Computational Fluid Dynamics Modelling for Discharge Rate Optimization in the Muda River Basin: A Case Study of Monsoon-Driven Flood Events

A detailed experimental and numerical investigation was carried out to examine the three-dimensional flow field, secondary flows, and vortex motion in an annular compressor cascade. Various flow visualizations near the blade surface and endwalls, wall static pressure and loss measurements, as well as hot-film and hot-wire measurements inside the blade boundary layers were performed at various flow rates to understand the complex flow phenomena. A Reynolds-averaged Navier–Stokes equation was solved to investigate the flow numerically. The detailed comparison between measurement and numerical prediction indicates that the complex three-dimensional flow phenomena (corner stall, vortex motion, radial mixing, etc.) are very well predicted with the numerical method.

A three-dimensional multiphase flow model for assessing NAPL contamination in porous and fractured media, 2. Porous medium simulation examples

Background: Floods are common global natural disasters that occur occasionally in Nepal. Numerous environmental damages and health impacts are known to occur due to direct effect of flooding. There are limited studies reporting the ocular diseases after the event of floods using a large population data. We aimed to present the spectrum of flood-related ocular diseases after monsoon flood in Nepal using a community-based database.
 Methods: A descriptive, cross sectional study design was used to collect the data from di­agnostic screening and treatment camps organized in flood-affected areas between one to three weeks after the monsoon flood in Nepal. The data was analyzed using the excel sheet and the results were presented in frequency and percentage in the frequency distribution tables.
 Results: The children and elderly constituted majority of patients. Females constituted 54% while males were 46% of the total patients presenting at the camps. The incidence of infec­tive ocular diseases was 23.8% and that of traumatic ocular injuries was 6.2% of total ocular diseases. Acute conjunctivitis was the most common infective ocular disease (21.1%) fol­lowed by keratitis (2.7%). Corneal abrasions (1.4%), subconjunctival hemorrhage (1.1%) and lid ecchymosis (0.7%) were common clinical diagnosis following traumatic ocular injury.
 Conclusions: The infective ocular diseases and traumatic ocular injuries are common flood-related ocular diseases seen within one to three weeks after flood receded. The knowledge of these expected ocular morbidities may help for proper planning and organization of such relief camps. Early diagnosis and treatment of these diseases can reduce the ocular morbidi­ties.

The Moscow agglomeration has a high proportion of impermeable surfaces, which leads to a specific water regime characterized by frequent short-term floods due to the rapid response of the catchment area to precipitation. For one of the largest tributaries of the Moskva River in the capital, the Setun River, the SWMM model was able to reproduce the passage of extreme flood events in 2020–2023. The model was calibrated using 30-minute monitoring data on water discharge and 10-minute precipitation rates obtained by interpolating to the center of the catchment. The performance of the model was assessed using relative error (RE) and coefficient of determination (R²). The calibration and verification results showed a good correlation between the modeled and measured maximum water discharges (R² = 0.77) with relative errors ranging from 2 to 56%. The most accurate results were obtained for flood events with peak flow rates exceeding 15 m³/s.

