Integration Of Computational Fluid Dynamics Research Articles

Computationally effective machine learning approach for modular thermal energy storage design

This research presents an innovative approach that integrates computational fluid dynamics (CFD) and machine learning (ML) for the design and optimization of thermal energy storage (TES) systems. Heat discharging parametric analyses conducted using CFD serve as the basis for training ML models, including linear regression, K-nearest neighbor (KNN) regression, gradient boost regression (GBR), XGBoost, LightGBM, and neural network (NN). NN emerges as the most suitable for predicting time-dependent variations of concrete and heat transfer fluid (HTF) temperatures. The trained ML models offer an efficient alternative to traditional CFD simulations, enabling the prediction of temperatures in concrete thermal energy storage (CTES) modules under varying inlet conditions, velocities, and time. Leveraging these ML models, the research demonstrates the design of modular CTES cascaded systems with multiple modules in series and parallel configurations, significantly reducing computational cost and time by over 99% compared to full-scale CFD simulations. For instance, in predicting 4-hour time-dependent thermal behavior, CFD takes 97 s per data point and 238,500 s for a single module, compared to ML models’ 16-20 ms per data point and around 290 s per module, indicating their efficiency and scalability in predicting thermal discharge, especially for modular CTES system design and optimization. ML models also demonstrate computational efficiency for designing CTES systems involving multiple modules, taking approximately 765 s - 1047 s for various CTES system configurations, indicating their effectiveness over CFD in predicting thermal discharge for modular CTES systems. The integration of CFD and ML provides a streamlined workflow for designing and optimizing CTES systems, reducing computational efforts, cost, and time. Moreover, this workflow can be updated with additional training data to implement it for unique modular designs with different conditions. Such a generalization of this ML-based approach makes it applicable to a wide range of thermal energy storage designs and geometries, offering a promising avenue for future research and development in the field of thermal energy storage.

Blood vessel wall shear stress determines regions of liposome accumulation in angiogenic vasculature.

Nanoparticles used for drug delivery often require intravenous administration exposing them to fluid forces within the vasculature, yet the impact of blood flow on nanoparticle delivery remains incompletely understood. Here, we utilized transgenic zebrafish embryos to investigate the relationship between the accumulation of fluorescently labeled PEGylated liposomes and various hemodynamic factors (such as flow velocity, wall shear stress (WSS), and flow pattern) across a wide range of angiogenic blood vessels. We reconstructed 3D models of vascular structures from confocal images and used computational fluid dynamics to calculate local WSS, velocities, and define flow patterns. The spatial distribution of fluorescently labeled liposomes was subsequently mapped within the same 3D space and correlated with local hemodynamic parameters. Through the integration of computational fluid dynamics and in vivo experimentation, we show that liposomes accumulated in vessel regions with WSS between 0.1-0.8 Pa, displaying an inverse linear correlation (R2 > 0.85) between time-averaged wall shear stress and liposome localization in vivo. Interestingly, flow pattern did not appear to impact liposome accumulation. Collectively, our findings suggest the potential of stealth liposomes for passive targeting of low-flow vasculature, including capillaries and intricate angiogenic vasculature resembling that of tumor vessel networks.

Optimization of olive pomace dehydration process through the integration of computational fluid dynamics and deep learning

ABSTRACT This study introduces an innovative approach to optimize the dehydration process of olive pomace by combining computational fluid dynamics (CFD) and deep learning. Through CFD, it identifies the optimal air inlet velocity in a prototype of a passive direct solar dehydrator for olive pomace, which allows for the reduction of its moisture content for subsequent use as biomass. The prototype was simulated in ANSYS software, and this simulation consisted of the following steps: prototype design, meshing, selection of physical models, material assignment, boundary condition simulation, and validation of results with data obtained from the prototype. Following this process, it was concluded that the optimal air inlet velocity to the dehydration chamber is 0.1 m/s. Concurrently, an artificial neural network model was used to analyze data from sensors in the physical prototype, revealing that solar radiation and ambient temperature are the most influential variables on the temperature of the dehydration chamber. This analysis resulted in a predictive model for the optimization of the dehydration process, with a correlation coefficient of 0.9699 for temp_up and 0.9710 for temp_down, and a Willmott coefficient of 0.9999, demonstrating a high concordance between the model’s predictions and the experimental data. The model’s input variables include solar radiation, ambient temperature, and both external and internal air humidity. This integration of CFD and deep learning offers a promising methodology for improving olive pomace dehydration systems and the industry in general.

