Article
Computer Science, Theory & Methods
Inaki Amatria-Barral, Jorge Gonzalez-Dominguez, Juan Tourino
Summary: This paper presents pRIblast, a highly efficient parallel application for comprehensive lncRNA-RNA interaction prediction. Compared to other tools, pRIblast has advantages in terms of computational speed and accuracy, and can handle large-scale datasets.
FUTURE GENERATION COMPUTER SYSTEMS-THE INTERNATIONAL JOURNAL OF ESCIENCE
(2023)
Article
Nanoscience & Nanotechnology
C. Paoletti, E. Santecchia, M. Cabibbo, M. Regev, S. Spigarelli
Summary: A physical model was used to study the dependence of minimum creep rate on stress and temperature for ETP copper. The model took into account the role of grain boundaries and the effect of grain growth. Experimental testing confirmed the model's predictions, showing that the grain size and dislocation density have an impact on creep behavior.
MATERIALS SCIENCE AND ENGINEERING A-STRUCTURAL MATERIALS PROPERTIES MICROSTRUCTURE AND PROCESSING
(2022)
Article
Computer Science, Interdisciplinary Applications
G. Houzeaux, R. M. Badia, R. Borrell, D. Dosimont, J. Ejarque, M. Garcia-Gasulla, V Lopez
Summary: This work presents an elastic computing methodology that adjusts the allocated resources to a simulation automatically based on the runtime measure of communication efficiency, resulting in efficient simulations.
COMPUTERS & FLUIDS
(2022)
Article
Chemistry, Multidisciplinary
Aleardo Manacero, Emanuel Guariglia, Thiago Alexandre de Souza, Renata Spolon Lobato, Roberta Spolon
Summary: Clustering is a classification method that groups objects based on similarity and extracts valuable information from large datasets. This paper proposes a parallel implementation on graphics processing unit for high-performance clustering at low cost, addressing the time-consuming issue caused by increasing data volumes.
APPLIED SCIENCES-BASEL
(2022)
Article
Computer Science, Hardware & Architecture
Renjiang Chen, Tao Liu, Zhaoyuan Liu, Li Wang, Min Tian, Ying Guo, Jingshan Pan, Xiaoming Wu, Meihong Yang
Summary: With the development of nuclear energy technology, it has become necessary to use high-performance computers for reactor simulation calculations. The method of characteristics (MOC) is the preferred method for simulating neutron transport in the nuclear reactor core. This paper proposes a fine-grained and universal two-level parallelization method based on the architecture of Sunway many-core processor and Sunway Bluelight II supercomputer. The parallelization achieved significant speedup and good scalability on Sunway Bluelight II with up to 18.6x performance improvement.
JOURNAL OF SUPERCOMPUTING
(2023)
Article
Materials Science, Multidisciplinary
Ronghai Wu, Michael Zaiser
Summary: This paper discusses the thermodynamic consistency of physically based crystal plasticity constitutive equations and proposes constraints to restore thermodynamic consistency.
JOURNAL OF THE MECHANICS AND PHYSICS OF SOLIDS
(2022)
Article
Mathematics, Applied
Alexander Heinlein, Oliver Rheinbach, Friederike Roever
Summary: This paper investigates the parallel performance of three-level fast and robust overlapping Schwarz (FROSch) preconditioners for linear elasticity. The paper describes the additional steps in the implementation of the recursive FROSch preconditioner and shows that explicit geometric information is not needed. Parallel results obtained on different supercomputers are discussed, highlighting the impact of hierarchical communication operations on computing time. Further analysis on large supercomputers with dragonfly interconnects is deemed necessary.
SIAM JOURNAL ON SCIENTIFIC COMPUTING
(2023)
Article
Computer Science, Interdisciplinary Applications
Vincent Delmas, Azzeddine Soulaimani
Summary: This paper presents a method for solving the Shallow-water equations in the study of floods, using domain decomposition and ghost-layer generation on large meshes. The performance of the method is evaluated through time and memory analysis on 2D and 3D meshes. The method is also applied to study the second-order resolution on real large-scale river meshes, and the impact of multiple layers of ghost cells on the execution times of different numerical methods is discussed.
JOURNAL OF COMPUTATIONAL PHYSICS
(2022)
Article
Computer Science, Theory & Methods
Zhiyong Xiao, Xu Liu, Jingheng Xu, Qingxiao Sun, Lin Gan
Summary: The study introduces a highly scalable hybrid parallel genetic algorithm based on Sunway TaihuLight Supercomputer, utilizing Cellular and Island models to achieve impressive performance for large-scale problems.
