역전기투석 장치 내 스페이서의 영향에 관한 수치해석적 연구

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.

The article is directly related to the authors’ previous study of the remotely operated underwater vehicle’s (ROV) motion, where the mathematical model of the ROV motion is derived for laminar flow of water. This solution of the ROV motion has been improved by using the three-dimensional computational fluid dynamics (CFD) simulation that allows simulating the turbulent flow, which is mainly generated by propellers. Both methods were coupled to obtain more accurate values of hydrodynamic damping coefficients for the improved determination of the ROV motion velocity. Results of the solution are discussed and some of them are compared with the experiment.

One of the major barriers for polymer electrolyte membrane (PEM) fuel cells to be commercially viable for stationary and transport applications is the durability of membranes undergoing chemical and mechanical degradation over the period of operation. Toward understanding the effects of operating parameters on membrane durability, this paper presents numerical simulations for a single channel PEM fuel cell undergoing changes in load, by subjecting the unit cell to step change in voltage. The objective is to elucidate the mechanical response of membrane, which is subjected to hygral (water) loading and unloading cycles at constant temperature. Detailed three-dimensional (3D) computational fluid dynamics (CFD) simulations are conducted, taking in to account the complex interactions of water transport dynamics and load changes, to accurately capture the water content in the membrane with changes in cell voltage. The water content obtained through CFD simulations is, in turn, used to carry out two-dimensional (2D) finite element (FE) analysis to predict the mechanical response of the membrane undergoing cyclic change in water content, as the operating voltage is cycled. The effects of cyclic changes in cell potential on the stresses induced, amount of plastic strain and their localization are analyzed for various inlet cathode humidity values for two sections along the length of the fuel cell.

Rationale: We previously found oropharyngeal enlargement and subglottal narrowing in asthma and chronic obstructive pulmonary disease (COPD) patients compared to healthy individuals. This study aimed to investigate the impact of oropharyngeal remodeling on the upper airway airflow dynamics in association with lung structure-function impairment in obstructive lung disease (OLD). Methods: Upper airway computed tomography (CT) scans, inspiratory (total lung capacity) and expiratory (functional residual capacity) lung CTs, and clinical data were collected from 185 participants (26 healthy, 56.6±5.1y; 73 COPD, 71.1±7.5y, and 86 asthma, 64.8±12.5y) enrolled in Jeonbuk National University Hospital, Jeonju, Korea. We conducted quantitative inspiratory-expiratory CT analysis and one-dimensional (1D) computational fluid dynamics (CFD) simulations of airflow in the lower airways, for all participants. Then, we selected ten representative subjects in each group based on key features of upper airways (healthy: 56±6.2y; COPD: 69.7±7.4y; asthma: 63.2±6.3y) and conducted three-dimensional (3D) CFD simulations of breathing (tidal volume: 6 ml/kg). Results: Patients with OLD showed 11 times greater CFD-derived lower airway resistance and 3.8% smaller transpulmonary pressure than healthy controls. 3D CFD simulations showed less pressure drop from the mouth to the level of the sixth cervical vertebra (C6) during inhalation in OLD (COPD: 37.3±41.9Pa; asthma: 32.01±18.4 Pa) than in healthy subjects (70.7±48.5) Pa. Upper (oropharyngeal) airway resistance was also smaller in OLD (COPD: 155.0±188.0 kPa·s/m³; asthma: 116.2±61.5 kPa·s/m³) than in healthy subjects (229.4±142.2 kPa·s/m³). This is attributable to oropharyngeal enlargement measured by greater oral space air fraction and larger supraglottal airway diameter at the level of the fourth cervical vertebra (C4). Less upper airway pressure drop to the level of C6 correlated with upper airway enlargement measured by normalized upper airway air volume (r=-0.60, p<0.001) and diameter at C4 (r=-0.56, p=0.001) as well as lung structure-function impairment such as functional small airways disease percent (fSAD%), less volumetric deformation (J), and less relative displacement (s[asterisk]) (r=-0.40, 0.47, 0.54; p=0.030, 0.009, 0.002) and worse spirometry outcomes such as post-BD FEV1% (r=0.54, p=0.002) and FEV1/FVC (r=0.54, p=0.002). Conclusion: CT-based CFD simulations showed patients with obstructive lung disease experienced elevated airway resistance in the lower airways but reduced pressure drop and airway resistance due to oropharyngeal enlargement during inhalation. Reduced aerodynamic effort may help compensate excessive respiratory effort required in the lower airways. Findings suggest that upper airway imaging and airflow analysis allows more complete interpretation of the entire respiratory function altered with airflow obstruction.

In industrial environments, boiler units are widely used to supply heat and electrical power. At an integrated steel mill, industrial boilers combust a variable mixture of metallurgical gases combined with additional fuels to generate high-pressure superheated steam. Most tangentially fired boilers have experienced water wall tube failures in the combustion zone, which are thought to be caused by some deficiency in the combustion process. The challenge faced in this present process is that there are very limited means to observe the boiler operation. In this study, a three-dimensional Computational Fluid Dynamics (CFD) modeling and simulation of an industrial tangentially fired boiler firing metallurgical gases was conducted. Simulation results obtained from the assembled CFD model were validated by industrial experiments. A quick comparison of the flame shape from the simulation to the actual flame in the boiler showed a good agreement. The flow field and temperature distribution inside the tangentially fired boiler were analyzed under the operation conditions, and a wall water tube overheating problem was observed and directly related to the flow characteristics.

