Electricity keeps modern systems moving, but it can also wear them down. A review from Southwest Jiaotong University showed how friction and electrical current combine to damage critical components and examined emerging tools to predict failures before they happen.
The study was published in Friction on May 11.
Current-carrying friction pairs are surfaces that slide or roll against each other while conducting electricity. (Think, for example, of the contact between trains and overhead power lines in electric railways.) These components are used in various applications, including rail transportation, power equipment or aerospace devices, given their high efficiency and reliability in energy transmission compared to traditional friction systems. However, the same conditions that make them effective can also accelerate damage.
“The problem is that electricity changes the way these surfaces wear: heat, electric arcs, mechanical forces, speed, load and the surrounding environment can interact, making damage much more complex than ordinary friction and wear,” said Guoqiang Gao, professor at Southwest Jiaotong University and lead author of the study.
Zeroing in on the mechanisms behind this problem, the researchers reviewed recent advances in current-carrying friction and wear while systematically evaluating how operating conditions shape degradation.
“The most important message is that current-carrying friction is not just mechanical wear plus electricity,” Gao said. “It is a coupled process in which mechanical contact, electrical current, heat, arc erosion and environmental conditions interact. If we want safer and longer-lasting electrified systems, we need to understand these interactions together and build prediction models that can work under real operating conditions.”
Studying three major categories of predictive models used to estimate friction and wear behavior, the researchers noted that mechanistic models often work only in limited settings because they depend on many specific conditions. Numerical simulation models can capture complex interactions but require intensive computation and careful parameter selection. Meanwhile, artificial intelligence approaches, including machine learning and deep learning, can rapidly predict wear behavior but depend heavily on high-quality data and often struggle to explain why failures happen.
The authors also pointed to a major limitation in the field: many laboratory experiments still fail to reflect the harsh conditions experienced by real systems, which often operate at high speeds, under heavy loads, and with strong electrical currents. They therefore argued that progress would require more realistic testing platforms, improved multi-physics models and stronger AI systems informed by both experimental and simulated data.
The review provides a comprehensive picture of the current understanding of wear in electrified moving contacts and outlines future directions for research. The findings could contribute to the development of more reliable and durable technologies in sectors ranging from transportation and aerospace to energy systems and advanced manufacturing.
“Our ultimate goal is to predict wear accurately before serious damage occurs, so engineers can design better materials, optimize operating conditions and plan maintenance more intelligently,” Gao said.
Funding
D OI Link:
https://doi.org/10.26599/FRICT.2025.9441151
Friction
Progress on current-carrying friction and wear and prediction models: A review
11-May-2026