The Bureau of Reclamation is developing a water-resource project, the Closed Basin Division, in the San Luis Valley of south-central Colorado that is designed to salvage unconfined ground water that currently is discharged as evapotranspiration. The water table in and near the 130,000-acre Closed Basin Division area will be lowered by an annual withdrawal of as much as 100,000 acre-feet of ground water from the unconfined aquifer. The legislation authorizing the project limits resulting drawdown of the water table in preexisting irrigation and domestic wells outside the Closed Basin Division to a maximum of 2 feet. Water levels in the closed basin in the northern part of the San Luis Valley historically have fluctuated more than 2 feet in response to water-use practices and variation of climatically controlled recharge and discharge. Declines of water levels in nearby wells that are caused by withdrawals in the Closed Basin Division can be quantified if water-level fluctuations that result from other water-use practices and climatic variations can be estimated. This study was done to evaluate water-level change at selected observation wells in and near the Closed Basin Division. Regression models of monthly water-level change were developed to predict monthly water-level change in 46 selected observation wells. Predictions of monthly water-level change are based on one or more of the following: elapsed time, cosine and sine functions with an annual period, streamflow depletion of the Rio Grande, electrical use for agricultural purposes, runoff into the closed basin, precipitation, and mean air temperature. Regression models for five of the wells include only an intercept term and either an elapsed-time term or terms determined by the cosine and sine functions. Regression models for the other 41 wells include 1 to 4 of the 5 other variables, which can vary from month to month and from year to year. Serial correlation of the residuals was detected in 24 of the regression models. These models also include an autoregressive term to account for serial correlation in the residuals. The adjusted coefficient of determination (Ra2) for the 46 regression models range from 0.08 to 0.89, and the standard errors of estimate range from 0.034 to 2.483 feet. The regression models of monthly water- level change can be used to evaluate whether post-1985 monthly water-level change values at the selected observation wells are within the 95-percent confidence limits of predicted monthly water-level change.

The impact of the street configurations on pollutants dispersion from vehicles exhausts within urban canyons was numerically investigated using a computational fluid dynamics (CFD) model. Three-dimensional flow and dispersion of gaseous pollutants were modeled using standard κ − e turbulence model, which was numerically solved based on Reynolds-averaged Navier–Stokes equations by the commercial CFD code FLUENT. The concentration fields in the urban canyons were examined in three cases of street configurations: (1) a regular-shaped intersection, (2) a T-shaped intersection and (3) a Skew-shaped crossing intersection. Vehicle emissions were simulated as double line sources along the street. The numerical model was validated against wind tunnel results in order to optimize the turbulence model. Numerical predictions agreed reasonably well with wind tunnel results. The results obtained indicate that the mean horizontal velocity was very small in the center near the lower region of street canyon. The lowest turbulent kinetic energy was found at the separation and reattachment points associated with the corner of the down part of the upwind and downwind buildings in the street canyon. The pollutant concentration at the upwind side in the regular-shaped street intersection was higher than that in the T-shaped and Skew-shaped street intersections. Moreover, the results reveal that the street intersections are important factors to predict the flow patterns and pollutant dispersion in street canyon.

Characteristics of convective heat transfer of a supersonic model combustor with variable inlet flow conditions were studied by numerical simulation in this paper. The three-dimensional flow and wall heat flux at different air inlet Mach numbers of 2.2, 2.8 and 3.2 were studied numerically with Reynolds-averaged Navier–Stokes equations with a shear-stress transport (SST) k − ω turbulence model and a three-step reaction model. Meanwhile, ethylene was chosen as the fuel, and the fixed fuel-to-air equivalence ratio is 0.8 in all cases in this paper. The results of the simulations indicate that wall heat flux distribution of the combustor is very non-uniform with several peaks of wall heat flux at varied locations. For the low inlet Mach number of 2.2, a shock train structure is formed in the isolator, and three peaks of wall heat flux are located respectively on the backward face of the cavity, on the side wall near the fuel injection and on the bottom wall near the injection holes, and a maximum wall heat flux reaches 5.4 MW/m2. For the medium inlet Mach number of 2.8, there exists a much shorter shock structure with three peaks of wall heat flux similar to that of Mach number 2.2. However, as the inlet Mach number increased to 3.2, there is no shock structure upstream of fuel injections, and the combustor flow is in a supersonic mode with different locations and values of wall heat flux peaks. The statistical results of wall heat loading show that the change of total wall heat is not monotonic with the increase of inlet Mach number, and the maximum appears in the case of Mach number being 2.8. Meanwhile, for all the cases, the bottom wall takes up more than 50% of the total heat loading.

