TY - JOUR A1 - Otto, Peter A1 - Lorenzis, L. A1 - Unger, Jörg F. T1 - Explicit dynamics in impact simulation using a NURBS contact interface N2 - In this paper, the impact problem and the subsequent wave propagation are considered. For the contact discretization an intermediate NURBS layer is added between the contacting finite element bodies, which allows a smooth contact formulation and efficient element‐based integration. The impact event is ill‐posed and requires a regularization to avoid propagating stress oscillations. A nonlinear mesh dependent penalty regularization is used, where the stiffness of the penalty regularization increases upon mesh refinement. Explicit time integration methods are well suited for wave propagation problems, but are efficient only for diagonal mass matrices. Using a spectral element discretization and the coupled FE‐NURBS approach the bulk part of the mass matrix is diagonal. KW - Impact simulation KW - Explicit dynamics KW - Isogeometric analysis KW - Spectral elements KW - Mortar method PY - 2019 UR - https://nbn-resolving.org/urn:nbn:de:kobv:b43-494947 DO - https://doi.org/10.1002/nme.6264 SP - 1 EP - 21 PB - Wiley Online Libary AN - OPUS4-49494 LA - eng AD - Bundesanstalt fuer Materialforschung und -pruefung (BAM), Berlin, Germany ER - TY - JOUR A1 - Robens-Radermacher, Annika A1 - Unger, Jörg F. T1 - Efficient reliability analysis coupling importance sampling using adaptive subset simulation and PGD model reduction N2 - One of the most important goals in civil engineering is to guaranty the safety of constructions. National standards prescribe a required failure probability in the order of 10−6 (e.g. DIN EN 199:2010-12). The estimation of these failure probabilities is the key point of structural reliability analysis. Generally, it is not possible to compute the failure probability analytically. Therefore, simulation-based methods as well as methods based on surrogate modeling or response surface methods have been developed. Nevertheless, these methods still require a few thousand evaluations of the structure, usually with finite element (FE) simulations, making reliability analysis computationally expensive for relevant applications. The aim of this contribution is to increase the efficiency of structural reliability analysis by using the advantages of model reduction techniques. Model reduction is a popular concept to decrease the computational effort of complex numerical simulations while maintaining a reasonable accuracy. Coupling a reduced model with an efficient variance reducing sampling algorithm significantly reduces the computational cost of the reliability analysis without a relevant loss of accuracy. T2 - GAMM - Gesellschaft für Angewandte Mathematik und Mechanik e.V. CY - Wien, Austria DA - 18.02.2019 KW - PGD model reduction KW - Numerical models KW - FE modelling PY - 2019 DO - https://doi.org/10.1002/pamm.201900169 VL - 19 IS - 1 SP - 1 EP - 2 PB - Wiley- VCH Verlag GmbH CY - Weinheim AN - OPUS4-49786 LA - eng AD - Bundesanstalt fuer Materialforschung und -pruefung (BAM), Berlin, Germany ER - TY - CONF A1 - Unger, Jörg F. T1 - Efficient reliability analysis with model reduction techniques N2 - One of the most important goals in civil engineering is to guaranty the safety of constructions. National standards prescribe a required failure probability in the order of 10-6 (e.g. DIN EN 199:2010-12). The estimation of these failure probabilities is the key point of structural reliability analysis. Generally, it is not possible to compute the failure probability analytically. Therefore, simulation-based methods as well as methods based on surrogate modelling or response surface methods have been developed. Nevertheless, these methods still require a few thousand evaluations of the structure, usually with finite element (FE) simulations, making reliability analysis computationally expensive for relevant applications. The aim of this contribution is to increase the efficiency of structural reliability analysis by using the advantages of model reduction techniques. Model reduction is a popular concept to decrease the computational effort of complex numerical simulations while maintaining a reasonable accuracy. Coupling a reduced model with an efficient variance reducing sampling algorithm significantly reduces the computational cost of the reliability analysis without a relevant loss of accuracy. T2 - MathMet2019 CY - Lisbon, Portugal DA - 20.11.2019 KW - Reliability method KW - Numerical example PY - 2019 AN - OPUS4-49989 LA - eng AD - Bundesanstalt fuer Materialforschung und -pruefung (BAM), Berlin, Germany ER - TY - JOUR A1 - Bayerlein, Bernd A1 - Hanke, T. A1 - Muth, Thilo A1 - Riedel, Jens A1 - Schilling, Markus A1 - Schweizer, C. A1 - Skrotzki, Birgit A1 - Todor, A. A1 - Moreno Torres, Benjami A1 - Unger, Jörg F. A1 - Völker, Christoph A1 - Olbricht, Jürgen T1 - A Perspective on Digital Knowledge Representation in Materials Science and Engineering N2 - The amount of data generated worldwide is constantly increasing. These data come from a wide variety of sources and systems, are processed differently, have a multitude of formats, and are stored in an untraceable and unstructured manner, predominantly in natural language in data silos. This problem can be equally applied to the heterogeneous research data from materials science and engineering. In this domain, ways and solutions are increasingly being generated to smartly link material data together with their contextual information in a uniform and well-structured manner on platforms, thus making them discoverable, retrievable, and reusable for research and industry. Ontologies play a key role in this context. They enable the sustainable representation of expert knowledge and the semantically structured filling of databases with computer-processable data triples. In this perspective article, we present the project initiative Materials-open-Laboratory (Mat-o-Lab) that aims to provide a collaborative environment for domain experts to digitize their research results and processes and make them fit for data-driven materials research and development. The overarching challenge is to generate connection points to further link data from other domains to harness the promised potential of big materials data and harvest new knowledge. KW - Data infrastructures KW - Digital representations KW - Digital workflows KW - Knowledge graphs KW - Materials informatics KW - Ontologies KW - Vocabulary providers PY - 2022 UR - https://nbn-resolving.org/urn:nbn:de:kobv:b43-546729 DO - https://doi.org/10.1002/adem.202101176 SN - 1438-1656 SP - 1 EP - 14 PB - Wiley-VCH GmbH CY - Weinheim AN - OPUS4-54672 LA - eng AD - Bundesanstalt fuer Materialforschung und -pruefung (BAM), Berlin, Germany ER - TY - GEN A1 - Rosenbusch, Sjard Mathis A1 - Balzani, Daniel A1 - Unger, Jörg F. T1 - Regularization of softening plasticity models for explicit dynamics using a gradient-enhanced modified Johnson-Holmquist model N2 - The behavior of concrete under high strain rates is often described by plasticity models with softening, which is modeled by a reduction of the yield surface as a function of the local equivalent plastic strain. Among these are the RHT model, the K\&C model and the Johnson-Holmquist concrete model. These models are however local and therefore produce mesh-dependent results. In this contribution, the gradient-enhancement of such models is investigated. First, the mesh-dependency of these local formulations based on the analysis with a modified JH2 model as a representative for these constitutive formulations is demonstrated using a one-dimensional benchmark example. The central difference method is used as solver with a diagonal mass matrix obtained from a Gauß-Lobatto integration. In the benchmark, the width of the damaged zone decreases upon mesh-refinement and the dissipated plastic energy tends to zero. It is further shown that a significantly small safety factor for the critical time step is needed in order to achieve accurate results for the benchmark example. Next, two gradient-enhancement approaches are investigated. The enhancement is based on the inclusion of inertia and damping to the additional Helmholtz equation which enables the use of the central difference method as an explicit solver. In the first formulation, the yield surface and therefore the softening is formulated in terms of a nonlocal equivalent plastic strain. In the second approach, a hardening term which depends on the local equivalent plastic strain is introduced to the modified JH2 model in addition to the nonlocal softening. This approach is inspired by results from gradient plasticity in quasi-static loading scenarios. It is shown that the approach without hardening can still lead to mesh-dependent results while the model that includes hardening successfully inhibits strain localization and leads to a converging dissipated plastic energy. This is further confirmed in a two-dimensional wedge-splitting experiment where the damage pattern produced