A sub-grid scale model for Burgers turbulence based on the artificial neural network method

The present study proposes a sub-grid scale model for the one-dimensional Burgers turbulence based on the neuralnetwork and deep learning method.The filtered data of the direct numerical simulation is used to establish thetraining data set,the validation data set,and the test data set.The artificial neural network(ANN)methodand Back Propagation method are employed to train parameters in the ANN.The developed ANN is applied toconstruct the sub-grid scale model for the large eddy simulation of the Burgers turbulence in the one-dimensionalspace.The proposed model well predicts the time correlation and the space correlation of the Burgers turbulence.

Stochastic Analysis and Modeling of Velocity Observations in Turbulent Flows

Highly turbulent water flows,often encountered near human constructions like bridge piers,spillways,and weirs,display intricate dynamics characterized by the formation of eddies and vortices.These formations,varying in sizes and lifespans,significantly influence the distribution of fluid velocities within the flow.Subsequently,the rapid velocity fluctuations in highly turbulent flows lead to elevated shear and normal stress levels.For this reason,to meticulously study these dynamics,more often than not,physical modeling is employed for studying the impact of turbulent flows on the stability and longevity of nearby structures.Despite the effectiveness of physical modeling,various monitoring challenges arise,including flow disruption,the necessity for concurrent gauging at multiple locations,and the duration of measurements.Addressing these challenges,image velocimetry emerges as an ideal method in fluid mechanics,particularly for studying turbulent flows.To account for measurement duration,a probabilistic approach utilizing a probability density function(PDF)is suggested to mitigate uncertainty in estimated average and maximum values.However,it becomes evident that deriving the PDF is not straightforward for all turbulence-induced stresses.In response,this study proposes a novel approach by combining image velocimetry with a stochastic model to provide a generic yet accurate description of flow dynamics in such applications.This integration enables an approach based on the probability of failure,facilitating a more comprehensive analysis of turbulent flows.Such an approach is essential for estimating both short-and long-term stresses on hydraulic constructions under assessment.

Artificial neural network-based one-equation model for simulation of laminar-turbulent transitional flow

A mapping function between the Reynolds-averaged Navier-Stokes mean flow variables and transition intermittency factor is constructed by fully connected artificial neural network(ANN),which replaces the governing equation of the intermittency factor in transition-predictive Spalart-Allmaras(SA)-γmodel.By taking SA-γmodel as the benchmark,the present ANN model is trained at two airfoils with various angles of attack,Mach numbers and Reynolds numbers,and tested with unseen airfoils in different flow states.The a posteriori tests manifest that the mean pressure coefficient,skin friction coefficient,size of laminar separation bubble,mean streamwise velocity,Reynolds shear stress and lift/drag/moment coefficient from the present two-way coupling ANN model almost coincide with those from the benchmark SA-γmodel.Furthermore,the ANN model proves to exhibit a higher calculation efficiency and better convergence quality than traditional SA-γmodel.

Artificial neural network-based subgrid-scale models for LES of compressible turbulent channel flow

Fully connected neural networks(FCNNs)have been developed for the closure of subgrid-scale(SGS)stress and SGS heat flux in large-eddy simulations of compressible turbulent channel flow.The FCNNbased SGS model trained using data with Mach number Ma=3.0 and Reynolds number Re=3000 was applied to situations with different Mach numbers and Reynolds numbers.The input variables of the neural network model were the filtered velocity gradients and temperature gradients at a single spatial grid point.The a priori test showed that the FCNN model had a correlation coefficient larger than 0.91 and a relative error smaller than 0.43,with much better reconstructions of SGS unclosed terms than the dynamic Smagorinsky model(DSM).In a posteriori test,the behavior of the FCNN model was marginally better than that of the DSM in predicting the mean velocity profiles,mean temperature profiles,turbulent intensities,total Reynolds stress,total Reynolds heat flux,and mean SGS flux of kinetic energy,and outperformed the Smagorinsky model.

