Fourier-Feature Neural Surrogate for Hemodynamic Field Reconstruction in Stenotic and Bifurcating Flows

While designing the power requirements of a ship, the most important factor to be considered is the ship resistance, or the sea drag forces acting on the ship. It is important to have an estimate of the ship resistance while designing the propulsion system since the power required to overcome the sea drag forces contribute to ‘losses’ in the propulsion system. There are three main methods to calculate ship resistance: Statistical methods like the Holtrop-Mennen (HM) method, numerical analysis or CFD (Computational Fluid Dynamics) simulations, and model testing, i.e. scaled model tests in towing tanks. At the start of the design stage, when only basic ship parameters are available, only statistical models like the HM method can be used. Numerical analysis/ CFD simulations and model tests can be performed only when the complete 3D design of the ship is completed. The present paper aims at predicting the calm water ship resistance using CFD simulations, using the Flow-3D software package. A case study of a roll-on/roll-off passenger (RoPax) ferry was investigated. Ship resistance was calculated at various ship speeds. Since the mesh affects the results in any CFD simulation, multiple meshes were used to check the mesh sensitivity. The results from the simulations were compared with the estimate from the HM method. The results from simulations agreed well with the HM method for low ship speeds. The difference in the results was considerably high compared to the HM method for higher ship speeds. The capability of Flow-3D to perform ship resistance analysis was demonstrated.

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.

Bench test and computational fluid dynamics (CFD ) simulation of the air pump loss on the 380 hydraulic retarder are completed in this paper . It turns out that the CFD simulation has nice reliability from the contrast of the air pump loss between the test and the simulation . And on this foundation , CFD simulation of two kinds of retarders with different turbulent structures is done . The contrast of the results of the simulation give the effects of different turbulent structures on cutting down the are pump loss .

The chemical composition and properties of fuels used in compression-ignition engines can influence engine performance significantly. Consequently, the modeling of fuel chemistry within computational fluid dynamics (CFD) simulations of diesel and other compression ignition engines is important. Modern detailed chemical mechanisms may provide predictive modeling of fuel chemistry; however, they are generally far too computationally expensive for use in CFD. We present simulations of diesel engine combustion, focusing on the prediction of ignition, using the CONVERGE CFD software package. A CFD simulation framework with models for turbulence and spray breakup and atomization is presented with a reduced global reaction model to describe fuel oxidation and ignition. The global reaction model incorporates a single parameter, the derived cetane number (DCN), to describe fuel reactivity variability. CFD simulations are compared to experiments carried out in a single-cylinder diesel engine for compositionally diverse conventional and alternative diesel and jet fuels. Model-experiment comparisons show general agreement for ignition timing and the influence of fuel variability on ignition timing. In addition, the sensitivity of CFD predictions on the chemistry, turbulence, and spray models is illustrated.

The clarity of seismic images has a great impact on the results of seismic data interpretation; therefore, high-resolution seismic images can provide finer geological information for seismic exploration. Coordinate-based Multilayer Perceptron(MLP) is easy for gradient-based optimization and machine learning, and represents a more compact order of magnitude than grid sampling. In this paper, we propose a combination of Random Fourier Features(RFF) and coordinate based MLP method to achieve super-resolution reconstruction of seismic images, taking one image coordinate as input and the output is the RGB value of the pixel at that location. Preprocessing the spatial coordinates of the image with Fourier feature mapping enables the MLP to learn higher frequency detail information and achieve super-resolution reconstruction of a single seismic image. Experiments show that the texture and edge details of the final processed seismic image are much clearer.

