Designing an aircraft
What does the process of initial design for an aircraft actually look like?
Apologies for the delay since the last issue – I was moving halfway across the country! Several trains, vans and a lot of shifting boxes later, I’m now settling into a new flat in Manchester, and I’m itching to dive back into the world of decarbonised aviation.
So, last time we looked at the potential of various key technologies to improve the efficiency of an aircraft, potentially improving electric/hydrogen aircraft significantly. But how do you go about actually designing an aircraft? How do you decide the key parameters of the design, and what influences those decisions? Today, let’s dive into the early stages of this process, and consider how that changes for novel technologies and designs.
As usual, I’ll caveat this right up front – I’m not an aviation expert, and as such I may miss things or make errors. Therefore, if you have suggestions, questions or spot areas where I’ve missed the mark, let me know!
An Overview
There are many ways to approach a complex topic like (aircraft) design, but however you break it down, the process resembles a funnel. At the top, you figure out what problem you’re trying to solve, and figure out the broadest shape of a solution. As you progress down the stages, the interrogation and analysis becomes ever more rigorous and in-depth, and the play in the design steadily smaller. For aircraft, that could well look something like this:
Scope the mission – what role does the aircraft need to perform, and what does it need to do that? What key constraints and requirements do those impose on the aircraft?
Preliminary sizing – using those constraints and requirements to come up with an initial design, choosing both an overall design as well as sizing the most vital aspects.
Conceptual design – translating the initial sizing into specifics, locking down the broad planform, choosing airfoils and so on.
Detail Design – iterating into progressively finer details, working all the way down to each individual component and making sure everything works as it should.
Different approaches may break the process down in slightly different ways, and use different naming, but the broad strokes are universal. For this post, we’ll focus more on the start of this design process, as well as the limits when applied to novel designs (which, after all, is why we’re here!).
What’s the Mission?
Aircraft are designed around particular missions and goals. They’re inherently specialised – a Cessna 172, a C-5 Galaxy and an F-22 Raptor are all vastly different aircraft, and share very little in terms of their design. Hence, the very first step when creating a novel aircraft is to figure out what role it will play. This in turn requires an understanding of what’s already out there, how well they perform, strengths and weaknesses as well as gaps in the market. Anyone who’s done market research and competitor analysis will be very familiar with this step!
Looking at commercial aviation, the market divides broadly into cargo and passenger markets (although many passenger airliners also carry cargo, which complicates things somewhat). To massively simplify, commercial designs are driven by two main key requirements – what do you want to carry, and how far do you want to take it?
Range is perhaps the simpler of the two. The shortest flights can be tens to hundreds of kilometres, whilst long haul flights can span half the planet. Range in turn influences many other requirements of the aircraft – longer range likely means flying higher, which means dealing with pressurisation, as well as flying in less dense air. Longer range also likely means flying faster, and far more of the flight will be spent at cruise versus climbing/descending. The further an aircraft travels, the more redundancy and safety measures required, as there’s more likelihood of being further from safety as and when any issues may arise. Range may also affect the size of aircraft that is financially viable, as fuel loads and overhead from supporting longer range may rule out smaller aircraft simply through scaling factors.
Payload is a little more complex, because what you want to transport can impose differing requirements. Cargo is perhaps simpler, but even that can come in a host of different sizes and densities. Some cargo will need pressurisation and certain temperature requirements, whereas other loads may not. Some cargo is containerised, and so a cargo aircraft may need to support loading and safe storage of specific dimensions. Heavy lift aircraft may need to transport vehicles, or very specific shapes of cargo.
Passengers impose a whole other basket of constraints. They need life support and safety equipment – seats, belts, life-jackets, emergency exits, slides and more. If the aircraft is flying high enough, they need cabin pressurisation and heating. And then there’s all the amenities – toilets, bag storage, galleys, food, entertainment systems and more! Not to mention, that if you have passengers, you need a cabin crew to look after them.
So, how do you pick a mission? If you’re looking at commercial aviation, then you’re presumably looking to make money with this aircraft. That means looking at the existing market and understanding trends past, present and future. What aircraft are flown on what routes? Are they full? Are airlines clamouring for a particular size of aircraft that doesn’t currently exist? Are there other pressures on the aviation market that could affect demand?
As with many areas of business, this is part rigorous analysis and part black magic/crystal ball gazing. Aviation is an industry with long timelines – it can take decades to bring a new aircraft to market, and in that time a lot can change. COVID-19 is a perfect example of how industry dynamics have shifted enormously, multiple times, in the space of little over a year. Climate change will likely shift things yet further, and probably in directions we can’t yet fully anticipate.
