Analysing Road Car Drag

This article was first published in 2008.

Reductions in the amount of energy consumed by vehicles will require that they become more aerodynamically slippery. It doesn’t matter if the motive power is pure electric, an internal combustion engine (either petrol or diesel) or a hybrid power source: aerodynamic drag reductions are vital if efficiency gains are to be made.

But how can that these improvements take place? Or to put it another way, what actually causes cars to experience drag and how can these factors be reduced?

Drag Components

Aerodynamic drag on road vehicles can be split into four different types. These are:

• Separation pressure drag

• Viscous drag

• Induced drag

• Interference drag

Don’t panic – we’ll take a look at them in turn and examine how they apply to real vehicles.

1. Separation Pressure Drag

In conventional road vehicles, separation pressure drag is the biggy – accounting for between 50 and 90 percent of the total aero drag. Separation pressure drag is also amongst the easiest of the drag components to envisage.

Airflow over cars can be basically split into two types: attached and non-attached. At the front of a modern car, the flow across the surface of the bonnet will almost always be attached. This can be seen if small lengths of wool are stuck on the body surface – the tufts all line up one after another.

However the real action is at the back. On a sedan, the airflow will generally be attached at the trailing edge of the roof - but then it has to make the transition down onto the rear glass. If this successfully occurs, and the flow remains attached right down the back window and onto the boot, the car's doing very well. Why? Because when the air finally leaves the trailing edge of the boot, the cross-sectional area of disturbed air being pulled along behind (called the "wake") will be small.

You can see that the wakes behind a hatchback or wagon (where the airflow separates at the end of the roof) will be much larger than for a well designed 3-box sedan.

The cross-sectional area of the wake is particularly important because the wake comprises a low pressure area – in other words, one that is literally pulling the car backwards. The bigger the wake, the greater the separation pressure drag.

Reducing the size of the wake is so important that most cars are boat-tailed, ie the sides of the car get closer together over the last metre or so of the body. If the flow remains attached down the sides of the car, this reduces the area of the wake.

A small wake is the easiest way of recognising which everyday cars are slippery and which are not.

In addition to the size of the wake, pressure drag also makes itself felt in trailing vortices. Looking a bit like whirling ribbons being pulled along by a running child, these spinning disturbances create drag.

They’re shed from rear edges, especially the C-pillars.

2. Viscous Drag

To understand Viscous Drag, two ideas need to be embraced.

The first is that air is a fluid, and like other fluids, has viscosity. Viscosity is the ease with which particles of the fluid move past one another. Honey is highly viscous – the particles don’t want to slide past one another. Engine oil is less viscous, and air is much less viscous again. However, air still has a degree of viscosity.

So why is the viscosity of air important? Because when the air passes over a car’s body, it is being constantly sheared.

Let’s take a look at why this happens.

If you park a car on a dirt road, the car will soon be covered with a fine layer of dust. When you drive off, watch the surface of the bonnet. Despite it being covered in dust, the dust particles will largely remain in place. But why doesn’t the dust get blown away by the airflow over the car? The answer is that the air in contact with the body isn’t moving. That is, the tiny molecules of air closest to the body are not sliding over the paintwork.

As you get further away from the car body surface, the airspeed gets greater and greater, until when you’re far enough away from the body, the airspeed is the same as the car’s forward speed.

This diagram shows the effect. Here the length of the arrows is proportional to the speed of air. You can see the red arrow, a little above the surface of the car, is much shorter than the green arrow, further from the car. Therefore, because the layers of air are travelling at different speeds, they are sliding over one another – or to put it another way, particles of the fluid are constantly moving past one another. This was our definition of viscosity - the ease with which particles of the fluid move past one another.

Now you can see why the viscosity of the air has an impact.

Incidentally, this area of low-speed flow attached to the surface of the car is called the boundary layer.