Today, many wells in heterogeneous carbonate reservoirs throughout the Middle East and Caspian Sea area are being completed with long intervals and complex completion equipment, such as limited entry liners and inflow control devices (ICDs). Accurate modeling and understanding of fluid placement during a stimulation operation through such long complex completions can be quite challenging. In this paper, an integrated approach of modeling, experiments and computational fluid dynamics (CFD) simulations will be presented. A computational tool based on a transient one-dimensional (1D) approach has been developed in-house to model stimulation through such completions. In order to validate the 1D approach for reliable fluid displacement predictions, annulus fluid displacement experiments in perforated liners supported by CFD simulations have been conducted. The experimental setup consists of an outer Lucite pipe representing the well-bore geometry and an inner aluminum pipe with multiple perforations of variable diameter. The setup mimics the real field application in a 1:1 scale with respect to the pipe and perforation diameters while the test section is limited to 27 ft. The experimental procedure is as follows: fluid is pumped into the annulus through the perforations until the annulus is completely filled,the fluid tank is switched "on-the-fly" to the tank with the displacing fluid resulting in the displacement of the annulus fluid by the displacing fluid through the perforations. The experiments and the CFD simulations indicate that even with a single open perforation inside the 27-ft test section, the three-dimensional (3D) flow field around the perforation rapidly transitions to uniform annular flow within a few feet uniformly displacing the annulus fluid. The model also provides insight into various well injection processes in sandstone and carbonate formation, such as scale squeezes, solvent treatments, and HF acid treatments. Introduction The length of production and injection intervals in a single well continues to increase in order to contact more reservoir rock with a single wellbore penetration and improve project economics. Multiple, thick stacked reservoirs are accessed by vertical and deviated wells while thinner reservoirs are being exploited with longer and longer horizontal wells1–3. To keep up with drilling capability, completion technology has advanced considerably in the past few years. Horizontal wells are "compartmentalized" using mechanical and/or swellable packers. Within a given compartment, injection and production can be controlled by use of inflow control devices (ICDs), limited entry (LE) liners, and/or sliding sleeves. The liners also make it easier to enter the completion interval with logging tools, wire line, and coiled tubing. While various tools are available to evaluate long-term, steady-state production and injection into such wells, modeling of short-term, transient injection into such completions is very challenging due to the long lateral lengths, limited entry points from the liner to each compartment, and the presence of an open annulus with occasional packer isolations. This is especially important for long wells in carbonate formations as all wells require acid stimulation upon initial completion, and occasionally throughout the well life.

A normally-open thermally-actuated microvalve was designed (using microfabrication/soft-lithography techniques involving 3D Printed molds), assembled and tested. The motivation of the research work is to develop an array of microvalves for precise delivery of water to individual plants in a field (with the goal of developing smart irrigation systems for high value cash-crops in the agricultural sector). It is currently impossible to control application of irrigation-water at the level of a single plant. If such a capability were practically available on farms, the result would be a step change in precision agriculture, such that the output of every plant in a farm field could be optimized (i.e., food-water-energy nexus in sustainability applications). The aim of this study is to develop and test a microfluidic system (consisting of a microvalve array) that could be controlled, capillary by capillary, to deliver the needed amount of water to individual plants in a large field. Two types of test fluids were leveraged for thermo-hydraulic actuation of the microvalves developed in this study: (a) Design-I: using air, and (b) Design-II: using Phase Change Material (PCM). The PCM used in this study is PureTemp29. The proposed approach enabled a simple and cheap design for microvalves that can be manufactured easily and are robust to weather conditions (e.g., when exposed to the elements in orchards and open fields). Other advantages include: safe and reliable operation; low power consumption; can tolerate anomalous pressure loads/fluctuations; simple actuation; affords easy control schemes; is amenable for remote control; provides long-term reliability (life-cycle duration estimated to be 3∼5 years); can be mass produced and is low maintenance (possibly requiring no maintenance over the life time of operation). The microvalve consists of two layers: a flow layer and a control layer. The control layer is heated from below and contains a microfluidic chamber with a flexible polymeric thin-membrane (200 microns in thickness) on top. The device is microfabricated from Poly-Di-Methyl-Siloxane (PDMS) using soft lithography techniques (using a 3D Printed mold). The control chamber contains either air (thermo-pneumatic actuation) or PCM (thermo-hydraulic actuation involving repeated melting/freezing of PCM). The flow layer contains the flow channel (inlet and outlet ports, horizontal section and valve seat). The experimental results from testing the efficacy of the two types of micro-valves show a 60% reduction (for thermo-pneumatic actuation using air) and 40% reduction (for thermo-hydraulic actuation using PCM) in water flow rates for similar actuation conditions (i.e., heater temperature values). PCM design is expected to consume less power (lower OPEX) for long-term actuation but may have slower actuation speed and have higher manufacturing costs (CAPEX). Air actuation design is expected to consume more power (higher OPEX) for longer-term operation but may have faster actuation speeds and lower manufacturing costs (CAPEX). Computational Fluid Dynamics (CFD) simulations were performed to investigate the effect of flowing water (in the microfluidic channel) on the average absolute pressure and temperature of air in the actuation chamber. The CFD simulations were performed using a commercial tool (Ansys™ 2019R1®). The results from the CFD simulations are presented in this study.

