Download or read book Data driven and Physics constrained Uncertainty Quantification for Turbulence Models written by Jan Felix Heyse and published by . This book was released on 2022 with total page pages. Available in PDF, EPUB and Kindle. Book excerpt: Numerical simulations are an important tool for prediction of turbulent flows. Today, most simulations in real-world applications are Reynolds-averaged Navier-Stokes (RANS) simulations, which average the governing equations to solve for the mean flow quantities. RANS simulations require modeling of an unknown quantity, the Reynolds stress tensor, using turbulence models. These models are limited in their accuracy for many complex flows, such as those involving strong stream-line curvature or adverse pressure gradients, making RANS predictions less reliable for design decisions. For RANS predictions to be useful in engineering design practice, it is therefore important to quantify the uncertainty in the predictions. More specifically, in this dissertation the focus is on quantifying the model-form uncertainty associated with the turbulence model. A data-free eigenperturbation framework introduced in the past few years, allows to make quantitative uncertainty estimates for all quantities of interest. It relies on a linear mapping from the eigenvalues of the Reynolds stress into the barycentric domain. In this framework, perturbations are added to the eigenvalues in that barycentric domain by perturbing them towards limiting states of 1 component, 2 component, and 3 component turbulence. Eigenvectors are permuted to find the extreme states of the turbulence kinetic energy production term. These eigenperturbations allow to explore a range of shapes and alignments of the Reynolds stress tensor within constraints of physical realizability of the resulting Reynolds stresses. However, this framework is limited by the introduction of a uniform amount of perturbation throughout the domain and by the need to specify a parameter governing the amount of perturbation. Data-driven eigenvalue perturbations are therefore introduced in this work to address those limitations. They are built on the eigenperturbation framework, but use a data-driven approach to determine how much perturbation to impose locally at every cell. The target amount of perturbation is the expected distance between the RANS prediction and the true solution in the barycentric domain. A general set of features is introduced, computed from the RANS mean flow quantities. The periodic flow over a wavy wall (for which also a detailed high-fidelity simulation dataset is available) serves as training case. A random forest machine learning model is trained to predict the target distance from the features. A hyperparameter study is carried out to find the most appropriate hyperparameters for the random forest. Random forest feature importance estimates confirm general expectations from physical intuition. The framework is applied to two test cases, the flow over a backward-facing step and the flow in an asymmetric diffuser. Both test cases and the training case exhibit a flow separation where the cross sectional area increases. The distribution of key features is studied for these cases and compared against the one from the training case. It is found that the random forest is not extrapolating. The results on the two test cases show uncertainty estimates that are characteristic of the true error in the predictions and give more representative bounds than the data-free framework does. The sets of eigenvectors from the RANS prediction and the true solution can be connected through a rotation. The idea of data-driven eigenvector rotations as a data-driven extension to the eigenvectors is studied. However, continuousness of the prediction targets is not generally achievable because of the ambiguity of the eigenvector direction. The lack of smoothness prevents the machine learning models from learning the relationship between the features and the targets, making data-driven eigenvector rotations in the discussed setup not practical. The last chapter of this dissertation introduces a data-driven baseline simulation, which corresponds to the expected value in the data-driven eigenvalue perturbation framework. The Reynolds stress is a weighted sum of the Reynolds stresses from the extreme states. A random classification forest trained to predict which extreme state is closest to the true Reynolds stress is used to compute these weights. It does so by giving a probabilistic meaning to the raw predictions of the constituent decision trees. On the test cases, the data-driven baseline predictions are similar but not equal to the data-free baseline. They complement the uncertainty estimates from the data-driven eigenvalue perturbations.
