Engineering specialist AVL RACETECH has several different divisions with state-of-the-art commercially available technology .
One division takes care of the powertrain and the vehicle engineering areas, including electrical or internal combustion engines, hybrids and various other variants.
The other major division in AVL is the instrumentation and test systems section. This includes a vast array of test beds that are used widely among top class motorsport teams and the road car industry.
Therefore, AVL is able to provide an excellent testing environment for its customers. One of those testing environments is the multi-functional test bed, known as TB503.
This allows either single components or the entire racecar to be run – no matter whether it has an internal combustion engine, hybrid drive, or electric drive. The high-tech facility is designed for extreme acceleration and deceleration values, meaning it covers all current and future motorsport requirements.
The interactions of the individual vehicle components are absolutely identical to use on any given racetrack. The option to directly connect the testbed to the AVL driving simulator makes the testbed even more realistic.
Key features of the bed are a complete vehicle test capability with wheel dynos with maximum outputs of up to 1200 kW per wheel, powerful cooling fans and battery emulators, providing a highly realistic environment for racetrack-specific setups and optimizations, and optional connection to AVL driving simulator with their AVL vehicle simulation software (VSM™ RACE).
As an example, for the adaptability of using the 503 testbed, it is not only a pure gearbox test or a pure engine test bed but AVL can also put in a full vehicle.
“That means, essentially, we roll in the real car, we remove the four wheels and we connect every corner to a dyno, and by then we can actually run the car on the dyno, which is quite close to track testing because we can run the real ECU, the real motor, the real battery and so on,” says AVL’s Department Leader Race Engineering, Michael Peinsitt.
“Typically, in Formula E, we run with a battery emulator but we can run with the real battery as well.
“So, you have no test bed specific solutions that you need to create and use, because you’re actually running with the real car and that provides a very realistic testing environment which is very interesting given the testing limitations that exist in various racing categories.”
What the 503 Test Bed Looks Like
The 503 has four wheel dynos that run at a power of 800 kilowatts in nominal power. It can go up to 1200 kilowatts which is needed for high performance race categories like Formula 1 when using the brakes of the vehicles on the test bed.
“The car is suspended on its own suspension, but we are not moving it vertically. So, there is no physical bumps where you can see the car shaking, etc.
“The bumps are there, but they are inside the vehicle dynamic simulation that’s running in the background. We use our own AVL vehicle dynamics simulation, it’s called VSM™ RACE.
“This has a track model inside which has all the track radiuses and also a representation of the curbs and all the bumps that exist on the race track.
“That means the car will be exercised in the simulation as if it was running on the real tracks in terms of the bumps and wheelspin when you go over a bump or wheel locking scenarios when braking too hard over the bumps, or if you’re running over a kerb, all that is simulated very realistically, and provides realistic inputs for the car on the dyno.
How 503 Deals with Oscillations
Oscillations are a big deal in powertrains and one of the key challenges of engineers to try and find solutions to maximise efficiency and pace.
The 503 will let the car freely decide how it oscillates and how it behaves in order to guarantee the same behaviours as on the race track.
According to Peinsitt to achieve that “we have several major ingredients: one is the dynos, which need to have very low inertia, in the magnitude of a real wheel, which allows those oscillations to occur, and another one is the vehicle dynamics simulation which reacts to every change that you impose on the car.
For example, if AVL were to increase the grip on the tyres by say three or four percent, then the car would drive differently on the dyno because it has virtually increased the grip.
“So, it will achieve different top speeds, different corner speeds,” adds Peinsitt.
In the same fashion, if AVL were to change just the downforce of the car on the rig, it would drive very differently. That means immediately the effects of the changes are seen virtually in the vehicle dynamic simulation.
“We have had feedback from customers, Formula E customers, they said they saw those oscillation problems at the track, tried to unsuccessfully replicate it on a test bed, and only when coming to Test Bed 503 it was the first time that they could actually replicate the same observations that they saw on the track.
“That’s simply because of those ingredients, the vehicle dynamic simulation and the low inertia dynos that we’re using.”
Why 503 is a Winner with Customers
“Actually, I think [the 503 Test bed] is one of the very few who can achieve this range of testing applications,” says Peinsitt.
Not only is AVL selling test beds to customers but it also selling the vehicle dynamic simulation to customers, “so it is out there on the market, it is available for customers to purchase,” says Peinsitt.
“The point is it’s not a cheap thing to purchase a test bed of this magnitude with this quality, it requires a substantial amount of money, and many customers say, ‘well, we cannot bring up that amount of money to purchase our own test bed, so we’d rather come to AVL for a couple of weeks per year and do our stuff there.’
Because then it’s AVL RACETECH’s responsibility to make sure the maintenance is done, the dyno is in good shape, the vehicle dynamic simulation works, it’s all on AVL´s shoulders, and AVL take that responsibility very seriously.
“The big advantage is on the one hand, we are very repeatable, we are in a very controlled environment here. On the other hand, due to the vehicle dynamic simulation, we are able to get changes to the lap time that we’re actually simulating on the test bed. If, for example, we change the grip level of the tyre, if we change the aero characteristics of the car, if we change the MGU power that we have on the car, all that will translate into a difference in how the car drives and you can see that on the data and on the lap time.”
Accuracy in Simulation
The 503 is truly the Next Best thing to being on track with engineers wowed by the scope of what is possible in a controlled environment.
Test beds have been common practice among engineers for decades, particularly engine and vehicle dynamic (so-called 7-poster rigs). But the 503 took all of those attributes and brought them to one product.
“The next step, really, to get even more realism, you need to be at the real track. But here we can load the whole car, the whole powertrain with the exact loads as it would do on the on the real circuit.
“There is the repeatability and that’s something quite unique, and we can connect what you see here with the driving simulator.”
That means a real driver will be in the DiL simulator, “driving this car on the test bed!”
That’s obviously very important because in the past, even though the simulation says a specific solution is the optimum one, if the driver doesn’t like it for whatever reason because his style is different, or his preference for oversteer or understeer is different, he will not go any quicker.
“That means you need to find a solution that also in the theory is quick, but the driver buys into it and accepts it and can handle what you throw at him,” remarks Peinsitt.
“That’s another step to combine the driving simulator with this full vehicle test bed. You can’t get any closer to running on the real racetrack.”
