Symposium 8: Beyond Similitude: Role of Multiscale Heterogeneity in Fracture Prognosis

Aluminum alloys commonly used in airframe structures have been observed to show orthotropy in fracture when processed through hot rolling or extrusion, while other properties such as yield are more isotropic. Fracture orthotropy is likely due to a competition between damage accumulation within the grains by void growth and cleavage along the grain boundaries. Analysis of the fracture surface indicates varying degrees of dimpled regions (indicating damage by void growth) and quasi-brittle flat regions coinciding with the grain boundary (indicating grain boundary failure). To help determine structure-property relations in such materials, this paper describes a computational model for fracture in ductile polycrystals accounting for both the damage mechanisms. The model is validated by comparing with experiments on a high strength aluminum alloy, AA2139.
EXTENDED ABSTRACT

The ability to rapidly predict the growth behavior of microstructurally small cracks (MSCs) has the potential to significantly advance fracture-based designs and structural prognosis. The difficulties associated with characterizing or predicting MSC growth using experimental and numerical techniques preclude the applicability of such techniques in industrial design approaches, despite their potential benefits. Here, we propose a framework to accelerate high-fidelity MSC growth predictions using deep-learning algorithms, viz. , convolutional neural networks (CNNs). The primary research aim is to train CNNs to predict the rules governing MSC growth and to subsequently apply the trained CNNs to make rapid forward predictions of local crack extension given microstructural neighborhood information along a crack front. The training data are acquired from a large number of “virtual” MSC growth observations enabled by high-fidelity finite-element-based simulations. The MSC-growth-simulation framework, data-extraction strategies, and application of deep-learning algorithms for data-driven model development will be presented, and the resulting advantages will be demonstrated.
EXTENDED ABSTRACT

This study proposed a forward analysis method to predict tensile strength and total elongation by considering the three-dimensional microstructure of dual-phase steels. By repeating the forward analysis, an inverse analysis was performed to search for a microstructure with higher tensile properties. The optimal microstructures found by the inverse analysis were consistent with conventional materials engineering findings, demonstrating that the proposed inverse analysis method is effective in solving the structure-properties linkages in the inverse direction.
EXTENDED ABSTRACT

The frictional crack is extensively observed in natural phenomena like earthquake fracture, fracture at rock
fault line and in fracture of other geological materials. The contact between the flaw faces alters the stress
field in comparison to the stress field in an open condition or when not in contact. The crack growth
initiation in an open condition have been investigated sufficiently in literature. However, the frictional crack
or closed crack in an anisostropic medium has hardly been addressed. The naturally occurring biological
materials such as bone, wood, cartilage, etc. are anisotropic as well as quasi-brittle in nature. Considering
the vital application of these naturally occurring composites, it calls for a thorough investigation. The
current research work performs compression test on wood with an embedded central pore for different
contact surface conditions and orientation of pore. In addition to that, it carries out numerical simulation of
crack growth initiation and propagation using cohesive zone model (CZM) considering quasi-brittle nature
of wood. Further, it verifies the applicability of classical fracture criteria for friction crack propagation in
wood.
EXTENDED ABSTRACT

The macroscopic behavior of plasma sprayed zirconia coatings is greatly affected by their microstructure, and phenomena that occur at this scale, such as micro-craking and cracks closure. The present study investigates the effect of micro-porosities and micro-craking on compressive waves mitigation, material compaction, and macroscopic fracture. A procedure to generate 3D digital twins that faithfully represents the microstructure is developped using the discrete element method (DEM) and analysis of scanning electron microscope (SEM) images. Static and dynamic compressive loadings are applied to 3D and 2D twins to identify their macroscopic behavior. Local damage mechanisms and their influence on the waves mitigation and the macroscopic damage are observed and discussed related to the current knowledge in the literature.
EXTENDED ABSTRACT

It is known that damage or inelastic softening can cause an ill-posed problem leading to localization and mesh-dependence in finite element simulations. Here, a nonlocal hardening variable is introduced in a finite deformation Eulerian formulation of inelasticity. This nonlocal variable is defined over an Eulerian region of nonlocality, which is a sphere with radius equal to a characteristic length, defined in the current deformed geometry of the material. The influence of the nonlocal hardening variable is studied using an example of a plate that is loaded by a prescribed boundary displacement causing formation of a shear band. Predictions of the applied load vs. displacement curves and contour plots of the total distortional deformation of the plate and the hardening variable are studied. It is shown that the characteristic material length controls the structure of the shear band developed in the plate.
EXTENDED ABSTRACT

We study the intermittent damage evolution preceding compressive failure using a non-local damage model accounting for material disorder and long-range elastic interactions. Our theoretical predictions are successfully compared with experiments carried on a model elasto-damageable 2D solid where damage evolution is tracked at both the global and local scale. Finally, we show how our understanding of these failure precursors can be harnessed for predicting the remaining lifetime of structures under compression.
EXTENDED ABSTRACT