Aerodynamics is not a precise science. If it were, then all F1 cars would look and perform the same (given the same engine etc), and there would be no need for the teams to spend millions of pounds on wind tunnel testing and computational fluid dynamics (CFD), and Adrian Newey would be out of a job. However, it isn’t, and that makes designing an aerodynamic piece of bodywork somewhat of an art form, where the intuition of the designer can make a great deal of difference. Just like a good artist knows how to paint a house to make it look, well, like a house, so a good designer knows how to draw a form that is aerodynamic.

So, when CUER recognised through our own wind tunnel tests that the 2009 canopy wasn’t very good (it was shedding two big vortices), a fourth year project student turned solar car designer sat down with the computer equivalent of a drawing board and sketched out some curves based on his own intuition and knowledge, and thus was born the first iteration of the 2011 canopy.

As pleased as no doubt he must have been with his work, you can’t just sketch some curves and call that a new design. The canopy was then honed through a series of wind tunnel tests and also with the use of CFD. With each run, things were learnt about the airflow over the bodywork. Patches of flow separation were found and the geometry would be subtly modified to try and correct it on the next run.

This iterative process underlies all aerodynamic development work, from the wings of 747s to the diffuser of a F1 car. Essentially, the process can continue indefinitely, ideally converging on a “perfect” design, but in the real world, time and budget constraints decide where the cut-off is.

This canopy was designed over the course of a fourth year project in a wind tunnel, and then integrated with the existing upper body geometry using a CFD based approach before a finished version of the canopy was assembled in Solidworks, our CAD package provided by Dassault Systems. It is purported to reduce the total drag of the car by nearly 10% compared to the previous canopy.

The canopy material was to be polycarbonate, chosen for its impact resistance (good for deflecting stray kangaroos) as well as its transparency. So, the next step of the production process was to make a mould for the vacuum forming.

The mould making is a two part process. First, a female mould is milled from modelling board, which is basically Wood+, which has very uniform density and stable material properties over a range of temperatures, allowing it to be easily milled to any shape.

In the past, these shapes would have been made by hand, no doubt by a very skilled carpenter. However, because Technology Is Awesome, all we have to do now is feed the CAD file into a computer connected to a big machine and it will cut the exact shape out that we need. This work was done by Jaguar Land Rover, who also generously provided the expensive modelling board needed for the process.

You’ll notice it comes in two halves; this is because we chose to hinge the front half of the canopy, whilst the rear half stays attached to the car, so we vacuum form the two halves separately. Due to the material property of polycarbonate, forming it directly onto Wood+ results in a translucent finish, which is not very conducive to being able to see properly. Therefore, we use a casting resin (ALWA Mould D), which was generously provided by JohnBurn. This resin is poured into the mould, which then sets and takes the exact form of the mould.

Once the resin sets, we then have a male mould and this is what we use to vacuum form the polycarbonate.

Et voila! We end up with a nice shiny new canopy. All that remains now is to attach it to the hinges and nest it into the upper shell and it will be done!

Age: 21

College: Fitzwilliam

Engineering Specialisation: Aerospace

Fourth year project title: Internal Flows of a Solar Vehicle

What you’re actually doing for your fourth year project: Underbody Aerodynamics of a Solar Vehicle, and an Integrated CFD Workflow.

Hah. No, I just run random meshes and then scarper to play badminton.

Height: 175cm (1.02 Tom Grimbles)

Anything else that you feel defines you: I’m CHINESE!

Describe your typical daily routine:

Alarm goes off at 7.30am. I snooze for over an hour, and then probably get up at around 8.45 to get to lectures, if I have any. Breakfast will be eaten after morning lectures are over, and then I will either hang out in the Ashby Lab doing CFD or go back to College to do work.

My days are really exciting.

Aside: If you’re interested, George provides a great deal more information about his daily activities on his blog. With graphs and pie charts. We kid you not.

If you could be any element on the periodic table, what would you be and why?

Francium – it’s rare and unstable.

What three things would you choose to take with you to a desert island (such as Australia…)?

Surely, it depends on the purpose?

For the WSC, I will be taking my MP3 Player, copious amounts of ibuprofen (because I can’t drive the car for more than 30 minutes without getting back pain) and something to annoy Lucy with – maybe a vuvuzela.

What does your ideal Saturday night involve?

