Filtern
Dokumenttyp
- Vortrag (9)
- Zeitschriftenartikel (3)
- Beitrag zu einem Tagungsband (2)
- Posterpräsentation (2)
Schlagworte
- Digital twin (16) (entfernen)
Organisationseinheit der BAM
- 7 Bauwerkssicherheit (8)
- 7.7 Modellierung und Simulation (5)
- 8 Zerstörungsfreie Prüfung (5)
- 3 Gefahrgutumschließungen; Energiespeicher (3)
- 3.1 Sicherheit von Gefahrgutverpackungen und Batterien (3)
- 8.2 Zerstörungsfreie Prüfmethoden für das Bauwesen (3)
- 7.1 Baustoffe (2)
- 7.2 Ingenieurbau (2)
- 1 Analytische Chemie; Referenzmaterialien (1)
- 1.4 Prozessanalytik (1)
Paper des Monats
- ja (1)
Eingeladener Vortrag
- nein (9)
In materials and component research, artificial intelligence methodologies will lead to massive upheavals in the coming years. The processes of material development, material processing, lifetime prediction and material characterization will change significantly. By combining AI methods and new forms of knowledge representation, the data-based management of product life cycles will take on new qualities. To address this emerging field of research Fraunhofer IWM set up the online workshop »AI Methods for Fatigue Behavior Assessment and Component Lifetime Prediction«
Numerical models are an essential tool in predicting and monitoring the behavior of civil structures. Inferring the model parameters is a challenging tasks as they are often measured indirectly and are affected by uncertainties. Digital twins couple those models with real-world data and can introduce additional, systematic sensor uncertainties related to the sensor calibration, i.e. uncertain offsets and calibration factors.
In this work, the challenges of data processing, parameter identification, model selection and damage detection are explored using a lab-scale cable stayed bridge demonstrator. By combining force measurements in the cables with displacement measurements from both laser and stereo-photogrammetry systems, the elastic parameters of a three-dimensional finite element beam model are inferred.
Depending on the number of sensors and the number of datasets used, parametrizing the sensor offsets and factors, leads to model with over 100 parameters. With a real-time solution of the problem in mind, a highly efficient analytical variational Bayesian approach is used to solve it within seconds. An analysis of the required assumptions and limitations of the approach, especially w.r.t. to the computed evidence, is provided by a comparison with dynamic nested sampling in a simplified problem.
Finally, by inferring the value of additional damage parameters along the bridge, the method is successfully used to detect the location of an artificially introduced weak spot in the demonstrator bridge.
The quality of a model - and thus its predictive capabilities - is influenced by numerous uncertainties. They include possibly unknown boundary and initial conditions, noise in the data used for its calibration and uncertainties in the model itself. Here, the latter part is not only restricted to uncertain model parameters, but also refers to the choice of the model itself. Inferring these uncertainties in an automatic way allows for an adaption of the model to new data sets and for a reliable, reproducible model assessment. Note that similar concepts apply at the structural level, where a continuously updated digital twin allows virtual measurements at inaccessible positions of the structure and a simulation based lifetime prediction.
This work presents an inference workflow that describes the difference of measured data and simulated model responses with a generic interface that is independent from the specific model or even the geometry and can easily incorporate multiple data sources. A variational Bayesian inference algorithm is then used to a) calibrate a set of models to given data and to b) identify the best fitting one. The developed concepts are applied to a bridge Demonstrator equipped with displacement sensors, force sensors and a stereophotogrammetry system to perform a system identification of the material parameters as well as a real-time identification of a moving load.