Flexibility is fundamental
This combination of uncertainty and long timelines, combined with huge development costs, is a major factor behind the current ‘family’ approach to designing aircraft. Most commercial aircraft designs introduced are designed from the outset to support multiple different variants. This can range from fuselage extensions/reductions, business jet or cargo variants, to re-engining and more. This not only allows a single design to capture more markets, but also hedges against market conditions changing and making a design ill-suited for its original purpose.
Sometimes these bets work, and sometimes they don’t. The Airbus A320 is an example of a wildly successful family design, morphed into the A318, A319 and A321, with long range versions of some models. The whole range has been refreshed with the neo variants (New Engine Option), and is even pushing into markets normally occupied by wide-body aircraft. With steady sales and increasing adoption, it is now neck-and-neck with the Boeing 737 for the title of highest-selling jet airliner ever made, despite launching around two decades later!
However, another Airbus creation, the A380, did not fare so well. Again, this was designed to be an extensible platform, with freighter, stretch and re-engined variants proposed at various points. However, unfortunately, the A380 was a bet on what had seemed a steady evolution to ever-larger airliners and a hub-and-spoke flight pattern. Unfortunately, the industry shifted towards more of a point-to-point approach, and bar a few key routes, the aircraft was just too large. With COVID-19 dramatically cutting passenger numbers, most airlines running the type mothballed them, and Airbus saw the writing on the wall and cancelled the program, possibly before ever turning a profit.
Preliminary sizing
Let’s say our crystal ball is working well, we get a glimpse into the future, and find a particular mission which looks to have strong demand for years to come, and fancy a crack at the prize. What now?
The first stage is to assemble a thorough set of requirements and constraints for our aircraft. These cover not only the specifics of the chosen mission (minimum range and payload, but also runway length, cruise speed and other factors), but also commercial constraints (is it cheaper to run?) and more. Finally, aviation regulations impose their own set of constraints on aircraft. FAA part 23/25 or EASA CS-23/CS-25 make sure that all certified aircraft meet various performance and safety criteria, from take-off and landing performance to crew numbers, aisle widths, fire safety and much, much more. Not all will apply at the earliest stages of design, but some, such as climb rates and fuel reserves, can make or break an idea.
Once we have those requirements and constraints, we need to map those into the aircraft design space. What limitations do they impose on our proposed plane? How many engines should it have? Should it use canards? A t-tail? Are there specific mission requirements that constrain the design, or would a particular design offer a boost to particular use-cases?
Once we have our configuration, we can then draw our lines in the sand for key parameters such as MTOW (maximum take-off weight), wingspan and wing loading. These numbers won’t be final at this point, but the closer we can get, the more effective and faster many of the later sizing calculations will be.
Given the interdependent nature of all the various systems in an aircraft, there are a plethora of ways to approach preliminary sizing. Each major manufacturer has their own techniques and specialised software for such tasks, and there are yet more amongst the many textbooks and academic courses on the topic.
Many of those methods use rules of thumb to approximate aircraft behaviour. Simulating the full physical processes at work is usually intensive and thus is used sparingly, but if an aircraft design is similar to existing designs, there are many parallels that can be drawn. Major basic factors from wing loading to cabin dimensions can be guesstimated at an early stage from aircraft already in production. Those existing aircraft also serve as validation points when creating initial designs – if your tube-and-wing aluminium airliner has an aspect ratio wildly different from the majority of such aircraft flying today, there’s a good chance that some part of your design needs re-evaluating.
If you want a better idea of how such a process works, I’d recommend this open course from the Hamburg University of Applied Sciences. It’s based on the earlier work of Loftin, and gives an excellent overview of different aircraft designs, as well as how one would approach preliminary design of an airliner.
Into the unknown
I could easily dedicate an entire series of posts to breaking down how various design processes approximate different aspects of an aircraft’s performance, but ultimately, such processes lose their utility the further from the beaten path one goes. They work well for designing conventional aircraft precisely because conventional designs follow a well-known pattern. This intuitively makes sense – we can’t reason easily about a novel design because of its very novelty – its behaviour and dynamics are inherently uncertain.
That leaves us with a choice. The first option is that we can choose to carefully figure out the behaviour and dynamics of our novel aircraft, building up a mass of data. Only then, once we understand the design fully do we attempt to start shaping it for a commercial mission. This is the option taken by traditional aerospace companies and research institutions. It’s slow, but careful and rigorous.
The second option is we learn by doing, and attempt to de-risk the design as we go. It’s a harder and riskier path, but if successful, could bring novel designs to market far faster than the traditional approach. Given the pressing nature of the climate crisis and the need for novel architectures, speed (if it can be done safely) is of the essence. We’re out of space for this post, and we’ve barely scratched the surface, but join me in a future edition where we’ll explore how to break down a complex and novel configuration into testable sections, and ways to manage risk and uncertainty when everything is new.