Computational Fluid Dynamics-Discrete Element Method Modeling and Analysis of Proppant Transportation Inside Different Degrees of Inclined Tortuous Hydraulic Fractures

Summary Optimal proppant transportation and distribution in inclined fractures play a critical role in maximizing the flow conductivity of complex fracture networks in unconventional gas/oil reservoirs. However, existing fracture models have ignored the effect of the tortuosity of inclined fractures, and more efforts are needed in this regard. To address this gap, a comprehensive integration of computational fluid dynamics and the discrete element method (CFD-DEM) is used in this study to numerically simulate the behavior of proppant transport in inclined fractures with tortuous shapes. The results reveal that, as the inclination angle varies, the proppant transport distance and packing height exhibit a nearly linear trend in straight fractures, while, in tortuous fractures, they follow a nonlinear pattern. Additionally, the proppant velocity, fluid velocity, and proppant force chain within inclined fractures with tortuous shapes exhibit significant disparities when compared with their straight counterparts.

Safety analysis of proton exchange membrane water electrolysis system

This study presents a novel causality model for the safety analysis of proton exchange membrane water electrolysis (PEMWE). The model is based on the integration of computational fluid dynamics (CFD)-based and Bayesian network (BN)-based modeling approaches. The model is aimed at analyzing evolving hazard scenarios, such as gas permeation/crossover (GP) during PEMWE based on fluid dynamics and electrochemistry of electrolysis. The probabilistic domain is employed to model these scenarios using a BN-based technique to assess the likelihood of hazardous scenario occurrence for the first time. A sensitivity analysis was performed, and it was found that operating pressure is the most significant contributor to GP. Operating cell voltage and pressure are two dominant parameters that can trigger GP with the largest likelihood of 87%. Therefore, gas combinators serve as the most effective safety barrier in avoiding the formation of hazardous H2/O2 gas mixtures. This work provides a robust framework for analyzing accident scenarios and developing safety management strategies. It is expected that this model will help in the development of safer PEMWE systems, which is of great importance for various industrial applications.

Integration of Computational Fluid Dynamics and Artificial Neural Network for Optimization Design of Battery Thermal Management System

The increasing popularity of lithium-ion battery systems, particularly in electric vehicles and energy storage systems, has gained broad research interest regarding performance optimization, thermal stability, and fire safety. To enhance the battery thermal management system, a comprehensive investigation of the thermal behaviour and heat exchange process of battery systems is paramount. In this paper, a three-dimensional electro-thermal model coupled with fluid dynamics module was developed to comprehensively analyze the temperature distribution of battery packs and the heat carried away. The computational fluid dynamics (CFD) simulation results of the lumped battery model were validated and verified by considering natural ventilation speed and ambient temperature. In the artificial neural networks (ANN) model, the multilayer perceptron was applied to train the numerical outputs and optimal design of the battery setup, achieving a 1.9% decrease in maximum temperature and a 4.5% drop in temperature difference. The simulation results provide a practical compromise in optimizing the battery configuration and cooling efficiency, balancing the layout of the battery system, and safety performance. The present modelling framework demonstrates an innovative approach to utilizing high-fidelity electro-thermal/CFD numerical inputs for ANN optimization, potentially enhancing the state-of-art thermal management and reducing the risks of thermal runaway and fire outbreaks.

Machine Learning Techniques in Structural Wind Engineering: A State-of-the-Art Review

Machine learning (ML) techniques, which are a subset of artificial intelligence (AI), have played a crucial role across a wide spectrum of disciplines, including engineering, over the last decades. The promise of using ML is due to its ability to learn from given data, identify patterns, and accordingly make decisions or predictions without being specifically programmed to do so. This paper provides a comprehensive state-of-the-art review of the implementation of ML techniques in the structural wind engineering domain and presents the most promising methods and applications in this field, such as regression trees, random forest, neural networks, etc. The existing literature was reviewed and categorized into three main traits: (1) prediction of wind-induced pressure/velocities on different structures using data from experimental studies, (2) integration of computational fluid dynamics (CFD) models with ML models for wind load prediction, and (3) assessment of the aeroelastic response of structures, such as buildings and bridges, using ML. Overall, the review identified that some of the examined studies show satisfactory and promising results in predicting wind load and aeroelastic responses while others showed less conservative results compared to the experimental data. The review demonstrates that the artificial neural network (ANN) is the most powerful tool that is widely used in wind engineering applications, but the paper still identifies other powerful ML models as well for prospective operations and future research.