Wind is a renewable energy resource but its intermittent nature poses some issues for transmission system operators, wind farms, and the electricity market. To address these issues, wind power forecasting is crucial. In this study, the data were obtained from a turbine anemometer in the 1-year time horizon of 10 min for wind power forecasting. Models are designed with univariate long short-term memory, multivariate long short-term memory, multilayer perceptron (MLP), extreme gradient boosting (XGBOOST), convolutional neural networks (CNNs), long short-term memory, gated recurrent unit (GRU), and random forest (RF) algorithms. Outliers in the data are cleaned with the isolation forest algorithm. Hyperparameter optimization of the models, designed with machine learning algorithms for a selected month from each season (January, April, July, and November) was made with Bayesian optimization. The success of the models was measured by mean absolute error, root mean square error, and determination coefficient. Among the designed models, for January, April, and November, the multivariate long short-term memory model gave the most successful results with 0.926, 0.961, and 0.946 determination coefficient values, respectively. For July, the XGBOOST model provided the most successful result with 0.916 value of the determination coefficient

Purpose: The purpose of this study was to compare calibration factors for deep dose equivalent Hp (10) and shallow dose equivalent Hp (0.07) between Cesium (Cs)-137 and X-ray sources when they are exposed to same dose and to determine uncertainties with MTS-N (LiF: Mg, Ti) chips when they are exposed to low dose ≤ 2mGy. Material and Methods: Thermoluminescent (TL) chips were annealed at 400oC for one hour and allowed to cool and were subjected to a temperature of 100oC for another two hours using a TLD Furnace Type LAB-01/400. They were then taken to a Secondary Standard Dosimetry Laboratory (SSDL) for irradiation using a Cs-137 source at known doses (0.2-2mGy). A RadPro Cube 400 manual TLD Reader was used to determine corresponding TL signal. The above process was replicated but with a calibrated X-ray unit as the source for calibration. Results: The calibration factors (CF) from the line graph of dose (mGy) against TL signal (count) for Cs-137 source with Hp (10) and Hp (0.07) were 3.72 x 10-6 and 5.97x10-6 mGy/count respectively. Those with X-ray source for Hp (10) and Hp (0.07) were 3.44x10-6 and 4.05x10-6 mGy/count respectively with an overall coefficient of determination (R2) = 0.99. The adjusted maximum percentage deviation between the actual and calculated dose for both sources was -2.74%. The percent (%) deviation of the mean with both sources for Hp (10) and Hp (0.07) was 3.9% and 19% respectively. Conclusion: Adjusted percent deviation from both sources were within the recommended dose limit of ±30% by the Radiological Protection Institute of Ireland (RPII) and within the International Commission on Radiological Protection (ICRP) limit respectively. Better accuracy was seen for Hp (10) with both sources compared to Hp (0.07). Calibration of the MTS-N chips using both sources was successful and can be used for personal dosimetry.

This study aims to optimize the impeller geometry of a backward-curved blade centrifugal fan in a helicopter oil cooler to improve the aerodynamic performance of the impeller. The study consists of three main parts: aerodynamic design and parametrization of the impeller, numerical analysis of impeller performance, and multi-objective design optimization to maximize fan static pressure and total to static isentropic efficiency. In the first part, a one-dimensional impeller design was performed using quasi-experimental methods and a direct optimization method with the multi-objective genetic algorithm, and the impeller geometry was parameterized through a commercial tool. The three-dimensional flow through the impeller was solved in the second part with a commercial computational fluid dynamics tool. The Reynolds-averaged Navier-Stokes equations were solved on a multi-block grid, and a second-order accurate finite volume method is employed. The most appropriate turbulence model and grid size were selected considering time, cost, and fidelity. In the third part, a sensitivity analysis was performed with the design-of-experiment method, and the parameters that affect the objective function most significantly were determined. A multi-objective design optimization based on a non-dominated sorting genetic algorithm was performed with the Kriging response surface method, and Pareto-optimal solutions were obtained. Results show that, at the design point, there is a 9.6% and 0.96% increase in the fan static pressure and total to static isentropic efficiency, respectively. Trained kriging response surface model predicted the fan static pressure and total to static isentropic efficiency with an error of 0.12% and 0.33%, respectively.

