Minimum Reynolds Number Research Articles

Enhancing thermal efficiency in twisted tri-lobe double pipe heat exchangers via integrated CFD and AI approaches

Industrial energy efficiency relies on precise heat exchanger design. Engineers aim to enhance heat transfer while minimizing pressure drop. Twisted tube Double Pipe Heat Exchangers (DPHEs) excel, inducing secondary flows that disrupt boundary layers and enhance fluid mixing. This investigation aims to provide insights into the relationships among specific factors and the thermal performance of a DPHE with twisted tri-lobe tubes, utilizing water as the working fluid. This is accomplished by employing an approach that integrates computational fluid dynamics with an artificial neural network, coupled with a single-objective genetic algorithm. Data inquiry was facilitated using response surface methodology, considering five design variables including the Reynolds numbers of the outer and inner twisted tubes, outer and inner twist ratios, and twisting direction. The simulation employed the turbulent k-ω shear-stress transport model, and the governing equations were solved using the finite volume method. The results underscored the efficacy of the developed multi-approach algorithm in maximizing the Performance Evaluation Criterion (PEC) both within the specified domain (10000 ≤ Re ≤ 20000 & 5≤twist ratio≤20) and beyond the domain (5000 ≤ Re ≤ 50000 & 2.5≤twist ratio≤25) of design variables. In the specified domain, a PEC of 1.16 was achieved, demonstrating the algorithm's effectiveness. Furthermore, the proposed method, extrapolated to the data space, yielded a noteworthy improvement in PEC of 16.35 %. In both domains, the optimized set of design variables for maximizing the PEC comprised a minimum outer Reynolds number, maximum inner Reynolds number, maximum outer twist ratio, and minimum inner twist ratio.

Effect of the kinematic viscosity on liquid flow hydrodynamics in vortex mixers

The effect of the kinematic viscosity of similar fluids on liquid flow hydrodynamics and the onset of the instability were investigated to acquire information on the mixing behavior of vortex mixers. To assess the influence of the kinematic viscosity, water and two glycerin aqueous solutions with varying weight percentages (25 wt % and 50 wt %) were used. The critical Reynolds number of the flow regime transitions was found using PLIF and PIV flow-field techniques. The liquid flow hydrodynamics and gas flow dynamics in the proposed vortex mixer geometry were compared, as well. It was observed that the kinematic viscosity has a linear relation with the critical Reynolds number; however, the coefficient is different for liquids and gases. Additionally, kinematic viscosity dictates the minimum Reynolds number where the fluids have a full turn inside the chamber. It was observed that fluids having higher kinematic viscosities have a full turn at lower Reynolds numbers, and the critical Reynolds number value decreases with increasing kinematic viscosity.

Hydrothermal performance of mini-channel heat sink using nanofluids/hybrid nanofluids: A numerical study

ABSTRACT We investigated hydrothermal performance of the mini-channel heat sink by using water, and nanofluids/hybrid nanofluids at 1% of volumetric concentration. The water-based nanofluids studied in this work were CuO, TiO 2 and Fe 2 O 3 with CuO-TiO 2 and CuO-Fe 2 O 3 hybrid combinations. The mean diameter of all the nanoparticles was fixed to be 60 nm. For numerical modeling, multiphase mixture model was used. The maximum heat transfer recorded for CuO and (CuO .75%+ Fe 2 O 3 .25%) based nanofluids was 59.12W and 58.27 W, respectively, at maximum Reynolds number of 1750. The maximum percentage increment in heat transfer recorded for CuO and hybrid (CuO .75%+Fe 2 O 3 .25%) water-based nanofluids was 29.24% and 24.55%, respectively, then water at minimum Reynolds number of 750. Results are evaluated in the form of the base temperature, thermal resistance, heat transfer, pressure drop and Nusselt number. The minimum base temperature was recorded for CuO-H 2 O as 35.4 at maximum Reynold number (Re) of 1750 among all the nanofluids/hybrid nanofluids. However, maximum percentage reduction in the base temperature was recorded for CuO-H 2 O as 7.04% compared with water at minimum Reynolds number of 750. It was observed that nanofluids lead toward high heat transfer.

Three dimensional numerical study of PV module cooling by using thermoelectric effects and nano-enhanced confined multiple slot jet impingement

In this study, 3D numerical study of photovoltaic (PV) module cooling is analyzed by combined utilization of thermoelectric generator (TEG) and nano-enhanced slot jet impingement system. For the relevant parameters of interest, such as Reynolds number (between 10 and 100), nanoparticle loading (solid volume fraction between 0 and 0.02), non-dimensional vertical distance between the slots and upper plate (between 2.5 and 8), non-dimensional spacing between the slots (between 2.5 and 8), and cold NF inlet temperature (between 20oC and 30oC), finite element method is utilized. Using only TEG and nano-jet assisted cooling with TEG result in PV-cell temperature reduction by about 43.33oC and 69.55oC as compared to un-cooled PV module. It is observed that cell temperature decreases by about 2.4oC-2.9oC when considering the minimum and maximum Reynolds number cases. When nanoparticle loading is increased to 0.02, output TEG power rises while cell temperature decreases by about 0.75oC. When varying the vertical distance between the slot to the impinging surface, cell temperature changes up to 1.9oC while distance between the slots has very slight impact on the temperature variation. Decreasing the inlet fluid temperature by about 10oC decreases the cell temperature by about 8.5oC.