Tear down some good roads in a Lambo. Play a game of Badminton. Drive quietly back up in Endeavour. Drink a few beers, pwn some noobs at CoD, shoot some flares and pimp some hos.

How many hairdryers would it take to power you?

42

If your subteam had to compete in an Olympic (or Winter Olympic or Paralympic) event, which would it be and why?

Bobsleigh. Those things are streamlined. Streamlined things are good. The aero subteam likes streamlined things. We can probably even use CFD to design an even faster bobsleigh than before!

What is your favourite joke?

If I told you I’d have to kill you.

If you were a Pokemon, which would you be?

Charizard. He don’t take shit from nobody.

What was the first thing you ever wanted to be when you grew up?

Racing driver.

What do you want to be now?

It’s changed. I now want to be an engineer.

Hah.

No, still the racing driver thing.

If you had to organize a Cambridge bop, what would the theme be, and what costume would you wear?

Anime party! I would wear a Goku costume, dye my hair bright yellow and pretend to be a Super Saiyan, which would fulfill my deepest teenage fantasy.

If your subteam were on the Titanic and there wasn’t enough room in the lifeboats for all of you, who would go down with the ship?

Me. I must have done something very wrong in order to have my sub-team end up in 1912 on a sinking superliner (or alternatively, I would have invented time travel and so should be saved so I can share this discovery with all of humankind).

If you were CUER’s Title Sponsor, what name would you give to the car and the team for the 2011 WSC?

Team: Super Robot Monkey Team Hyperforce Go!

Car: Bob

What is your favourite xkcd cartoon?

Angular momentum

What should your subteam’s/CUER’s theme tune be?

CUER: Power Rangers theme tune.

Sub-team: Jay-Z – 99 Problems

It is probably because of this everyday familiarity of the concept – or at least the language – of aerodynamics that BMW chose a slightly different approach in their latest ad campaign, in an attempt to sound hi-tech:

Transcript:

Is this the car of the future?

It’s made of lightweight materials.

It uses fuel…intelligently.

And…

It manipulates…the wind.

Really? It manipulates the wind? That’s clever.

Although most people are indeed familiar with words like ‘aerodynamic’, ‘streamlined’, ‘air resistance’, ‘drag’ (we’ll leave ‘wind manipulation’ for a bit) – what do they actually mean? Why should a torpedo be more aerodynamic than a brick? What is the science behind our instinct to streamline things?

We’ll start with ‘air resistance’, otherwise known as ‘aerodynamic drag’. There are actually two main sources of drag – skin friction drag and pressure drag. Both can ultimately be traced back to a single phenomenon known as the boundary layer.

The boundary layer exists as a result of the viscosity of air. At the surface of a body, the velocity of the air molecules must be zero (relative to the body), due to frictional forces. This is known as the “no-slip condition”. In the free stream, the air is moving at a (mostly) uniform velocity. Therefore, somewhere between the surface and the free stream, there must be a gradual increase in velocity. This region is known as the boundary layer.

The boundary layer characteristics (thickness, type) are related to a measurement known as the Reynolds number, which itself is derived from three factors:

- the density of the fluid

- the speed at which the fluid travels relative to the body

- the distance along the body the fluid has travelled.

A boundary layer may be laminar or turbulent. Briefly, a laminar boundary layer occurs at low speeds and densities, and can be thought of as consisting of horizontal motion of air molecules, in neat ‘layers’ of increasing velocity. Turbulent boundary layers are associated with high speeds and densities and – obviously – are more turbulent, displaying large-scale vertical and horizontal unsteady motion. It is very difficult to keep a boundary layer laminar, as even very small discontinuities in a surface can be sufficient to cause transition to turbulence.

Skin friction drag arises from friction – shear forces – between the molecules in the boundary layer and the surface of the car. It depends on factors such as the total surface area of the car and the type of surface – a smooth surface will produce less drag than a rougher surface. Why? Because a rough surface will cause protrusions into the boundary layer, disrupting it and increasing the frictional force. It is for this reason that solar car teams go to such lengths to encapsulate their cars, forming a smooth uninterrupted surface. Additionally, turbulent boundary layers result in higher skin friction drag, so maintaining laminar flow for as long as possible is also important. Encapsulation helps with this, too, since with many solar vehicles, the discontinuity presented by a solar array can often cause transition to turbulence.