Download or read book Data driven Approach for Turbulence Modeling in Rotating Flows and Stratified Flows written by Xinyi Huang and published by . This book was released on 2023 with total page 0 pages. Available in PDF, EPUB and Kindle. Book excerpt: Turbulence modeling, including wall models in large-eddy simulations (LESs) and RANS models in Reynolds-averaged Navier-Stokes (RANS) simulations, is usually not considered for non-canonical flows, including rotating flows and stratified flows. Modeling non-canonical flows encounters difficulties. Some of the main difficulties lie in the fact that these flows have multiple flow controlling parameters (FCPs), and thus, the flow behavior is hard to explore, let alone get accurate modeling. The data-driven approach is considered a possible solution to this. The increasing computational resources and shared turbulence data allow another way to utilize the data other than pure human analyses of the physics. However, pure data-driven methods are often criticized for their weak interpretability and generalizability. In this work, multiple data-driven techniques are applied to some persistent problems in turbulence modeling under the circumstances of rotating flows and stratified flows. The problems include not only the accurate modeling of the flow but also the efficient FCP space exploration, model selection, uncertainty quantification, etc. Both the dataset and existing knowledge of physics are utilized, and then data-driven approach shows the interpretability and generalizability. They show how these traditionally difficult problems can be tackled through physics-informed data-driven approach, which significantly saves human labor. To be more specific, data-driven approach to wall modeling is compared to physics-based approach for a spanwise rotating channel, and it shows a more accurate yet still generalizable behavior. When modeling is extended to an arbitrarily directional rotating channel, a surrogate model is efficiently developed through the utilization of Bayesian optimization, when such behavior is never understood in the existing literature. Data-driven approach is also applied to RANS modeling. The diverse modeling makes model selection awkward for a newbie, and we train a recommender system to provide guidelines. Modeling itself for non-canonical cases, e.g., stratified flows, is also troublesome, because the multi-stage behavior of the flow requires automated switching of modeling between different models as the flow develops. A linear logistic regression is developed for automating the classification. The models can then be evaluated through a global epistemic uncertainty quantification (UQ) method, which allows the exploration of dominating terms in a RANS model and determining a priori if a calibration can generalize to other flow conditions. In general, data-driven approach has been used for multiple applications in turbulence modeling, and they show their capability and interpretability.
Download or read book Physics based Uncertainty Quantification of Reynolds averaged navier stokes Models for Turbulent Flows and Scalar Transport written by Zengrong Hao and published by . This book was released on 2020 with total page pages. Available in PDF, EPUB and Kindle. Book excerpt: Numerical simulations for turbulent flows and scalar (e.g. temperature, concentration and humidity) transport is one of the most challenging topics in urban wind engineering. For the design and optimization of configurations in cities, the Reynolds-averaged-Navier-Stokes (RANS) method for turbulence modeling has evident superiority over the turbulence-resolving methods (e.g. directly-numerical-simulation (DNS), large-eddy-simulation (LES), or RANS-LES hybrid approaches) in terms of efficiency and robustness. However, because "all models are wrong" (Box (1976)), the predictions of a RANS simulation always have uncertainties that originate in the inherent inadequacies of various physical hypotheses in the RANS models. To quantify these model uncertainties is not only significant for improving the practicability of RANS method in wind engineering, but also potentially help us understand the physics of turbulence in a broader sense. The objective of this thesis is to develop physics-based, data-free methods for RANS model uncertainty quantification (UQ) in engineering turbulent flows and scalar transport. These UQ methods are expected to estimate the appropriate bounds of quantities of interest (QoIs) at the cost of O(10) or fewer individual steady RANS simulations without any a priori data. The development of each method generally follows two principles: i) relaxing a well-established baseline model to address some inherent inadequacies in its physical assumptions; and ii) perturbing the released degrees-of-freedom (DOFs) based on some conceptual "limiting conditions" in physics. The studies of UQ methodologies in this thesis are divided into four separate parts as follows, of which Parts I and II are on the models for Reynolds stress, and Parts III and IV on the models for scalar flux. Part I addresses the uncertainty in the linear-eddy-viscosity (LEV) assumption that results in incorrect shape and orientation of Reynolds stress. This part directly applies the method previously proposed by Emory et al. (2013) and Gorlé et al. (2012), named Reynolds-stress-shape-perturbation (RSSP), to examine its bounding behaviors for QoIs in complex problems. The investigation reveals that the RSSP method's incapability in bounding the turbulence-related QoIs in separation and backflow regions essentially does not originates in the LEV assumption but in the dissipation determination. Part II proposes the double-scale double-LEV (DSDL) model to address the uncertainty in the energy dissipation determination, which specifically overpredicts the dissipation rates in the turbulence with vortex shedding behind bluff bodies. The model uncertainty is represented by one or two uncertain parameters that roughly indicate the intensity of the interaction between coherent structures and stochastic turbulence. The applications of the DSDL model in several problems show promising performance in terms of bounding the turbulent energies behind bluff bodies and meanwhile maintaining appropriate mean-flow predictions. Part III proposes the one-equation (OE) method to quantify the uncertainty in scalar flux models. The method is designed from the perspective of ordinary vector field, aiming at optimizing the local productions of scalar flux magnitudes. It shows some favorable bounding behaviors for scalar-related QoIs, although the ignorance of uncertainty in the modeled pressure-scrambling effect limits its performance to some extent. Alternative to OE, Part IV proposes the pressure-scrambling-perturbation (PSP) method for scalar flux model UQ by addressing the uncertainty in the pressure-scrambling effect in scalar flux dynamics. It is based on two conceptual "limits" for the pressure-scrambling directions indicated by two classical phenomenological theories. The PSP method exhibits superior bounding behaviors over the OE method for the cases in this thesis. The works in this thesis are expected to contribute to the physical foundations of both the data-free and data-driven approaches for RANS model UQ.