The role of buoyancy and thermal acceleration in heat transfer deterioration and thermal oscillation for the CO2 around pseudo-critical region

Trans-critical CO2 Rankine cycle has great potential in converting low-grade heat into electricity, because of its good temperature glide matching between working medium and heat source. However, the drastic variation in thermo-physical properties of CO2 around the pseudo-critical region causes the abnormal flow and heat transfer behaviors. In this paper, the heat transfer deterioration and thermal oscillation in the heat addition of trans-critical CO2 cycle are investigated by experiments. The test section is heated uniformly by a DC power (0∼6 kW) and the minimum Reynolds number in the test section inlet is about 8000 (it is much larger than 2300) for the all cases. A shear stress reconstruction model is corrected by introducing the velocity profile of turbulence in the boundary layer to attempt to provide a quantitative criterion for deterioration and oscillation. The role of buoyancy and thermal acceleration in heat transfer deterioration and thermal oscillation is elucidated by connecting the shearing-stress reconstruction model and experimental results. It is found that, the thermal acceleration in the near-wall zone provides the key motivation to drive the heat transfer deterioration. The thermal oscillation is driven by the transition between partial and full re-laminarization flows in the local zone of heat transfer deterioration. When (Δτa/τw)<1&(Δτ/τw)tot>1 or 0.1<(Δτ/τw)tot<1, the turbulence is partially re-laminarized. When(Δτa/τw)>1&(Δτa/τw)>>(Δτbo/τw), the intense thermal acceleration causes the turbulence to be fully re-laminarized in the near-wall zone and it maintains the new laminar boundary . The turbulence is partially re-laminarized in the lower boundary and it is almost full re-laminarization in the upper boundary of the oscillation wall temperatures. The results confirm that the emergence of heat transfer deterioration and oscillation instability in mixed turbulent convective relate with the change of flow states caused by acceleration and buoyancy effects.

Predicting the energy stability limit of shear flows using weighted velocity components

Predicting the subcritical transition in fluid dynamic systems remains a challenging task, but recent advancements utilizing edge tracking methods, polynomial Lyapunov functions, and various energy norms have shown promise. In this study, we propose a novel approach by defining the general kinetic energy through weighted velocity components. The minimal Reynolds number is determined, where the derivative of this generalized energy with respect to time is zero. The procedure is similar to that of the well-known Reynolds–Orr equation. Unlike traditional methods, our approach does not necessitate the monotonic decay of the classic perturbation kinetic energy, resulting in a larger critical Reynolds number and reduced conservativeness of the Reynolds–Orr equation. However, the energy production of the pressure is not negligible, in contrast to the classical Reynolds–Orr equation. The pressure's implicit dependence on the velocity field complicates the variation process. To address this, a method is presented to handle the problem effectively. Our approach is then applied to analyze parallel flows, specifically the plane Couette and plane Poiseuille flows, wherein the problem can be further simplified using the complex Fourier transformation. The weights of velocity components are optimized to maximize the critical Reynolds number, resulting in a significant increase.

Towards a finite-time singularity of the Navier–Stokes equations. Part 3. Maximal vorticity amplification

An exact analytical solution is obtained for the dynamical system derived in Part 1 of this series (Moffatt &amp; Kimura, J. Fluid Mech, vol. 861, 2019a, pp. 930–967), which describes the approach of two initially circular vortices of finite but small cross-section symmetrically located on inclined planes. This exact solution, applicable in the inviscid limit, allows determination of the amplification $\mathcal {A}_{\omega }$ of the axial vorticity within the finite time $T$ during which the basic assumptions of the model continue to apply. It is first shown that, for arbitrarily prescribed $\mathcal {A}_{\omega }$ , it is possible to specify smooth initial conditions of finite energy such that, in the inviscid limit, this amplification is achieved within the time $T$ . When viscosity is included, an estimate is provided for the minimum vortex Reynolds number that is sufficient for the same result to hold. The predictions are broadly compatible with results from direct numerical simulations at moderate Reynolds numbers. Moreover, it is shown that one may come arbitrarily close to a finite-time singularity of the Navier–Stokes equation by appropriate choice of an initial, smooth, finite-energy velocity field; however, this approach to a singularity is ultimately thwarted through breach of the assumptions on which the dynamical system is based. Thus we make no claim here concerning realisation of a Navier–Stokes singularity. Moreover, we find that the conditions required to attain a large amplification $\mathcal {A}_{\omega }\gg 1$ during the time $T$ are far beyond those that can be realised in either experiment or direct numerical simulation.

