From the Arctic Circle to the Nürburgring: a conversation with quattro developers
Back to overviewAudi has been testing its quattro systems under extreme conditions for many years now – with considerable effort, a deft touch and a meticulous approach. We brought together experts from various development areas to talk about the process. Dieter Weidemann (drive) and William Wijts (suspension) explain how they test the purely mechanical center differential on snow and ice. Roland Waschkau from Audi Sport GmbH is the sport differential specialist, while Stefan Lehner (suspension) has a very different responsibility – he develops and tests software for the electronically controlled quattro systems. And his colleague Marc Baur (suspension) specializes in electric torque vectoring.
Mr. Wijts, Mr. Weidemann, your testing work takes you all around Europe – including winter testing on snow and ice. Why is that particularly important for you?
Weidemann: We do tuning work on ice and snow because this allows the subtle differences between the different quattro variants to be clearly distinguished. Routes with a low coefficient of friction are the most important challenge for us. Although this involves operating at the limits of driving dynamics, one of our main aims is to ensure that the car is controllable for everyday drivers, i.e. people who might unintentionally find themselves at the limits on slippery roads in ordinary driving. They are the reason why we develop and test the quattro drive, particularly on snow, in order to guarantee safe and comfortable handling.
What happens on these test drives?
Wijts: Our name for the site is “Kalt 1.” It’s located near the Arctic Circle. To tune the center differential, we typically travel to Sweden with four or six cars, supported by one or two mechanics. We tend to go in February, when the terrain is covered in deep snow and the weather conditions are at their most stable. The private test site has a wide range of different routes, some of them with outward banking corners and other intricacies. There are multiple handling courses on the lake and on land that have been specifically designed to allow us to test different cornering profiles and cornering speeds. The site even has a serpentine road for traction tuning.
Weidemann: We work best at temperatures below 5 °C (41 °F). This makes the conditions broadly constant and allows us to reliably compare the various configurations. At the same time, a reference vehicle is always needed in order to properly judge how the conditions are changing. The best conditions of all are -25 °C (-13 °F) with really hard snow that offers good grip, but it never stays like that for the whole day. The second test run is already different from the first, the snow gets flattened, and the ice gets polished, and so the conditions gradually deteriorate. But we are familiar with this effect and know how to take it into account.
How risky is your work?
Wijts: The worst that can happen on the lake is you fly off the edge and have to be dug out of the snow. When it comes to land handling at the private test site, we know the routes so well by now that we can analyze the intricacies of tuning without difficulty.
At the same time, we take the limits of physics very seriously. Within these limits, we drive as fast as necessary for our work – at 30 km/h (18.6 mph) in some sections, at 160 km/h (99.4 mph) in others.
Weidemann: When we are tuning center differentials, we deactivate as many of the suspension control systems as possible, including ESC, so that we can evaluate the quattro drive in isolation. Our aim is for the basic driving characteristics to be harmonious even without the control systems, so the car does exactly what the driver wants. Incidentally, this strategy once surprised a high-ranking colleague from technical development who had moved to Audi from a direct competitor and was accompanying us in Sweden for the first time. After a test run in an A6 with quattro drive, he told us enthusiastically about how smoothly and harmoniously the control systems worked. But they had all been deactivated.
You are well practiced when it comes to winter testing. How long have you been involved?
Wijts: My first winter test was in 1998. Back then, the requirements in terms of the Torsen differential were far lower than they are today – the priority was that it had to be robust and capable of generating as little tension as possible in tight corners and when maneuvering. Then Dieter came along. He was already very interested in how the cars drive. The suspension and drive development teams have been pushing each other forward ever since.
Weidemann: Our department took on responsibility for the quattro drive and restructured the development process. Together with our colleagues from the suspension team, we gradually developed a shared philosophy. We selected very different basic distributions and locking values for the front and rear axle and examined their functionality, initially on the test bench. We then took the most promising differentials with us to Sweden. The most intensive period of development was in the late 2000s, when we were investigating fundamental concepts for the self-locking center differential using entirely different technologies. It wasn’t unusual for us to set off with 30 different components or more.
How long did winter testing take back then?
Wijts: Ten years ago, it took around two weeks – with no weekends off. And every day followed the same pattern: Driving, letting the mechanics make modifications, driving again… After that, we developed a concept that we still use today. We define the different models that we intend to use in the test drives in advance – an S4* and an SQ7*, for example. This is important because the differentials are used in very different classes of vehicle, after all.
Weidemann: We then take two identical vehicles of each model type with us, both containing the standard self-locking center differential. The car that performs slightly better becomes our first reference car. We install one of the prototype differentials in the other car and see whether it improves the driving characteristics. If it does, that car becomes the new reference, and the next development variant is installed in the first car. This allows us to optimize the driving characteristics iteratively until we have worked through all of the variants.
Do you always agree in your assessment?
Wijts: Dieter and I discuss the performance of the car in detail, like how it behaves during load changes or whether there is too much or too little understeer. In the end, we always agree. Just like an old married couple [laughs].
What are the properties of the ideal self-locking center differential?
Wijts: Every customer should be able to drive an Audi quickly and safely on snow and ice. This means we place great importance on ensuring that the car responds the same way in every situation where possible, that it steers spontaneously and takes corners neutrally. And when the rear does break away during a burst of acceleration, for example, it is important that the movement is not too dramatic so the car remains easy to handle. It also needs to have very good traction, of course.
Weidemann: It goes without saying that Audi has made immense progress with the mechanical quattro system since 1980. Today’s basic distribution of 40 percent to the front axle and 60 percent to the rear axle, combined with the right locking values, means there is an excellent overlap in terms of strong grip and good handling. However, its main strength is and remains the traction it provides on surfaces with a low coefficient of friction.
