Paolo Moretti’s research while affiliated with Friedrich-Alexander-University Erlangen-Nürnberg and other places

We investigate interface failure of model materials representing architected thin films in contact with heterogeneous substrates. We find that, while systems with statistically isotropic distributions of impurities derive their fracture strength from the ability to develop rough detachment fronts, materials with hierarchical microstructures confine failure near a prescribed surface, where crack growth is arrested and crack surface correlations are suppressed. We develop a theory of network Green’s functions for the systems at hand, and we find that the ability of hierarchical microstructures to control failure mode and locations comes at no performance cost in terms of peak stress and specific work of failure and derives from the quenched local anistotropy of the elastic interaction kernel.

We investigate interface failure of model materials representing architected thin films in contact with heterogeneous substrates. We find that, while systems with statistically isotropic distributions of impurities derive their fracture strength from the ability to develop rough detachment fronts, materials with hierarchical microstructures confine failure near a prescribed surface, where crack growth is arrested and crack surface correlations are suppressed. We develop a theory of network Green's functions for the systems at hand, and we find that the ability of hierarchical microstructures to control failure mode and locations comes at no performance cost in terms of peak stress and specific work of failure and derives from the quenched local anistotropy of the elastic interaction kernel.

Creep failure of hierarchical materials is investigated by simulation of beam network models. Such models are idealizations of hierarchical fibrous materials where bundles of load-carrying fibers are held together by multi-level (hierarchical) cross-links. Failure of individual beams is assumed to be governed by stress-assisted thermal activation over local barriers, and beam stresses are computed by solving the global balance equations of linear and angular momentum across the network. Disorder is mimicked by a statistical distribution of barrier heights. Both initially intact samples and samples containing side notches of various length are considered. Samples with hierarchical cross-link patterns are simulated alongside reference samples where cross-links are placed randomly without hierarchical organization. The results demonstrate that hierarchical patterning may strongly increase creep strain and creep lifetime while reducing the lifetime variation. This is due to the fact that hierarchical patterning induces a failure mode that differs significantly from the standard scenario of failure by nucleation and growth of a critical crack. Characterization of this failure mode demonstrates good agreement between the present simulations and experimental findings on hierarchically patterned paper sheets.

Quasibrittle materials endowed with (statistically) self-similar hierarchical microstructures show distinct failure patterns that deviate from the standard scenario of damage accumulation followed by crack nucleation and growth. Here we study the failure of paper sheets with hierarchical slice patterns as well as nonhierarchical and unpatterned reference samples, considering both uncracked samples and samples containing a macroscopic crack. Failure is studied under displacement-controlled tensile loading as well as under creep conditions. Acoustic emission records and surface strain patterns are recorded alongside stress-strain and creep curves. The measurements demonstrate that hierarchical patterning efficiently mitigates against strain localization and crack propagation. In tensile loading, this results in a significantly increased residual strength of cracked samples. Under creep conditions, for a given range of lifetimes, hierarchically patterned samples are found to sustain larger creep strains at higher stress levels; their creep curves show unusual behavior characterized by multiple creep rate minima due to the repeated arrest of emergent localization bands.

Hierarchical microstructures are often invoked to explain the high resilience and fracture toughness of biological materials such as bone and nacre. Biomimetic material models inspired by such hierarchical biomaterials face the obvious challenge of capturing their inherent multiscale complexity, both in experiments and in simulations. To study the influence of hierarchical microstructure on fracture properties, we propose a large-scale three-dimensional hierarchical beam-element simulation framework, in which we generalize the constitutive framework of Timoshenko beam elasticity and maximum distortion energy theory failure criteria to the complex case of hierarchical networks of up to six self-similar hierarchical levels, consisting of approximately 5 million elements. We perform a statistical study of stress-strain relationships and fracture surface morphologies and conclude that hierarchical systems are capable of arresting crack propagation, an ability that reduces their sensitivity to preexisting damage and enhances their fault tolerance compared to reference fibrous materials without microstructural hierarchy.