Gradual or sudden changes in the state of structural systems caused, for example, by deterioration or accidental load events can influence their load-bearing capacity. Structural changes can be inferred from static and/or dynamic response data measured by structural health monitoring systems. However, they may be masked by variations in the structural response due to varying environmental conditions. Particularly, the interaction of nominally load-bearing components with nominally non-load bearing components exhibiting characteristics that vary as a function of the environmental conditions can significantly affect the monitored structural response. Ignoring these effects may hamper an inference of structural changes from the monitoring data. To address this issue, we adopt a probabilistic model-based framework as a basis for developing digital twins of structural systems that enable a prediction of the structural behavior under varying ambient condition. Within this framework, different types of data obtained from real the structural system can be applied to update the digital twin of the structural system using Bayesian methods and thus enhance predictions of the structural behavior. In this contribution, we implement the framework to develop a digital twin of a simply supported steel beam with an asphalt layer. It is formulated such that it can predict the static response of the beam in function of its temperature. In a climate chamber, the beam was subject to varying temperatures and its static response wass monitored. In addition, tests are performed to determine the temperature-dependent properties of the asphalt material. Bayesian system identification is applied to enhance the predictive capabilities of the digital twin based on the observed data.
Gradual or sudden changes in the state of structural systems caused, for example, by deterioration or accidental load
events can influence their load-bearing capacity. Structural changes can be inferred from static and/or dynamic response data
measured by structural health monitoring systems. However, they may be masked by variations in the structural response due to
varying environmental conditions. Particularly, the interaction of nominally load-bearing components with nominally non-load
bearing components exhibiting characteristics that vary as a function of the environmental conditions can significantly affect the
monitored structural response. Ignoring these effects may hamper an inference of structural changes from the monitoring data. To
address this issue, we adopt a probabilistic model-based framework as a basis for developing digital twins of structural systems
that enable a prediction of the structural behavior under varying ambient condition. Within this framework, different types of data
obtained from real the structural system can be applied to update the digital twin of the structural system using Bayesian methods
and thus enhance predictions of the structural behavior. In this contribution, we implement the framework to develop a digital
twin of a simply supported steel beam with an asphalt layer. It is formulated such that it can predict the static response of the beam
in function of its temperature. In a climate chamber, the beam was subject to varying temperatures and its static response wass
monitored. In addition, tests are performed to determine the temperature-dependent properties of the asphalt material. Bayesian
system identification is applied to enhance the predictive capabilities of the digital twin based on the observed data.
Digital twin for hydrogen refueling stations: An analytical study for safe and efficient operations
(2023)
This analytical overview is the initial study in developing the digital sensors twin for the hydrogen refueling station. The research includes a comprehensive analysis of the current literature, refueling protocol, standards, mathematical model, simulation model, and trends in digital twins and H2 technologies. Furthermore, this study has explored the various HRS attributes, such as leakage, temperature, pressure, and hydrogen flow, and identified different use cases of the hydrogen fuel station to build a digital sensor twin. Various real gas equations, such as the Van der Waals and Able-Noble equation of state (EOS), are applied to analyze the key factors and parameters affecting hydrogen and the MATLAB simulation model to validate the key elements and parameters. This study also focuses on identifying and addressing research gaps in hydrogen technology, contributing to advancing hydrogen infrastructure, and supporting the global transition to cleaner energy solutions.
The main challenge in using numerical models as digital twins in real applications for prognosis purposes, such as reliability analysis, is the calibration and validation of the models based on uncertain measurement data. Uncertainties are not limited to the measurement data, but the numerical model itself will not be perfect due to the modelling assumptions.
In this contribution, a probabilistic inference method for model calibration, based on the Bayes’ Theorem, is used to face that issue. Such inference approaches include uncertainties on the data as well as on the model parameters, allowing to compute an a posteriori distribution for the model parameters as well as a noise term reflecting the measured data. However, such probabilistic inference methods require a lot of evaluations of the numerical forward model for different model parameters. An improvement of the efficiency is obtained by replacing the forward model with a reduced model. Model reduction, e.g. the proper generalized decomposition (PGD) method, is a popular concept to decrease the computational effort, where each evaluation of the reduced forward model is a pure less costly function evaluation.
