Keeping Autonomous Vehicles Moving
The old saying goes an army marches on its stomach. In a parallel scenario, a fleet of autonomous vehicles operate on their power source. The ability to extend or efficiently use this energy is crucial to maintain critical monitoring, communication, and protection equipment. Research teams at University of Michigan are focusing on ways to extend the operating life of their power supply.
Anna G. Stefanopoulou, professor of Mechanical Engineering and the William Clay Ford Professor of Manufacturing at UMich and the director of the UMich Energy Institute, and her team are working to make the most accurate estimate of the state of batteries in a fleet of autonomous vehicles. Their goal is to ensure the vehicles can operate efficiently even at environmental extremes.
Lithium-ion batteries degrade when its components, like the electrolyte solution, separator, or current collectors, are compromised. These batteries also ‘age’ prematurely when stored for long periods at higher temperatures. The ‘aging’ process creates internal particle cracking that can accelerate degradation.
Stefanopoulou and her team turned this challenge into an opportunity by using battery ‘aging’ process as a health indicator.
“Think of it as ‘a health certificate’ for the battery,” said Stefanopoulou. “Aged batteries have a waist line that is expanding like many of us aging urban dwellers.”
After some preliminary work, they found that the degree of cell swelling was correlated to cell degradation. With this discovery, they were off to the races. The team has designed multiple experiments to evaluate the ‘aging’ process. They hope to characterize the temperature, depth of discharge, and other chemical properties of the battery to develop ways to predict the energy left within.
According to Stefanopoulou, these results have a large range of applications from autonomous robots to scooters to large buses to even power grid storage.
UMich researchers are also optimizing how conventional and autonomous ground vehicles burn fuel. André L. Boehman, professor of Mechanical Engineering, and a team of researchers have been developing methods for representing fuels in simplified ways to make computer predictions of combustion within the engine more accurate.
“When designing the engine of the future, it is important to use computer technology to simulate an optimal design,” said Boehman. “In particular, we need to describe the chemistry of combustion in a manner that is concise mathematically and tractable.”
This is no easy task. The complexity of the problem parallels the complexity of the compounds that comprise the fuel. For instance, it takes more than 8,000 steps to characterize all of the individual chemical reactions that occur as one molecule of jet fuel burns, while also simulating the fluid flow and thermal phenomena within the engine accurately. According to Boehman, the fastest super-computers on the planet would be required to crunch this problem.
To avoid this complication, this team of researchers team has simplified how to represent the fuel chemistry. By winnowing down the fuel to a simplified ‘surrogate’ mixture and then describing the combustion chemistry of that simplified fuel mixture, the team aims to describe the ignition behavior to quantify fuel reactivity and develop a validated predictive model of the combustion process. In addition, they are developing and validating methods to reduce the kinetics as the fuel burns.
“We are taking on the project to meet a military need, but there are larger civilian benefits from this work,” said Boehman. “If we can develop pathways from alternate and sustainable feedstocks, we could achieve fossil carbon reductions and make dramatic change in our carbon footprint.”
###
Stefanopoulou was joined by Jason Siegel and Peyman Mohtat at UMich, Aaron Knobloch at GE Global Research, and Yi Ding and Matt Castanier at the U.S. Army Ground Vehicle System Center on the project, titled “Advanced Battery Diagnostics: Decode the information in Electrode Swelling.”
Boehman was joined by Margaret Wooldridge, Angela Violi, Doohyun Kim, Shuqi Cheng, and Dongil Kang at UMich, Marcis Jansons of Wayne State University, Peter Schihl at the U.S. Army Ground Vehicle System Center, J. Timothy Edwards at the Air Force Research Laboratory, and James Anderson with Ford Motor Company on the project, titled “Ignition Studies for Kinetic Mechanism Development and Validation.”
Stacy W. Kish