by the local model is mesh-dependent as well and the dissipated plastic energy tends to zero with mesh-refinement. The proposed nonlocal model with hardening results in a consistent damage pattern and the dissipated plastic energy converges. Furthermore, the nonlocal model with hardening is less sensitive to time step refinement, such that computational efficiency can be improved compared to the local model. The numerical experiments are implemented using the free and open-source tool FEniCSx. KW - Concrete modeling KW - Explicit Dynamics KW - FEniCS KW - Gradient plasticity KW - JH2 model KW - Mesh convergence PY - 2024 DO - https://doi.org/10.5281/zenodo.13983859 PB - Zenodo CY - Geneva AN - OPUS4-65083 LA - eng AD - Bundesanstalt fuer Materialforschung und -pruefung (BAM), Berlin, Germany ER - TY - CONF A1 - Unger, Jörg F. T1 - From scientific models to reliable decision making – RDM as a backbone for validated, FAIR, and reproducible engineering simulations N2 - Engineering simulations play a central role in predicting system behavior and supporting design decisions. Their impact will increase due to the concepts of virtual lab, digital twins and predictive maintenance, where expensive prototyping is reduced with simulations or where real measurement data of the structures or parts are used in combination with simulation models to improve decision making. The credibility of a simulation model depends on the trust it earns within the scientific and engineering community. This trust can only be established through rigorous verification and validation. Verification and validation are inherently hierarchical processes. At the highest level, global approaches—such as standardized benchmarks and shared validation datasets—provide a common foundation for assessing model correctness and applicability across domains. These global resources enable regulatory bodies, industry, and research communities to evaluate whether models and their implementations meet agreed-upon standards. Research Data Management (RDM) provides the infrastructure to support these processes by enabling formalized model definitions, structured validation datasets, and transparent verification procedures and complete provenance descriptions of the underlying workflows. Currently, many engineering models are published as descriptive text or embedded in specific software implementations, which hinders reproducibility and standardization. Numerical results presented in publications must be fully reproducible. For regulatory assessment and interoperability, models require software-agnostic, machine-readable definitions that include all information needed for implementation and execution, such as scope, assumptions, parameter identification procedures, and uncertainty bounds. Formal definitions alone do not guarantee consistency. Verification benchmarks are essential to confirm that different implementations of the same mathematical or numerical model produce equivalent results. Likewise, validation requires structured experimental data that is publicly available and documented according to community-agreed protocols for tests, measurements, and metadata. Robust calibration and uncertainty quantification frameworks must provide systematic, reproducible procedures for parameter estimation and uncertainty propagation, while the complete provenance of these steps has to be documented. Validation should extend beyond controlled laboratory conditions to diverse real-world scenarios. The presentation will explore practical strategies for achieving reproducibility and reusability in engineering simulations through structured workflows, provenance tracking, and standardized data management practices. It will address the integration of experimental data for validation, the role of verification benchmarks, uncertainty quantification frameworks, and tool-independent model descriptions—including VMAP for result exchange—within collaborative platforms that support joint projects and regulatory compliance. T2 - GAMM Activity Group on Research Software Engineering and Research Data Management in Mathematics & Mechanics (GAMM FA RSE&RDM) CY - Braunschweig, Germany DA - 04.12.2025 KW - Experimental data KW - Verification and validation KW - UQ KW - Reproducibility (workflows) KW - Reusability (provenance tracking of simulation results) PY - 2025 AN - OPUS4-65267 LA - eng AD - Bundesanstalt fuer Materialforschung und -pruefung (BAM), Berlin, Germany ER - TY - CONF A1 - Unger, Jörg F. T1 - Methodological prerequisites