Data-Driven Failure Prediction in Brittle Materials: A Phase Field-Based Machine Learning Framework

Abstract

Failure in brittle materials led by the evolution of micro-to macro-cracks under repetitive or increasing loads is often catastrophic with no significant plasticity to advert the onset of fracture. Early failure detection with respective location are utterly important features in any practical application, both of which can be effectively addressed using artificial intelligence. In this paper, we develop a supervised machine learning (ML) framework to predict failure in an isothermal, linear elastic and isotropic phase-field model for damage and fatigue of brittle materials. Time-series data of the phase-field model is extracted from virtual sensing nodes at different locations of the geometry. A pattern recognition scheme is introduced to represent time-series data/sensor node responses as a pattern with a corresponding label, integrated with ML algorithms, used for damage classification with identified patterns. We perform an uncertainty analysis by superposing random noise to the time-series data to assess the robustness of the framework with noise-polluted data. Results indicate that the proposed framework is capable of predicting failure with acceptable accuracy even in the presence of high noise levels. The findings demonstrate satisfactory performance of the supervised ML framework and the applicability of artificial intelligence and ML to a practical engineering problem, i.e., data-driven failure prediction in brittle materials.

Review of “Integrated Computer-Aided Process Engineering Session in the International Symposium on Innovation in Materials Processing (ISIMP, 26–29 October 2021)”

“…Recently, many researchers have been using AI to discover new materials or optimize the composition and process conditions to achieve desirable performances, from DoE [43][44][45][46][47] to ML. 36,38,56) Lee et al 47) optimized the parameters for synthesizing nickel phosphide as a catalyst using the Taguchi method in the DoE and established correlations between the modeling results and those of the electrochemical reactions. The Taguchi method finds the optimal point with minimum experiments 43,44,47) and can solve various distributions such as experimental equipment and worker proficiency.…”

“…Recent advancements in computational methods, such as phase-field modeling (PFM) and molecular dynamics (MD) simulations, have enabled researchers to study microstructural evolution with high accuracy and efficiency. In addition, optimization techniques such as machine learning (ML), [29][30][31][32][33][34][35][36] artificial intelligence (AI), [37][38][39] and design of experiments (DoE) [43][44][45][46][47] may significantly accelerate process development by rapidly exploring and identifying optimal process conditions based on computational and/or experimental data, reducing the need for costly and time-consuming trial-and-error approaches. Hence, integrating these optimization techniques with advanced simulation methods can further enhance the accuracy and efficiency of material property prediction and process optimization.…”

Atomistic-to-Meso Multi-Scale Data-Driven Graph Surrogate Modeling of Dislocation Glide

“…In general, this is accomplished by recomputing the elements of W as p ij , representing the probability of a jump per unit of time, in units of s −1 . Probabilities are obtained through (6) p ij = w i→j q j j w i→j q j , where p ij is now the walker's probability to go from node i to node j, per unit time .…”

“…At that stage, the collective behavior of dislocations would intrinsically incorporate stochastic effects of lower scales that would be propagated to the continuum (i.e., through dislocation density and plastic strains), therefore providing efficient multi-scale coupling starting in the MD domain. This feature is essential to the development of predictive models at the component level, whether the interest is on visco-elasto-plasticity [43,45,44], fracture [7,6] or fatigue [8].…”

Nonlocal Machine Learning of Micro-Structural Defect Evolutions in Crystalline Materials

“…With this mindset, we obtain a fast alternative to simulate dislocation dynamics through the nonlocal surrogate model, while still maintaining the underlying physics of micro-structural processes. This leads to a more efficient connection to macroscale problems such as visco-elasticity [45] and fracture [7,9]. This paper is organized as follows: in Section 2, we present the high-fidelity twodimensional DDD simulation setting for single crystals under creep for three canonical conditions.…”