[LUM#22] Rock-Solid Predictions
Reactor containment structures, cooling towers, spent fuel storage pools… concrete is everywhere in a nuclear power plant. And it is subjected to numerous stresses: radiological, thermal, chemical, hydraulic, and mechanical. To predict the watertightness of these facilities, Yann Monerie, a researcher at the Mechanics and Civil Engineering Laboratory, uses artificial intelligence.
More than 20 million tons of concrete are produced in France each year to build buildings, dams, bridges, and even… nuclear power plants. These are high-risk facilities where this mixture of sand, cement, gravel, and water acts as a containment barrier. But how can we be sure that this barrier will remain watertight over time? “The impermeability of concrete is determined by how it cracks. The more tortuous the path of these cracks, the less easily fluids flow through them, and therefore the more impermeable it is, ” explains Yann Monerie, a researcher at LMGC1.
Tortuosity
To determine this tortuosity, one must first understand the properties of the mixture’s various components, as well as how the aggregates are distributed within the mortar. “Concrete is not homogeneous, and the spatial distribution of its different phases determines how it cracks, ” explains the specialist in materials micromechanics, who is collaborating on this issue with the Institute for Radiation Protection and Nuclear Safety.
This is because a crack does not propagate randomly, but moves step by step from one gravel particle to another . “The interfaces between the aggregates and the matrix are weak zones that determine how cracks propagate.” It is this path that researchers aim to predict in order to better understand the mechanical properties of this material.
From 4 months to 3 seconds
And to make these predictions, Yann Monerie and his colleagues are using artificial intelligence. “We have developed a model that allows for the statistical reconstruction of the three-dimensional microstructure from two-dimensional cross-sections, ” explains the researcher.
After analyzing thousands of photos of concrete cross-sections and their corresponding 3D models, the AI enables them to determine, based on simple images, the size and shape of the aggregates as well as how they are distributed within the cement. This modeling allows researchers to create a numerical simulation of crack propagation. “Without machine learning, it takes about four months of calculations to perform these simulations; with AI, it takes two to three seconds!”
This predictive artificial intelligence makes it possible to estimate a power plant’s permeability based on simple cross-sectional views of concrete samples. And these applications extend even beyond the nuclear sector: “Using this method, we can also assess the watertightness of homes built on contaminated soil, for example, ” adds Yann Monerie.
UM podcasts are now available on your favorite platform (Spotify, Deezer, Apple Podcasts, Amazon Music, etc.).
- LMGC (CNRS, UM)
↩︎