In response to a solicitation for Computational Fluid Dynamic (CFD) simulations of flow in a T-junction from the Nuclear Energy Agency of the Organization for Economic Co-operation and Development (OECD/NEA) [1], Structural Integrity (SI) Associates of San Jose, CA, USA has completed a simulation using the commercially available CFD code ANSYS-CFX. The simulation was carried out for geometry and flow conditions matching an experiment underway at Vattenfall Research and Development in Sweden where two streams of differing temperature mix in a T-junction. The benchmark was to be a blind test of CFD simulations. Experimental results were not released until after benchmark participants submitted the results of their simulation to the OECD/NEA. The CFD simulation done by SI was carried out on an all hexahedral mesh of 1,881,504 elements generated using the ANSYS Mechanical Finite Element program. The mesh was then transferred into the CFX program and, after applying proper boundary conditions, a time-dependent simulation was done using the Large Scale Eddy Simulation (LES) turbulence model [2] with Wall Adjusting Local Eddy Viscosity (WALE) sub-grid scale (SGS) turbulence model [2, 3]. Comparisons of the CFD results to the Vattenfall data show close correspondence for mean and root-mean-square (rms) turbulent temperature fluctuations.

CFD simulation and experimental validation of a GM type double inlet pulse tube refrigerator

Recently, image-based computational fluid dynamic (CFD) simulations have been proposed to investigate the local hemodynamics inside human cerebral aneurysms. It was suggested that the knowledge of the computed three-dimensional flow fields can be used to assist clinical risk assessment and treatment decision making. Therefore, it was desired to know the reliability of CFD for cerebral blood flow simulation, and be able to provide clinical feedback. However, the validations were not yet comprehensive as they lack either patient-specific boundary conditions (BCs) required for CFD simulations or quantitative comparison methods. In this study, based on a recently proposed in-vitro quantitative CFD evaluation approach via virtual angiography, the CFD evaluation was extended from phantom to patient studies. In contrast to previous work, patient-specific blood flow rates obtained by transcranial color coded Doppler ultrasound measurements were used to impose CFD BCs. Virtual angiograms (VAs) were constructed which resemble clinically acquired angiograms (AAs). Quantitative measures were defined to thoroughly evaluate the correspondence of the detailed flow features between the AAs and the VAs, and thus, the reliability of CFD simulations. The proposed simulation pipeline provided a comprehensive validation method of CFD simulation for reproducing cerebral blood flow, with a focus on the aneurysm region. Six patient cases were tested and close similarities were found in terms of spatial and temporal variations of contrast agent (CA) distribution between AAs and VAs. For patient #1 to #5, discrepancies of less than 11% were found for the relative root mean square errors in time intensity curve comparisons from characteristic vasculature positions. For patient #6, where the CA concentration curve at vessel inlet cannot be directly extracted from the AAs and given as a BC, deviations about 20% were found. As a conclusion, the reliability of the CFD simulations was well confirmed. Besides, it was shown that the accuracy of CFD simulations was closely related to the input BCs.

<p indent="0mm">Freshwater resources are increasingly scarce around the world. Reverse osmosis (RO) is one of the most advanced desalination technologies, accounting for more than 50% of the desalination market. High permeability reverse osmosis membranes can significantly improve water production efficiency. However, concentration polarization due high flux on membrane surfaces leads to the aggravation of fouling, which greatly hinders the application of high permeability RO membranes. Therefore, developing a more efficient optimal design method based on deep learning methods other than traditional manual iterative design methods for membrane modules can accelerate the commercialization of high permeability RO membranes and has great research value. In this paper, high-dimensional nonlinear transport models in consideration of complex feed channels for seawater RO desalination with a high permeability membrane are established with respect to various working conditions. A three-dimensional (3D) multi-physics computational fluid dynamics (CFD) model with high fidelity is established for the fluid flow and salt transport in the feed channel of the RO membrane module. Using a Latin-like hypercube sampling method, 726 CFD models with respect to different design parameter combinations are generated and calculated on the “Tianhe-2” supercomputer. Theoretically, the maximum computational scale in parallel is over 20000 cores. Then we establish multilayer perceptron (MLP) models incorporating with the CFD simulation results, which are randomly divided into training, validation, and test sets according to the ratio of 650:66:10. The established MLP models using the deep learning method are applied to predict the distribution of 3D local velocity, pressure, and concentration with respect to large-scale parameter combinations within the considered design parameter space, as surrogate transport models for physical models. In order to appraise the accuracy of MLP models, this paper compares the results of 10 working condition models in the test set by calculating structural similarity, root mean squared error (RMSE), and regression coefficient of determination between results using CFD and MLP methods. With calculating RMSE, the prediction accuracy reaches 93.5%, 98.3%, and 95.1%. The calculation efficiency is increased by 1–2 orders of magnitude than that of using the traditional finite element method. To evaluate the accuracy of the results predicted by MLP in terms of macroscopic physical parameters, we calculate the Sherwood number and drag coefficient using the CFD and MLP methods with respect to various working conditions corresponding to the test set. The average prediction errors of the drag coefficient and Sherwood number are 2.68% and 6.23% respectively. Furthermore, the reliability of the MLP models is verified for predicting the 3D local velocity, pressure, and concentration in the reverse osmosis channel under different working conditions. The highly scalable parallel calculation method proposed in this work is suitable for the massive screening and optimal design of multi-type feed spacers. This paper provides a computable model and data support to reveal the complex mechanism of the flow and transport process in the RO membrane module, and has critical significance for the factory-scale application of high permeability RO membranes in the future.

