Cast your mind back to the 2007 edition of BMW-Sauber's F1 entrant. Those of you with sharp memories will recall a sleek looking car with nice, simple livery that looked rather classy next to the bright red of Ferrari and the silvery McLaren.

What about 2008? Visually things have gone rapidly down hill — looking at the car head on it is impossible not to get distracted by ugly viking horns plastered to the front bodywork — glance elsewhere over the car and you'll see various other appendages sprouting up (eg, the front wing is heavily detailed with a more complex design than most).

BMW-Sauber have gone from having the best looking car on the grid to the back row — only Honda and the infamous elephant ears come close to posing a challenge.

Fortunately for BMW success in F1 does not come down to looks. These changes reveal how the boffins in BMW's Hinwil research bunker are thinking about car design and development.

Generally a large part of improving performance in F1 is by being innovative in aerodynamics. For a primer on aero read part 1 and part 2 of F1-Pitlane's introduction to the subject.

And this is exactly where the BMW team decided to take a leap of faith and fully embrace the field of Computer Fluid Dynamics, or CFD for short. BMW-Sauber weren't the first team to embrace CFD — that honour goes to the old Benetton team &mdash but they were the first to differentially invest in computer technology at the expense of wind tunnels, which were the old fashioned way of improving car aero.

Let's cast our mind back 40 years ...

A short history lesson

In the days of Colin Chapman life was much simpler. If you had the means and motivation you'd mock up a mini version of your car, borrow a wind tunnel for a few hours and test various wing designs to see which added the most downforce for the least drag. Based on some approximate results you'd then take your most promising aero innovations to the track (or sometimes a race) to try.

Because of the vagaries wind tunnels, often what happened in the lab was not replicated on the track.

As technology evolved teams started to build their own wind tunnels. Because of the complexity of managing airflow 20-40% scale tunnels were common (increasing to 60% by the late 1990s and most of this decade). Now teams can and do run full-scale wind tunnels to test aero performance of cars.

However, wind tunnels only do so much. The biggest issue is accurately modelling tyre to air interactions. Since the wheels account for around 40% of the drag on an F1 car they exert a large aerodynamic influence. This is especially true in an open wheeled series were airflow coming off the tyres will interact with the aero devices behind it. Measuring this in a wind tunnel can be fraught.

Also working out aero performance through corners, or over bumps, or when following a car was difficult with a tunnel — this is known as transients and their research is at the forefront of CFD technology.

Let's fast forward back to today ...

The future is CFD

With the growing power of computers it has become possible to study the effect of airflow on F1 cars with some accuracy and in different situations.

Want to know what how tyres interact with the rear wing? CFD can tell you. What about when two cars follow each other? Yup, CFD has the answer.

So how does it work?

Effectively the CFD engineer loads the computer with a drawing of the car, usually using CAD software. A big 3D mesh is constructed over which the governing equations of fluid dynamics are solved. These equations are the conservation equations, the most common being the Navier-Stokes equation. The only problem is that this is a non-linear partial differential equation, which means it is darn difficult to solve.

Unsurprisingly these equations can take a long time to solve, especially if a high resolution is required or the model is complex. Indeed it is only in recent years that the computing power has been available to model a full car.

The data can then be analysed (eg, turned into graphical images — see main picture) to understand pressure points or how air streams interact with the car.

The beauty of using computers is that it gives the aerodynamicist much greater freedom to develop the car. By turning an art into more of a science we are seeing more innovative aero solutions pop up. CFD really comes into its own when modelling thermal and aero interactions — brake ducts would be a classic example.

Despite the rise of CFD, wind tunnels will not disappear overnight. Tunnels play an important role. CFD is used more for concept design and to aid understanding of aero interactions but wind tunnels provide experimental corroboration of CFD. Successful aero programmes combine world class CFD with an expensive wind tunnel programme.

Computers may seem like a silver bullet but they are not — plenty of hard work is still needed. For a start there are lots of things that can go wrong; this is a model after all. It's important that data from the computer correlates well with real world data and teams spend a lot of time ensuring that everything is properly calibrated.

Back to BMW

Last year BMW were facing a decision with respect to their aero capability. Do they invest in a second wind tunnel or plough the cash into a supercomputer?

Given the subject of this article you won't be surprised they decided to invest in Albert-2 (their first computer was called Albert-1). This went live in the middle of 2007 so it is no surprise to see the fruits of its labours appearing on the 2008 car.

Albert-2 has allowed BMW to significantly reduce their aero design cycles. Given the ease with which new parts can be tested and existing solutions optimised, it isn't hard to see why BMW have taken such an aggressive aero approach this year.

Let's revisit those ghastly viking wings. Their purpose is to control airflow over the car by generating vortices (downforce generation is minimal), which allows the back of the car to be more efficient and so reducing drag.

Had BMW invested in a wind tunnel instead its unlikely they'd have settled on such a radical solution.

By making sure they had the fastest and most powerful computer on the grid BMW were able to close the gap to their competitors. Indeed since then we've seen other teams, such as Renault, invest heavily in CFD.

Although wind tunnels won't disappear overnight make no mistake, the future is binary.

Up next

We'd really like to showcase a couple of CFD studies on F1-Pitlane. Obviously I don't have Albert-2 sitting in my living room so it is difficult. For those of you who want to read more on this subject I encourage you to read Computer Car Aerodynamics, a book that runs through a series of CFD studies by BAR-Honda's old CFD consultancy, Advantage-CFD.

Hopefully, over the next few months we at F1-Pitlane may get the opportunity to support some CFD studies being done. There are no promises and I certainly don't want to give too much away at this point but watch this space.