for reliable simulation results in the Virtual Lab and Digital Twins N2 - Mit dem zunehmenden Einsatz von Methoden wie Künstlicher Intelligenz, Digitalen Zwillingen und datengetriebener Modellierung erweitert sich das Anwendungsspektrum der virtuellen Produktentwicklung kontinuierlich. In sicherheitsrelevanten Anwendungsfeldern steigen dabei die Anforderungen an die Qualität, Nachvollziehbarkeit und externe Prüfbarkeit von Simulationsergebnissen – insbesondere im Hinblick auf Reproduzierbarkeit und Transparenz. Der Beitrag analysiert zentrale methodische und strukturelle Herausforderungen für den zuverlässigen Einsatz numerischer Simulationen in solchen Kontexten. Im Mittelpunkt stehen folgende Aspekte: • Die formale, softwareunabhängige Spezifikation mathematischer und numerischer Modelle, einschließlich einer standardisierten Beschreibung sowohl der Problemstellung als auch der Ergebnisgrößen. • Die Validierung sämtlicher Modellkomponenten anhand von Anwendungsszenarien mit hoher Ähnlichkeit zur realen Systemumgebung. • Die systematische Quantifizierung von Unsicherheiten, insbesondere im Hinblick auf Sim2Real-Modellabweichungen und deren Einfluss auf die Aussagekraft der Simulation. • Die Verifikation der Softwareimplementierung durch standardisierte Benchmark-Vergleiche. Hierzu wird ein Konzept föderierter Benchmarks mit dokumentierter Herkunft („Provenance“) vorgestellt, das eine transparente, reproduzierbare und öffentlich zugängliche Verifikation von Simulationswerkzeugen ermöglicht. Die Benchmarks folgen definierten Formaten, können dezentral veröffentlicht und in eine zentrale, abfragbare Datenbank integriert werden. Durch diese standardisierten Schnittstellen können dann Metriken zum Vergleich der Benchmarks visualisiert und in Analysen bereitgestellt werden. Der Beitrag zielt darauf ab, zentrale Herausforderungen und offene Fragestellungen bei der methodischen Absicherung simulationsbasierter Aussagen zu identifizieren und zur Diskussion zu stellen – insbesondere im Spannungsfeld zwischen Modellbildung, Validierung, Verifikation und Unsicherheitsquantifizierung. T2 - Symbiose von Simulation und Test CY - Wiesbaden, Germany DA - 12.11.2025 KW - Structured validation data KW - Model calibration KW - UQ PY - 2025 AN - OPUS4-65270 LA - eng AD - Bundesanstalt fuer Materialforschung und -pruefung (BAM), Berlin, Germany ER - TY - INPR A1 - Andrés Arcones, Daniel A1 - Weiser, M. A1 - Koutsourelakis, P-S. A1 - Unger, Jörg F. T1 - Embedded Model Bias Quantification with Measurement Noise for Bayesian Model Calibration N2 - A key factor in ensuring the accuracy of computer simulations that model physical systems is the proper calibration of their parameters based on real-world observations or experimental data. Inevitably, uncertainties arise, and Bayesian methods provide a robust framework for quantifying and propagating these uncertainties to model predictions. Nevertheless, Bayesian methods paired with inexact models usually produce predictions unable to represent the observed datapoints. Additionally, the quantified uncertainties of these overconfident models cannot be propagated to other Quantities of Interest (QoIs) reliably. A promising solution involves embedding a model inadequacy term in the inference parameters, allowing the quantified model form uncertainty to influence non-observed QoIs. This paper introduces a more interpretable framework for embedding the model inadequacy compared to existing methods. To overcome the limitations of current approaches, we adapt the existing likelihood models to properly account for noise in the measurements and propose two new formulations designed to address their shortcomings. Moreover, we evaluate the performance of this inadequacy-embedding approach in the presence of discrepancies between measurements and model predictions, including noise and outliers. Particular attention is given to how the uncertainty associated with the model inadequacy term propagates to the QoIs, enabling a more comprehensive statistical analysis of prediction’s reliability. Finally, the proposed approach is applied to estimate the uncertainty in the predicted heat flux from a transient thermal simulation using temperature observations. KW - Model bias KW - Bayesian inference KW - Noise KW - Model updating KW - Quantity of Interest PY - 2025 UR - https://nbn-resolving.org/urn:nbn:de:kobv:b43-652720 DO - https://doi.org/10.48550/arXiv.2410.12037 SP - 1 EP - 37 PB - arXiv.org AN - OPUS4-65272 LA - eng AD - Bundesanstalt fuer