Simulation Analysis of New Energy Vehicle Engine Cooling System Based on K-E Turbulent Flow Mathematical Model

New energy vehicles have better clean and environmental protection characteristics than traditional fuel vehicles.The new energy engine cooling technology is critical in the design of new energy vehicles.This paper used oneand three-way joint simulation methods to simulate the refrigeration system of new energy vehicles.Firstly,a k-εturbulent flow model for the cooling pump flow field is established based on the principle of computational fluid dynamics.Then,the CFD commercial fluid analysis software FLUENT is used to simulate the flow field of the cooling pump under different inlet flow conditions.This paper proposes an optimization scheme for new energy vehicle engines’“boiling”phenomenon under high temperatures and long-time climbing conditions.The simulation results show that changing the radiator’s structure and adjusting the thermostat’s parameters can solve the problem of a“boiling pot.”The optimized new energy vehicle engine can maintain a better operating temperature range.The algorithm model can reference each cryogenic system component hardware selection and control strategy in the new energy vehicle’s engine.

Constrained re-calibration of two-equation Reynolds-averaged Navier–Stokes models

Machine-learned augmentations to turbulence models can be advantageous for flows within the training dataset but can often cause harm outside.This lack of generalizability arises because the constants(as well as the functions)in a Reynolds-averaged Navier–Stokes(RANS)model are coupled,and un-constrained re-calibration of these constants(and functions)can disrupt the calibrations of the baseline model,the preservation of which is critical to the model's generalizability.To safeguard the behaviors of the baseline model beyond the training dataset,machine learning must be constrained such that basic calibrations like the law of the wall are kept intact.This letter aims to identify such constraints in two-equation RANS models so that future machine learning work can be performed without violating these constraints.We demonstrate that the identified constraints are not limiting.Furthermore,they help preserve the generalizability of the baseline model.

Evaluation of Nonbreaking Wave-Induced Mixing Parameterization Schemes Based on a One-Dimensional Ocean Model

Surface waves have a considerable effect on vertical mixing in the upper ocean.In the past two decades,the vertical mixing induced through nonbreaking surface waves has been used in ocean and climate models to improve the simulation of the upper ocean.Thus far,several nonbreaking wave-induced mixing parameterization schemes have been proposed;however,no quantitative comparison has been performed among them.In this paper,a one-dimensional ocean model was used to compare the performances of five schemes,including those of Qiao et al.(Q),Hu and Wang(HW),Huang and Qiao(HQ),Pleskachevsky et al.(P),and Ghantous and Babanin(GB).Similar to previous studies,all of these schemes can decrease the simulated sea surface temperature(SST),increase the subsurface temperature,and deepen the mixed layer,thereby alleviating the common thermal deviation problem of the ocean model for upper ocean simulation.Among these schemes,the HQ scheme exhibited the weakest wave-induced mixing effect,and the HW scheme exhibited the strongest effect;the other three schemes exhibited roughly the same effect.In particular,the Q and P schemes exhibited nearly the same effect.In the simulation based on observations from the Ocean Weather Station Papa,the HQ scheme exhibited the best performance,followed by the Q scheme.In the experiment with the HQ scheme,the root-mean-square deviation of the simulated SST from the observations was 0.43℃,and the mixed layer depth(MLD)was 2.0 m.As a contrast,the deviations of the SST and MLD reached 1.25℃ and 8.4 m,respectively,in the experiment without wave-induced mixing.

Utilizing Bayesian Modeling and MCMC for Accurate Characterization of Naturally Occurring Radionuclides Reference Background Levels in Mining Areas

Statistical biases may be introduced by imprecisely quantifying background radiation reference levels. It is, therefore, imperative to devise a simple, adaptable approach for precisely describing the reference background levels of naturally occurring radionuclides (NOR) in mining sites. As a substitute statistical method, we suggest using Bayesian modeling in this work to examine the spatial distribution of NOR. For naturally occurring gamma-induced radionuclides like 232Th, 40K, and 238U, statistical parameters are inferred using the Markov Chain Monte Carlo (MCMC) method. After obtaining an accurate subsample using bootstrapping, we exclude any possible outliers that fall outside of the Highest Density Interval (HDI). We use MCMC to build a Bayesian model with the resampled data and make predictions about the posterior distribution of radionuclides produced by gamma irradiation. This method offers a strong and dependable way to describe NOR reference background values, which is important for managing and evaluating radiation risks in mining contexts.