Quasi-brittle materials endowed with (statistically) self-similar hierarcical microstructures show distinct failure patterns that deviate from the standard scenario of damage accumulation followed by crack nucleation-and-growth. Here we study the failure of paper sheets with hierarchical slice patterns as well as non-hierarchical and unpatterned reference samples, considering both uncracked samples and samples containing a macroscopic crack. Failure is studied under displacement-controlled tensile loading as well as under creep conditions. Acoustic emission records and surface strain patterns are recorded alongside stress-strain and creep curves. The measurements demonstrate that hierarchical patterning efficiently mitigates against strain localization and crack propagation. In tensile loading, this results in a significantly increased residual strength of cracked samples. Under creep conditions, for a given range of lifetimes hierarchically patterned samples are found to sustain larger creep strains at higher stress levels; their creep curves show unusual behavior characterized by multiple creep rate minima due to the repeated arrest of emergent localization bands.

Spatial and temporal features are studied with respect to their predictive value for failure time prediction in subcritical failure with machine learning (ML). Data are generated from simulations of a novel, brittle random fuse model (RFM), as well as elasto-plastic finite element simulations (FEM) of a stochastic plasticity model with damage, both models considering stochastic thermally activated damage/failure processes in disordered materials. Fuse networks are generated with hierarchical and nonhierarchical architectures. Random forests - a specific ML algorithm - allow us to measure the feature importance through a feature’s average error reduction. RFM simulation data are found to become more predictable with increasing system size and temperature. Increasing the load or the scatter in local materials properties has the opposite effect. Damage accumulation in these models proceeds in stochastic avalanches, and statistical signatures such as avalanche rate or magnitude have been discussed in the literature as predictors of incipient failure. However, in the present study such features proved of no measurable use to the ML models, which mostly rely on global or local strain for prediction. This suggests the strain as viable quantity to monitor in future experimental studies as it is accessible via digital image correlation.

By introducing hierarchical patterns of load-parallel cuts into axially loaded brittle sheets, the resistance to propagation of mode-I cracks is very significantly enhanced. We demonstrate this effect by simulation of two-dimensional beam network models and experimentally by testing paper and polystyrene (PS) sheets that are sliced with a laser cutter to induce load-perpendicular hierarchical cut patterns. Samples endowed with nonhierarchical reference patterns of the same cut density and nonsliced sheets are considered for comparison. We demonstrate that hierarchical slicing can increase failure load, apparent fracture toughness, and work of fracture of notched paper and PS sheets by factors between 2 and 10.

Spatial and temporal features are studied with respect to their predictive value for failure time prediction in subcritical failure with machine learning (ML). Data are generated from simulations of a novel, brittle random fuse model (RFM), as well as elasto-plastic finite element simulations (FEM) of a stochastic plasticity model with damage, both models considering stochastic thermally activated damage/failure processes in disordered materials. Fuse networks are generated with hierarchical and nonhierarchical architectures. Random forests - a specific ML algorithm - allow us to measure the feature importance through a feature's average error reduction. RFM simulation data are found to become more predictable with increasing system size and temperature. Increasing the load or the scatter in local materials properties has the opposite effect. Damage accumulation in these models proceeds in stochastic avalanches, and statistical signatures such as avalanche rate or magnitude have been discussed in the literature as predictors of incipient failure. However, in the present study such features proved of no measurable use to the ML models, which mostly rely on global or local strain for prediction. This suggests the strain as viable quantity to monitor in future experimental studies as it is accessible via digital image correlation.

Hierarchical microstructures are often invoked to explain the high resilience and fracture toughness of biological materials such as bone and nacre. Biomimetic material models inspired by those structural arrangements face the obvious challenge of capturing their inherent multi-scale complexity, both in experiments and in simulations. To study the influence of hierarchical microstructural patterns in fracture behavior, we propose a large scale three-dimensional hierarchical beam-element simulation framework, where we generalize the constitutive behavior of Timoshenko beam elasticity and Maximum Distortion Energy Theory failure criteria to the complex case of hierarchical networks of approximately 5 million elements. We perform a statistical study of stress-strain relationships and fracture surface mophologies, and conclude that hierarchical systems are capable of arresting crack propagation, an ability that reduces their sensitivity to pre-existing damage and enhances their fault tolerance compared to reference generic fibrous materials.