Article
Thermodynamics
Xinzi Tang, Keren Yuan, Nengwei Gu, Pengcheng Li, Ruitao Peng
Summary: This study presents a novel approach for uncertain analysis and aerodynamic robustness optimization of wind turbine airfoil considering the uncertainties of turbulence and geometric errors. An interval method coupled with the Kriging model is used to quantify the uncertain influence and integrate it in the optimization process. The optimized airfoil shows a reduced fluctuation range and maintains the average lift to drag ratio, demonstrating its robustness against uncertainties.
Article
Mechanics
Minghan Chu, Xiaohua Wu, David E. E. Rival
Summary: This study utilizes a physics-based uncertainty quantification framework to estimate the uncertainty in Reynolds-averaged Navier-Stokes models by introducing eigenvalue, eigenvector, and turbulence kinetic energy perturbations. A regression-based marker function is introduced for simulating laminar-turbulent transitional flows, and a monotonic behavior of the predicted uncertainty bounds is observed with respect to turbulence kinetic energy perturbation. The predicted uncertainty bounds show a synergy behavior when eigenvalue perturbations are augmented with the marker function, significantly increasing the size of the uncertainty bounds.
Article
Engineering, Industrial
Zihan Wang, Mohamad Daeipour, Hongyi Xu
Summary: This paper proposes a new methodology to quantify and propagate aleatoric uncertainties distributed in complex topological structures. It introduces a random field-based uncertainty representation approach that captures the topological characteristics using the shortest interior path distance. Parameterization methods and non-intrusive uncertainties propagation methods are employed to propagate the uncertainties. Engineering case studies demonstrate the effectiveness of the proposed methodology.
RELIABILITY ENGINEERING & SYSTEM SAFETY
(2023)
Article
Polymer Science
Mathew C. Celina, Erik Linde, Estevan Martinez
Summary: Carbonyl formation is a key indicator for oxidative processes and degradation of plastics, with infrared spectroscopy being a commonly used method for identification and quantification. However, there are significant error margins in carbonyl quantification, with various factors influencing the accuracy of measurements.
POLYMER DEGRADATION AND STABILITY
(2021)
Article
Engineering, Multidisciplinary
J. Parekh, R. W. C. P. Verstappen
Summary: Despite its limitations, Reynolds-averaged Navier-Stokes (RANS) based modeling remains the most recognized approach for computational fluid dynamics (CFD) applications. Quantification of model-form uncertainties in RANS has recently gained significant interest in the turbulence modeling community. A stochastic RANS solver with an efficient implementation of the intrusive polynomial chaos (IPC) method is presented in this study, which quantifies and propagates uncertainties associated with the output of the RANS model. The stochastic solver has shown better performance than traditional uncertainty quantification (UQ) methods in benchmark problems.
INTERNATIONAL JOURNAL FOR UNCERTAINTY QUANTIFICATION
(2023)
Article
Engineering, Multidisciplinary
Zhengyuan Mao, Linna Fan, Pinliang Dong
Summary: This article addresses the transmission problem of distance uncertainties and proposes a new method for measuring distance uncertainties, which is more efficient and robust than traditional Monte Carlo methods.
Article
Mechanics
Yu-Eop Kang, Sunwoong Yang, Kwanjung Yee
Summary: This study proposes a physics-aware reduced-order modeling (ROM) approach that utilizes interpretable and information-intensive latent variables (LVs) extracted by a beta-variational autoencoder. The independence and information intensity of these LVs are quantitatively scrutinized, and their physical meanings are thoroughly investigated. The proposed physics-aware ROM is compared with conventional ROMs, and its validity and efficiency are successfully verified.
Article
Mechanics
Yu-Eop Kang, Sunwoong Yang, Kwanjung Yee
Summary: This study proposes a physics-aware ROM that utilizes interpretable and information-intensive LVs, which correspond to the generating factors of the training dataset, Mach number, and angle of attack, as confirmed through quantitative analysis in a two-dimensional transonic flow benchmark problem.
Article
Geochemistry & Geophysics
Moritz O. Ziegler, Oliver Heidbach
Summary: The distance to failure of upper crustal rock in the prevalent stress field is important for understanding fault reactivation and managing georeservoirs. However, stress magnitude data for model calibration are sparse, leading to large model uncertainties. To reduce uncertainties, additional constraints on stress magnitudes are incorporated to assess the plausibility of different data-based stress states. A case study in southern Germany demonstrates the effectiveness of this approach in identifying plausible stress states and reducing model uncertainties.
JOURNAL OF GEOPHYSICAL RESEARCH-SOLID EARTH
(2023)
Article
Meteorology & Atmospheric Sciences
Maryam Ilyas, Douglas Nychka, Chris Brierley, Serge Guillas
Summary: Instrumental global temperature records are derived from in situ measurements of land and sea surface temperatures, with accuracy being crucial for climate science. Spatial gaps in the distribution of instrumental temperature measurements introduce coverage error. A new method is developed to quantify this coverage error while accounting for uncertainties in advanced spatial statistics model parameters.