Numerical prediction of inner turbulent flow in conical diffuser by using a new five-point scheme and DLR k-ε turbulence model

The internal turbulent flow in conical diffuser is a very complicated adverse pressure gradient flow.DLR k-ε turbulence model was adopted to study it.The every terms of the Laplace operator in DLR k-ε turbulence model and pressure Poisson equation were discretized by upwind difference scheme.A new full implicit difference scheme of 5-point was constructed by using finite volume method and finite difference method.A large sparse matrix with five diagonals was formed and was stored by three arrays of one dimension in a compressed mode.General iterative methods do not work wel1 with large sparse matrix.With algebraic multigrid method(AMG),linear algebraic system of equations was solved and the precision was set at 10-6.The computation results were compared with the experimental results.The results show that the computation results have a good agreement with the experiment data.The precision of computational results and numerical simulation efficiency are greatly improved.

Assessment of k–ε models using tetrahedral grids to describe the turbulent flow field of a PBT impeller and validation through the PIV technique

In turbulence modeling, the RNG and Realizable models have important improvements in the turbulent production and dissipation terms in comparison to the Standard. The selection of the appropriate turbulence model has an impact on the convergence and solution in STRs, and they are used in mixing, multiphase modeling or as starting solution of transient models as DES and LES. Although there are several studies with the pitched blade turbine(PBT) impeller, most of them used the Standard model as representative of all k–ε models, using structured hexahedral grids composed of low number of cells, and in some cases under axial symmetry assumptions.Accordingly, in this work the assessment of the Standard, RNG and Realizable models to describe the turbulent flow field of this impeller, using the Multiple Reference Frame(MRF) and Sliding Mesh(SM) approaches with tetrahedral domains in dense grids, is presented. This kind of cell elements is especially suitable to reproduce complex geometries. Flow velocities and turbulent parameters were verified experimentally by PIV and torque measurements. The three models were capable of predicting fairly the pumping number, the power number based on torque, and velocities. Although the RNG improved the predictions of the turbulent kinetic energy and dissipation rate, the Realizable model presented better performance for both approaches. All models failed in the prediction of the total dissipation rate, and a dependence of its value on the number of cells for the MRF was found.

Heat Transfer and Flow Characteristics Predictions with a Refined k-ε-f_u Turbulent Model in Impinging Jet

Local heat transfer and flow characteristics in a round turbulent impinging jet for Re≈23 000 is predicted numerically with the RANS approach and a k-ε-fu turbulence model. The heat transfer predictions and turbulence parameters are verified against the axis-symmetric free jet impingement measurements and compared with previous other turbulence models, and results show the k-ε-fu model has a good performance in predictions of the local wall heat transfer coefficient, and in agreement with measurements in mean velocity profiles at different radial positions as well. The numerical model is further used to examine the effect of the fully confined impingement jet on the local Nusselt number. Local Nusselt profiles in x and y-centerlines for the target plate over three separation distances are predicted. Compared with the experimental data, the numerical results are accurate in the central domain around the stagnation region and present a consistent structure distribution.

Numerical analysis of turbulent mixed convection air flow in inclined plane channel with k-εtype turbulence model

Numerical study on turbulent mixed convection in inclined plane channels,from 15° to 90° (vertical),was carried out to examine the effect of inclination on fluid flow and heat transfer distributions.The turbulent air flows upward or downward into the duct with one wall heated from bottom.Calculation results with several kinds of k-εtype turbulence models were used to compare the experimental data with those in literatures to determine suitable model.The dependents of Nusselt number on the inclination angle of both the buoyancy-aided and buoyancy-opposed flow are discussed.

