Layout Optimization of Synchronously Operating Rotor Sails Using CFD and ANN Surrogate Modeling

Simulation of the dynamics and disposition of inhaled particles within human lungs is an invaluable tool in both the development of inhaled pharmacologic drugs and the risk assessment of environmental particulate matter (PM). The goal of the present focused study was to assess the utility of three-dimensional computational fluid dynamics (CFD) models in studying the local deposition patterns of PM in respiratory airways. CFD models were validated using data from published experimental studies in human lung casts. The ability of CFD to appropriately simulate trends in deposition patterns due to changing ventilatory conditions was specifically addressed. CFD simulations of airflow and particle motion were performed in a model of the trachea and main bronchi using Fluent Inc.'s FIDAP CFD software. Particle diameters of 8 microm were considered for input flow rates of 15 and 60 L/min. CFD was able to reproduce the observed spatial heterogeneities of deposition within the modeled bifurcations, and correctly predicted the "hot-spots" of particle deposition on carinal ridges. The CFD methods also predicted observed differences in deposition for high-versus-low flow rates. CFD models may provide an efficient means of studying the complex effects of airway geometry, particle characteristics, and ventilatory parameters on particle deposition and therefore aid in the design of human subject experiments.

Performance improvement of solid oxide fuel cells by combining three-dimensional CFD modeling, artificial neural network and genetic algorithm

Optimization methods have been widely applied to the aerodynamic design of gas turbine blades. While applying optimization to high-fidelity computational fluid dynamics (CFD) simulations has proven capable of improving engineering design performance, a challenge has been overcoming the prolonged run-time due to the computationally expensive CFD runs. Reduced-order models and, more recently, machine learning methods have been increasingly used in gas turbine studies to predict performance metrics and operational characteristics, model turbulence, and optimize designs. The application of machine learning methods allows for utilizing existing knowledge and datasets from different sources, such as previous experiments, CFD, low-fidelity simulations, 1D or system-level studies. The present study investigates inserting a machine learning model that utilizes such data into a high-fidelity CFD driven optimization process, and hence effectively reduces the number of required evaluations of the CFD model. Artificial Neural Network (ANN) models were trained on data from over three thousand two-dimensional (2D) CFD analyses of turbine blade cross-sections. The trained ANN models were then used as surrogates in a nested optimization process alongside a full three-dimensional Navier–Stokes CFD simulation. The much lower evaluation cost of the ANN model allows for tens of thousands of design evaluations to guide the search of the best blade profiles to be used in the more expensive, high-fidelity CFD runs, improving the progress of the optimization while reducing the required computation time. It is estimated that the current workflow achieves a five-fold reduction in computational time in comparison to an optimization process that is based on three-dimensional (3D) CFD simulations alone. The methodology is demonstrated on the NASA/General Electric Energy Efficient Engine (E3) high pressure turbine blade and found Pareto front designs with improved blade efficiency and power over the baseline. Quantitative analysis of the optimization data reveals that some design parameters in the present study are more influential than others, such as the lean angle and tip scaling factor. Examining the optimized designs also provides insight into the physics, showing that the optimized designs have a lower amount of pressure drop near the trailing edge, but have an earlier onset of pressure drop on the suction side surface when compared to the baseline design, contributing to the observed improvements in efficiency and power.

Study of wake characteristics of a vertical axis wind turbine by two- and three-dimensional computational fluid dynamics simulations

The marine large-bore low-speed engines face challenges such as complex testing and large computational workload. A feasible solution is to adopt a combined optimization approach using one-dimensional performance simulation and three-dimensional computational fluid dynamics (CFD) simulation. This paper utilizes a one-dimensional full-cycle simulation model of a marine large-bore low-speed engine with a complete scavenging and combustion coupling process. Combined with the Non-dominated Sorting Genetic Algorithm II (NSGA-II), it conducts multi-objective optimization based on the controlled variables such as the closing time of the exhaust valve, the injection timing, and the injection duration. Within a short period of time, it obtains the optimized boundary conditions. In the three-dimensional CFD simulation, the indicated thermal efficiency was increased by 1.139% and 1.264% respectively, and the maximum error compared with the one-dimensional simulation results was 1.33% and in the three-dimensional CFD simulation, the CO2 emissions were reduced by 2.2052% and 2.4886% respectively, with a maximum error of 1.69% compared to the one-dimensional simulation results. Ultimately, the results of the one-dimensional optimization were consistent with those of the three-dimensional CFD simulation. Moreover, the computing time required was significantly less than that of the traditional optimization based on three-dimensional CFD simulation.