Improving the accuracy of daily solar radiation prediction by climatic data using an efficient hybrid deep learning model: Long short-term memory (LSTM) network coupled with wavelet transform

Data assimilation techniques are widely used to predict complex dynamical systems with uncertainties, based on time-series observation data. Error covariance matrices modeling is an important element in data assimilation algorithms which can considerably impact the forecasting accuracy. The estimation of these covariances, which usually relies on empirical assumptions and physical constraints, is often imprecise and computationally expensive, especially for systems of large dimensions. In this work, we propose a data-driven approach based on long short term memory (LSTM) recurrent neural networks (RNN) to improve both the accuracy and the efficiency of observation covariance specification in data assimilation for dynamical systems. Learning the covariance matrix from observed/simulated time-series data, the proposed approach does not require any knowledge or assumption about prior error distribution, unlike classical posterior tuning methods. We have compared the novel approach with two state-of-the-art covariance tuning algorithms, namely DI01 and D05, first in a Lorenz dynamical system and then in a 2D shallow water twin experiments framework with different covariance parameterization using ensemble assimilation. This novel method shows significant advantages in observation covariance specification, assimilation accuracy, and computational efficiency.

In this recent study we have simulated free surface flow over submerged weirs using the computational fluid dynamics code, Fluent, to forecast flow near the structure where three-dimensional (3D) flows were dominant. The fully 3D, Reynolds-averaged Navier—Stokes equations were solved using the model. To understand the complex flows, the volume of fluid method with genetic reconstruction scheme was adopted and the standard k-ε equations were used for turbulence model. The numerical results obtained using the compressed mesh systems were in accordance with the experimental data.

To analyze the fluid–solid coupling stress characteristics of the Francis turbine runner comprehensively, based on the Reynolds-averaged Navier–Stokes equations and shear stress transport k-ω turbulence model, this article performs numerical simulation of three-dimensional steady incompressible turbulent flow through the whole passage of a certain large Francis turbine under multiple operating conditions with Computational Fluid Dynamic software CFX and contrasts the result with that model test conversion. With the help of ANSYS workbench platform, equivalent stress, deformation distribution, and variation of the runner under multiple operating conditions are obtained through loading the water pressure on the runner blade as structural plane load to blades by the method of unidirectional fluid–solid coupling. The results show that under small flow operating conditions, flow patterns in the runner are disordered, the stress on the blade distributes unevenly, and the maximum stress lies on the influent side of the blades connected to the band; as the flow increases, the stress appears intensively around the effluent side of the blades connected to the runner crown. The maximum deformation first decreases and then increases as the flow increases. The deformation area expands from the middle effluent side of blades to the band. The results can be found in the researches on the structure design and the safety and stability of the Francis turbine runner.

Aerodynamic shape optimization of aircraft configurations often ignores stability considerations. To address this, a method for the computation of static, dynamic, and transient aircraft stability derivatives and their sensitivities for use in gradient-based optimization is introduced and evaluated. Computational fluid dynamics in the form of a three-dimensional structured-grid multiblock flow solver with both Euler and Reynolds-averaged Navier–Stokes equations is used. To compute the stability derivatives, a time-spectral formulation is used to compute an oscillating solution for the configuration of interest. From this oscillating solution, a series of linear regressions is performed to calculate the various stability derivatives. Because the solution is time dependent, it contains the information required to compute the transient, or “dot,” derivatives for the configuration. An adjoint method is used to compute the gradients of the stability derivatives of interest, enabling gradient-based stability-constrained aerodynamic shape optimization with respect to a large number of design variables. The computed stability derivatives are verified for an airfoil and validated for a generic unmanned combat aerial vehicle. The stability-constrained optimization of a wing demonstrates the viability and usefulness of the method for aircraft design optimization.