检索结果-内蒙古大学图书馆

Computational fluid dynamics analysis of hybrid foam-mat freeze drying process

学校读者我要写书评

暂无评论

AIP Conference Proceedings 2024年第1期2991卷

作者： Retno Wulandari Evi Susanti Melisa Surya Andini Bagus Anggoro Dwicahyo Arvin Dwi Nanda Mechanical Engineering Department Faculty of Engineering Universitas Negeri Malang Semarang Street No. 5 Malang 65145 Jawa Timur Indonesia Biotechnology Program Department of Applied Science Faculty of Mathematics and Natural Science Universitas Negeri Malang Semarang Street No. 5 Malang 65145 Jawa Timur Indonesia

To produce halal, safe, and affordable food collagen are factors that need to be resolved in food processing industry. Collagen was made from milkfish scales and B. megaterium TR-10. Freeze drying method has been around for a long time and usually use in food processing, including collagen production. Freeze drying techniques need in this separation of protein and water as a solvent due addition of acetone and NaCl so that the protein has not to denaturation due to processing above 50ᵒC. Computational method to simulate heat and mass transfer freeze drying in the collagen formation at -60ᵒC with a pressure less than 10 Pa to obtain optimal efficiency. When compared to conventional dryers, hybrid freeze dryers excel in several key areas: drying efficiency, energy efficiency, moisture removal procedure, product quality, and nutrient or bioactive ingredient preservation. The simulation using a foam mat with variation 1mm, 3mm, and 5mm. Changing product into a foam to expand the surface area and mass transfer increase thus drying process take a short time. This simulation begins with designing use Ansys R21 and the next process is numerical calculations to analyze heat and mass transfer in freeze drying process to obtain optimal efficiency result. The optimal value is obtained because effectiveness heat and mass transfer process when the collagen converted into foam causing surface largen with thinnest layers. Drying time with hybrid technique is between 1/4 to 1/7 times faster compared non-foam drying.

关键词：

Neural FIM for learning Fisher information metrics from point cloud data

学校读者我要写书评

暂无评论

arXiv 2023年

作者： Fasina, Oluwadamilola Huguet, Guillaume Tong, Alexander Zhang, Yanlei Wolf, Guy Nickel, Maximilian Adelstein, Ian Krishnaswamy, Smita Applied Math Program Yale University New HavenCT United States Mila - Quebec AI Institute MontrealQC Canada Department of Mathematics and Statistics Université de Montréal MontrealQC Canada Department of Computer Science and Operations Research Université de Montréal MontrealQC Canada FAIR Meta AI United States Department of Math Yale University New HavenCT United States Department of Computer Science Yale University New HavenCT United States Department of Genetics Yale University New HavenCT United States

Although data diffusion embeddings are ubiquitous in unsupervised learning and have proven to be a viable technique for uncovering the underlying intrinsic geometry of data, diffusion embeddings are inherently limited due to their discrete nature. To this end, we propose neural FIM, a method for computing the Fisher information metric (FIM) from point cloud data - allowing for a continuous manifold model for the data. Neural FIM creates an extensible metric space from discrete point cloud data such that information from the metric can inform us of manifold characteristics such as volume and geodesics. We demonstrate Neural FIM's utility in selecting parameters for the PHATE visualization method as well as its ability to obtain information pertaining to local volume illuminating branching points and cluster centers embeddings of a toy dataset and two single-cell datasets of IPSC reprogramming and PBMCs (immune cells). Copyright © 2023, The Authors. All rights reserved.

关键词： Embeddings

DP Compress: a Model Compression Scheme for Generating Efficient Deep Potential Models

学校读者我要写书评

暂无评论

arXiv 2021年

作者： Lu, Denghui Jiang, Wanrun Chen, Yixiao Zhang, Linfeng Jia, Weile Wang, Han Chen, Mohan HEDPS CAPT College of Engineering Peking University Beijing100871 China Songshan Lake Materials Laboratory Guangdong Dongguan523808 China Institute of Physics Chinese Academy of Sciences Beijing100190 China Program in Applied and Computational Mathematics Princeton University PrincetonNJ United States Beijing Institute of Big Data Research Beijing100871 China Institute of Computing Technology Chinese Academy of Sciences Beijing100190 China University of Chinese Academy of Sciences Beijing100049 China Laboratory of Computational Physics Institute of Applied Physics and Computational Mathematics Fenghao East Road 2 Beijing100094 China

Machine-learning-based interatomic potential energy surface (PES) models are revolutionizing the field of molecular modeling. However, although much faster than electronic structure schemes, these models suffer from costly computations via deep neural networks to predict the energy and atomic forces, resulting in lower running efficiency as compared to the typical empirical force fields. Herein, we report a model compression scheme for boosting the performance of the Deep Potential (DP) model, a deep learning based PES model. This scheme, we call DP Compress, is an efficient post-processing step after the training of DP models (DP Train). DP Compress combines several DP-specific compression techniques, which typically speed up DP-based molecular dynamics simulations by an order of magnitude faster, and consume an order of magnitude less memory. We demonstrate that DP Compress is sufficiently accurate by testing a variety of physical properties of Cu, H2O, and Al-Cu-Mg systems. DP Compress applies to both CPU and GPU machines and is publicly available online. Copyright © 2021, The Authors. All rights reserved.

