An important member of the DC Structures Studio wider team is Peter Coysh, our UK-based aerodynamics engineer. He’s got a fascinating career and unbelievable experience, including designing Formula 1 championship-winning cars.
Peter has also worked on DC Structures Studio projects including the Longford 90-metre cable stay bridge, Manganui Gorge 100-m suspension bridge, and the Fork Stream 70-m truss bridge.
Q&A
Hi Peter, tell us a bit about you!
I spent 18 years in Formula One across multiple different teams. Quite a few wins and world championships, designing and testing. Now I’m applying all those transferable skills to lots of different things.
What’s your engineering speciality?
I’m an aerospace engineer. My background is in what’s termed “low-speed aerodynamics”. So, Generally considering sort of up to about 200 miles an hour.
That doesn’t sound low-speed…
Everything up to Mach 0.3 is low speed. A commercial airliner doing Mach 0.8 would be transonic, and high-speed is Mach 2 and above. So that’s fighter jets, spacecraft, and experimental vehicles.
For so-called “low speed”, it doesn’t really matter whether it’s a car doing 200, or a bridge or a building, or a plane or helicopter, or whatever. It’s all the same techniques.
You’re trained as an aerospace engineer, so did you work on planes before Formula One cars?
I went through university doing a wide range of things, but mostly focused on commercial aerospace. I did a six-month placement at what’s now Leonardo Helicopters in the southwest of England, working on the aerodynamics and aeroacoustics of their current rotor blades – they were top secret back then. They’re known for high-performance rotor blades, very high lift, very high top speed, but very loud.
What’s a particular detailed aerodynamics project you’re proud of?
I did a project for someone who wanted to have an aftermarket add-on to the Tesla Cybertruck. This is before it came out, so we had to do surface reconstruction because no one had the CAD for it. So all we had was images on the internet, we turned them into an interactive 3D design, and then simulated that and made their package work with it. Later, they took the actual Cybertruck through a wind tunnel, and our simulations were within 1% of real-world performance.
What kinds of projects do you work on now, other than performance cars and bridges?
Basically, the ones that interest me, I’m lucky in that I can pick and choose. I do, for example, some electronics cooling work, so bike lights and industrial fish tank lights. I was just drawing up, in a spare moment for an architect in Italy, a sort of Arabic dome for someone’s house. And I was a motorsport consultant for a computer game.
How did you get into car racing?
I was always a Formula One fan, and I had in the back of my mind that if I do this, the opportunity might be there. Then after I finished my placement I put in some effort applying to teams.
What did you do in Formula One?
I ended up doing a really wide range of things, from performance modelling to trackside support to designing parts, testing parts in wind tunnels, working with the engine suppliers and tyre companies to maximise performance.
What’s your favourite story from the Formula One days?
The best one is probably the Brawn GP year, 2009. So I found out that my employer Honda were pulling out of Formula One from CNN while I was on holiday. There was a buyout of the team, and we had no money. So it was like, literally If you don’t need it, turn it off.
If we need to build something, we’re going to recycle anything we can. We’d spent 18 months, already designing the car. And then building it. And because we were no longer Honda, anything that was coming from Japan wasn’t coming. So they literally had to cut the back off the chassis, cut the front off the gearbox and make new bits to shoehorn a new engine and cooling system in.
You won the World Championship that year, right?
Yeah, because it was such a good car even after all that. I remember phoning up suppliers because we only had seven water radiators because they were being supplied by a Honda affiliated company in Japan, and they’d already sent seven. And that’s all we got, to last the whole year for two cars.
So they knew if they crashed they wouldn’t race again.
High stakes!
And we had to do last-minute changes of air horn, changes of cooling layout, to make it run. And we’re sitting there while everyone else is doing preseason testing as normal and thinking “They’re not very quick. What’s wrong?” But yeah, when our car came out, the others were just not that quick.
We managed to struggle on through the second half of the year with the car that was running out of bits and no money to develop it.
What a great underdog story. What happened next?
The same company then was bought out to become what is now Mercedes. There was a long lead-up. I was working on the Championship-winning 2014 Mercedes car, two years before the new regulations and hybrid engines. We stole a march on everyone. New era, new everything. I spent two and a half years just working on that one car.