ATMOSPHERIC MEASUREMENT TECHNIQUES
(2021)
Article
Engineering, Mechanical
Xiao-Wei Liu, Da-Gang Lu
Summary: There are two main challenges in fatigue load probability modeling. Firstly, it is difficult to measure the estimation errors, especially in the tails with low-probability and high-stress levels. Secondly, the component number of the mixture model cannot be observed. To address these challenges, this research introduces the hierarchical Bayesian mixture model and the Dirichlet process prior. A relative error measure is proposed to reveal the errors of the density tails. The results of an illustrative example show significant relative errors in the tails and a discrete distribution of the mixture number.
INTERNATIONAL JOURNAL OF FATIGUE
(2023)
Article
Engineering, Electrical & Electronic
Yan Pan, Tonghai Wu, Yunteng Jing, Pujian Wang
Summary: This study establishes a knowledge-guided three-layer model to characterize the multiattribute oil state. Data dispersion is considered by assigning fuzzy probability among the attribute layer. Improved evidential reasoning is used to solve inconsistent decisions. The effectiveness of the proposed approach is verified using real-world monitoring data.
IEEE TRANSACTIONS ON INSTRUMENTATION AND MEASUREMENT
(2022)
Article
Astronomy & Astrophysics
R. Ballhausen, T. R. Kallman, L. Gu, F. Paerels
Summary: Fitting plasma models to high-quality spectra is essential for studying the physical conditions of astrophysical sources. However, current models often fall short in describing the observed data. This study investigates the reasons behind the failure of fitting photoionized plasma models to high-resolution X-ray spectra and examines the effects of systematic uncertainties in atomic data and model uncertainties on the fitting results.
ASTROPHYSICAL JOURNAL
(2023)
Article
Mechanics
Ali Eidi, Navid Zehtabiyan-Rezaie, Reza Ghiassi, Xiang Yang, Mahdi Abkar
Summary: This study quantifies the model-form uncertainties in RANS simulations using a data-driven machine-learning technique. By applying a two-step feature-selection method and the extreme gradient boosting algorithm, more accurate representations of the Reynolds stress anisotropy are obtained. The proposed framework provides optimal estimation of uncertainty bounds for the RANS-predicted quantities of interest.
Review
Computer Science, Interdisciplinary Applications
Erdem Acar, Gamze Bayrak, Yongsu Jung, Ikjin Lee, Palaniappan Ramu, Suja Shree Ravichandran
Summary: This paper reviews the practices of uncertainty treatment in design optimization of structural and multidisciplinary systems under uncertainties, covering uncertainty modeling, uncertainty analysis, and design optimization methods incorporating various techniques.
STRUCTURAL AND MULTIDISCIPLINARY OPTIMIZATION
(2021)
Article
Computer Science, Interdisciplinary Applications
N. Guillaud, G. Balarac, E. Goncalves
COMPUTERS & FLUIDS
(2018)
Article
Mechanics
Jean Decaix, Matthieu Dreyer, Guillaume Balarac, Mohamed Farhat, Cecile Muench
EUROPEAN JOURNAL OF MECHANICS B-FLUIDS
(2018)
Article
Mechanics
Nicolas Odier, Guillaume Balarac, Christophe Corre
INTERNATIONAL JOURNAL OF MULTIPHASE FLOW
(2018)
Article
Green & Sustainable Science & Technology
N. Guillaud, G. Balarac, E. Goncalves, J. Zanette
Article
Mechanics
Aurelien Vadrot, Alexis Giauque, Christophe Corre
JOURNAL OF FLUID MECHANICS
(2020)
Article
Physics, Fluids & Plasmas
Anastasiia Gorbunova, Guillaume Balarac, Mickael Bourgoin, Leonie Canet, Nicolas Mordant, Vincent Rossetto
PHYSICAL REVIEW FLUIDS
(2020)
Article
Physics, Fluids & Plasmas
Hugo Frezat, Guillaume Balarac, Julien Le Sommer, Ronan Fablet, Redouane Lguensat
Summary: In this paper, a new strategy is presented to model subgrid-scale scalar flux in a three-dimensional turbulent incompressible flow using physics-informed neural networks (NNs). By incorporating classical transformation invariances and symmetries into the model as hard and soft constraints, the proposed model outperforms purely data-driven models and parametric state-of-the-art subgrid-scale models in simulation-based experiments. The considered invariances act as regularizers on physical metrics during prior evaluation, improving stability and performance of the model in large-eddy simulations and enabling generalization to unseen regimes.
PHYSICAL REVIEW FLUIDS
(2021)
Article
Mechanics
Alexis Giauque, Aurelien Vadrot, Paolo Errante, Christophe Corre
Summary: This study analyzes the influence of the physical complexity of real gases on the amplitude of subgrid-scale terms in large eddy simulations, finding that SGS turbulent stress and pressure terms are significant in the momentum equation. Additionally, significant SGS pressure work and fluxes are found in the inertial zone of the turbulent kinetic energy spectrum, highlighting the need for specific models in this area.