Waschkau: For customers looking for dynamic driving on dry roads, we offer our sport differential, the quattro sport, for the rear axle. This makes the car agile by actively shifting the drive torque between the rear wheels as required. Race tracks and high coefficients of friction also play an important role in tuning, particularly for the RS models. Unlike the traditional center differential, the sport differential is an electronically controlled system. It consists of two clutches and a superimposed transmission, while the software is specially adjusted and fine-tuned for each vehicle.
Another controlled system is the quattro with ultra technology. What inspired its development?
Weidemann: A survey around ten years ago found that many Audi customers who did not drive a quattro believed the additional consumption to be far higher than it actually is, and had not yet experienced the strengths of the all-wheel drive. The quattro is for more than just icy and snowy conditions – it is already beneficial when turning or moving off on dry asphalt. In other words, there are many situations where customers can briefly enjoy the benefits of the quattro drive, but there are also many driving scenarios where no difference between the quattro and front-wheel drive are visible. Consequently, we decided to develop an entirely new system with a switchable rear axle that combines the high efficiency of a front-wheel drive with all of the benefits of permanent all-wheel drive.
Wijts: Our objective was to make it so there was no noticeable difference compared with the permanent quattro. We created 15 computational models describing the differences in areas like the feel of the steering. In other words, we were able to identify exactly when the steering on a front-wheel drive car feels lighter than on a quattro as the load increases, or the point at which it begins to tramline. The rear-axle drive is temporarily engaged in all of these situations. This always happens preemptively, so the all-wheel drive is already active by the time it is needed.
How difficult was it to tune and test the quattro with ultra technology?
Weidemann: We had the advantage that the legal safety requirements for the test were low – even if the system had failed completely, the car would have continued with the normal front-wheel drive. This meant we could begin road testing extremely early in development, which was very good news in terms of maturity. As we often do, we drove a 120-kilometer (74.6 mi) loop around Ingolstadt on public roads. This allowed us to continuously double-check the progress of the software versions, which is very useful and important. It meant we were able to work through all the different driving situations, from southern Italy to the Arctic Circle, like a customer would in real life and adjust our operating strategy to ensure that the driving characteristics were the same as for the permanent quattro.
Speaking of the operating strategy and software: Mr. Baur and Mr. Lehner, this is where you come in.
Lehner: Our department looks after the mechatronic and fully electric quattro systems. We develop software functions such as clutch actuation under changing conditions and demands. We are involved in every controlled system. In the mechanical world, this means the quattro drive for the R8, for vehicles with longitudinally mounted engines and the quattro ultra, and the quattro for cars based on the MQB platform. They use a hydraulic multi-plate clutch where efficiency of control is also extremely important. Under normal, smooth driving conditions, the drive is largely provided via the axle that is driven directly. The drive torque is smoothly distributed to all four wheels as necessary.
Baur: And then we have the e-tron models with the electric all-wheel drive that no longer contains any physical all-wheel components – nothing mechanical or hydraulic at all. Each axle has one or even two motors which combine to produce the electric quattro.
How does your testing process work? You must have a lot of digital screws and levers at your disposal…
Lehner: Yes, there are thousands of potential parameters for the software, depending on how finely they are tuned and combined. We gradually narrow them down to a reasonable number, ending up with around 100 to 200 property parameters. We use them to tune the traction and the handling individually for each model derivative, for example. Testing takes place the traditional way during a test drive: One of the team sits in the passenger seat with a laptop and repeatedly loads different software versions into the system. We take the same corners, lap after lap, until we find the ideal end result.
Do you do more testing on the road or on the test track?
Lehner: When a project begins, there are various reasons why we are not allowed to take cars on the road yet, so we have no choice but to use the test site. What that does mean is conditions can be reliably replicated and we can take things to the limit when driving. Sometimes we only drive for an hour, then spend five hours at a desk analyzing the measurements. Then, in the second phase of development when things are heading toward series production, we spend a lot of time in the car simulating various customer situations. That involves winter roads in Sweden, Alpine roads with cracked and crumbling surfaces, and also the famous Nordschleife loop at the Nürburgring – because it packs a lot of specific challenges into a small space, like changing surfaces and banked corners. In direct contrast, we also do test drives in parking garages, because concrete has a very low coefficient of friction and the turns are very tight.
The more complex the all-wheel system, the more extensive its development and testing. Is that right, Mr. Baur?
Baur: Yes, increased complexity means more functions and parameters. In the e-tron, the clutch between the front and rear axle is fully emulated by the software, resulting in tens of thousands of lines of code and countless parameters. The big challenge is tuning the systems involved – particularly in the S* model with its three electric motors. In this particular case, we worked in close cooperation with our colleagues who developed the drive control unit and the control units for the power electronics and the brake control system. We have to develop a large number of functions while always taking into account the various repercussions they can have on other systems. We have the very same aim as our colleagues: to ensure that the car demonstrates predictable and reproducible behavior in every situation.
How does your testing work?
Baur: When we go to the test site, we often have a team of 20 or 30 people working in close cooperation, from various departments of drive and suspension development. We all sit at the same table and, when it comes to testing, in the same car – figuratively speaking, at least.
Weidemann: What I find endlessly intriguing about the quattro is its diversity. I am responsible for all of the components that distribute torque between the axles in the models with longitudinally mounted front engines – and that is just one part of the big picture. However the quattro drive is realized from a technical perspective, it is always synonymous with a high degree of safety, strong traction, and dynamic handling. These are our great strengths. Audi has gained experience in these areas over many generations of quattro models. Our objective is to work together to leverage all of the available potential so that we can give our customers the greatest possible benefit and ensure that they are fascinated by Audi.