Back to the brick. It has a pretty rough surface, so we know that it will have a fairly high level of skin friction. We can smooth that out, so that it has a perfect surface, but it’s still unlikely that it will match up to a streamlined torpedo. A brick is not streamlined. It’s a brick. It’s got lovely perpendicular sides and right angles, which is great for house-building but won’t get it a cameo in Das Boot. This is where pressure drag comes in.

Sharp corners on a body present an ‘obstacle’ to boundary layers. The pointier the corner, the more of a problem it can be. This is to do with the pressure distribution around a moving body, caused by those corners. As an example, here is the pressure distribution around a sphere in inviscid (frictionless – so no boundary layer) flow:

Essentially, a concave curve away from the body represents an area of high pressure, and a convex curve towards the body represents an area of low pressure. This is due to Bernoulli’s Principle, since a directional change in airflow causes a change in velocity. The greater the degree of curvature, the greater the pressure difference, so on a body with sharp corners, such as a brick, the pressure variation over the body will be more extreme.

Now, a boundary layer on the front half of such a body has no problem, because the air is travelling from an area of high pressure to an area of low pressure – along a positive pressure gradient – and so is ‘pushed’ along. On the back half of the sphere/brick, however, this changes. The air is now travelling from low to high pressure – an adverse pressure gradient that opposes the airflow – and this has the effect of reducing the velocity of the air in the boundary layer – so much so that, eventually, it starts to go backwards. The boundary layer separates from the body. Instead, an area of unsteady, chaotic flow is formed – the wake.

This causes a problem. Looking back at the pressure diagrams, we can see that the area of high pressure at the nose of the body is balanced out by an equivalent area at the rear. If the boundary layer separates, this area is no longer present, replaced by a lower-pressure wake. The result? Pressure drag:

We can’t eliminate pressure drag, but we can reduce it. This is done by encouraging a much gentler adverse pressure gradient (APG). How? By reducing the sharpness of corners, blending everything into everything else – by streamlining. The boundary layer will separate further along the body and the area of the wake is reduced, along with the corresponding drag.

So, let’s streamline our sphere-brick:

Unfortunately for us, it no longer looks like a brick. Instead, it looks quite similar to the body of a solar car or a cross section of a wing – both of which belong to a class of objects known as aerofoils. It has lost its symmetry because it is the back half – the APG – that is causing the most problems.

Pressure drag tends to be the dominant source of drag, so teams will trade off a greater surface area for a smaller wake.

Now, aerodynamics does not end there. Even more complex things happen when you duct-tape glue fairings and a canopy onto that solar car body shape. The presence of the road has a strong aerodynamic effect on the vehicle. There are books written about the aerodynamics of spinning wheels. Wings are a whole other story entirely. Once the sound barrier is broken (unlikely, in a solar car) the rules are almost completely rewritten.

However, hopefully this has opened a window onto the challenges faced by the Aero Team. Air (and any other fluid) is such a wonderfully complex substance that the mathematics involved is incredibly nasty:

As a result, most of our aerodynamic research is qualitative – not dealing in numerical values other than experimental ones. CFD modelling is one tool that allows us to predict how exactly the air will behave around a car, but – with so many factors involved – it can never be completely relied on. Wind tunnel testing is absolutely fraught with interfering variables – the most obvious being that while we can scale down our car, we can’t scale down the boundary layer that goes with it, so our results will never be completely accurate.

Finally, from reading this article, you should have concluded (rightly) that it is very difficult to build a vehicle that doesn’t manipulate the wind in some way. It’s very difficult to build anything that doesn’t. And that therefore, BMW has wasted three very expensive seconds of airtime on informing us that its all-new car-of-the-future range has the same aerodynamic qualities as a brick.

****

For those who are keen to explore further:

• Stunning flow visualisation photographs
• Vortex streets (for those wondering what the circular pattern at the rear of the aerofoil image is – see Satellite Eye on Earth for some really amazing examples.)
• Introduction to Reynolds Numbers
• More on boundary layers

—————————————————————————————–

——————————————-

Image Credits

1. Nasa Glenn Research Centre
2. Aerospace Web
3. Aerospace Web
4. Coilgun Systems
5. Plus Magazine, Supersonic Bloodhound - Ben Evans

Ok, so there’s nothing here yet. But coming soon…