AEAC fixed-wing concept proposal

My concept proposal to UVic AERO for the 2027 AEAC competition aircraft: a fast twin-motor fixed-wing. One option for the team to consider, not a final design.

This is a proposal, not the team’s aircraft. I put it forward for UVic AERO’s design review on 1 October 2026. Nothing here has been built.

The problem

UVic AERO is entering the 2027 AEAC student UAS competition. The first task is a herd survey: sprint 5 km to the survey area, fly a grid, then orbit and read 3 cm ear tags, all in 30 minutes on one battery. Time to the survey area is ranked.

The rules cap the aircraft at 15 kg, electric only, with no parachute. It has to fly in wind up to 10 knots, gusting 20.

This is the fast option: dash at 45 m/s, survey at 20 m/s, with enough battery that endurance is never the limit.

Span: 2.30 m
Wing area: 0.736 m²
Flying mass: 7.15 kg
Design mass: 9.5 kg
Dash: 45 m/s
Stall, clean: 11.4 m/s

Everything is sized for 9.5 kg, about 2.3 kg more than it flies at today, so payload can be added later without a redesign.

General arrangement drawing with top, side and front views, dimensions in millimetres, and a summary of wing, tail, balance, power and speeds — General arrangement. Dimensions in millimetres from the wing-root leading edge.

How this was made

I set the concept and requirements, decided what to compare, and chose between the results. I built the analysis code (AeroSandbox, plus a mission simulation) and the Fusion models AI-assisted. I then checked the aerodynamics by building the aircraft in flow5 and comparing values, and ran a modal study of the tail in Fusion.

Getting to this layout

I went through six layouts in three rounds. Each round I described what I wanted to compare, generated the models AI-assisted, and chose a direction.

Twin-boom pusher layout with a long wing — 1a. First pass, endurance

Twin-boom pusher layout with a shorter wing — 1b. First pass, speed

Twin tractor motors on the wing, camera in the nose, T-tail — 2a. Nose camera, twin tractor, T-tail

Single tractor motor in the nose, camera under the chin, conventional tail — 2b. Chin camera, single tractor, conventional tail

Twin tractor layout with a T-tail, third round — 3a. Easy cornering

Single tractor layout with a conventional tail, third round — 3b. Fast

The first round was a twin-boom pusher. I dropped it because the team wants to hand-launch, and a propeller behind the wing is in the way of the throw. The second round put the motors in front and split on the camera: in the nose with a motor on each wing, or under the chin behind a nose motor. The proposal keeps the nose camera and twin motors from 2a and the low tail from 2b.

Key decisions

Airfoil

	RG15	HQ 2.5/12 ✓
Thickness	8.9%	12.0%
Drag at dash (cd)	0.0051	0.0060
Max lift at landing	1.20	1.30
Fits the 25 mm spar	No	Yes

Chose HQ 2.5/12 because it has the lowest dash drag of any section thick enough to hold the 25 mm spar. The thin speed sections (RG15, S2048, MH32) were faster on paper but too thin at the root.

Lift against drag for HQ 2.5/12, RG15 and S2048 at cruise and dash Reynolds numbers — Section polars from NeuralFoil at cruise and dash. The shaded bands are the lift coefficients the aircraft actually flies at. Eight sections were compared at landing, cruise and dash.

Wing size

	2.1 m span	2.3 m span ✓	2.5 m span
Cruise drag	Higher	Baseline	About 4 W lower
Stall speed	Above 12 m/s	11.4 m/s	Lower
Weight and roll rate	Lighter, quicker	Baseline	Heavier, slower

Chose 2.3 m span because it sits at the knee of the curves. Going to 2.5 m saves about 4 W in cruise for a heavier wing that rolls more slowly; going smaller pushes stall above 12 m/s.

Three plots of dash drag power, cruise drag power and stall speed against wing area for four spans — Planform sweep at flying weight. The chosen wing is 0.736 m² with an aspect ratio of 7.2.

Motor layout

	Single nose motor	A motor on each wing ✓
View from the nose camera	Behind the propeller	Clear
Slipstream over the stabiliser	Yes	No, passes outboard
If one motor fails	Glide	Strong yaw, needs managing

Chose A motor on each wing because the camera needs a clear view from the nose to read the ear tags. The cost is the one-motor-out case, covered below.

Tail

	T-tail	Low tail ✓
Flying mass	7.28 kg	7.15 kg
Fin area	About 25% larger	Baseline
Elevator	Clean air, above the wake	Behind the wing
Mode near prop frequency	Yes, 68 Hz	Not in the estimate

Chose Low tail because the T-tail's modes land where the propellers turn at loiter and cruise. Its clean-air elevator isn't worth that, or the extra weight.