What other F1 teams have you worked with?
Before that, I had spent five years with McLaren Racing. After Mercedes I was team leader at Toro Rosso, now known as VCARB. And then I went freelance but also did some work supporting Hitech Motorsport’s potential entry to Formula One.
What did you take away from those years?
The Formula One Method.
What’s the Formula One Method?
Whether it’s a bridge or something else, the first thing in a full project will be requirements capture and performance analysis. So, For me, the F1 method is not to dive in and start designing parts, but to actually work out what parts you need. What should you be focusing on? What’re the important characteristics?
Can you give us an example?
A couple years ago, I did an outboard motor. I was working with it with an electric motor company. So we wanted to design the cooling system, we did a bespoke propeller for it.
So starting with performance analysis. How do we make the best propeller to suit these kind of boats, this electric motor? What are the characteristics that are going to work? Once you’ve worked out those kind of details, then you take it to CAD design. You know how big it’s gonna be, how many blades, what kind of profile the blades are going to want.
And then you get into fine detail of the hydrodynamic sections, CADing it up. That week or two of performance analysis will save you six months of design.
What are the F1 Method steps?
In F1 you’ve got no time. It’s a good environment to learn in. So, as much as possible, early simulation to try and work out what the key performance parameters are. It’s all about performance.
Step 1: Requirements capture. Understanding what the limitations are. What are the pinch points you’re going to face?
Step 2: At those points, you’re building computer simulations of various different kinds. Everything from just a simple spreadsheet, calculation all the way up to high-fidelity simulations if required. Take all of this different information to understand what’s important in attacking those limiting points.
Step 3: Feedback loop. You’re taking those performance simulations and the targets to inform what you design. When you find a solution to a pinch point, it might create a new pinch point. You go around the loop.
Does the feedback loop ever end?
For a car? No. You just keep getting better and better, even during a season. You might have designed and tested four full engines before it even gets to a race, and then the end-of-season engine is four steps on from that, with 70 or 80 horsepower more.
For bridges, though? Or a project that has an endpoint? Yes, you loop until you’ve got your parameters right, the performance and the cost right, and you’ve got a limited time until you need to start building.
What does that look like, for example, at Mercedes with the new 2014 F1 regulations?
In requirements capture, we looked at drivability, we looked at how to make the petrol engine, the batteries and the two electric motors all work together. We just needed to keep adding more power, which made everything bigger. So there are the pinch points: fitting everything in the car, and cooling it. And as it gets bigger, it gets heavier.
So we simulated. Where do we put all the weight? So, that actually informed how they laid out the whole car, how they laid out the engine. They moved the compressor to the front of the engine. Normally, the compressor and the turbo go next to each other, and would get each other hot. But if you put the compressor on the front of the engine, it’s better for the car in terms of layout and weight distribution. And it’s better for temperatures.
But you’ve now got a 500-millimetre long shaft spinning at 125,000 rpm. Okay, you’ve gone around the loop, you’ve found a new pinch point.
You go ahead and find out how to fix that. We found a problem. We’ve understood it, we’ve simulated it, we’ve tested it. We found a new problem. Okay, hit that again in the same way.
So the F1 Method is a good environment for adapting to new regulations?
Right. How do you take this new regulation or this new opportunity, and turn it into the fastest car? All the teams generate that information as early as possible and we’ll be refining it as things go along. It’s the same for bridge engineering, or anything else where things change over time.
Let’s talk about bridges then! Starting with the basics: why are aerodynamics important for bridges?
For bridges, it’s always about the edge-of-envelope. It’s always: does it break, or is it heavy, is it expensive?
It’s slightly different to a car, where every time it moves, you’re gonna have to put more fuel or more electricity in, or it’s gonna handle wrong and crash more often if you’ve got the aerodynamics wrong.
But yeah, on a bridge or building, you’re thinking is it safe? Is it going to have to have modifications later on? Is it going to be buildable? Is it going to start to flutter at part-completion? All these things can occur.
And also, it’s about user comfort. If you’re crossing a very large valley, the wind speeds and buffeting can put people off using it.
So you’re not just thinking about the performance of the completed bridge, but also making sure nothing goes wrong with the wind partway through the build?
That’s right, and there’s lots of different ways to handle that.
