Testing Drivetrain Parts with Advanced Technology
Drivetrains are essential to get vehicles moving. They take power from the engine and transmit it to the wheels. If a component is weak, it can break just like a chain, causing vehicle problems or even accidents. To keep machines and vehicles running smoothly and safely, it is essential that drivelines are subjected to extensive testing. This involves simulating all relevant operational conditions and factors. This test data can then also contribute to early detection of abnormalities and problems in vehicles and machines. With this data, an AI model can be trained to detect deviations and develop a smart maintenance regime, resulting in efficiency gains and higher quality.
With our Modular Multi-Load Drivetrain Test Cell, we have an advanced, safe test bench that allows powertrain components to be dynamically tested. Algorithms are then developed for smart defect detection. In this article, we take a closer look at how this works.
Vehicle drivetrain example - part of our Range Rover Evoque
Component testing: the lab emulates the road
We have an advanced test bench that allows us to test various drivetrain components - both for (hybrid) EVs. This setup includes components like the differential unit (which helps your car turn), driveshaft and wheel axles. It is a Hardware-in-the-Loop setup, software is used to control everything safely. Realistic driving situations are simulated: going up a hill, for example, is simulated by adjusting how much torque and speed the test motors use.
To test components also in non-standard situations, anomalies can be mechanically introduced or emulated. Anomalies we have tested include misaligning components or letting the clutch slip. This allows us to see how the system deals with these. This helps us figure out how to detect these problems early in practical scenarios.
The tests provide a wealth of data. Regular data such as engine torque and engine speed can be recorded anyway. In addition, an additional customised data acquisition system can be added to log data from specific sensors, such as subframe vibration or bearing housing loads.
Example of a damaged component. Image Source
Digital Twin
From this test setup, we also developed a Digital Twin via a Simscape model. This allows us to test even more scenarios safely via simulation. We validated the reliability of this Digital Twin by comparing the test results with the digital model with those of the test setup. The simulation results correlate with the experimental data.
The Digital Twin is particularly valuable for testing severe anomalies that would be fatal to the tested components. For safety reasons, our test set-up will shut down, but from the generated simulation data we can learn a lot from this about risk prevention.
Moreover, we manage to generate a lot of data that can be used for AI training. We can make our detection tools smarter without risking actual parts.
From vehicles to industrial machines: improving safety and reliability
This kind of testing doesn't just help cars; it can be used for all sorts of machines. The technology allows for more improvements in how machines are controlled and how they perform, making everything from driving to industrial machinery safer and more reliable. We can therefore offer the possibility to detect, mitigate and analyze components failure modes in vehicles and machines for our customers.