05 July 2006

Interview with JCB DIESELMAX Project Aerodynamicist

The man behind the successful Bristol Siddeley Bloodhound missile, Ron Ayers came into the national consciousness in 1994 when he conducted the arduous aerodynamic test programme for ThrustSSC. The project was able to proceed only after Ayers simultaneously conducted two separate aerodynamic investigations – one via computational fluid dynamics (CFD), the other with empirical full-scale speed testing of a model on a rocket railway – and was able to correlate the findings from the two so closely that he felt confident a supersonic programme could safely be achieved.

One of Ayers’ personal projects has been an intensive investigation of why a number of land speed record cars that have run at the Bonneville Salt Flats have apparently underperformed, given their various performance parameters. This study, together with his CFD investigation, has proved invaluable in the development of the JCB DIESELMAX project.


You are known for designing the shape of ThrustSSC, which achieved 763 mph. JCB DIESELMAX is much slower. Does this make it an easier project?

No. JCB DIESELMAX is just as difficult, and just as interesting, but the challenges are very different. If designing a 300mph diesel car were easy, someone would already have done it.

What makes JCB DIESELMAX difficult?

Unlike ThrustSSC, JCB DIESELMAX is driven through its wheels. This means that the car must be driven at the limit of wheel adhesion, which introduces problems of stability and control. Unlike ThrustSSC, which had the advantage of an enormous excess of power from its jet engines, we must make the very best use of the power from our JCB engines.

On such a design, where do you start?

First, we identify the internal features that define the minimum frontal area for the car. These clearly would include the engines, wheels, driver’s safety cell, transmission and suspension. A minimum slender skin shape is fitted around these items. Other internal features, such as turbochargers, cooling system, parachutes, fuel tank etc are, where possible, housed within this shape. At this stage, a great deal of iteration takes place between Visioneering (designing the structure and interior of the car), Mike Turner of JCB (defining the skin shape) and myself, in order to arrive at the final low-drag shape. Note that on such a slender vehicle, the skin friction drag is actually greater than the pressure drag, so minimising surface area is as important as minimising frontal area.

What other factors do you consider?

At the speeds we are designing for, controlling the vertical force is most important. We must avoid excessive aerodynamic downforce, as this would compress the suspension and it would also lead to an increased rolling resistance. You might think that an aerodynamic upforce would be beneficial, as this would reduce rolling resistance and hence lead to increased speed. Unfortunately this strategy would also lead to a reduced traction, with a corresponding reduction in performance, and to an increased risk of control loss. Thus, both on front wheels and on rear wheels, we have aimed to achieve minimum aerodynamic up/download.

Yaw stability is clearly important. Indeed, when the car is near limiting tyre adhesion, the yaw stability becomes crucial. However, too much yaw stability may introduce problems of controllability, and may make the car oversensitive to crosswind gusts, so careful consideration of fin size is essential.

I hear that you are also studying rolling resistance. How does this influence your design?

It is well known that rolling resistance increases with speed, but this is even more true on a Bonneville-type salt flat than on tarmac or concrete. This is because the front wheels dislodge salt particles, which then impinge on the downstream surfaces and the rear wheels, creating extra drag. I have designed the front wheel fairings so that the wake from the front wheels is gently diverted away from the rest of the vehicle. I hope this will significantly reduce rolling resistance. To the best of my knowledge, this has never been tried before on a speed record vehicle, so I await with interest to see if it works as planned. I hope so, because my current estimates of rolling resistance show that it is significantly greater than the aerodynamic drag, so considerable performance benefits could accrue.

You have been quoted as saying that the main aerodynamic problems are related to the underside of the car. What are these problems?

The low ground clearance means that the airflow is severely constrained. Thus, in the high pressure regions the pressures are higher and in the low pressure regions the pressures are lower, resulting in a significantly increased pressure drag. Also, the underside contains the wheels and wheel-bays. These are impossible to streamline properly, so make a considerable contribution to the drag. Indeed, about 50% of all of the drag comes from the small space underneath. The relatively large frontal area above the height of the front intake accounts for the other 50%.

Have you conducted any wind tunnel tests?

No. All our research has been carried out using computational fluid dynamics (CFD). This work was carried out by the Motor Industries Research Association. In discussion with M.I.R.A, we agreed that it would be difficult to achieve a sufficiently high rolling-road speed and a sufficiently large model scale to represent the flow accurately. In particular, the boundary layer on the car would not be well represented by the wind tunnel tests. This could lead to serious errors, as the ground clearance is not much greater than the boundary layer thickness.

Also, although the car is travelling at less than one half of the speed of sound, compressibility effects are becoming significant. Indeed, at top speed the flow around the wheel/ground contact points is sonic. We can simulate this on the CFD but not in a wind tunnel with rolling road.

The fact that the skin friction drag is significantly greater than the pressure drag shows that we were correct in concentrating on getting the correct representation of viscous phenomena. This adds weight to the decision to use CFD and not a wind tunnel.

Is this the first project to rely totally on CFD?

I cannot honestly say that I have broken new ground by relying totally on CFD, as I do not know about the research programmes of all the others. However, I suspect it is a first as CFD has only recently matured to a degree that makes it relatively trustworthy. Of course, the speed record pioneers used neither CFD nor wind tunnels. Incidentally, on many other shapes, such as Formula 1 cars, I would prefer to trust the wind tunnel instead of the CFD.

The car looks stunning. How much did aesthetics control the design?

Not at all. The design was controlled by function. The important point was to minimise the drag and make the car stable. This controlled all of the principal dimensions. Then Mike Turner did a brilliant job in drawing the final shape. There is an old saying concerning cars and aircraft that “…if it looks right, it is right”. On that basis, I think that we have designed a classic.