Session: 14-04-01: Reliability and Safety in Transportation Systems I

Paper Number: 147100

147100 - Assessing the Resilience of 77 Ghz Radar Sensors in Autonomous Vehicles: A Multiphysics Simulation Approach for Low-Speed Collisions Leading to Significant Radome Deformations 

The transportation landscape is undergoing a significant transformation with the emergence of autonomous vehicles (AVs). These self-driving cars hold immense promise for revolutionizing transportation by offering enhanced safety, efficiency, and accessibility. However, the widespread adoption of AVs hinges on their ability to navigate complex environments reliably and safely. This necessitates a robust suite of sensors that can accurately perceive the surroundings and provide real-time data for decision-making algorithms. Among these sensors, 77 GHz Radar sensors play a critical role by providing long-range object detection and tracking capabilities in various weather conditions, making it a vital component of the AV sensory system.

 

One of the main advantages of radar over other sensor technologies is that the radar sensor can be placed behind a low-profile radome or even behind the vehicle bumper or facia. While this allows car designers to retain the aesthetics of the car, the radomes and bumpers should be designed to not significantly distort the electromagnetic waves coming from the sensor antennas. Furthermore, any deformation to the radomes can significantly compromise the functionality of a sensor previously calibrated for a non-deformed radome. Therefore, there is a need to quantify the impact of crash-induced deformation on the operation of the radar sensor. Physical testing through controlled crash scenarios is a traditional approach for evaluating sensor performance in collisions. However, this method has limitations, particularly during the initial design phases of AVs. The high cost and time associated with physical testing can hinder the exploration of numerous design iterations. Additionally, replicating the vast array of potential real-world collision scenarios becomes impractical.

To address these limitations, high-fidelity multiphysics simulation frameworks offer a powerful alternative for assessing sensor resilience. This research leverages such a framework, employing an explicit finite element solver to simulate the structural deformations experienced by the AV bumpers and radomes during a collision. The framework is further enhanced by incorporating a 3D full-wave electromagnetics field solver to predict changes in the radiation characteristics of the radar antenna due to radome deformation. By incorporating material properties, sensor characteristics, and real-world impact data, these simulations can accurately predict the deformations that sensors might experience in various low-speed collisions and their effects on radar detection.

This research focuses on investigating the impact of low-speed collisions on the performance of 77 GHz radar sensors in autonomous vehicles. Low-speed collisions, though seemingly less severe than high-impact crashes, can still cause significant deformations in the radomes and bumpers that house the radar sensor. These deformations can potentially compromise the sensor's functionality, impacting the vehicle's ability to perceive its surroundings accurately.

The research will analyze the performance of 77 GHz radar sensors positioned at different zones of an autonomous vehicle after experiencing low-speed collisions that result in substantial bumper and radome deformations. Through the multiphysics simulations, we aim to assess the extent to which the sensor's performance, such as detection range and field of view, is affected by the deformations. This information is crucial for understanding the sensor's effectiveness in real-world situations where replacing or repairing a damaged sensor might not be immediately feasible. The simulations will analyze the sensor's functionality under deformed conditions, informing decisions on whether the sensor can provide sufficient data for continued safe operation until repairs can be made.

By achieving these objectives, this research using multiphysics simulations will contribute significantly to enhancing the safety and reliability of autonomous vehicles. The comprehensive understanding of sensor resilience under deformed conditions will inform design improvements for both the sensors themselves and their placement on the vehicle. Additionally, the ability to simulate various collision scenarios will expedite the development process by allowing for the virtual testing of numerous design iterations. Ultimately, this research has the potential to pave the way for the development of more robust and reliable AVs, fostering a future of safer and more efficient transportation.

Presenting Author: Satish Kumar Meenakshisundaram Ansys

Presenting Author Biography: Satish holds a Master's degree in Mechanical Engineering from Michigan Technological University. Satish's expertise lies in applying advanced computational methods to solve complex engineering problems. Satish is highly skilled in Finite Element Analysis (FEA), Structural mechanics, and Multiphysics modeling. Satish is a strong advocate for democratizing simulation, making these powerful tools accessible to all engineers, not just specialists. He believes this will empower engineers across disciplines to design and develop innovative products more efficiently. His background also includes experience in automating processes, further enhancing efficiency.

Authors:

Satish Kumar Meenakshisundaram Ansys
Ushe Chipengo Ansys
Amogh Shejwal Ansys