arXiv subjects

Alexander Hartmaier

Publications and source records attributed to Alexander Hartmaier.

6 recordsLinked to original sources

arXiv2026

Diversity-Aware Batch-Mode Active Learning for Efficient Sampling in Data-Driven Constitutive Modeling

The constitutive behavior of materials is modeled through relationships between stress, strain, and possibly additional internal variables. This results in relatively high-dimensional feature spaces for machine learning models rendering the efficient generation of informative datasets essential as brute force methods suffer from the curse of dimensionality. This work introduces a diversity-aware batch-mode query-by-committee active-learning strategy to generate datasets of maximum information content at minimum cost. In contrast to existing methods, this novel method selects multiple informative, non-redundant queries per iteration, enabling concurrent generation of informative datasets and reducing the number of machine-learning retraining cycles. A central component of this method is a cosine-similarity-based metric that complements the uncertainty criterion based on committee variance by promoting within-batch diversity. The query selection is guided by committee variance and a diversity-promoting criterion. The approach is benchmarked for efficient stress-space sampling in data-driven constitutive modeling. In this setting, a committee of support vector classifiers approximates the so-called yield surface, which is a manifold dividing the six-dimensional stress space into an elastic and plastic domain. We demonstrate that the method handles different batch sizes robustly, maintains high within batch diversity, and rapidly reduces committee uncertainty. The resulting machine learning yield surfaces achieve predictive accuracy comparable to sequential active learning, while requiring substantially fewer retraining cycles. This makes the proposed approach an efficient strategy for stress space sampling in data driven constitutive modeling and for reducing time to solution via concurrent data collection in each iteration.

physics.comp-ph↗

arXiv2024

A Workflow-Centric Approach to Generating FAIR Data Objects for Computationally Generated Microstructure-Sensitive Mechanical Data

From a data perspective, the materials mechanics field is characterized by sparsity of available data, mainly due to the strong microstructure-sensitivity of properties like strength, fracture toughness, and fatigue limit. This requires testing specimens with different thermo-mechanical histories, even when the composition is similar. Experimental data on mechanical behavior is rare, as mechanical testing is destructive and requires significant material and effort. Furthermore, mechanical behavior is typically characterized in simplified tests under uniaxial loading conditions, whereas a complete characterization requires multiaxial testing. To address this data sparsity, simulation methods like micromechanical modeling can contribute to microstructure-sensitive data collections. This work introduces a novel data schema integrating both metadata and mechanical data, following the workflows of the material modeling processes by which the data has been generated. Each workflow run produces unique data objects by incorporating user, system, and job-specific information correlated with mechanical properties. This approach can be applied to any type of workflow as long as it is well-defined. This integrated format provides a sustainable way of generating Findable, Accessible, Interoperable, and Reusable (FAIR) data objects. The metadata elements focus on key features required to characterize microstructure-specific data, simplifying the collection of purpose-specific datasets by search algorithms.

physics.comp-ph↗

arXiv2015

Free energy function of dislocation densities by large scale atomistic simulation

This paper discusses the free energy of complex dislocation microstructures, which is a fundamental property of continuum plasticity. In the past, multiple models of the self energy of dislocations have been proposed in the literature that partially contradict each other. In order to gain insight into the relationship between dislocation microstructures and the free energy associated with them, instead of deriving a model based on theoretical or phenomenological arguments, here, these quantities are directly measured using large scale molecular dynamics simulations. Plasticity is induced using nanoindentation that creates an inhomogeneous distribution of dislocations as the result of dislocation nucleation and multiplication caused by the local deformation. Using this approach, the measurements of dislocation densities and free energies are ab-initio, because only the interatomic potential is defining the reaction of the system to the applied deformation. The simulation results support strongly a linear relation between the scalar dislocation density and the free energy, which can be related very well to the classical model of mechanical energy of straight dislocations, even for the complex dislocation networks considered here.

cond-mat.mtrl-sci↗

arXiv2010

Ab-initio tensile tests of Al bulk crystals and grain boundaries: on the universality of mechanical behaviour

We have performed ab-initio tensile tests of bulk Al along different tensile axes, as well as perpendicular to different grain boundaries to determine mechanical properties such as interface energy, work of separation and theoretical strength. We show that all the different investigated geometries exhibit energy-displacement curves that can be brought into coincidence in the spirit of the well known universal binding energy relationship curve. This simplifies significantly the calculation of ab-initio tensile strengths for the whole parameter space of grain boundaries.

cond-mat.mtrl-sci↗

arXiv2008

Influence of crystal anisotropy on elastic deformation and onset of plasticity in nanoindentation -- a simulational study

Using molecular-dynamics simulation and finite-element modelling, we simulate nanoindentation into the three principal surfaces -- the (100), (110) and (111) surface -- of Cu and Al. In the elastic regime, the simulation data agree fairly well with the linear elastic theory of indentation into an elastically anisotropic substrate. With increasing indentation, the effect of pressure hardening becomes visible. When the critical stress for dislocation nucleation is reached, even the elastically isotropic Al shows a strong dependence of the surface orientation on the force-displacement curves. After the load drop, when plasticity ahs set in, the influence of the surface orientation is lost, and the contact pressure (hardness) becomes independent of the surface orientation.

cond-mat.mtrl-sci↗

arXiv2008

Pair vs many-body potentials: influence on elastic and plastic behavior in nanoindentation

Molecular-dynamics simulation can give atomistic information on the processes occurring in nanoindentation experiments. In particular, the nucleation of dislocation loops, their growth, interaction and motion can be studied. We investigate how realistic the interatomic potentials underlying the simulations have to be in order to describe these complex processes. Specifically we investigate nanoindentation into a Cu single crystal. We compare simulations based on a realistic many-body interaction potential of the embedded-atom-method type with two simple pair potentials, a Lennard-Jones and a Morse potential. We find that qualitatively many aspects of nanoindentation are fairly well reproduced by the simple pair potentials: elastic regime, critical stress and indentation depth for yielding, dependence on the crystal orientation, and even the level of the hardness. The quantitative deficits of the pair potential predictions can be traced back (i) to the fact that the pair potentials are unable in principle to model the elastic anisotropy of cubic crystals; (ii) as the major drawback of pair potentials we identify the gross underestimation of the stable stacking fault energy. As a consequence these potentials predict the formation of too large dislocation loops, the too rapid expansion of partials, too little cross slip and in consequence a severe overestimation of work hardening.

cond-mat.mtrl-sci↗