Designing a Pavilion that Generates Electricity

This paper demonstrates a design method that integrates various computational tools such as Rhinoceros, the artificial neural network from MATLAB, and computational fluid dynamics from Eddy3D to search for the optimal aerodynamic geometries for a wind turbine. It introduces a site-specific microclimate analysis method that can maximize site-specific wind energy potential. Through the integration of computational fluid dynamics (CFD) and artificial neural networks (ANN), the study was able to find the optimized shape to maximize the wind potential for the specific test site. These ANN models use fewer computational resources and less time with reasonable average regression values up to 0.96. The result shows improvement of the annual hourly wind speed around the wind turbine up to 13.24 m/s. It would be beneficial to test the proposed method with actual performance to improve the proposed method.

A simulation-based design framework to iteratively analyze and shape urban landscapes using point cloud modeling

The topic of this paper evolves on the discourse of digital modeling in landscape design. Current design methods stagger to address physical forms and dynamics present in the environment. This status quo limits possibilities to integrate scientific evidence when developing spatial and aesthetic configurations in urban landscapes. Remote sensing technology such as laser scanning measures physical forms to reproduce them as geo-specific digital 3D models, while dynamic simulation is widely used to predict how scenarios will perform under given conditions. However, there is still a need for a holistic design process that is capable of integrating both the measured physical forms and physical dynamics. This paper presents a novel framework using point cloud modeling to shape design scenarios that are iteratively evaluated for their performance.The proposed framework is demonstrated through a case study in Singapore. New spatial configurations are tested for the site through an iterative and comparative analysis of the design performance. The case study exposes (1) a site-specific design approach by iteratively modeling a laser-scanned point cloud model, (2) a workflow to convert the geometric data from the point cloud models into voxels and meshes, (3) an integration of computational fluid dynamics (CFD) simulation during design development as per-point attributes, and (4) a comparison of the configurations to identify best performing scenarios.This design framework can support city managers, planners, urban and landscape designers to better inform their decision-making process by relying on accurate scientific feedback. By guiding the design process with the consideration of the built environment as a complex adaptive system, it will be possible to improve how open spaces and ecosystem services perform in cities, and to design landscapes that can mitigate dynamic events such as urban heat islands.

Simulation and Experiment of Gas-Solid Flow in a Safflower Sorting Device Based on the CFD-DEM Coupling Method

To study the movement characteristics and separation mechanism of safflower petals and their impurities under the action of airflow and lower the impurity rate in the cleaning operation process, integration of computational fluid dynamics (CFD) and discrete element method (DEM) codes was performed to study the motion and sorting behavior of impurity particles and safflower petals under different airflow inclination angles, dust removal angles and inlet airflow velocities by establishing a true particle model. In this model, the discrete particle phase was applied by the DEM software, and the continuum gas phase was described by the ANSYS Fluent software. The Box-Behnken experimental design with three factors and three levels was performed, and parameters such as inlet airflow velocity, airflow inclined angle, and dust remover angle were selected as independent variables that would influence the cleaning impurity rate and the cleaning loss rate. A mathematical model was established, and then the effects of various parameters and their interactions were analyzed. The test results show that the cleaning effect is best when the inlet airflow velocity is 7 m/s, the airflow inclined angle is 0°, and the dust remover angle is 25°. Confirmatory tests showed that the average cleaning impurity rate and cleaning loss rate were 0.69% and 2.75%, respectively, which dropped significantly compared with those from previous optimization. An experimental device was designed and set up; the experimental results were consistent with the simulation results, indicating that studying the physical behavior of safflower petals-impurity separation in the airflow field by using the DEM-CFD coupling method is reliable. This result provides a basis for follow-up studies of separation and cleaning devices for lightweight materials such as safflower petals.