Article
Mechanics
Pedro Veras, Guillaume Balarac, Olivier Metais, Didier Georges, Antoine Bombenger, Claire Segoufin
Summary: The proposed approach utilizes machine learning techniques to reconstruct proper mean and fluctuating inlet boundary conditions based on known downstream flow information. It significantly improves numerical predictions compared to basic inlet boundary conditions, and successfully simulates the turbulent field in swirling flows inside a conical diffuser.
Article
Physics, Fluids & Plasmas
Anastasiia Gorbunova, Carlo Pagani, Guillaume Balarac, Leonie Canet, Vincent Rossetto
Summary: We conducted numerical simulations to study the spatiotemporal two-point correlation function of passively advected scalar fields in three-dimensional homogeneous isotropic turbulence. Our aim was to test analytical results obtained using functional renormalization group (FRG). The simulations showed decay of Eulerian correlations of the scalar as a Gaussian at small time delays and a predicted crossover to exponential decay at large time delays, which could not be observed due to numerical noise. We accurately confirmed FRG results by introducing finite time correlations in the synthetic velocity field and studying the crossover between two regimes.
PHYSICAL REVIEW FLUIDS
(2021)
Article
Mechanics
A. Grenouilloux, J. Leparoux, V. Moureau, G. Balarac, T. Berthelon, R. Mercier, M. Bernard, P. Benard, G. Lartigue, O. Metais
Summary: With the increasing computational power, Computational Fluid Dynamics (CFD) has become essential in complex industrial processes. Large-Eddy Simulation (LES) is a valuable tool for simulating complex unsteady flows, but faces limitations in mesh generation and 'time-to-solution'. This work proposes an automatic mesh definition procedure and a numerical technique to reduce the time required for LES. By addressing these challenges, the goal is to develop an accurate LES strategy at an optimized computational cost.
JOURNAL OF TURBULENCE
(2023)
Article
Mechanics
Thomas Berthelon, Guillaume Sahut, Julien Leparoux, Guillaume Balarac, Ghislain Lartigue, Manuel Bernard, Vincent Moureau, Olivier Metais
Summary: The strong increase in computational power has made it possible to use Large Eddy Simulation (LES) for industrial configurations. However, the time-to-solution is still too long for daily use in the design phases. This work aims to develop a new time integration method to reduce the time-to-solution of LES by allowing the use of larger time steps. The proposed method, based on the Backward Differentiation Formula (BDF) scheme, provides a linearized implicit time advancement that significantly reduces the solution time while maintaining accuracy.
JOURNAL OF TURBULENCE
(2023)
Article
Computer Science, Interdisciplinary Applications
Manuel Bernard, Ghislain Lartigue, Guillaume Balarac, Vincent Moureau, Guillaume Puigt
INTERNATIONAL JOURNAL FOR NUMERICAL METHODS IN FLUIDS
(2020)
Proceedings Paper
Construction & Building Technology
Flavia Turi, Regiane Fortes-Patella
29TH IAHR SYMPOSIUM ON HYDRAULIC MACHINERY AND SYSTEMS
(2019)
Proceedings Paper
Construction & Building Technology
F. Doussot, G. Balarac, J. Brammer, O. Metais, C. Segoufin
29TH IAHR SYMPOSIUM ON HYDRAULIC MACHINERY AND SYSTEMS
(2019)
Article
Computer Science, Interdisciplinary Applications
Ashish Bhole, Herve Guillard, Boniface Nkonga, Francesca Rapetti
Summary: Finite elements of class C-1 are used for computing magnetohydrodynamics instabilities in tokamak plasmas, and isoparametric approximations are employed to align the mesh with the magnetic field line. This numerical framework helps in understanding the operation of existing devices and predicting optimal strategies for the international ITER tokamak. However, a mesh-aligned isoparametric representation encounters issues near critical points of the magnetic field, which can be addressed by combining aligned and unaligned meshes.
INTERNATIONAL JOURNAL FOR NUMERICAL METHODS IN FLUIDS
(2024)
Article
Computer Science, Interdisciplinary Applications
Federico Vismara, Tommaso Benacchio
Summary: This paper introduces a method for solving hyperbolic-parabolic problems on multidimensional semi-infinite domains. By dividing the computational domain into bounded and unbounded subdomains and coupling them using numerical fluxes at the interface, accurate numerical solutions are obtained. In addition, computational cost can be reduced by tuning the parameters of the basis functions.
INTERNATIONAL JOURNAL FOR NUMERICAL METHODS IN FLUIDS
(2024)
Article
Computer Science, Interdisciplinary Applications
Keigo Enomoto, Takato Ishida, Yuya Doi, Takashi Uneyama, Yuichi Masubuchi
Summary: We have developed a novel Moving Particle Simulation (MPS) method to accurately reproduce the motion of fibers in sheared liquids. By introducing the micropolar fluid model, we address the issue of fibers being aligned with the flow direction in conventional MPS simulations. Our method is capable of accurately reproducing the fiber motion predicted by Jeffery's theory.
INTERNATIONAL JOURNAL FOR NUMERICAL METHODS IN FLUIDS
(2024)