FUTURE GENERATION COMPUTER SYSTEMS-THE INTERNATIONAL JOURNAL OF ESCIENCE
(2021)
Article
Computer Science, Interdisciplinary Applications
Sandeep Kumar, Pierre Gosselet, Dengpeng Huang, Christian Weissenfels, Peter Wriggers
Summary: This paper presents a parallel implementation of the Optimal Transportation Meshfree (OTM) method on large CPU clusters. The paper introduces new concepts and techniques to reduce computational efforts, such as dynamic halo regions, efficient data management strategies, and nearest neighborhood communication. The Parallel performance analysis is carried out for challenging multiphysics applications and adequate scalability is reported.
JOURNAL OF COMPUTATIONAL SCIENCE
(2022)
Article
Engineering, Mechanical
Chaitali S. Patil, Supriyo Chakraborty, Stephen R. Niezgoda
Summary: This study compared the effects of the Voce hardening law and dislocation density based hardening law on crystal plasticity simulations of tensile deformation. The results showed that the choice of hardening law influences stress distribution, with average texture characteristics predicted similarly by both laws. However, micro-texture varied with increasing strain.
INTERNATIONAL JOURNAL OF PLASTICITY
(2021)
Article
Engineering, Mechanical
Zixu Guo, Dawei Huang, Xiaojun Yan
Summary: In this study, a physics-based model is proposed to predict the gamma/gamma' microstructure evolution of single crystal superalloy at medium temperature and high stress level. The model considers multiple hardening mechanisms and is validated through creep fracture and interrupted tests, showing good agreement with the experimental results.
INTERNATIONAL JOURNAL OF PLASTICITY
(2021)
Article
Computer Science, Hardware & Architecture
Jorge Gonzalez-Dominguez, Jose M. Martin-Martinez, Roberto R. Exposito
Summary: Tandem Repeats (TRs) are DNA segments that occur multiple times in a sequence and have gained attention for their potential relation with human diseases. Dot2dot, a tool for identifying TRs, has been improved with MPI-dot2dot, which combines MPI and OpenMP to reduce execution time in multicore clusters. This new version performs more than 100 times faster on a 16-node multicore cluster while providing the same biological results as Dot2dot.
JOURNAL OF SUPERCOMPUTING
(2022)
Article
Computer Science, Theory & Methods
Niranda Perera, Arup Kumar Sarker, Mills Staylor, Gregor von Laszewski, Kaiying Shan, Supun Kamburugamuve, Chathura Widanage, Vibhatha Abeykoon, Thejaka Amila Kanewela, Geoffrey Fox
Summary: The Data Science domain has witnessed significant expansion in the past decade, driven largely by the Big Data revolution. The use of Artificial Intelligence (AI) and Machine Learning (ML) in data engineering applications has led to the integration of data processing pipelines for terabytes of data. However, the commonly used serial Dataframes (e.g., R, pandas) face performance limitations when working with moderately large datasets. This paper introduces a cost model for evaluating parallel processing patterns and evaluates the performance of Cylon on the ORNL Summit supercomputer.
FUTURE GENERATION COMPUTER SYSTEMS-THE INTERNATIONAL JOURNAL OF ESCIENCE
(2023)
Article
Computer Science, Artificial Intelligence
Patricia Gonzalez, Roberto Prado-Rodriguez, Attila Gabor, Julio Saez-Rodriguez, Julio R. Banga, Ramon Doallo
Summary: Understanding the deregulation of cell signaling networks is crucial for studying diseases. Computational models, such as CellNOpt, provide a systematic tool to analyze these complex biochemical networks. In this paper, the use of ant colony optimization is proposed as a novel method to improve the limitations of the existing genetic algorithm in CellNOpt, and its performance is demonstrated in liver cancer therapy research.
EXPERT SYSTEMS WITH APPLICATIONS
(2022)
Article
Materials Science, Multidisciplinary
Russell E. Marki, Kyle A. Brindley, Rodney J. McCabe, Marko Knezevic
Summary: This paper presents a mathematical procedure for invertible microstructure-property linkages for orthorhombic polycrystalline metals using the generalized spherical harmonics (GSH) spectral basis. The procedure allows for the computation of property closures and enables the simulation of microstructurally heterogeneous components under thermo-mechanical loadings in a computationally efficient manner. The developed framework has been demonstrated using alpha-uranium as a case study.