As part of the Hypersonic International Flight Research Experimentation (HIFiRE) Direct-Connect Rig (HDCR) test and analysis activity, three-dimensional computational fluid dynamics (CFD) simulations were performed using two Reynolds-Averaged Navier Stokes solvers. Measurements obtained from ground testing in the NASA Langley Arc-Heated Scramjet Test Facility (AHSTF) were used to specify inflow conditions for the simulations and combustor data from four representative tests were used as benchmarks. Test cases at simulated flight enthalpies of Mach 5.84, 6.5, 7.5, and 8.0 were analyzed. Modeling parameters (e.g., turbulent Schmidt number and compressibility treatment) were tuned such that the CFD results closely matched the experimental results. The tuned modeling parameters were used to establish a standard practice in HIFiRE combustor analysis. Combustor performance and operating mode were examined and were found to meet or exceed the objectives of the HIFiRE Flight 2 experiment. In addition, the calibrated CFD tools were then applied to make predictions of combustor operation and performance for the flight configuration and to aid in understanding the impacts of ground and flight uncertainties on combustor operation.

Investigation on the drainage condition within the ballast layer based on 3D CFD simulations

CFD simulations of the wind environment around an airport terminal building

High pressure ratio, radial-inflow turbines typically experience supersonic expansion in the nozzle section. Accurate estimation of the flow conditions and velocity triangle at the nozzle outlet is of critical importance in correctly predicting the overall turbine performance. The meanline modeling of such a nozzle requires special attention, due to the significantly altered flow field downstream of the throat. In this study, the flow field of a supersonic expansion nozzle is investigated, using a three-dimensional (3D) computational fluid dynamics (CFD) simulation calibrated with test data. Three different CFD configurations are explored: the nozzle alone, the nozzle plus rotor coupled with a mixing plane, and the nozzle plus rotor coupled with the nonlinear harmonic (NLH) method. These configurations are compared to each other to gauge the effect of the rotor and stator interaction and the potential for error in establishing the velocity triangles. The exit vane angle, number of vanes, and expansion ratio across the nozzle are systematically varied to provide the data as the base for nozzle modeling. Finally, a meanline method is proposed to calculate the pressure loss and flow deviation at the nozzle outlet and is compared with CFD results.

This paper studies the injection of gas in a stagnant solid bed via a horizontal nozzle from experimental and theoretical perspectives, with the analysis of permanent jet plumes and subsequent bubble detachment. Advanced 3D X-Ray imaging techniques have been developed and used to observe the phenomenon, as well as to quantify important process parameters (gas distribution, jet length, bubble size, etc.). The experimental results are compared with those obtained from three-dimensional computational fluid dynamics (CFD) simulations of a gas-solid fluidized bed with primary air injection performed through a two-orifice horizontal nozzle. The proposed model correctly predicts the jet penetration at all the jet velocities tested. Uneven distribution of air through the cross-section of the bed is also observed, with the presence of the same two distinct regions experimentally observed, namely a fluidized core region and a compacted peripheral region. However, the model fails in predicting jets interaction and curvature due to asynchronous bubble detachment from the jets. This study demonstrates the usefulness of X-ray radiography in noninvasively visualising detailed internal features of industrial fluidized beds, thus providing an excellent benchmark to challenge the assumptions made in computational and theoretical models.

Effect of geometry on the thermal behavior of catalytic micro-combustors

Vertical open display cabinets are widely used in shopping mall, supermarkets, retail stores. Maintaining the temperature of foods in the display cabinet is vitally important to retailers to ensure optimal food quality and safety. The purpose of this study is to reduce the infiltration of air and heat loss from ambient space to display cabinet. The three-dimensional Computational Fluid Dynamics(CFD) simulation is used for the analysis of air flow patterns and temperature distribution in refrigerated display cabinets. Under several operating conditions which vary both the inner and outer jet velocities in the range from 0.3 to 1.1 m/s, simulations were carried out. This paper presents a performance of display cabinets with single jet and double jet. The energy consumption due to thermal entrainment ratio is plotted with varying Re. It was found that the double jet system is better than single jet system in terms of temperature distribution and energy saving.

Performance enhancement of finned annulus using surface interruptions in double-pipe heat exchangers

CFD-based optimization and design of multi-channel inorganic membrane tubes

The present study aimed to numerically establish a new metamodel for predicting the propagation distribution of styrene, which is one of the hazardous and noxious substances (HNSs) spilled from ships. Three-dimensional computational fluid dynamics (CFD) simulations were conducted for 80 different scenarios to gather large amounts of data on the spatial distribution of the change in concentration over time. We used the commercial code of ANSYS Fluent (V.17.2) to solve the Reynolds-averaged Navier–Stokes equations, together with the scalar transport equation. Based on the CFD results, we adopted the well-known kriging model to create a metamodel that estimated the propagation velocity and spatial distributions by considering the effect of the current surface velocity, deep current velocity, surface layer depth, and crack position. The results show that the metamodel accurately predicted the changes in the local distribution of styrene over time. This model was also evaluated using the hidden-point test.