关键词： Molecular dynamics

Preconditioned BFGS-based uncertainty quantification in elastic full waveform inversion

学校读者我要写书评

暂无评论

arXiv 2020年

作者： Liu, Qiancheng Beller, Stephen Lei, Wenjie Peter, Daniel Tromp, Jeroen Department of Geosciences Princeton University PrincetonNJ08544 United States Thuwal Saudi Arabia Program in Applied and Computational Mathematics Princeton University PrincetonNJ08544 United States

Full Waveform Inversion (FWI) has become an essential technique for mapping geophysical subsurface structures. However, proper uncertainty quantification is often lacking in current applications. In theory, uncertainty quantification is related to the inverse Hessian (or the posterior covariance matrix). Even for common geophysical inverse problems its calculation is beyond the computational and storage capacities of the largest high-performance computing systems. In this study, we amend the Broyden-Fletcher-Goldfarb-Shanno (BFGS) algorithm to perform uncertainty quantification for large-scale applications. For seismic inverse problems, the limited-memory BFGS (L-BFGS) method prevails as the most efficient quasi-Newton method. We aim to augment it further to obtain an approximate inverse Hessian for uncertainty quantification in FWI. To facilitate retrieval of the inverse Hessian, we combine BFGS (essentially a full-history L-BFGS) with randomized singular value decomposition to determine a low-rank approximation of the inverse Hessian. Setting the rank number equal to the number of iterations makes this solution efficient and memory-affordable even for large-scale problems. Furthermore, based on the Gauss-Newton method, we formulate different initial, diagonal Hessian matrices as preconditioners for the inverse scheme and compare their performances in elastic FWI applications. We highlight our approach with the elastic Marmousi benchmark model, demonstrating the applicability of preconditioned BFGS for large-scale FWI and uncertainty quantification. Copyright © 2020, The Authors. All rights reserved.

关键词： Uncertainty analysis

Deep learning of accurate force field of ferroelectric HfO2

学校读者我要写书评

暂无评论

Physical Review B 2021年第2期103卷 024108-024108页

作者： Jing Wu Yuzhi Zhang Linfeng Zhang Shi Liu Fudan University Shanghai 200433 China School of Science Westlake University Hangzhou Zhejiang 310024 China Institute of Natural Sciences Westlake Institute for Advanced Study Hangzhou Zhejiang 310024 China Yuanpei College Peking University Beijing 100871 China Program in Applied and Computational Mathematics Princeton University Princeton New Jersey 08544 USA Key Laboratory for Quantum Materials of Zhejiang Province Hangzhou Zhejiang 310024 China

The discovery of ferroelectricity in HfO2-based thin films opens up new opportunities for using this silicon-compatible ferroelectric to realize low-power logic circuits and high-density nonvolatile memories. The functional performances of ferroelectrics are intimately related to their dynamic responses to external stimuli such as electric fields at finite temperatures. Molecular dynamics is an ideal technique for investigating dynamical processes on large length and time scales, though its applications to new materials are often hindered by the limited availability and accuracy of classical force fields. Here we present a deep neural network–based interatomic force field of HfO2 learned from ab initio data using a concurrent learning procedure. The model potential is able to predict structural properties such as elastic constants, equation of states, phonon dispersion relationships, and phase transition barriers of various hafnia polymorphs with accuracy comparable with density functional theory calculations. The validity of this model potential is further confirmed by the reproduction of experimental sequences of temperature-driven ferroelectric-paraelectric phase transitions of HfO2 with isobaric-isothermal ensemble molecular dynamics simulations. We suggest a general approach to extend the model potential of HfO2 to related material systems including dopants and defects.

关键词： Interatomic & molecular potentials Structural properties Ferroelectrics Artificial neural networks Molecular dynamics

Inferring solar differential rotation through normal-mode coupling using bayesian statistics

学校读者我要写书评

暂无评论

arXiv 2021年

作者： Kashyap, Samarth Ganesh Bharati Das, Srijan Hanasoge, Shravan M. Woodard, Martin F. Tromp, Jeroen Department of Astronomy and Astrophysics Tata Institute of Fundamental Research Mumbai India Department of Geosciences Princeton University PrincetonNJ United States NorthWest Research Associates Boulder Office 3380 Mitchell Lane BoulderCO United States Department of Geosciences and Program in Applied & Computational Mathematics Princeton University PrincetonNJ United States

Normal-mode helioseismic data analysis uses observed solar oscillation spectra to infer perturbations in the solar interior due to global and local-scale flows and structural asphericity. Differential rotation, the dominant global-scale axisymmetric perturbation, has been tightly constrained primarily using measurements of frequency splittings via "a-coefficients". However, the frequency-splitting formalism invokes the approximation that multiplets are isolated. This assumption is inaccurate for modes at high angular degrees. Analysing eigenfunction corrections, which respect cross coupling of modes across multiplets, is a more accurate approach. However, applying standard inversion techniques using these cross-spectral measurements yields a-coefficients with a significantly wider spread than the well-constrained results from frequency splittings. In this study, we apply Bayesian statistics to infer a-coefficients due to differential rotation from cross spectra for both f-modes and p-modes. We demonstrate that this technique works reasonably well for modes with angular degrees ` = 50 − 291. The inferred a3−coefficients are found to be within 1 nHz of the frequency splitting values for ` > 200. We also show that the technique fails at ` © 2021, CC BY.