JOURNAL OF NUCLEAR MATERIALS
(2022)
Article
Astronomy & Astrophysics
Jeffersson A. Agudelo Rueda, Daniel Verscharen, Robert T. Wicks, Christopher J. Owen, Georgios Nicolaou, Kai Germaschewski, Andrew P. Walsh, Ioannis Zouganelis, Santiago Vargas Dominguez
Summary: Energy dissipation in collisionless plasmas, particularly the energy transfer and transport associated with 3D small-scale magnetic reconnection events, is still not well understood. This study investigates the spatial energy transfer during a highly dynamic and asymmetric reconnection event in anisotropic and decaying Alfvenic turbulence. The findings suggest that electron bulk flow transports thermal energy density more efficiently than kinetic energy density, and the energy density transfer is dominated by plasma compression.
ASTROPHYSICAL JOURNAL
(2022)
Article
Engineering, Mechanical
Zhangxi Feng, Reeju Pokharel, Sven C. Vogel, Ricardo A. Lebensohn, Darren Pagan, Eloisa Zepeda-Alarcon, Bjorn Clausen, Ramon Martinez, George T. Gray, Marko Knezevic
Summary: This paper presents crystallographically-based phase transformation models and deformation mechanism models for predicting strain-induced austenite to martensite transformation. The models can predict the strain-path sensitive, strain-rate and temperature sensitive deformation of stainless steels. The deformation of constituent grains is modeled as a combination of anisotropic elasticity, crystallographic slip, and phase transformation, while the hardening is based on the evolution of dislocation density and phase fractions. The models are calibrated and validated using experimental data and are used to simulate the deformation processes of stainless steel materials. The simulation results are compared and analyzed with experimental results in terms of geometry, mechanical response, phase fractions, and texture evolution.
INTERNATIONAL JOURNAL OF PLASTICITY
(2022)
Article
Materials Science, Multidisciplinary
Adnan Eghtesad, John D. Shimanek, Shun -Li Shang, Ricardo Lebensohn, Marko Knezevic, Zi-Kui Liu, Allison M. Beese
Summary: This study successfully integrates first-principles calculations based on density functional theory (DFT) into the dislocation density hardening law of the crystal plasticity fast Fourier transform (CPFFT) model, improving the robustness of the model and reducing the uncertainties in calibrating the macroscopic flow response.
COMPUTATIONAL MATERIALS SCIENCE
(2022)
Article
Engineering, Multidisciplinary
R. Sancho, V. Rey-de-Pedraza, P. Lafourcade, R. A. Lebensohn, J. Segurado
Summary: An FFT-based algorithm is proposed to simulate the propagation of elastic waves in heterogeneous domains. The method incorporates the application of Dirichlet boundary conditions and uses a stable beta-Newmark approach for time discretization. By solving the equilibrium equations in Fourier space and employing a preconditioned Krylov solver, the method achieves high accuracy and computational efficiency. Numerical examples demonstrate its effectiveness in simulating wave propagation in different mediums.
COMPUTER METHODS IN APPLIED MECHANICS AND ENGINEERING
(2023)
Article
Materials Science, Multidisciplinary
Miroslav Zecevic, Ricardo A. Lebensohn, Laurent Capolungo
Summary: This paper presents a new formulation and numerical implementation of a strain-gradient crystal plasticity model within a large-strain elasto-viscoplastic fast Fourier transform-based micromechanical model. The model is used to study the formation of kink bands during layer-parallel compression of nano-metallic laminates. The interaction between dislocations and interfaces is considered in the model to accurately simulate the behavior of the layered composites.
JOURNAL OF THE MECHANICS AND PHYSICS OF SOLIDS
(2023)
Article
Engineering, Mechanical
C. K. Cocke, H. Mirmohammad, M. Zecevic, B. R. Phung, R. A. Lebensohn, O. T. Kingstedt, A. D. Spear
Summary: This study extends a large-strain FFT-based crystal plasticity model to simulate ductile fracture of polycrystalline materials. By incorporating a triaxiality-based continuum damage mechanics (CDM) formulation into a large-strain elasto-viscoplastic FFT (LS-EVPFFT) framework and using an integral-based nonlocal regularization approach, the model is able to accurately predict the macroscopic stress-strain response and necking behavior of ductile polycrystals.