Solid Boundary as Energy Source and Sink in a Dry Granular Dense Flow: A Comparison between Two Turbulent Closure Models

Solid boundary as energy source and sink of the turbulent kinetic energy of the grains, and its influence on the mean and turbulent features of a dry granular dense flow, are investigated by using the proposed zero- and first-order turbulent closure models. The first and second laws of thermodynamics are used to derive the equilibrium closure relations, with the dynamic responses postulated by a quasi-static theory for weak turbulent intensity. Two closure models are applied to analyses of a gravity-driven flow down an inclined moving plane. While the calculated mean porosity and velocity correspond to the experimental outcomes, the influence of the turbulent eddy evolution can be taken into account in the first-order model. Increasing velocity slip on the inclined plane tends to enhance the turbulent dissipation nearby, and the turbulent kinetic energy near the free surface. The turbulent dissipation demonstrates a similarity with that of Newtonian fluids in turbulent boundary layer flows. While two-fold roles of the solid boundary are apparent in the first-order model, its role as an energy sink is more obvious in the zero-order model.

Solution of Self-similar Equations of the k-ε Model in the Shear Turbulent Mixing Problem and Its Numerical Simulation

The paper presents the k-ε model equations of turbulence with a single set of constants chosen by the authors, which is appropriate to simulate a wide range of turbulent flows. The model validation has been performed for a number of flows and its main results are given in the paper. The turbulent mixing of flow with shear in the tangential velocity component is discussed in details. An analytical solution to the system of ordinary differential equations of the k-ε model of turbulent mixing has been found for the self-similar regime of flow. The model coefficients were chosen using simulation results for some simplest turbulent flows. The solution can be used for the verification of codes. The numerical simulation of the problem has been performed by the 2D code EGAK using this model. A good agreement of the numerical simulation results with the self-similar solution, 3D DNS results and known experimental data has been achieved. This allows stating that the k-ε model constants chosen by the authors are acceptable for the considered flow.

基于邻域K-shell分布的关键节点识别方法

复杂网络中关键节点的精准识别对于网络结构稳定和信息传播起着至关重要的作用。传统K-shell方法仅通过节点在网络中所处位置对节点的重要性进行评估,导致区分度不高。基于此,综合考虑了节点的全局信息和局部信息对节点重要性的影响,提出一种基于邻域K-shell分布的关键节点识别方法。该方法通过节点邻域Ks值定义节点的熵,从而反映邻居节点的K-shell分布特征。通过11个网络数据集上的仿真实验,验证了所提方法能够更准确地识别并区分复杂网络中的关键节点。

基于K-means算法的建筑群震害分析模型缩减方法

基于建筑群模型和弹塑性时程分析的精细化城市震害模拟技术能够为防震减灾及应急救援决策提供必要的依据和参考。为了减小城市建筑群震害模拟的计算量和计算时间,本文提出一种基于聚类算法的建筑群模型缩减方法。该方法采用K-means聚类算法,首先基于建筑结构属性向量对建筑群进行聚类,将相似的建筑结构聚为一组;然后从每组选取一个代表建筑组成建筑群缩减模型,通过减少需要分析的建筑结构数量来减少建筑群震害模拟的计算量。本文对传统的K-means算法进行改进,通过设定组内建筑结构的差异上限自动调整聚类分组数量;提出将具体地震动作用下结构地震损伤指数作为结构属性向量进行聚类,并通过算例对比分别采用两种缩减模型,即基于损伤指数聚类的缩减模型与基于结构力学模型参数聚类的缩减模型,计算结构损伤状态准确程度。对比结果表明:在聚类分组数量相同的情况下,基于损伤指数的分组明显优于基于模型参数的分组,采用模型缩减方法能够在保证足够计算精度前提下显著减少建筑群震害模拟计算量和计算时间。

基于改进N-K模型的地铁盾构掘进安全风险耦合研究

为预防和控制地铁盾构掘进施工过程中的关键安全风险,并精准判断哪些风险耦合情境是导致事故发生的显著情境,提出改进N-K模型研究地铁盾构掘进安全风险耦合;综合运用文献研究、事故案例、专家访谈等方法,辨识地铁盾构掘进的关键安全风险因素;基于N-K模型提出新的地铁盾构掘进安全风险耦合评估模型,并选用安全事故案例验证该模型的适用性。结果表明:辨识得到地铁盾构掘进关键安全风险因素清单,包括4类一级风险因素,21个二级风险因素;地铁盾构掘进施工安全风险随着耦合因素种类的增加而变大,4因素风险耦合值最高,3因素风险耦合值次之,双因素风险耦合值最低,作业人员安全意识薄弱和机械故障参与作用的耦合情境更容易发生安全事故。

