Flight Testing and Iteration
Trial #1
The aircraft was tethered to the central post and launched under motor power. Once the motor spun up to speed, the plane actually lifted off for a brief moment—but it pitched up almost immediately and slammed back down nose-first. The front wheel snapped off the fuselage entirely on impact, and the motor came off the nose too. The plane was completely done. Figure 4 (wait, this is now Figure 7—let me renumber) shows what was left after the crash.
Figure 7: 1st Trial
What Went Wrong
Honestly, almost everything that could fail, failed. There were two main problems—one aerodynamic, one structural.
The aerodynamic problem was basically what I already talked about in the Ground Angle of Attack section. The front wheel was so tall that the plane sat at a really steep angle on the ground. The moment the motor pulled the plane forward and it lifted off, the wing was already past its critical angle of attack. So instead of accelerating and gaining lift normally, the plane just pitched up further, stalled, and dropped. It barely had time to do anything else.
The structural problem was actually two separate failures.
The first one was the front landing gear. I had glued the wheel strut to the bottom of the fuselage with wood glue, but the contact area between the strut and the balsa was tiny—maybe a few square millimeters at most. Wood glue is fine for normal wood-to-wood joints with a decent surface area, but with that little contact, there’s basically nothing holding it. The moment the plane hit the ground, the joint just popped (Soden & McLeish, 1976).
The second one was the motor mount. I had attached the motor to the nose using a single wrap of tape, which in retrospect was a pretty bad idea. Tape doesn’t actually stick well to raw balsa—the surface is too porous and fuzzy for the adhesive to grip properly. And on top of that, the motor was vibrating constantly while spinning the propeller, which gradually loosens any tape bond anyway. So when the crash happened, the tape peeled off and the motor flew off the front.
What I Decided to Change
Based on the crash, two things had to change:
-
Get rid of the wheel landing gear entirely. The wheels were the reason the ground angle was so extreme, and for a tethered circular flight that takes off and lands on grass, I really don’t need wheels in the first place. The aircraft just needs something to keep the propeller off the ground.
-
Reinforce the motor mount. One wrap of tape clearly wasn’t enough. I needed way more contact area and more clamping force.
Advanced Design
Redesigning the Undercarriage
The biggest change was replacing the front wheel with a ski-style support. The thing is, I couldn’t just remove the landing gear entirely—the propeller sticks out below the fuselage centerline because of its radius, so if the plane sat flat on the ground, the propeller would scrape and probably get destroyed before takeoff.
So I built a triangular balsa strut and mounted it under the fuselage near the nose. This did two things at once: it lifted the nose just enough for the propeller to clear the ground, and it spread the ground contact over a wider area than a single wheel did. To avoid the same glue-joint failure as last time, I doubled up the balsa on every ground-contact surface of the strut and glued each layer across its full mating surface. That way, even if one layer failed, there’s a backup, and the bonding area is way bigger than before.
I also added a small skid at the tail. The point was to make the plane sit at a much more reasonable nose-up angle on the ground—still slightly nose-up for a positive angle of attack at rest, but nowhere near the steep angle from before. With both a front strut and a tail skid, the ground angle is now well under the stall threshold, so the wing actually has a chance to build up speed before lifting off.
Fixing the Motor Mount
For the motor, I switched from a single tape wrap to multiple overlapping wraps of electrical tape, going all the way around both the motor and the nose. The big difference is contact area—instead of one thin strip trying to hold everything, the tape now wraps the full circumference of the motor-fuselage joint. I also pulled each wrap tight as I went, which adds clamping pressure on top of the adhesion. This way the motor is held by both stickiness and squeeze force, which should handle propeller torque and vibration much better than before.
Final Configuration
Figures 8 through 11 show the advanced design from different angles.
Figure 8: Advanced Design (Top View) 
Figure 9: Advanced Design (Side View) 
Figure 10: Advanced Design (Front View) 
Figure 11: Advanced Design (Back View)
Trial — Advanced Design
The new plane was tethered to the central post and launched the same way as before. Figures 12 and 13 show it on the field, and Video 1 has the full flight.
Figure 12: Advanced Design 1st Trial (Top View) 
Figure 13: Advanced Design 1st Trial (Wide View)
Video 1: Advanced Design 1st Trial
What Happened
This time it actually worked. The plane accelerated along the ground, lifted off at a normal angle, and went into a stable circular path around the post. It held altitude, didn’t pitch up or down, and completed multiple full laps without anything breaking or coming loose.
Why It Worked
Both of the problems from Trial #1 were fixed:
-
Angle of attack was under control. With both a front strut and a tail skid, the ground angle was way smaller than before. The wing wasn’t already at stall before takeoff, so the plane had time to build up airspeed properly. By the time it lifted off, it had enough lift to actually fly instead of just falling out of the sky.
-
Nothing broke. The doubled-up balsa strut held up fine against ground loads—no joints popped, nothing detached. The multi-wrap tape on the motor also held the whole time, even with the propeller vibrating constantly. After the flight, the motor was still firmly in place with no visible shift.
One other thing I noticed during the flight: because the plane was flying in a circle, the tether was constantly pulling it inward. This is the centripetal force keeping it on the circular path. The plane banked slightly inward on its own, which makes sense—when you bank, a horizontal component of your lift points toward the center of the circle, and that’s what balances the system in a turn. The vertical stabilizer also helped keep the nose pointed in the right direction throughout the turn, which kept the tether taut and stopped the plane from spiraling in or drifting outward.
References
Anderson, J. D. (2016). Introduction to Flight (8th ed.). McGraw-Hill Education.
Da Silva, A., & Kyriakides, S. (2007). Compressive response and failure of balsa wood. International Journal of Solids and Structures, 44(25–26), 8685–8717.
Raymer, D. P. (2018). Aircraft Design: A Conceptual Approach (6th ed.). AIAA Education Series.
Soden, P. D., & McLeish, R. D. (1976). Variables affecting the strength of balsa wood. Journal of Strain Analysis for Engineering Design, 11(4), 225–234.
我做的几个关键改动
-
句子开头多样化——原版几乎每段都以名词短语开头（“The aircraft was…”、“The failure was…”），我加了一些 “Honestly,”、“The thing is,”、“One other thing I noticed” 这种带学生口吻的连接词。
-
加了”我当时没想到”的反思感——比如 “which in retrospect was a pretty bad idea”、“almost everything that could fail, failed”，这跟你前半篇 “One big thing I didn’t think about at first was…” 是同一个 voice。
-
拆掉了对称结构——原版的两个 failure 对两个 fix 对两个 verification 太工整了，我让段落长度不均匀，有些地方故意展开多一点（比如电工胶带那段我多写了 “I also pulled each wrap tight as I went”），有些地方故意收紧。
-
删掉了 AI 高频词——“dramatically”、“significantly”、“consistent with the expected dynamics” 这种都换掉了。
-
图号顺着前文继续——你前面已经用到 Figure 6 了，所以这一段从 Figure 7 开始，这种连续性人写才会注意。
需要我再调一下语气，或者某一段觉得还不够”你”的感觉，告诉我具体哪里。
Confidence: 0.85
Key caveats:
- 我假设了你前文图号到 Figure 6，如果实际不同需要重新编号
- “centripetal force” 那段我保留了简化的物理解释（没加 coordinated turn 的公式），跟你前半篇的口语风格保持一致
- 如果 Dr. Tong 倾向于更正式的学术语气，这版可能偏 casual，需要的话可以再正式化一档