INTERNATIONAL JOURNAL OF PLASTICITY
(2023)
Article
Engineering, Multidisciplinary
Valentin Gallican, Miroslav Zecevic, Ricardo A. Lebensohn, Martin I. Idiart
Summary: Approximations for the elastic properties of dilute solid suspensions with imperfect interfacial bonding are derived and assessed. Two approximations are generated using a variational procedure, with one dependent on an arithmetic mean and the other dependent on a harmonic mean for averaging the interfacial compliance. The harmonic approximation is found to be more accurate than the arithmetic approximation, which has practical relevance given the widespread use of the latter in existing descriptions.
JOURNAL OF ELASTICITY
(2023)
Article
Materials Science, Multidisciplinary
Aritra Chakraborty, Ricardo A. Lebensohn, Laurent Capolungo
Summary: At moderate-to-high temperatures and below the yield strength, the inelastic deformation of metals is mainly controlled by vacancy diffusion-mediated processes. Vacancies (or atoms) can diffuse preferentially along grain boundaries or along dislocations, resulting in climb and self-climb. The proposed thermodynamically-consistent model considers the coupling between grain boundary and grain bulk diffusion-mediated plasticity mechanisms and predicts the strain rate dependencies and steady-state creep rate scaling with respect to grain size, temperature, and stress.
JOURNAL OF THE MECHANICS AND PHYSICS OF SOLIDS
(2023)
Article
Materials Science, Multidisciplinary
Yifan Zhang, Miroslav Zecevic, Aritra Chakraborty, Rodney J. McCabe, Thomas J. Nizolek, Ricardo A. Lebensohn, John S. Carpenter, Nan Li, Laurent Capolungo
Summary: This study investigates the dependence of kinking on microstructural attributes in NMLs through in situ micropillar compression, microstructure characterization, simulations, and modeling. The development of internal stresses during loading activates local layer-parallel glide triggering kinking in NMLs. The effect of key microstructural features on kink band formation in NMLs is also revealed.
Article
Engineering, Mechanical
Adnan Eghtesad, Qixiang Luo, Shun -Li Shang, Ricardo A. Lebensohn, Marko Knezevic, Zi-Kui Liu, Allison M. Beese
Summary: This study combines a full-field crystal plasticity model with a first principles-informed dislocation density hardening law and a machine learning approach to investigate the microstructural features correlated with micromechanical field localization in polycrystalline Ni. The results show that regions near grain boundaries, higher Schmid factors, low slip transmissions, and high intergranular misorientations are more prone to being micromechanical hotspots. The integration of physics-based crystal plasticity with machine learning provides insights into the initiation zones of micromechanical damage in polycrystalline metals.
INTERNATIONAL JOURNAL OF PLASTICITY
(2023)
Article
Chemistry, Physical
Dongil Shin, Ryan Alberdi, Ricardo A. Lebensohn, Remi Dingreville
Summary: Recent developments in micromechanics and neural networks have provided promising paths for accurately predicting the response of heterogeneous materials. The deep material network, with its multi-layer design and trained micromechanics building blocks, offers the ability to extrapolate material behavior to different constitutive laws without retraining. However, the random initialization of network parameters in current training methods leads to unavoidable errors. In this study, we propose a visualization technique using an analogous unit cell to initialize deeper networks and improve the accuracy and calibration performance, while also providing a more intuitive representation of the network for explainability.
NPJ COMPUTATIONAL MATERIALS
(2023)
Article
Geography, Physical
Maria-Gema Llorens, Albert Griera, Paul D. Bons, Ilka Weikusat, David J. Prior, Enrique Gomez-Rivas, Tamara de Riese, Ivone Jimenez-Munt, Daniel Garcia-Castellanos, Ricardo A. Lebensohn
Summary: This study investigates the influence of ice deformation history on the development of crystallographic preferred orientations (CPOs) using full-field numerical simulations. The results show that the second deformation event tends to destroy the first inherited fabric, but the transition is slow when crystallographic axes are critically oriented with respect to the second imposed regime. Therefore, caution must be exercised when interpreting observed CPOs in areas with complex deformation histories.
Proceedings Paper
Computer Science, Artificial Intelligence
John Donaghy, Kai Germaschewski
Summary: This article introduces the use of normalizing flows to learn an approximate model of the noisy particle distribution function. It balances the trade-off between computational cost and intrinsic noise, and models the distribution function that contains noise, is temporally dynamic, and can be non-Gaussian and multi-modal.
MACHINE LEARNING, OPTIMIZATION, AND DATA SCIENCE (LOD 2021), PT I
(2022)