Download or read book Data Driven Computational Methods written by John Harlim and published by Cambridge University Press. This book was released on 2018-07-12 with total page 171 pages. Available in PDF, EPUB and Kindle. Book excerpt: Describes computational methods for parametric and nonparametric modeling of stochastic dynamics. Aimed at graduate students, and suitable for self-study.
Download or read book Data Driven Analysis and Modeling of Turbulent Flows written by Karthik Duraisamy and published by Academic Press. This book was released on 2025-03-01 with total page 0 pages. Available in PDF, EPUB and Kindle. Book excerpt: Data-driven Analysis and Modeling of Turbulent Flows explains methods for the analysis of large fields of data, and uncovering models and model improvements from numerical or experimental data on turbulence. Turbulence simulations generate large data sets, and the extraction of useful information from these data fields is an important and challenging task. Statistical learning and machine learning have provided many ways of helping, and this book explains how to use such methods for extracting, treating, and optimizing data to improve predictive turbulence models. These include methods such as POD, SPOD and DMD, for the extraction of modes peculiar to the data, as well as several reduced order models. This resource is essential reading for those developing turbulence models, performing turbulence simulations or interpreting turbulence simulation results.
Download or read book Model Reduction and Approximation written by Peter Benner and published by SIAM. This book was released on 2017-07-06 with total page 421 pages. Available in PDF, EPUB and Kindle. Book excerpt: Many physical, chemical, biomedical, and technical processes can be described by partial differential equations or dynamical systems. In spite of increasing computational capacities, many problems are of such high complexity that they are solvable only with severe simplifications, and the design of efficient numerical schemes remains a central research challenge. This book presents a tutorial introduction to recent developments in mathematical methods for model reduction and approximation of complex systems. Model Reduction and Approximation: Theory and Algorithms contains three parts that cover (I) sampling-based methods, such as the reduced basis method and proper orthogonal decomposition, (II) approximation of high-dimensional problems by low-rank tensor techniques, and (III) system-theoretic methods, such as balanced truncation, interpolatory methods, and the Loewner framework. It is tutorial in nature, giving an accessible introduction to state-of-the-art model reduction and approximation methods. It also covers a wide range of methods drawn from typically distinct communities (sampling based, tensor based, system-theoretic).?? This book is intended for researchers interested in model reduction and approximation, particularly graduate students and young researchers.
Download or read book Predicting Structured Data written by Neural Information Processing Systems Foundation and published by MIT Press. This book was released on 2007 with total page 361 pages. Available in PDF, EPUB and Kindle. Book excerpt: State-of-the-art algorithms and theory in a novel domain of machine learning, prediction when the output has structure.
Download or read book Handbook of Uncertainty Quantification written by Roger Ghanem and published by Springer. This book was released on 2016-05-08 with total page 0 pages. Available in PDF, EPUB and Kindle. Book excerpt: The topic of Uncertainty Quantification (UQ) has witnessed massive developments in response to the promise of achieving risk mitigation through scientific prediction. It has led to the integration of ideas from mathematics, statistics and engineering being used to lend credence to predictive assessments of risk but also to design actions (by engineers, scientists and investors) that are consistent with risk aversion. The objective of this Handbook is to facilitate the dissemination of the forefront of UQ ideas to their audiences. We recognize that these audiences are varied, with interests ranging from theory to application, and from research to development and even execution.