This research aims to create an artificial neural network (ANN) regression model for predicting the performance parameters of the perforated micro-pin fin (MPF) heat sinks for various geometric parameters and inflow conditions. A three-dimensional computational fluid dynamics (CFD) simulation system is developed to generate dataset samples under different operational conditions, which are specified using Latin hypercube sampling (LHS). An ANN model is first obtained by optimizing the model hyper-parameters, which are then deployed to learn from the input feature space that consists of perforation diameter, perforation location, and inflow velocity. For accurate training of the ANN, the model is trained over a range of uniformly distributed data points in the input feature space. The developed multi-layer model predicted Nusselt number and friction factor with the mean absolute percentage error of 4.45% and 1.80%, respectively. Subsequently, the developed surrogate model is used in the optimization study to demonstrate the application of the surrogate model. A multi-objective non-dominated sorting genetic algorithm (NSGA-II) is used to perform the optimization of the perforation location, diameter, and inflow conditions. Negative of the Nusselt number and friction factor are chosen as objectives to minimize. A Pareto front is obtained from the optimization study that shows a set of optimal solutions. Thermal performance of the perforated MPF is increased between 11.5% and 39.77%. The optimizer selected a significantly smaller hole diameter at a higher location and a faster speed to maximize the Nusselt number and minimize the friction factor.

A stern slamming analysis based on three-dimensional computational fluid dynamics (CFD) simulation is presented with an application to a liquefied natural gas (LNG) carrier with twin skegs. This study includes; seakeeping analysis, statistical analysis for relative motions and velocities, three-dimensional slamming simulation by a CFD software, and structural assessment for plates and stiffeners. The stern areas are divided into panels in which relative velocity/motion and pressure coefficients are to be calculated. Seakeeping calculations are carried out in full load and ballast loading conditions at ship speeds of 0 and 5 knots. A series of equivalent 20-year return sea states in a wave scatter diagram are selected for environmental conditions. Extreme velocities are then evaluated from the loading conditions and the speeds considered with reference to the probability of slamming occurrence. Slamming simulations are carried out in a three-dimensional domain with a CFD software to calculate pressure coefficients. Two-phase flow with water and air is to be adopted in conjunction with free surface capturing method. Viscous laminar flow is assumed in simulation. Slamming design pressure is calculated by the pressure coefficients and the extreme velocities. Based on computed design pressure, an ultimate strength analysis is performed for the determination of required plate thickness. Also, required stiffener dimensions are determined by analytic formulas. As mentioned above, this approach has been applied to an LNG carrier with twin skegs. In the application, two-phase flow with water and air was adopted in conjunction with the volume-of-fluid method for free surface capturing. Mixed hexahedral and tetrahedral grids were employed. The computational case was determined from simulations of global ship motion. Maximum slamming pressure was found near the end of a skeg. Large pressure also can be observed in the stern overhang area. Generally slamming pressure decreases away from the stern.

ABSTRACTSedimentation is one of the most popular wastewater treatment processes, and is used to separate solid particles from carrier fluid in settling tanks known as clarifiers. The clarifier, as the last major facility in wastewater treatment plants (WWTPs), can limit or define the performance of the overall WWTP. This paper presents a novel three-dimensional unsteady computational fluid dynamics (CFD) model to improve the efficiency of an industrial clarifier that had been experiencing underperformance and reduction in wastewater handling capacity. We propose a numerical technique to address the transient process of removing sludge from the floor of clarifiers by using rotating rakes. The CFD model was first applied to analyzing the ramifications of the current clarifier geometry on performance. The results show that the root causes for underperformance are related to the unconventional top side feed design of the clarifier, which leads to significant asymmetry in the flow distribution. The CFD model was next used to investigate various design modifications with the goal of improving the clarifier performance. A few geometry modification ideas such as an inward baffle, dissipating inlets, and a submerged skirt were found to create a more uniform flow distribution in the clarifier, significantly reducing the backflow into the feedwell and the velocity of the flow exiting the feedwell, which helps the solid particles to settle in the clarifier. These three designs were found to reduce the effluent total suspended solids (TSS) by more than 80% and thus significantly improve clarifier performance. It is believed that the CFD model developed in this study can become a computationally efficient tool for investigating the performance of industrial clarifiers with complex geometries and rotating rakes.

