Decoding material wear with supercomputers

Decoding material wear with supercomputers
A small piece of metal is simulated on a computer—atom by atom. Credit: TU Wien

Wear and friction are crucial issues in many industrial sectors: What happens when one surface slides across another? Which changes must be expected in the material? What does this mean for the durability and safety of machines?

What happens at the cannot be observed directly. However, an additional scientific tool is now available for this purpose: For the first time, complex have become so powerful that wear and friction of real materials can be simulated on an atomic scale.

The tribology team at TU Wien (Vienna), led by Prof. Carsten Gachot, has now proven that this new research field now delivers reliable results in a current publication in the renowned scientific journal ACS Applied Materials & Interfaces. The behavior of surfaces consisting of copper and nickel was simulated with high-performance computers. The results correspond amazingly well with images from electron microscopy—but they also provide valuable additional information.

Friction Changes Tiny Grains

To the , it does not look particularly spectacular when two surfaces slide across each other. But on the , highly complicated processes take place: "Metals, as they are used in technology, have a special microstructure," explains Dr. Stefan Eder, first author of the current publication. "They consist of small grains with a diameter of the order of micrometers or even less."

When one metal slides over the other under high shear stress, the grains of the two materials come into intense contact with each other: they can be rotated, deformed or shifted, they can be broken up into smaller grains or grow due to increased temperature or mechanical force. All these processes, which take place on a microscopic scale, ultimately determine the behavior of the material on a large scale—and thus they also determine the service life of a machine, the amount of energy lost in a motor due to friction, or how well a brake works, in which the highest possible friction force is desired.

A metal surface (copper and nickel) under stress. different kinds of deformation are visible. Changes in the granular structure of the metal are simulated on a supercomputer with atomic resolition. Credit: TU Wien

Computer Simulation and Experiment

"The result of these microscopic processes can then be examined in an electron microscope," says Stefan Eder. "You can see how the grain structure of the surface has changed. However, it has not yet been possible to study the time evolution of these processes and explain exactly what causes which effects at which point in time."

This gap is now being closed by large molecular dynamics simulations developed by the tribology team at TU Wien in cooperation with the Excellence Centre of Tribology (AC²T) in Wiener Neustadt and the Imperial College in London: Atom by atom, the surfaces are simulated on the computer. The larger the simulated chunk of material and the longer the simulated time period, the more computer power is needed. "We simulate sections with a side length of up to 85 nanometers, over a period of several nanoseconds," says Stefan Eder. That doesn't sound like much, but it's remarkable: Even the Vienna Scientific Cluster 4, Austria's largest supercomputer, may sometimes be busy with such tasks for months at a time.

The team investigated the wear of alloys of copper and nickel—and did so using different mixing ratios of the two metals and different mechanical loads. "Our computer simulations revealed exactly the variety of processes, microstructural changes and wear effects that are already known from experiments," says Stefan Eder. "We can use our simulations to produce images that correspond exactly to the images from the electron microscope. However, our method has a decisive advantage: We can then analyze the process in detail on the computer. We know which atom changed its place at what point in time, and what exactly happened to which grain in which phase of the process."

Understanding Wear—Optimizing Industrial Processes

The new methods are already met with great interest from industry: "For years, there has been an ongoing discussion that tribology could benefit from reliable simulations. Now we have reached a stage where the quality of the simulations and the available computing power are so great that we could use them to answer exciting questions that would otherwise not be accessible," says Carsten Gachot. In the future, they also want to analyze, understand, and improve industrial processes on the atomic level.


Explore further

Gently caressing atoms

More information: Stefan J Eder et al, Unraveling and Mapping the Mechanisms for Near-surface Microstructure Evolution in CuNi Alloys under Sliding, ACS Applied Materials & Interfaces (2020). DOI: 10.1021/acsami.0c09302
Citation: Decoding material wear with supercomputers (2020, June 30) retrieved 13 July 2020 from https://phys.org/news/2020-06-decoding-material-supercomputers.html
This document is subject to copyright. Apart from any fair dealing for the purpose of private study or research, no part may be reproduced without the written permission. The content is provided for information purposes only.
24 shares

Feedback to editors

User comments