A novel nonlinear system identification approach is presented based on the coupling of a neuro-fuzzy model (NFM) with a multilayer perceptron (MLP) neural network. Therefore, the recurrent NFM is employed for multi-step ahead predictions, whereas the MLP is subsequently used to perform a nonlinear quasi-static correction of the obtained time-series output. In the present work, the proposed method is applied as a reduced-order modeling (ROM) technique to lower the effort of unsteady motion-induced computational fluid dynamics (CFD) simulations, although it could be utilized generally for any nonlinear system identification task. For demonstration purposes, the NLR 7301 airfoil is investigated at transonic flow conditions, while the pitch and plunge degrees of freedom are simultaneously excited. In addition, the sequential model training process as well as the model application is presented. It is shown that the essential aerodynamic characteristics are accurately reproduced by the novel ROM in comparison to the full-order CFD reference solution. Moreover, by examining the results of the NFM without MLP correction it is indicated that the new approach leads to an increased fidelity regarding nonlinear ROM-based simulations.

Not only air temperature and humidity, but also radiant temperature affects indoor thermal comfort. Therefore, radiant heat transfer should be considered in computational fluid dynamic (CFD) simulation. In this paper, thermal environment in a room equipped with cooling radiant ceiling panel (CRCP) system was estimated by radiation-coupled CFD. In order to introduce radiant heat transfer into CFD, the distribution factor was used. Two calculation cases with different emissivity were evaluated using the predicted mean vote (PMV), a measure of thermal comfort. Case of this paper, the value of PMV increases by 1.19 in the case of emissivity of 1.0 compared to the case of emissivity of 0.9. Therefore, it is important to properly consider the effect of emissivity when evaluating thermal comfort by CFD simulation.

The main objective of the present work is to study the effect of an external store on a subsonic fighter aircraft. Generally most modern fighter aircrafts are designed with an external store installation. In this study, a subsonic fighter aircraft model has been manufactured using a computer numerical control machine for the purpose of studying the effect of the aerodynamic interference of the external store on the flow around the aircraft wing. A computational fluid dynamic (CFD) simulation was also carried out on the same configuration. Both the CFD and the wind tunnel testing were carried out at a Reynolds number 1.86×105 to ensure that the aerodynamic characteristic can certify that the aircraft will not be face any difficulties in its stability and controllability. Both the experiments and the simulation were carried out at the same Reynolds number in order to verify each other. In the CFD simulation, a commercial CFD code was used to simulate the interference and aerodynamic characteristics of the model. Subsequently, the model together with an external store was tested in a low speed wind tunnel with a test section sized 0.45 m×0.45 m. Measured and computed results for the two-dimensional pressure distribution were satisfactorily comparable. There is only a 19% deviation between pressure distribution measured in wind tunnel testing and the result predicted by the CFD. The result shows that the effect of the external storage is only significant on the lower surface of the wing and almost negligible on the upper surface of the wing. Aerodynamic interference due to the external store was most evident on the lower surface of the wing and almost negligible on the upper surface at a low angle of attack. In addition, the area of influence on the wing surface by the store interference increased as the airspeed increased.

Evaluation and verification of patient-specific modelling of type B aortic dissection