Download or read book Data Driven Uncertainty Quantification for Large Scale Simulations written by Fabian Franzelin and published by . This book was released on 2018 with total page pages. Available in PDF, EPUB and Kindle. Book excerpt:
Download or read book Advances in Electromagnetics Empowered by Artificial Intelligence and Deep Learning written by Sawyer D. Campbell and published by John Wiley & Sons. This book was released on 2023-08-03 with total page 596 pages. Available in PDF, EPUB and Kindle. Book excerpt: Advances in Electromagnetics Empowered by Artificial Intelligence and Deep Learning Authoritative reference on the state of the art in the field with additional coverage of important foundational concepts Advances in Electromagnetics Empowered by Artificial Intelligence and Deep Learning presents cutting-edge research advances in the rapidly growing areas in optical and RF electromagnetic device modeling, simulation, and inverse-design. The text provides a comprehensive treatment of the field on subjects ranging from fundamental theoretical principles and new technological developments to state-of-the-art device design, as well as examples encompassing a wide range of related sub-areas. The content of the book covers all-dielectric and metallodielectric optical metasurface deep learning-accelerated inverse-design, deep neural networks for inverse scattering, applications of deep learning for advanced antenna design, and other related topics. To aid in reader comprehension, each chapter contains 10-15 illustrations, including prototype photos, line graphs, and electric field plots. Contributed to by leading research groups in the field, sample topics covered in Advances in Electromagnetics Empowered by Artificial Intelligence and Deep Learning include: Optical and photonic design, including generative machine learning for photonic design and inverse design of electromagnetic systems RF and antenna design, including artificial neural networks for parametric electromagnetic modeling and optimization and analysis of uniform and non-uniform antenna arrays Inverse scattering, target classification, and other applications, including deep learning for high contrast inverse scattering of electrically large structures Advances in Electromagnetics Empowered by Artificial Intelligence and Deep Learning is a must-have resource on the topic for university faculty, graduate students, and engineers within the fields of electromagnetics, wireless communications, antenna/RF design, and photonics, as well as researchers at large defense contractors and government laboratories.
Download or read book Machine Learning in Modeling and Simulation written by Timon Rabczuk and published by Springer Nature. This book was released on 2023-11-04 with total page 456 pages. Available in PDF, EPUB and Kindle. Book excerpt: Machine learning (ML) approaches have been extensively and successfully employed in various areas, like in economics, medical predictions, face recognition, credit card fraud detection, and spam filtering. There is clearly also the potential that ML techniques developed in Engineering and the Sciences will drastically increase the possibilities of analysis and accelerate the design to analysis time. With the use of ML techniques, coupled to conventional methods like finite element and digital twin technologies, new avenues of modeling and simulation can be opened but the potential of these ML techniques needs to still be fully harvested, with the methods developed and enhanced. The objective of this book is to provide an overview of ML in Engineering and the Sciences presenting fundamental theoretical ingredients with a focus on the next generation of computer modeling in Engineering and the Sciences in which the exciting aspects of machine learning are incorporated. The book is of value to any researcher and practitioner interested in research or applications of ML in the areas of scientific modeling and computer aided engineering.
Download or read book Bayesian Reinforcement Learning written by Mohammad Ghavamzadeh and published by . This book was released on 2015-11-18 with total page 146 pages. Available in PDF, EPUB and Kindle. Book excerpt: Bayesian methods for machine learning have been widely investigated, yielding principled methods for incorporating prior information into inference algorithms. This monograph provides the reader with an in-depth review of the role of Bayesian methods for the reinforcement learning (RL) paradigm. The major incentives for incorporating Bayesian reasoning in RL are that it provides an elegant approach to action-selection (exploration/exploitation) as a function of the uncertainty in learning, and it provides a machinery to incorporate prior knowledge into the algorithms. Bayesian Reinforcement Learning: A Survey first discusses models and methods for Bayesian inference in the simple single-step Bandit model. It then reviews the extensive recent literature on Bayesian methods for model-based RL, where prior information can be expressed on the parameters of the Markov model. It also presents Bayesian methods for model-free RL, where priors are expressed over the value function or policy class. Bayesian Reinforcement Learning: A Survey is a comprehensive reference for students and researchers with an interest in Bayesian RL algorithms and their theoretical and empirical properties.
Download or read book Data Driven Science and Engineering written by Steven L. Brunton and published by Cambridge University Press. This book was released on 2022-05-05 with total page 615 pages. Available in PDF, EPUB and Kindle. Book excerpt: A textbook covering data-science and machine learning methods for modelling and control in engineering and science, with Python and MATLAB®.