And also making sure once it’s built you don’t need to come back and do more work for unforeseen wind conditions?
They often have to come back and retrofit parts under bridges to deal with buffeting. So, there’s always those extreme cases of turbulence, of wind speed. Unusual events that we’re always looking for, with bridges.
So the main thing that you consider is wind?
For bridges, it’s all about airflow, wind flow. I also have done some acoustic work with DC Structures Studio.
What does acoustic work entail?
It’s pretty complicated. In this case, we were designing a cable stay bridge. One of the local residents was concerned that the cables would make humming noises in the wind. I assessed the cable, but that’s only the first step, because most people can’t easily understand what an engineering report actually means.
So I was able to say that it’s going to be 28 decibels, equivalent to a soft whisper, it’s going to sound like an E5 string on a violin, and it will be drowned out by the sound of the wind. So when we sent it to the planning authority, everyone could know straight away what kind of sound we’re talking about without having to decipher technical jargon. It was a two-line sentence and it resolved the issue quickly.
Do you do that sort of thing often?
I’ve done acoustic work for helicopters, so I’ve got knowledge of that area. It’s a nice thing to have up our sleeves, but it’s not an area I focus on.
What are some of the innovations DC Structures Studio has used in aerodynamics?
One of the things we’ve done is using laser-cut aesthetic panels to break up the aerodynamic structures. Instead of having the whole bridge, all having the air flowing in unison, the more we can break that up, the smaller any problems will be. They’re a low-cost design that improves the aerodynamic performance and also because we can put laser-cut designs on them they add some architectural merit.
There are other benefits to these laser-cut panels too, which you’ll see on the new Manganui Gorge bridge and others. They also break up the balustrade system and sections of tensioned wires.
Does that have an effect on people walking on the bridge?
Yes, it moves less in the wind. Another thing we innovated was introducing porous decking that air can easily move though. If you imagine the deck of a bridge is a bit like an aeroplane wing, it has different air pressures on top pushing down and on bottom pushing up as air flows around it, that causes up-and-down movement. By making the deck porous, the air can travel through it which gets rid of those pressure differences. The difference that made was something like a magnitude of 10, huge.
And since we specified that for DCSS footbridges, there are now universities out there that are doing work looking at how they might put porous sections into big road bridges.
What’s the process you use for aerodynamic design on a bridge?
Yeah, it’s not a standard process. It’s a non-coupled aero structure analysis. We do a dynamic CFD [computational fluid dynamics] analysis on a part or a whole bridge to understand the loads and frequencies. CFD is basically a wind tunnel generated on a computer.
That’s then processed and fed into a simple computerised spring-damper model of the bridge to understand how large the oscillations will be and more importantly, whether they all naturally damp away or whether they will become unsafe.
That’s the requirements capture and the simulation. Then we do the feedback loop of design.
We’re looking at performance in those 1000-year extreme high wind events, and also whether the bridge will move around in very light wind, because we want it to be comfortable to walk on every day.
Do you also do higher levels of modelling?
I can do what’s called a localised model. So for example I’ve been working on a football stadium for the 2030 FIFA World Cup. We’ve done testing on that, looking at the aerodynamics of the cantilevered roof. The next step is to broaden that out to a 10-kilometre radius around the stadium. We’re looking at the topography, the coast, the buildings surrounding it. We’re looking at the wind inputs from different directions. We’re trying to make sure that we’ve captured all the potential problems, such that it will be built right first time. I can absolutely do that for larger bridges where we need to consider broader inputs.
What is the standard for this type of analysis?
I don’t think anywhere in the world has got a CFD wind engineering policy yet, but when it does happen, this type of analysis will be up to or ahead of the standard. It’s state-of-the-art. I’m in touch with a former colleague who is working with the Australasian Wind Engineering Society, who are working on a quality assurance manual for wind engineering, and there’s nothing in there that we’re not already doing for these bridges.
How accurate is your modelling?
When I compare my CFD computer analysis with real-world performance of completed structures, my correlation is usually within 1% of the real-world performance. Easily within 5%, which is the normal standard in structural engineering.
To find out more about the DC Structures Studio team, and our specialised expertise in pedestrian and cycle bridges, why not contact head engineer & architect Dan Crocker to have a chat about our capabilities?
