Tradespace exploration and the Pareto front: a solar-electric aircraft
A morphological matrix turns a few design choices into hundreds of candidate systems, a simple physical model evaluates each, and the Pareto front identifies the designs worth discussing. The method is applied to a solar-electric aircraft at different latitudes and seasons.
Engineers often select a design early. This is efficient, but it usually means that relevant alternatives are never evaluated. A system often has a particular wingspan simply because the first sketch had it.
Tradespace exploration avoids this. The design choices are listed, combined systematically and evaluated with a model, and only then are the designs identified that no other design outperforms on every objective. This article shows how a few choices generate a tradespace and what the Pareto front shows that a single "best" design does not.
A buyer choosing an electric car applies the same logic, usually without naming it. A car that is both more expensive and shorter in range than another is never the right choice, whatever the buyer's priorities. Removing every such car leaves the Pareto front: the cars for which more range can only be had at a higher price. Personal circumstances, such as the length of the daily commute, then choose among them.
The systematic combination of options is Fritz Zwicky's morphological analysis: list each design dimension and its possible values, and consider every combination; his Discovery, Invention, Research through the Morphological Approach (1969) is the classic account. The notion of a set of solutions in which no objective can be improved without worsening another goes back to Vilfredo Pareto's Manuale di economia politica (1906). De Weck's roadmapping method makes "morphological matrix and tradespace" the technical-model section of every roadmap, and its sample roadmap is a solar-electric aircraft, the same subject as below. For the physics, André Noth's doctoral thesis (2008) on the design of solar airplanes for continuous flight sets out the energy balance this kind of model rests on.
The morphological matrix
A solar-electric aircraft meant to stay up day and night has to collect enough solar energy during the day to fly and to charge the batteries that carry it through the night. Four choices dominate:
| Dimension | Options |
|---|---|
| Wingspan | 15, 20, 25, 30, 40, 50 m |
| Battery specific energy | 250, 350, 450 Wh/kg |
| Solar cell efficiency | 20, 25, 30% |
| Payload | 5, 10, 25, 50 kg |
That is aircraft.
The model
The model is my own and deliberately simple. Level flight needs power
which grows with mass faster than linearly and falls with wing area . The batteries must store times the night length, so their mass grows with , so the total mass grows, so grows. If this loop converges, the design closes; if not, it is the well-known mass spiral and the aircraft cannot exist. A design that closes must then pass the energy test: solar energy collected in the day must cover the day's flight plus recharging the battery for the night.
The figure below evaluates all 216 aircraft. Only a few combinations are worth discussing, the Pareto front identifies them, and it changes considerably with the operating environment.
- The reference case. At the equator (12 hours of night, full sun), 129 of the 216 designs close and 112 fly day and night. Only 9 are Pareto-optimal: for each, no other design is both lighter and carries more payload. Hovering over them shows that every one uses the 450 Wh/kg battery and the smallest wing that works for its payload, and the solar-cell efficiency hardly matters. At the equator sunlight is plentiful; the battery, which has to carry the aircraft through the night, is the binding technology.
- Summer at mid-latitudes. Short nights make batteries lighter. More designs close and fly (174) and the front shifts to lighter aircraft.
- Winter at mid-latitudes. Long nights and weak sun. Not a single design flies: they close, but the sun cannot recharge them. This explains why high-altitude solar aircraft operate near the tropics or in summer.
- Battery improvement. Stay in winter and raise the battery improvement factor. At ×1.4 three designs fly; at ×1.6 seven, with up to 10 kg of payload. That number, the battery specific energy needed to open a market, is what a roadmap's figure-of-merit target should be.
The Pareto front as a summary. Let design dominate design if is at least as good on every objective and strictly better on one. Whatever weights a decision maker puts on payload and mass, a dominated design is never the unique best choice: the design that dominates it scores at least as well under every weighting with non-negative weights. So the best choice for any such preference lies on the front, and the 216 aircraft reduce to the few worth considering. The weights, meaning the business case, decide among them.
A tradespace changes the nature of a design review. Instead of defending one design, the team shows the frontier of what today's technology allows, and how it would move if a figure of merit improved. That second picture connects engineering to technology strategy, and it cannot be derived from a single point design.
References
- V. Pareto, Manuale di economia politica, Società Editrice Libraria, Milan, 1906.
- F. Zwicky, Discovery, Invention, Research through the Morphological Approach, Macmillan, 1969.
- A. Noth, Design of Solar Powered Airplanes for Continuous Flight, doctoral thesis, ETH Zürich, 2008.
- D. Knoll, A. Golkar and O. de Weck, "A concurrent design approach for model-based technology roadmapping", 12th Annual IEEE International Systems Conference (SysCon), 2018.
- O. L. de Weck, Technology Roadmapping and Development: A Quantitative Approach to the Management of Technology, Springer, 2022.