The heterogeneous spatial distribution of material parameters in the forward model is described by a lognormal random field. This allows identifying a variable stiffness over the spatial directions by identifying the random field variables with given measurement data. These changes can e.g. be caused by damage. The lognormal field is approximated as series expansion for the PGD problem.
The derived efficient model identification procedure is shown using a real reinforced prestress demonstrator bridge and stereophotogrammetry measurement data. A digital twin for that demonstrator bridge is build up using a set of measurement data and verified by testing additional measurement data. PGD model error against the FEM model is discussed based on an importance sampling analysis computing the Bayes Factor.
The ability of industrial X-ray computed tomography (CT) to scan an object with several internal and external features at once causes increasing adoption in dimensional metrology. In order to evaluate the quality of a measurement value, the task-specific measurement uncertainty has to be determined. Currently, VDI/VDE 2630 part 2.1 gives a guideline to determine the uncertainty of CT measurements experimentally by conducting repeated measurements. This is costly and time-consuming. Thus, the aim is to determine the task-specific measurement uncertainty numerically by simulations (e. g. according to the guide to expression of uncertainty in measurement (GUM) Supplement 1). To achieve that, a digital twin is necessary. This contribution presents a simple first approach how a digital twin can be built. In order to evaluate this approach, a study comparing measurements and simulations of different real CT systems was carried out by determining the differences between the measurement results of the digital twin and of the measurement results of the real-world CT systems. The results have shown a moderate agreement between real and simulated data. To improve on this aspect, a standardized method to characterize CT systems and methods to implement CT parameters into the simulation with sufficient accuracy will be developed.
To provide better means for a safe and effective monitoring of cemented waste packages including prediction tools to assess the future integrity development during pre-disposal activities, several digital tools are evaluated and improved in the frame of the EC funded project PREDIS. Safety enhancement (e. g. less exposure of testing personnel) and cost effectiveness are part of the intended impact.
The work includes but is not limited to inspection methods such as muon imaging, wireless sensors integrated into waste packages as well as external package and facility monitoring such as remote fiber optical sensors. The sensors applied will go beyond radiation monitoring and include Proxy parameters important for long term integrity assessment (e. g. internal pressure). Sensors will also be made cost effective to allow the installation of much more sensors compared to current practice.
The measured data will be used in digital twins of the packages for specific simulations (geochemical, integrity) providing a prediction of future behavior. Machine Learning techniques trained by the characterization of older packages will help to connect the models to the actual data.
All data (measured and simulated) will be collected in a joint data base and connected to a decision framework to be used at actual facilities.
The presentation includes detailed information about the various tools under consideration, their connection and first results of our research.
Multifaceted developments for pre-disposal management of low and intermediate level radioactive waste are undertaken in the EC funded project PREDIS. In work package 7, innovations in cemented waste handling and pre-disposal storage are advanced by testing and evaluating. To provide better means for safe and effective monitoring of cemented waste packages including prediction tools to assess the future integrity development during pre-disposal activities, several monitoring and digital tools are evaluated and improved. Both safety enhancement (e. g. less exposure of testing personnel) and cost effectiveness are part of the intended impact. Current methods to pack, store, and monitor cemented wastes are identified, analysed and improved. Innovative integrity testing and monitoring techniques applied to evaluate and demonstrate package and storage quality assurance are further developed. The work includes but is not limited to inspection methods such as muon imaging, wireless sensors integrated into waste packages as well as external package and facility monitoring such as remote fiber optical sensors. The sensors applied will go beyond radiation monitoring and include proxy parameters important for long term integrity assessment (e. g. internal pressure). The measured data will be used in digital twins of the packages for specific simulations (geochemical, integrity) providing a prediction of future behaviour. Machine Learning techniques trained by the characterization of older packages will help to connect the models to the actual data. As data handling, processing and fusion are crucial for both the monitoring and the digital twin model, all data (measured and simulated) will be collected in a joint data base and connected to a decision framework. Finally, the implementation of the improved techniques will be tested at actual facilities. An overview about various relevant tools, their interconnections, and first research results will be shown.