Download or read book Turbulence Modelling Approaches written by Konstantin Volkov and published by BoD – Books on Demand. This book was released on 2017-07-26 with total page 252 pages. Available in PDF, EPUB and Kindle. Book excerpt: Accurate prediction of turbulent flows remains a challenging task despite considerable work in this area and the acceptance of CFD as a design tool. The quality of the CFD calculations of the flows in engineering applications strongly depends on the proper prediction of turbulence phenomena. Investigations of flow instability, heat transfer, skin friction, secondary flows, flow separation, and reattachment effects demand a reliable modelling and simulation of the turbulence, reliable methods, accurate programming, and robust working practices. The current scientific status of simulation of turbulent flows as well as some advances in computational techniques and practical applications of turbulence research is reviewed and considered in the book.
Download or read book Knowledge Guided Machine Learning written by Anuj Karpatne and published by CRC Press. This book was released on 2022-08-15 with total page 442 pages. Available in PDF, EPUB and Kindle. Book excerpt: Given their tremendous success in commercial applications, machine learning (ML) models are increasingly being considered as alternatives to science-based models in many disciplines. Yet, these "black-box" ML models have found limited success due to their inability to work well in the presence of limited training data and generalize to unseen scenarios. As a result, there is a growing interest in the scientific community on creating a new generation of methods that integrate scientific knowledge in ML frameworks. This emerging field, called scientific knowledge-guided ML (KGML), seeks a distinct departure from existing "data-only" or "scientific knowledge-only" methods to use knowledge and data at an equal footing. Indeed, KGML involves diverse scientific and ML communities, where researchers and practitioners from various backgrounds and application domains are continually adding richness to the problem formulations and research methods in this emerging field. Knowledge Guided Machine Learning: Accelerating Discovery using Scientific Knowledge and Data provides an introduction to this rapidly growing field by discussing some of the common themes of research in KGML using illustrative examples, case studies, and reviews from diverse application domains and research communities as book chapters by leading researchers. KEY FEATURES First-of-its-kind book in an emerging area of research that is gaining widespread attention in the scientific and data science fields Accessible to a broad audience in data science and scientific and engineering fields Provides a coherent organizational structure to the problem formulations and research methods in the emerging field of KGML using illustrative examples from diverse application domains Contains chapters by leading researchers, which illustrate the cutting-edge research trends, opportunities, and challenges in KGML research from multiple perspectives Enables cross-pollination of KGML problem formulations and research methods across disciplines Highlights critical gaps that require further investigation by the broader community of researchers and practitioners to realize the full potential of KGML
Download or read book Modeling Complex Turbulent Flows written by Manuel D. Salas and published by Springer Science & Business Media. This book was released on 2012-12-06 with total page 385 pages. Available in PDF, EPUB and Kindle. Book excerpt: Turbulence modeling both addresses a fundamental problem in physics, 'the last great unsolved problem of classical physics,' and has far-reaching importance in the solution of difficult practical problems from aeronautical engineering to dynamic meteorology. However, the growth of supercom puter facilities has recently caused an apparent shift in the focus of tur bulence research from modeling to direct numerical simulation (DNS) and large eddy simulation (LES). This shift in emphasis comes at a time when claims are being made in the world around us that scientific analysis itself will shortly be transformed or replaced by a more powerful 'paradigm' based on massive computations and sophisticated visualization. Although this viewpoint has not lacked ar ticulate and influential advocates, these claims can at best only be judged premature. After all, as one computational researcher lamented, 'the com puter only does what I tell it to do, and not what I want it to do. ' In turbulence research, the initial speculation that computational meth ods would replace not only model-based computations but even experimen tal measurements, have not come close to fulfillment. It is becoming clear that computational methods and model development are equal partners in turbulence research: DNS and LES remain valuable tools for suggesting and validating models, while turbulence models continue to be the preferred tool for practical computations. We believed that a symposium which would reaffirm the practical and scientific importance of turbulence modeling was both necessary and timely.
Download or read book Clouds and Climate written by A. Pier Siebesma and published by Cambridge University Press. This book was released on 2020-08-20 with total page 421 pages. Available in PDF, EPUB and Kindle. Book excerpt: Comprehensive overview of research on clouds and their role in our present and future climate, for advanced students and researchers.
