and various multi-tools. In a few hours of final assembly we added carbon support, attached the control surfaces, cut out locations for the electronics, and slapped together some landing gear to keep everything off the floor. We weighed in the craft (will all flight equipment) at significantly under our target weight!
One lesson reiterated from the assembly process is the importance of solid interfaces and error-checking with calculations. A few quick hand calculations can save a fair amount of guesswork any-day, and interface design is the key to holding all of the parts together. For our next go-around, better interface designs are in the works and structural members will be as light as possible without sacrificing functionality.
After completing construction, it was time to fire up those lift motors! As expected, there were plenty of things to iron out to get these to work. The rear lifting motor had some issues with poor electrical connection and its pivot mechanism for yaw control.
We knew that the complexity of pivot mechanisms would be difficult to prototype, a fact that directed us towards this minimal-complexity design, but this one mechanism is necessary in order to achieve yaw control in the three-fan system. In this iteration, we pivoted the motor itself through a bearing shaft using a servo. The problem occurred at the connection to the motor mount, since we had to cut the shaft to allow clearance. Epoxy just doesn’t do so great for small contact areas with high stress (big surprise right?).
The main rotors spun up amazingly one by one, but together they just could not get past ⅓ throttle and thus did not generate their specified lift to get the craft off the ground. (The motors were spec’d at 13lbs thrust each, which should have been more than enough to lift our oversized prototype.) For the next full integration test, rigs are being constructed to fully characterize the motors’ thrust separately and together in order to pinpoint the problem. After this first test, it was speculated that our limit could be the power draw capacity of the battery itself, in which case a solution of using two smaller batteries (instead of a single large one for both motors) was proposed.
Our action items from the motor test include designing new light-weight mechanisms for the yaw pivot and investigating the root cause of our main rotor failure. We are confident that both can be accomplished before our next iteration!
Overall, the big takeaway from this integration test: Prototype early and often! Unforeseen problems always come up, and tackling those early creates more reliability moving forward.
Until next time,
eyes on the sky,