Model behaviour
Racecar Engineering|Design of a Racecar
Understanding the benefits and limitations of scale model wind tunnel programmes
JAHEE CAMPBELL-BRENNAN

Wind tunnel testing has a fascinating history, beginning in the aerospace industry. As testing methods became more sophisticated and laboratory testing evolved, a need was identified to provide a method of validating design concepts. The Wright brothers were the first to achieve controlled flight as they built a tunnel to calibrate the characteristics of wing profiles and showed that the accepted characteristics were wrong.

However, few tunnels existed that would be capable of managing a 1:1 scale prototype. Aerospace engineers quickly figured out there were many advantages to preparing scale models. Namely, it enabled testing to take place at all, but furthermore cost, materials use and build times were slashed when compared to full-size prototypes.

Cars by their very nature operate close to, and in contact with, the ground, so automotive wind tunnels differ substantially in design to aerospace ones. A moving ground plane and effective boundary layer control are critical features but, in the early days of automotive wind tunnel testing, these weren’t available to full-size models. The concept of scale models within automotive and motorsport testing was therefore born out of similar necessity to aerospace.

A modern automotive wind tunnel’s principal function is to accurately recreate the motion of a car driving on a road surface. The key difference experimentally, however, is that you are moving the air over the vehicle, rather than moving the vehicle through the air. At a minimum this requires the presence of a relative air velocity over the car body, and a moving ground plane to simulate the asphalt passing beneath.

Matched relationship

The relationship between tunnel and model is a valuable point of learning as they must be matched in order to create a suitable test environment. So, from an engineering point of view, the particular tunnel facilities available for a project tend to decide the maximum scale of model used. Let’s explore what that means in practice.

When testing in any wind tunnel you have a fixed cross-sectional area. Naturally, the presence of a test model introduces a flow constriction into the tunnel. The presence of the flow constriction can manifest in two ways – either as a solid blockage, attributed to the frontal area of the vehicle, or wake blockage, which occurs when the dispersion of turbulent wake from the body is impinged by the roof and side walls of the tunnel. The design of the tunnel does have influence into how significant these effects are, but they are inescapable.

Solid blockage is problematic as it causes local increases in velocity around the vehicle body. Due to the principles of continuity (Bernoulli), flow travelling through a constriction must increase in velocity to preserve mass flow. The static pressure around the car body therefore lowers and the measured forces around the body become skewed. The result are therefore invalid.

Wake blockage is a little different as it’s not directly a result of the model’s area but, due to its geometry, will generate a turbulent wake in the tunnel. Interfering with the wake flow affects the pressure distribution across the whole body. The rear wing and diffuser can be particularly sensitive to this as pressure distribution and mass flow are impacted, which means the point of flow separation can be altered.

‘Wake blockage can be an issue with larger test models. When the wake is very large, the flow can affect the breather, where tunnel pressure equilibrium is maintained. If the wake is too big it can also affect your pressure distribution across the test section,’ says René Hilhorst, chief of aerodynamics at Toyota Gazoo Racing. ‘Depending on the design and whether the splitter, rear wing or diffuser are strong in generating downforce, it can affect those to a greater degree, but it really does affect the whole car.’

The section, or blockage ratio (tunnel cross section vs test piece frontal area) is usually maintained at below 10 per cent. Although a blockage as low as is practically possible is desired, it’s always necessary to incorporate blockage corrections into any measured lift and drag coefficients to adjust for local velocity and pressure gradient influences. The actual correction arithmetic can be fairly complex and proprietary to a particular tunnel, but it’s important to remember this is not a problem specific to scale model testing.

As we’ll discover a little further into the article, the scale of the model ultimately dictates the air speed around it, and for this reason it’s usually a case of ‘the bigger the better’ when referring to scale.

‘The tunnel at Toyota Gazoo Racing [TGR] was built for F1,’ adds Hilhorst. ‘The main target was to have exceptional conditions for scale model testing up to 60 per cent, and capability to do very good full-scale testing, but we don’t do this so often.’

Levels of complexity

So now we understand some of the elementary concerns of wind tunnel testing, let’s take a look into the detail of the scale model and what’s required from it to produce meaningful data for the aerodynamicist.

The model itself is a pretty elaborate component as it must be intricate to be representative. In all but the simplest test cases, the model must incorporate all the degrees of freedom [DoF] a real car has – wheel rotation, steering, body and suspension movement (with accurate kinematics), ride heights, chassis pitch, roll and so on. And that doesn’t even touch on secondary effects such as aeroelasticity and tyre squish due to aero load. It can quickly become an extremely complex item.

Of course, the actual racecar is the perfect instrument in this sense, so how do you incorporate these features of a real car into a model? And more importantly, what determines the level of complexity required of a model for a particular project?

‘There are clear differences in model complexity as you get to more high profile championship levels, and therefore bigger budgets. The more basic models will not have any kind of ride height actuation or steering capability, and will likely have solid tyres,’ comments Dominic Harlow, a motorsport engineer director of engineering consultancy, Dominic Harlow Consulting. ‘When you get to Formula 1 territory, however, you’ve got pneumatic tyres, full DoF of suspension movement and ride height actuation, right down to the pre-loading of tyres to represent the effect of aero load.

‘Prices then can go from £50k [approx. $63,425 / €55,425] for something like a 30-40 per cent [scale] F3-style model up to around £500k [approx. $634,250 / €554,250] for an F1 model.’

Hilhorst adds: ‘When we were competing with Audi and Porsche in the WEC, the level of detail required was close to F1 level. We had to go into very refined models with lots of data acquisition and sensing to find the maximum information possible. When you are doing development for a project with less intense competition, you might focus more on styling, where the model detail might be less important, so it’s chosen accordingly.’

Added detail

Continue reading your story on the app

Continue reading your story in the magazine

MORE STORIES FROM RACECAR ENGINEERINGView All

The art of noise (reduction)

As racecars rely ever more heavily on electrical devices, shielding them from interference becomes critical. Racecar asks if graphene-based coatings are the answer

10+ mins read
Racecar Engineering
July 2021

Storm in a teacup

The introduction of Hypercar has meant the FIA and ACO have compromised every other category in their portfolio and led to total confusion for all on track

10+ mins read
Racecar Engineering
July 2021

Fluids and solids

On Hydrolastic and live leaf spring suspensions

5 mins read
Racecar Engineering
July 2021

FIA seat upgrade

The FIA has delivered an update to the 8855-1999 seat that is used as standard in closed-cockpit racecars. The update is designed to significantly improve the safety of the device.

1 min read
Racecar Engineering
July 2021

Stop right there

AP Racing started out as a skunkworks project and grew into a globally recognised leader in brake technology

8 mins read
Racecar Engineering
July 2021

Mad 4 it

BMW says its M4 GT3 car is a step change improvement over the outgoing M6, and will be proven in competition this year before customers take over for 2022

10+ mins read
Racecar Engineering
July 2021

New old Stock

In early May, NASCAR launched its Gen-7 car online. Known as the Next Gen, it is looking to recreate the series’ production car values

10+ mins read
Racecar Engineering
July 2021

Two steps beyond

Michelin and Goodyear have sliced up the FIA World Endurance Championship and become sole tyre suppliers to individual classes within it, but with very different targets

10+ mins read
Racecar Engineering
July 2021

Channel hopping

A window into the possibilities available using advanced simulation channels

5 mins read
Racecar Engineering
July 2021

H uncovered

Toyota entered a hydrogen-powered racecar in the Super Taiku 24-hour race at Fuji as it seeks to prove the long-term future of the ICE

5 mins read
Racecar Engineering
July 2021