关键词： Eigenvalues and eigenfunctions

Kinetic Frustration Effects on Dense Two-Dimensional Packings of Convex Particles and Their Structural Characteristics

学校读者我要写书评

暂无评论

arXiv 2021年

作者： Maher, Charles Emmett Stillinger, Frank H. Torquato, Salvatore Department of Chemistry Princeton University PrincetonNJ08544 United States Department of Physics Princeton University PrincetonNJ08544 United States Princeton Institute for the Science and Technology of Materials Princeton University PrincetonNJ08544 United States Program in Applied and Computational Mathematics Princeton University PrincetonNJ08544 United States

The study of hard-particle packings is of fundamental importance in physics, chemistry, cell biology, and discrete geometry. Much of the previous work on hard-particle packings concerns their densest possible arrangements. By contrast, we examine kinetic effects inevitably present in both numerical and experimental packing protocols. Specifically, we determine how changing the compression/shear rate of a two-dimensional packing of noncircular particles causes it to deviate from its densest possible configuration, which is always periodic. The adaptive shrinking cell (ASC) optimization scheme maximizes the packing fraction of a hard-particle packing by first applying random translations and rotations to the particles and then isotropically compressing and shearing the simulation box repeatedly until a possibly jammed state is reached. We use a stochastic implementation of the ASC optimization scheme to mimic different effective time scales by varying the number of particle moves between compressions/shears. We generate dense, effectively jammed, monodisperse, two-dimensional packings of obtuse scalene triangle, rhombus, curved triangle, lens, and "ice cream cone" (a semicircle grafted onto an isosceles triangle) shaped particles, with a wide range of packing fractions and degrees of order. To quantify these kinetic effects, we introduce the kinetic frustration index K, which measures the deviation of a packing from its maximum possible packing fraction. To investigate how kinetics affect short- and long-range ordering in these packings, we compute their spectral densities χV (k) and characterize their contact networks. We find that kinetic effects are most significant when the particles have greater asphericity, less curvature, and less rotational symmetry. This work may be relevant to the design of laboratory packing protocols. Copyright © 2021, The Authors. All rights reserved.

关键词： Kinetic theory

Uniformly Accurate Machine Learning Based Hydrodynamic Models for Kinetic Equations

学校读者我要写书评

暂无评论

arXiv 2019年

作者： Jiequn, H.A.N. Chao, M.A. Zheng, M.A. Weinan, E. Department of Mathematics Princeton University Program in Applied and Computational Mathematics Princeton University Department of Mathematics Purdue University Beijing Institute of Big Data Research

A new framework is introduced for constructing interpretable and truly reliable reduced models for multi-scale problems in situations without scale separation. Hydrodynamic approximation to the kinetic equation is used as an example to illustrate the main steps and issues involved. To this end, a set of generalized moments are constructed first through an autoencoder to optimally represent the underlying velocity distribution. The well-known closure problem is then solved with the aim of best capturing the associated dynamics of the kinetic equation. The issue of physical constraints such as Galilean invariance is addressed and an active learning procedure is introduced to help ensure that the data set used is representative enough. The reduced system takes the form of the conventional moment systems and works regardless of the numerical discretization used. Numerical results are presented for the BGK model. We demonstrate that the reduced model achieves a uniform accuracy in a wide range of Knudsen numbers spanning from the hydrodynamic limit to free molecular flow. Copyright © 2019, The Authors. All rights reserved.

关键词： Hydrodynamics

Bayesian Testing Of Granger Causality In Functional Time Series

学校读者我要写书评

暂无评论

arXiv 2021年

作者： Sen, Rituparna Majumdar, Anandamayee Sikaria, Shubhangi Applied Statistics Unit Indian Statistical Institute 8th Mile Mysore Rd RVCE Post KA Bangalore560059 India Stock Assessment Program Inter-American Tropical Tuna Commission San DiegoCA United States Department of Mathematics Indian Institute of Technology Madras TN Chennai India

We develop a multivariate functional autoregressive model (MFAR), which captures the cross-correlation among multiple functional time series and thus improves forecast accuracy. We estimate the parameters under the Bayesian dynamic linear models (DLM) framework. In order to capture Granger causality from one FAR series to another we employ Bayes Factor. Motivated by the broad application of functional data in finance, we investigate the causality between the yield curves of two countries. Furthermore, we illustrate a climatology example, examining whether the weather conditions Granger cause pollutant daily levels in a city. Copyright © 2021, The Authors. All rights reserved.

关键词： Statistical tests