A three-dimensional (3-D) computational fluid dynamics (CFD) model has been developed to analyze the liquid poison injection phenomenon of shutdown system 2 (SDS-2) of a Canada deuterium uranium (CANDU) reactor. Because the SDS-2 injects highly pressurized liquid poison into the moderator in a very short time, it is a major safety priority to confirm the effectiveness of the SDS-2 as one of the shutdown systems. In general, it is difficult to directly measure the velocity and concentration of the poison jet during an injection because of the complex nature of the injection system and the process. Therefore, a series of investigations has been performed to develop a CFD model for liquid poison injection phenomenon with limited validations. In this study, the validation of the existing CFD model for the poison injection phenomenon of the CANDU SDS-2 is extended to be applicable to a CANDU-6 reactor as well as a larger CANDU reactor. The analyses showed that the poison jet growth for those experiments simulated by the 3-D CFD model agrees reasonably with the experimental results. Therefore, it is concluded that the proposed 3-D CFD model can be used to assess the effectiveness of a liquid poison injection in compliance with the intended functional design requirements of the CANDU SDS-2.

CFD study on onset of liquid entrainment through ADS-4 branch line in AP1000

In this research, both numerically and experimentally, the thermal efficiency of a Printed Circuit Board (PCB) with two Light Emitting Diode (LED) chips was examined. The two LED lighting systems, which are single-cell LEDs, including PCB and copper plates, were manufactured, and tested under laboratory conditions to achieve this goal. The three-dimensional Computational Fluid Dynamics (CFD) model with natural convection effects prepared using the FloEFD software package to predict PCB surface temperature distributions. The goal was to perform comprehensive circuit board simulation and validate the numerical model built in this study using the experimental data during the studies. From the results, we can easily claim that higher temperature gradients are calculated and predicted near the LED chip because of heat generation. Data paths have played an essential role in the LED circuit board's temperature distribution. High-temperature variations are observed at short distances around the LED when the experimental and simulation results are compared. Temperature changes are minimized as they travel away from the LED chip. It is found that the error rate is below 5 percent overall between the experimental and simulation results. The numerical results were in proper alignment with numerical data obtained from the three-dimensional (3D) CFD model. Given thermal efficiency and using such models, this model can design and analyze Automotive Lighting Systems.

Three-dimensional computational fluid dynamics simulation of stirling engine

Experimental study and three-dimensional (3D) computational fluid dynamics (CFD) analysis on the effect of the convergence ratio, pressure inlet and number of nozzle intake on vortex tube performance–Validation and CFD optimization

To achieve high efficiency and low degradation of a polymer electrolyte fuel cell (PEMFC), it is necessary to maintain an appropriate level of humidification in the fuel cell membrane. Thus, membrane humidifiers are typically used in PEMFC systems. Parameter studies are important to evaluate membrane humidifiers under various operating conditions to reduce the amount of physical tests. However, simulative studies are computationally expensive when using detailed models. To reduce the computational cost, surrogate models are set up. In our study, a 3D computational fluid dynamics (CFD) model of a hollow fibre membrane humidifier is presented and validated using measurement data. Based on the results of the validated CFD model, a surrogate model of the humidifier is constructed using proper orthogonal decomposition (POD) in combination with different interpolation methods. To evaluate the surrogate models, their results are compared against reference solutions from the CFD model. Our results show that a Halton design combined with a thin-plate-spline interpolation results in the most accurate surrogate humidifier model. Its normalised mean absolute error for 18 test points when predicting the water mass fraction in the membrane humidifier is 0.58%. Furthermore, it is demonstrated that the solutions of the POD model can be used to initialise CFD calculations and thus accelerate the calculation of steady state CFD solutions.

<div class="section abstract"><div class="htmlview paragraph">The stable operation of turbocharger compressor at low flow rates is important to provide low end engine torque for turbocharged automotive engines. Therefore, it is important to be able to predict the lowest flow rates at different turbocharger speeds at which the surge phenomenon occurs. For this purpose, three-dimensional Computational Fluid Dynamics (CFD) simulations were performed on the turbocharger compressor including the entire compressor wheel and volute. The wheel consisted of six main and six splitter blades.</div><div class="htmlview paragraph">Historically, flow bench and engine testing has been used to detect surge phenomenon. However a complete 3D CFD analysis can be performed upfront in the design to calculate low end compressor surge performance. The analyses will help understand the fundamental mechanisms of stalled flow, the surge phenomenon, and impact of compressor inlet conditions on surge. To reduce computational time a small volume of compressed air system, essentially a small ‘B’ system as introduced by Greitzer [<span class="xref">1</span>], is used in the CFD model. CFD results have been compared with gas stand compressor maps and have good agreements for the compressor map points.</div><div class="htmlview paragraph">This paper presents a CFD analysis near the low flow region at constant turbo speed to predict automotive centrifugal compressor surge phenomenon.</div></div>