基于DEMATEL/N-K的机坪管制系统运行风险因素耦合分析

为了分析机坪管制系统风险因素的耦合性,将决策实验室分析(DEMATEL)模型与N-K模型相结合形成新的风险因素耦合分析模型。首先,利用N-K模型对2017—2019年机坪管制运行不安全事件进行耦合分析,计算不同耦合方式的耦合值;然后,利用DEMATEL模型计算安全风险因素的影响度和被影响度,确定要素的中心度排序,通过DEMATEL模型得到的综合影响矩阵计算可达矩阵,用来分析风险因素的可达性,最后,利用风险因素耦合值对各风险节点的中心度进行修正,获得关键的风险因素。基于实际运行数据,通过DEMATEL/N-K模型分析机坪管制系统的关键风险因素。结果表明,人为因素和机械因素之间的耦合程度明显更高,即更容易导致事故的发生,其中地面保障人员操作出错或未有效观察航空器造成航空器或地面保障设备故障是重点防控的对象。此外,在夜晚或雨天等视线条件下工作,更容易发生不安全事件,应当采取措施控制此类风险。

老年人家庭代际关系模式及其影响因素研究——基于代际双向报告的K均值聚类分析

在传统的代际关系模式研究中,数据往往来自家庭中父母或子女一方的报告,然而仅凭借“一面之词”对代际关系进行测量真的准确吗?这一问题的背后透视出当前国内相关研究中双向报告视角的缺失。利用2020年中国家庭追踪调查(CFPS)数据,运用K均值聚类分析及多层Logistic回归方法,探究当前中国老年人家庭中代际关系的双向报告状况、代际关系模式及其影响因素。研究结果表明:首先,老年父母的代际关系报告平均得分要略低于成年子女的报告平均得分,但两者间并不存在显著差异,代际利益假说未获支持。其次,基于双向报告的聚类结果,可将老年人家庭代际关系具体划分为结构维度上的“亲子一致型”、“亲近子疏型”、“亲疏子近型”模式与水平维度上的“亲子和睦型”、“亲子有隙型”、“亲子疏离型”模式。最后,从代际关系模式的影响因素来看,代际结构、老年父母特征及成年子女特征均会在结构维度与水平维度上产生显著影响,而家庭特征仅在水平维度上产生显著影响。简言之,两代人的年龄差距越大、父母年龄越小、子女年龄越大,越可能形成评价一致的代际关系;两代人的年龄差距越大、两代人的性别一致、父母年龄越小、父母获得子女支持以及在城镇家庭中,越可能形成和睦的代际关系。

融合N-K模型的复杂网络船舶自沉事故风险因素耦合分析

为定量分析船舶自沉风险因素间的影响关系,识别导致船舶自沉事故的关键因素,科学预防事故的发生,引入融合N-K模型的复杂网络研究船舶自沉事故风险耦合。首先结合中国海事局公布的136起船舶自沉事故案例,分析事故致因,将船舶自沉事故风险因素归纳为4个一级风险因素和15个二级风险因素,运用N-K模型计算出一级风险因素风险耦合的发生概率和风险值;然后,以二级风险因素为节点、致因关联为边,构建危险因子的关联网络,通过风险可达性分析和网络节点中心度分析,探究危险因子的作用机制,对危险因子进行初步识别,并以N-K模型计算的耦合值对节点中心度进行改进,获得最终的关键风险因素;最后,挖掘船舶自沉事故致因网络的凝聚子群并进行分析,得到密度矩阵,确定风险关联性最强的二级风险因素,以期从事故源头上采取有效措施,为船舶自沉事故的科学预防提供有益参考。结果表明:船舶自沉事故的发生与风险耦合值成正比,耦合因素越多则风险值越大;人的因素和船舶因素风险耦合易导致船舶自沉事故;导致船舶自沉事故的关键风险因素为安全意识淡薄、公司未履责、船舶管理不到位、公司管理不到位、船舶故障、船舶不适航,其中安全意识淡薄与其他风险关联性最大,须重点防范。