Thinking about the two factors that we’ve discussed so far, you can see that the longer that flow stays attached to the body, the smaller will be the likely size of the wake, and so the lower the separation pressure drag. So that’s good! However, the longer the flow stays attached to the body, the greater the amount of viscous shearing that is occurring in the boundary layer around the body – and that’s bad. However, overall, it is still much better to have as much attached flow as possible.

The boundary layer has another impact on drag. As you move towards the rear of the car – even one with attached flow from nose to tail – the boundary layer gets thicker. In other words, the number of air molecules getting dragged along with the car become more numerous.

So? The reason it’s relevant relates to pressure on the bodywork. We’ve already said that the wake behind the car is a region of low pressure, pulling the car backwards. But what about the fact that the car is running into air molecules at the front? Doesn’t that cause a pressure build-up? It sure does.

Here the red area shows the region of highest air pressure – at the stagnation point, where the air is literally stopped dead. Clearly, if we could apply this same pressure at the back of the car, the two pressures would cancel each other, reducing drag.

But, even on a car that is so streamlined it has attached flow from nose to tail, the thicker the boundary layer, the harder it is to get the air to apply this rearwards pressure. If the boundary layer didn’t exist, it would be possible to get the air to apply the same pressure on the back as on the front!

3. Induced Drag

Induced drag is a by-product of lift or downforce.

Lift or downforce occur only because there is a pressure differential. For example, a higher pressure on the underside of a car than the top surfaces will cause lift. A wing with a higher pressure on the upper surface than the lower surface will develop downforce.

Trouble is, these pressure differences always result in vortices of airflow between the two surfaces. The energy needed to create these vortices results in increased drag.

4. Interference Drag

Interference drag is that that comes about because vehicles are not single bodies without appendages. And unfortunately, when different shapes are combined to form a practical vehicle, the total drag is always greater than the sum of its parts.

For example, the drag of a wheel (even when covered in a fairing) and that of the car body will be greater when they are joined that when they are measured individually.

In effect , the flow over one surface interferes with the flow over another surface.

Relative Effects

Taking the above factors into account, it is an inescapable conclusion that all road vehicles will have aero drag. As described in the breakout box above, reducing separation pressure drag by keeping the flow attached will result in a greater viscous drag. Interference drag will always exist because it is impossible to produce a car shape that doesn’t have appendages sticking out here and there.

However, by the same token, it is possible to make vast aerodynamic drag improvements over current, conventional cars.

The leaders in the vehicular field are solar race cars, machines that have to achieve low drag figures if they’re to be competitive on such little total motor power. A conventional sedan (even a very good one like the Toyota Prius) might have a drag coefficient (Cd) of 0.26, but solar race cars have Cds of 0.1 or even lower.

Clearly the solar racing cars are an impractical shape for normal use, but that’s primarily because we expect current cars to be able to do everything, rather than specialise. (Consider what shape an electric, single person dedicated commuter car could take.)

So in terms of the above drag contributors, how do solar race cars compare with traditional passenger cars?

	Separation Pressure Drag	Viscous & Boundary layer Drag	Induced Drag	Interference Drag
Production Car	Large percentage	Small percentage	Medium – large percentage	Small – Large percentage
Solar Race Car	Zero percentage	Large percentage	Zero percentage	Small – Large percentage

Goro Tamai, in his book The Leading Edge (from which the above table is taken), suggests that the priority order for reducing drag should be as follows:

1. Reduce separation pressure drag (ie size of wake)

2. Reduce Induced Drag (ie downforce/lift drag – normally, it would be reducing lift)

3. Reduce Interference Drag (eg sealing holes, contouring junctions, etc)

4. Reduce Viscous Drag

Conclusion

The major reduction in aero drag of road car designs that occurred between the 1960s and the 1990s was through reduced separation pressure drag. There is still plenty of aerodynamic advantage to be gained by reducing this still further, but once that limit has been reached, attention will need to turn to reducing induced, interference and (finally) viscous drags.