Comprehensive study on boil-off gas generation from LNG road tankers under simultaneous impacts of heat leakage and transportation vibration

The minimization and utilization of boil-off gas (BOG) during the production, storage, and transportation of liquefied natural gas (LNG) have drawn wide interests. However, current studies on quantification of BOG generation still need to be improved because the mechanical vibration impact on BOG generation has seldom been considered in most of LNG studies. In this paper, the dynamic BOG generation under both heat leakage and vibration impact during the LNG on-road transportation have been quantitatively investigated via the integration of computational fluid dynamics (CFD) and process dynamic simulations. First, a multiphase LNG model is developed to characterize the thermal energy increment due to on-road vibration effect by CFD simulations, where a user-defined function (UDF) is employed with the dynamic mesh algorithm. Next, the vibration-induced thermal energy is introduced into a process dynamic model, where a heat-leakage considered LNG tanker is simulated to quantify the total BOG generation during the LNG on-road transportation. This work for the first time addresses two root causes of BOG generation with both CFD and process dynamic simulations, which lays a solid foundation for future BOG minimization and utilization during various LNG transportation, storage, and applications.

Experimental and computational investigation of energy ball wind turbine aerodynamic performance

Small-scale wind turbines with innovative design are introduced for small applications, providing clean renewable energy to rural homes, street lighting, and hybrid systems. Energy ball wind turbine, known as Venturi wind turbine, has untraditional blades’ shape and special aerodynamic behavior that creates a venturi effect on the air stream passing through its aspherical shape. This article represents an integration of computational fluid dynamics and wind tunnel experimentation to study the aerodynamic performance of a manufactured model of energy ball wind turbine. Physical models with different twist angles were fabricated and tested in a small wind test section. In these experiments, dynamic torque, angular velocity, and coefficient of performance values were measured at different speeds. The experimental power coefficient results were discussed showing the best-tested twist angle. Fluid flow simulation has been developed in ANSYS FLUENT software. The findings of these numerical simulations have provided pressure contour, velocity contour, and torque values which help to study the solidity effect on turbine’s power coefficient. Nevertheless, the velocity contours provided from the computational analysis ensure the Venturi effect of the energy ball wind turbine design.

High Resolution Urban Air Quality Modeling by Coupling CFD and Mesoscale Models: a Review

According to World Health Organization, 9 out of 10 people breathe polluted air and the ambient air pollution accounts for nearly 4.2 million early deaths worldwide. There is an urgent need for scientific management of urban air systems. Mathematical modeling of air quality helps the researchers and urban authorities in devising scientific management plans for mitigation of the associated impacts. We present an organized review of the broad aspects related to urban air quality modeling such as – urban microclimate, geospatial data, chemical transport models, computational fluid dynamics (CFD) models and integration of CFD and mesoscale models. The paper also discusses about the influence of urban land scape features on air quality, accuracy of emission inventory and model validation methods. The present review provides a vantage point to the researchers in the emerging field of high resolution urban air quality modeling for devising the location specific mitigation plans for the scientific management of the clean air.

A Study on the Statistical Calibration of the Holtrop and Mennen Approximate Power Prediction Method for Full Hull Form, Low Froude Number Vessels

