Mechanical systems, in which moving parts regularly come into contact, are susceptible to damage from the effects of friction. Tohoku University researchers have developed a contact control system controlled by artificial intelligence to greatly reduce contact with damaged areas. Although it’s only currently being tested in laboratory experiments, they believe it could ultimately help many types of machines run more smoothly.
“This could shift the design strategy of mechanical systems away from the traditional approach of developing new and superior materials towards developing surfaces that can actively adapt to reduce damage,” says Professor Motoyuki Murashima.
The work was a collaboration between Murashima at Tohoku University’s Faculty of Mechanical Systems Engineering and colleagues at Nagoya University and the Korea Photonics Technology Institute in South Korea.
Research focuses on the potential of innovative materials with ‘morphing surfaces’ that can be modified depending on the environment in which they are deployed. These materials are being developed by several research groups to mimic a shared flexibility found in living systems such as leaves, which have surfaces that change in response to changes in humidity. An engineering example previously developed by Murashima and colleagues is a surface composed of a membrane supported by a hard substrate, where changes in stress pressure alter the surface morphologies.
The team developed an artificial intelligence method in which sensors analyze the friction between two surfaces. After determining where damage is occurring, the process can take advantage of the surface’s “morphing” ability to minimize frictional contact with damaged areas.
“This is the first research in the world to use artificial intelligence to control the shape of changing surfaces and successfully detect the location of damage on interacting surfaces,” says Murashima.
In the course of analysis and adjustment in simulated test cases, the researchers were able to achieve a steady reduction in fluctuating friction caused by contact between affected parts of the material under study.
The proof-of-concept system used disks rotating in a cylinder. The crucial next step will be to approach situations where the method could be applied to real engineering challenges, such as B. Industrial machinery. The ultimate goal is to operate a wide range of machines with less routine wear and damage, longer service life and cost savings through fewer parts replacements.
“An important next step is the development of more sophisticated learning and control algorithms that reduce the time needed to learn the properties of the analyzed surfaces and therefore achieve more refined and faster control that prevents damage,” says Murashima.