Materialforschung und -pruefung (BAM), Berlin, Germany ER - TY - INPR A1 - Villani, P. A1 - Andés-Arcones, Daniel A1 - Unger, Jörg F. A1 - Weiser, M. T1 - Posterior sampling with Adaptive Gaussian Processes in Bayesian parameter identification N2 - Posterior sampling by Monte Carlo methods provides a more comprehensive solution approach to inverse problems than computing point estimates such as the maximum posterior using optimization methods, at the expense of usually requiring many more evaluations of the forward model. Replacing computationally expensive forward models by fast surrogate models is an attractive option. However, computing the simulated training data for building a sufficiently accurate surrogate model can be computationally expensive in itself, leading to the design of computer experiments problem of finding evaluation points and accuracies such that the highest accuracy is obtained given a fixed computational budget. Here, we consider a fully adaptive greedy approach to this problem. Using Gaussian process regression as surrogate, samples are drawn from the available posterior approximation while designs are incrementally defined by solving a sequence of optimization problems for evaluation accuracy and positions. The selection of training designs is tailored towards representing the posterior to be sampled as good as possible, while the interleaved sampling steps discard old inaccurate samples in favor of new, more accurate ones. Numerical results show a significant reduction of the computational effort compared to just position-adaptive and static designs. KW - Adaptive Gaussian Processes KW - Bayesian parameter identification KW - AGP PY - 2025 UR - https://nbn-resolving.org/urn:nbn:de:kobv:b43-652731 DO - https://doi.org/10.48550/arXiv.2411.17858 SP - 1 EP - 23 PB - arXiv.org AN - OPUS4-65273 LA - eng AD - Bundesanstalt fuer Materialforschung und -pruefung (BAM), Berlin, Germany ER - TY - CONF A1 - Unger, Jörg F. T1 - Methodological Prerequisites for Reliable Simulation Results in the Virtual Lab and Digital Twins N2 - Simulationen übernehmen zunehmend eine zentrale Rolle in sicherheitskritischen Entscheidungsprozessen, etwa im Bauwesen, bei digitalen Zwillingen oder in der prädiktiven Instandhaltung. Damit diese Entscheidungen auf zuverlässigen numerischen Modellen basieren können, müssen die zugrunde liegenden Teilmodelle eindeutig beschrieben, reproduzierbar und in der Community validiert sein. Dieses Manuskript skizziert die notwendigen Voraussetzungen für vertrauenswürdige, FAIR-konforme und entscheidungsfähige Simulationsprozesse. Zunächst wird die Notwendigkeit einer formalen, softwareunabhängigen Beschreibung mathematischer Modelle und numerischer Implementierungen diskutiert. Standards wie VMAP, Ontologien wie MathModDB sowie BIM- und STEP-Formate bieten erste Ansätze zur semantischen Modellbeschreibung und zur automatisierten Generierung simulationsfähiger Eingabedaten. Im zweiten Teil wird die Rolle experimenteller Daten für die Validierung von Modellen beleuchtet. Es wird gezeigt, dass hochwertige, strukturierte und maschinenlesbare Validierungsdatensätze essenziell sind, insbesondere für die Bewertung konstitutiver Modelle. Ein weiteres Kapitel widmet sich der Unsicherheitsquantifizierung. Neben klassischen Fehlermaßen werden aleatorische und epistemische Unsicherheiten sowie deren Einfluss auf die Modellbewertung behandelt. Besondere Aufmerksamkeit gilt der Bayes’schen Kalibrierung, der Trennung von Trainings und Testdaten und der Notwendigkeit, auch Begleitversuche aus Materialtests zu dokumentieren. Abschließend wird die Idee einer Benchmarking-Plattform vorgestellt, die auf reproduzierbaren Workflows, automatisierter Provenienzverfolgung und der Nutzung von Research Object (RO) Crates basiert. Diese Plattform erlaubt die dezentrale Veröffentlichung und zentrale Abfrage von Benchmark-Ergebnissen und fördert eine gemeinschaftliche Verifikation und Validierung. Ziel ist es, die Vergleichbarkeit von (Open-Source-)Simulationstools zu verbessern um dadurch komplexe experimentelle Setups zunehmend durch zuverlässige Simulationen ersetzen zu können. T2 - NAFEMS - Symbiose von Simulation und Test CY - Wiesbaden, Germany DA - 12.11.2025 KW - Verification and Validation, KW - Digital twins KW - Virtual lab KW - Uncertainty quantification PY - 2025 SN - 978-1-83979-242-7 SP - 23 EP - 28 PB - NAFEMS CY - Wiesbaden, Germany AN - OPUS4-65271 LA - eng AD - Bundesanstalt fuer Materialforschung und -pruefung (BAM), Berlin, Germany ER -