The article herein presents a statistical calibration study of the approximate power method of Holtrop and Mennen focused on adapting the method to vessels characterized as "full" hull forms and low design and operating speeds and, thus, low Froude numbers. The fitting of the method is done by adjusting the constants, coefficients, and components of the method's equations by a systematic variation process controlled by genetic algorithms. The database that the method is calibrated against is consisted by model test results from modern (built between 2010 and 2016) bulk carriers and tankers, the KVLCC2, and the method follows a multistage approach, calibrating first the model for the prediction of total resistance and applying the selfpropulsion equations afterward. The uncertainty of the new improved method is assessed and modeled with a nonlinear regression equation to enable the use of the calibrated method in the early ship design and optimization process. 1. Introduction One of the most important aspects of the design and study of vessels and ocean structures is the prediction of resistance and, thus, powering requirements for a range of operational speeds. Given the constant societal and legislative pressure for the reduction of the carbon footprint of shipping and maritime activities, the accurate and reliable powering prediction from an early stage and tightly integrated within the process itself as the respective optimization studies is imperative. Although towing tank model testing has been since early times the most reliable method for power predictions, because of its high cost and demand for a fixed and given geometry it can be used only in the basic design stage when the design parameters related to the vessel's hull geometry are fixed and serves as a final validation and benchmark to be used afterward in the conclusion of the shipbuilding process during sea trials. On the other hand, the last decade has seen the exponential growth of computational fluid dynamic solvers that solve Reynolds Averaged Navier Stokes equations over the hull form in finite volume approaches (Schneekluth & Bertram 1998; Specialist Committee on CFD 2014). Although originally the computational cost was penalizing its application in early design stages, the advances in computing hardware and software allowed the integration of Computational Fluid Dynamics (CFD) in the early ship hull form design and optimization (Specialist Committee on CFD 2014). Such methods, however, can be used in applications where variables such as the principal particulars have been fixed or have a small variance, with the focus being on the variation in local geometry and topological characteristics that have a significant effect on local flow phenomena (Specialist Committee on CFD 2014). However, when the application in question is a global optimization study, where there is a large variance in principal particulars (vessel's dimensions) and structural and cargo arrangements, there is an apparent need for a large number of optimization variants and, thus, evaluations (usually in the order of thousands). The application of CFD in such cases is impractical as the optimization algorithm will eventually require hundreds of thousands computational hours even in high-performance computers; therefore, empirical or statistical methods are better suited. The most prominent of these is the approximate power prediction method by Holtrop and Mennen (1982) together with its revision (Holtrop 1984). Although this methodology provides sufficient accuracy, the statistical sample of the hull forms on which it is based dates back to the 1970s and 1980s. Such hulls, although roughly similar, have some distinct deviations from modern commercial vessels. In the article presented herein, the authors attempt to make a calibration of the Holtrop and Mennen methodology via a systematic variation with the use of genetic algorithms. The calibration is based on a statistical sample of model test results of low Froude number (Fn) full hull forms (with Cb greater than 0.7) of modern commercial bulk carriers and tankers, and its focus is on the integration of the new coefficients in a holistic methodology for the optimization of large bulk carriers. The uncertainty of the new coefficients is also taken into account based on the sea trial results of an expanded statistical sample.

Prediction of multi-inputs bubble column reactor using a novel hybrid model of computational fluid dynamics and machine learning

ABSTRACTThe combination of machine learning and numerical methods has recently become popular in the prediction of macroscopic and microscopic hydrodynamics parameters of bubble column reactors. Such numerical combination can develop a smart multiphase bubble column reactor with the ability of low-cost computational time when considering the big data. However, the accuracy of such models should be improved by optimizing the data parameters. This paper uses an adaptive-network-based fuzzy inference system (ANFIS) to train four big data inputs with a novel integration of computational fluid dynamics (CFD) model of gas. The results show that the increasing number of input variables improves the intelligence of the ANFIS method up to , and the number of rules during the learning process has a significant effect on the accuracy of this type of modeling. Furthermore, the proper selection of model’s parameters results in higher accuracy in the prediction of the flow characteristics in the column structure.

Advanced integration of fluid dynamics and photosynthetic reaction kinetics for microalgae culture systems

BackgroundPhotosynthetic microalgae have been in the spotlight of biotechnological production (biofuels, lipids, etc), however, current barriers in mass cultivation of microalgae are limiting its successful industrialization. Therefore, a mathematical model integrating both the biological and hydrodynamical parts of the cultivation process may improve our understanding of relevant phenomena, leading to further optimization of the microalgae cultivation.ResultsWe introduce a unified multidisciplinary simulation tool for microalgae culture systems, particularly the photobioreactors. Our approach describes changes of cell growth determined by dynamics of heterogeneous environmental conditions such as irradiation and mixing of the culture. Presented framework consists of (i) a simplified model of microalgae growth in a culture system (the advection-diffusion-reaction system within a phenomenological model of photosynthesis and photoinhibition), (ii) the fluid dynamics (Navier-Stokes equations), and (iii) the irradiance field description (Beer-Lambert law). To validate the method, a simple case study leading to hydrodynamically induced fluctuating light conditions was chosen. The integration of computational fluid dynamics (ANSYS Fluent) revealed the inner property of the system, the flashing light enhancement phenomenon, known from experiments.ConclusionOur physically accurate model of microalgae culture naturally exhibits features of real system, can be applied to any geometry of microalgae mass cultivation and thus is suitable for biotechnological applications.

