Puller UAV

It flew once. It took off under its own power, climbed, slowed, stalled, and came down. That was my first time ever flying a fixed-wing aircraft, and I put it into the one corner of the envelope a one-metre airframe has no room to recover from. I stopped the project there and moved to something else.

I came back to it later with CFD, because I wanted to know whether the aeroplane had let me down or whether I had. The answer is the second one. At 17.5 m/s and 4° the simulation puts lift at 15.43 N. The airframe weighs 1.6 kg — 15.69 N. That is agreement to 1.7 %, from a solver that never saw the scale reading, and it means 17.5 m/s was very close to the speed this thing wanted to cruise at.

Work the stall from the same numbers and it lands near 10–11 m/s, so cruise sat about 1.7× above it. A conventional margin. It was there, and I flew straight out of it.

What the aircraft is: a one-metre span, NACA 4412 wing tapering from a 240 mm root to a 200 mm tip, 3.5° of dihedral and 2° of washout so the tips keep flying after the root gives up — which is to say I designed for stall behaviour and then out-flew the design. Everything is driven from a master skeleton part: one definition of spans, chords, and stations that every other part references. That is what made eleven assembly revisions survivable instead of eleven rebuilds.

The next one gets a buddy box, a bigger field, and someone next to me who has done it before.