Two routes nobody flies direct. One aircraft for both.
The brief came with two randomly assigned routes and a weekly passenger count for each. Neither has a direct flight today; both need at least one change.
The primary route, Mexico City to Saint Lucia, carries the passengers. The secondary route, Rio de Janeiro to Washington, is nearly twice as long with a third of the demand. Designing one aircraft for both meant every choice had to survive the harder of the two.
Design for the worst airport, and the rest take care of themselves.
- Runway
- 3,900 m
- Altitude
- 2,238 m
- Avg temp
- 16 °C
The highest airport. Air at 2,238 m is thin, so engines make less thrust and wings make less lift.
- Runway
- 2,745 m
- Altitude
- 4 m
- Avg temp
- 26.1 °C
The hottest airport. Hot air is thin air too, which hurts take-off the same way altitude does.
- Runway
- 1,320 m
- Width
- 42 m
- Avg temp
- 23.6 °C
The limit on everything. By far the shortest runway, and the narrowest: main landing gear had to sit within 35 m.
- Runway
- 2,103 m
- Altitude
- 4.6 m
- Avg temp
- 4.3 °C
The easy one. Cold, dense air and a mid-length runway: if the aircraft works anywhere else, it works here.
Take off and land on Rio's runway, at Mexico City's altitude, in Saint Lucia's heat. No real airport is that bad, which is the point: an aircraft that clears that combination clears all four.
Every number traces back to a route or a runway.
A narrow-body. Rio's 1,320 m runway rules out anything wide-body, and all four runways take aircraft under 90 tonnes.
Sized by constraint analysis across take-off, a 1.5° minimum climb, cruise at 35,000 ft and landing, all at the worst-case airport.
Big enough that the 4,104 km primary route, with a full cabin, fits inside the weight limit with margin to spare.
A standard airliner cruise. Both long routes reward a fuel-efficient cruise over a fast one.
More stable on the ground than a tail-dragger, tucked away for cruise, and spaced to fit Rio's 42 m-wide runway.
Worked out from the weight and position of every major part, from the wings and engines to the passengers and bags.
Why the CFM56-5A1, and not the better engines
Bypass ratio drives fuel burn, so more is better. The 3C2 has more, but costs $1.05M extra per engine, $2.1M for the pair, for a 0.4 gain. The newer LEAP-1A would be ideal, at around $14M per engine. We took the middle option: good efficiency at a price the aircraft could carry.
Per engine. All three cruise at Mach 0.80 at 35,000 ft.
140 seats, sized for demand, not for bragging rights.
260 passengers a week on the primary route works out at two flights of 130. So the cabin holds 140: a little headroom, and no empty metal flying across the Caribbean.
Low demand also made room for comfort. Business has 5 rows of 2-2 seats, 65 cm wide, with dividers. Economy has 20 rows of 3-3, 45 cm wide, above the narrow-body average, with extra legroom at the two exit rows by the wings. All-economy, the same aircraft could take 160.
Hover a seat
One route fits easily. The other only just flies.
Mexico City to Saint Lucia sits well inside the envelope: 130 passengers, 4,104 km, comfortably under the maximum take-off weight.
Rio to Washington is the problem. At 7,720 km it sits past the aircraft's maximum economic range, so payload has to make way for fuel. Our numbers put it at 44 passengers a flight. Physically possible, commercially weak.
The obvious fix, a bigger aircraft, is the one thing Rio's runway won't allow. So we set two flights a week on each route, carrying the 260 and the 86.
You don't design for the average case. You design for the binding one.
One short runway set the size, one high airport set the engines, and one long route showed where the design ran out. Finding the binding constraint first made every other decision quicker.
Same habit at work: find the one thing that actually limits a client's problem before optimising anything else.
Group design project (team of three), AERO20400 Aircraft Performance & Design, University of Manchester, May 2023.