Integration of CFD and polymerization for an industrial scale cis-polybutadiene reactor

ABSTRACTA computational fluid dynamics (CFD) approach, coupled with anionic polymerization kinetics, was used to investigate the solution polymerization in a 12 m3 industrial scale cis-polybutadiene reactor. The kinetic model with double catalytic active sites was integrated with CFD by a user-defined function. The coupled model was successfully validated by the plant data and then used to investigate the key operating variables. Also, predictions of CFD model were compared with those of continuous stirred tank reactor (CSTR) model. Although the reaction mixture is well mixed in the middle and at the top of the reactor, there exists a poor mixing feeding zone at the bottom, which leads to serious deviations from the ideal CSTR. The polymerization process with nonideal mixing is very sensitive to the inlet temperature and the feeding rate. Enhancing the mixing performance in the feeding zone could be an effective way to improve the product quality.

Coronary CT Angiography-Derived Fractional Flow Reserve

Purpose of Review Coronary CT angiography has been shown to be highly diagnostically accurate as compared with invasive coronary angiography and help clinical decision making that affords improved clinical outcomes. Unfortunately, routine coronary CTA does not allow for the discrimination of the hemodynamic significance of stenosis. Recently, through the integration of computational fluid dynamics, Fractional Flow Reserve CT (FFRct) has been developed which allows for the determination of the hemodynamic significance of specific lesions from a resting coronary CTA without additional imaging, a change in protocol or the administration of adenosine.

Integration of CFD and RTD analysis in flow pattern and mixing behavior of rotary pressure exchanger with extended angle

In this study, the computational fluid dynamics (CFD) approach combined with the residence time distribution (RTD) analysis was implemented to examine the mixing performance and flow pattern of rotary pressure exchanger (RPE). Based on oscillatory Reynolds number, a flow regime classification was established for RPE. A concept of extended angle of RPE was proposed, and then, its effects on mixing behavior were evaluated by CFD simulation in laminar model. Meanwhile, flow pattern in RPE was quantified by RTD study, and was well captured in the flow field analysis. In addition, the effects of operating conditions on the mixing and flow pattern were discussed. According to the results, it was shown that the extended angle of RPE is beneficial for mixing control, and a minimum volumetric mixing rate was achieved when the extended angle is ±30° compared with other configurations. In different operating conditions, the mixing rate was minimized at an oscillatory Reynolds number of about 178. Moreover, the smaller RTD variance, the closer was the flow pattern to an ideal plug flow, leading to a lower volumetric mixing rate of RPE. This study indicates that the RTD and CFD simulation are capable for mixing study and flow analysis in RPE device, and they are complementary and verifiable with each other.

Modeling analysis of laser cladding of a nickel-based superalloy

A novel numerical modeling simulation, involving an integration of computational fluid dynamics (CFD), a physical energy function and the finite element method (FEM) to investigate the thermal history experienced by the powder, deposited coating and substrate during the laser cladding of a nickel-based superalloy is performed. The laser cladding process is simulated in three sub-steps through different models. First, a CFD model is developed to track the trajectory of the powder during the feeding procedure before the powder particles merge into a molten pool of the substrate. Then, a physical energy function is used to compute the absorbed energy of the powder under laser irradiation. Finally, an FEM model is employed to investigate the addition of material and thermal history in the cladding structure. The simulation results are used to predict the solidification microstructure of the deposited coating and the dependence of the microstructure on the processing parameters, such as powder size, laser power and cladding speed. It is found that powder that is small in size is favorable for obtaining fine solidification microstructure in the laser deposited coating. However, the degree of influence of the powder particle size on cooling rate and resultant secondary dendrite arm spacing is weakened by a combined increase of the laser power and cladding speed at a fixed ratio of the laser power to cladding speed. This work contributes to the comprehensive understanding of laser cladding, as well as provides a fundamental basis for controlling the microstructure in laser deposited nickel-based superalloy coatings.