A spatiotemporal deep learning framework for prediction of crack dynamics in heterogeneous solids: efficient mapping of concrete microstructures to its fracture properties

TL;DR

This work tackles the computational burden of multiscale fracture simulations in heterogeneous concrete by introducing a spatiotemporal UNet surrogate that predicts full-field crack initiation and propagation, starting in the ITZ and advancing through the mortar matrix. The model ingests three spatial material-property maps plus the current phase-field damage index and predicts the damage field at a future step, while also estimating the corresponding stress-strain point via an attached FFNN. Grounded in cohesive phase-field fracture simulations from Abaqus, the approach uses a fully automated pipeline to map irregular FE data onto a regular grid, enabling efficient training with a relatively small dataset (≈470 simulations). Results show the surrogate can reproduce crack paths with a reasonable F1 score (~0.6) and accurately capture key features of the stress-strain response, while delivering orders-of-magnitude speedups over direct FE analyses. This framework enables microstructure-aware fracture analysis, facilitating materials optimization and structural health assessment for concrete systems, with potential extensions to neural-operator learning and PDE-informed surrogates.

Abstract

A spatiotemporal deep learning framework is proposed that is capable of 2D full-field prediction of fracture in concrete mesostructures. This framework not only predicts fractures but also captures the entire history of the fracture process, from the crack initiation in the interfacial transition zone to the subsequent propagation of the cracks in the mortar matrix. In addition, a convolutional neural network is developed which can predict the averaged stress-strain curve of the mesostructures. The UNet modeling framework, which comprises an encoder-decoder section with skip connections, is used as the deep learning surrogate model. Training and test data are generated from high-fidelity fracture simulations of randomly generated concrete mesostructures. These mesostructures include geometric variabilities such as different aggregate particle geometrical features, spatial distribution, and the total volume fraction of aggregates. The fracture simulations are carried out in Abaqus, utilizing the cohesive phase-field fracture modeling technique as the fracture modeling approach. In this work, to reduce the number of training datasets, the spatial distribution of three sets of material properties for three-phase concrete mesostructures, along with the spatial phase-field damage index, are fed to the UNet to predict the corresponding stress and spatial damage index at the subsequent step. It is shown that after the training process using this methodology, the UNet model is capable of accurately predicting damage on the unseen test dataset by using 470 datasets. Moreover, another novel aspect of this work is the conversion of irregular finite element data into regular grids using a developed pipeline. This approach allows for the implementation of less complex UNet architecture and facilitates the integration of phase-field fracture equations into surrogate models for future developments.

Figures (16)

• Figure 1: SEM imaging of interface zone in concrete materials bonifazi2015itz. The interface zone in concrete materials has a thickness between 20-100 µm, which, compared to other phases, is quite small and has relatively weak material properties.
• Figure 2: Effect of ITZ fracture properties on the fracture toughness of the concrete microstructure koopas2023comparative.
• Figure 3: The solid domain $\Omega$ exhibits a diffuse crack, where the width of the diffuse band is quantified by the length scale parameter $l_{c}$.
• Figure 4: Fully-automated computational pipeline for generating input data for cohesive phase-field fracture simulations at the mesoscale for concrete materials.
• Figure 5: The generation of an aggregate particle is achieved by computing the convex Hull of a set of random points.