Rear view of the aircraft with the stabiliser mounted low on the boom — Low tail (chosen)

Rear view of the aircraft with the stabiliser on top of the fin — T-tail variant

Chart of estimated tail vibration frequencies against propeller frequency. The T-tail's fin torsion at 70 Hz and fin bending at 90 Hz fall inside the bands around the propeller frequency at loiter and cruise. The low tail's boom torsion at 31 Hz is clear of them. — Estimated tail modes against the propeller's once-per-revolution frequency. Moving the stabiliser to the top of the fin also drops boom torsion from about 31 Hz to about 17 Hz.

The chart shows the first estimates. The modal study in Fusion agreed: the T-tail has a mode at 68 Hz, and its carbon spars on their own have one at 77 Hz. The propellers turn at about 70 Hz at loiter and 78 Hz at cruise, and the target is to keep every mode at least 15% away from both.

Fusion modal result for the T-tail on its boom, coloured by displacement, with a list of eight modes from 15.1 Hz to 102.7 Hz — T-tail modal study in Fusion. Mode 1 is at 15.1 Hz and mode 6 at 68.0 Hz.

Fusion modal result for the T-tail's fin spar and stabiliser spar, showing mode 5 at 76.951 Hz — The T-tail's carbon spars on their own: mode 5 at 77.0 Hz.

Battery

	12S2P pack ✓	12S3P pack
Flying mass	7.15 kg	8.06 kg
Endurance to 20% reserve	70 min	104 min
Charge left at 30 min	60%	72%

Chose 12S2P pack because it already gives 70 minutes against a 40-minute target. The larger pack stays as the option for heavier payloads.

Drag and power

Total drag is 6.4 N at 20 m/s and 18.1 N at 45 m/s, including a 1.35 factor for student build quality.

Bar charts of drag by component at cruise and at dash — Drag build-up. The wing dominates at both speeds. The pod is the next place to clean up.

The dash draws about 1,560 W at 76% throttle and cruise about 258 W. The motors have about three times the power the dash needs.

Thrust power needed and available against airspeed, and battery power drawn against airspeed — Left: thrust power needed against power available at full throttle. Right: battery power drawn.

Structure

The design case is 9.5 kg at +6 and −3 g, with a 1.5 factor on top for ultimate load. A 5 m/s vertical gust at dash speed gives 4.5 g and a 7.5 m/s gust gives 6.2 g, so the dash is only flown when gusts are moderate.

V-n diagram showing the manoeuvre envelope at 9.5 kg with gust lines at 5 and 7.5 metres per second — V-n diagram at the 9.5 kg design mass, with 5 and 7.5 m/s gust lines.

The main spar is a 25 × 1.5 mm carbon tube at 28% chord. At ultimate load the root bending moment is 218 N·m, which puts the tube at 355 MPa against an allowable of about 600 MPa. With 2° of washout the root reaches its maximum lift first, so the aircraft should drop its nose in a stall, not a wing.

Wing bending moment along the span at ultimate load, and local lift ratio along the semi-span — Left: ultimate bending moment along the span. Right: local lift ratio, highest at the root.

The biggest risk: one motor out

If one motor stops while the other is at high power, the aircraft yaws hard. The table compares that yawing moment with what full rudder can hold.

Phase	Speed	One motor, full throttle	One motor, trim thrust	Full rudder holds
Just after launch	14.3 m/s	30.2 N·m	3.7 N·m	5.1 N·m
Cruise	20.0 m/s	27.9 N·m	3.5 N·m	10.1 N·m
Dash	45.0 m/s	16.3 N·m	10.0 N·m	50.9 N·m

At trim power the rudder copes everywhere. At full power it only copes at dash speed, and a fin big enough to fix that would need about six times the area. So the plan is:

Cap the throttle at 60% for launch and climb, and 85% for the sprint.
Feed each motor’s RPM to the flight controller, and cut the good motor within about 0.3 s if the two diverge. The aircraft then glides to a landing.

Mission simulation

The simulation flies the whole task: launch, climb to 50 m, 5 km of sprint laps, a survey grid, six tag orbits, then loiter to the reserve.

Airspeed, battery power and state of charge against time through the simulated mission — Airspeed, battery power and state of charge through the task. The dashed line is the end of the 30-minute window.

Even with 60% more drag than the model predicts, the pack lasts over an hour. Endurance is not what limits this design. Gust loads, flutter and the motor-out case are.

Result

5 km sprint: 129 s
Top speed: about 54 m/s
Charge left at 30 min: 60%
Endurance to 20% reserve: 70 min
Static margin: 12%
Spar stress at ultimate: 355 of 600 MPa

On paper the proposal reaches 45 m/s, finishes the task in about 9 minutes and has more than twice the endurance it needs. These are simulation numbers, not test results. The list below is what has to happen before I would trust them.