Application of novel double-layer micro-jet heat sinks for energy-efficient thermal management of motor inverters in electric vehicles

The relentless advancement in electronics miniaturization and performance has birthed a tremendous amount of heat flux, demanding effective dissipation techniques. Thermal effects pose a significant hurdle in the thermal management of electric vehicles (EVs) and having advanced cooling techniques in place is vital. To tackle this, the present study numerically evaluates the cooling attributes of two novel double-layer micro-jet heat sinks (DLMJHSs) to address the thermal management of insulated-gate bipolar transistors (IGBTs) in the motor inverters of EVs. The numerical analyses are carried out for Reynolds numbers ranging from 5000 to 25,000, IGBT heat fluxes of 100, 200, and 300 W/cm2, and different configurations of DLMJHSs. The potential of using nanofluids is assessed by employing Al2O3–water nanofluid with different volume fractions of 0.8% and 1.7%. The finite volume technique is implemented to solve the equations and the turbulence closure is modeled using the k-ԑ model. Results show that HS #2 provides a 10.2% higher convective heat transfer coefficient leading to a 7.44 ℃ lower baseplate temperature than HS #1. Besides, HS #2 prevents the occurrence of high-temperature spots owing to a more uniform temperature distribution in comparison with HS #1. HS #2 demonstrates the most uniform temperature distribution, with a Coefficient of Variation of only 0.00993 and a maximum heat transfer coefficient of 80.5 kW/m2 ℃. The heat sink’s baseplate temperature drops by increasing either the Reynolds number or the volume fraction. The baseplate temperature declines by up to 6.94 ℃ using a nanofluid volume fraction of 1.7% compared with only the base fluid. A minimum Reynolds number of 15,700 is required to prevent overheating at the peak heat flux. The performance evaluation criterion (PEC) is higher than 1 for Re ≤ 15,000.

A global inertial permeability for fluid flow in rock fractures: Criterion and significance

Fluid flow in rock fractures is described by Darcy's law and its nonlinear extension, the Forchheimer equation in form of a quadratic polynomial. Since the nonlinearity of the Forchheimer equation, determining its key coefficient of inertial permeability (i.e., the quadratic coefficient representing inertial losses) requires a series of flow tests under different hydraulic gradients. However, how to choose the volumetric flux range for deriving inertial permeability is never scrutinized. Here we report a seemly irrational dependency of inertial permeability on volumetric flux range by massive direct numerical simulations of fluid flow in rock fractures. This variation presents traceability and is closely related to the evolution of microscale eddies and macroscale flow regimes. A concept of global inertial permeability is thus proposed from this traceable variation, which can reproduce the whole flow regime in rock fractures. Two parametrized criterion models are subsequently established for calculating global inertial permeability using the lowest cost and ensuring its accuracy: one in terms of minimum Reynolds number for practical purposes, and the other in terms of minimum eddy volume ratio for mechanistic interpretation. Further discussions indicate that the proposed global inertial permeability is of great importance to non-Darcian flow prediction and flow regime assessment. Moreover, the parameterized criterion models in the form of power-law function for determining global inertial permeability, drawn from 2D fracture simulations, can apply to 3D fractures and large-scale fractured rock aquifers. The findings and results in this study are of great significance for accurately evaluating the hydraulic conductivity of rock fractures, especially at relatively high flow rates, and are useful for many geophysical and industrial applications.

Blocking dead zones to avoid plugs in pipes

Slurries of solid particles in a liquid medium are common in chemical and petroleum industries. They possess significant clogging risks when transported in pipes during a process. This paper presents a simple and low-cost method to decrease the plugging risk in a flow system. For this, we streamline the pipes’ internal geometry, significantly reducing the volume of stagnation zones in pipes‘ dead legs. We test the approach experimentally and consider how the plugs are formed before and after the geometry modification. The experiments were carried out in a flow loop with the cohesive ice particles suspended in decane. The particle concentration was 2.2…15 % vol., and Reynolds numbers were below 20,000. After we reduced the dead zones with inserts, the particle concentration of a plug-free regime was increased by 8 % vol., and the minimum transport Reynolds number decreased by 74 %.

Experimental verification of the additivity between the barrel droplet's drag force and the fiber number

AbstractResearch on the single fiber scale has an essential application value in the coalescence field. The study of the droplets' behavior on the fiber junctions can provide support data for optimizing the coalescer's drainage performance on the macroscopic scale. However, relevant research focusing on the force change of droplets during the transition from a single fiber to multifiber is relatively scarce. The droplet migration platform was built using the Lavision particle image velocimetry system, which can observe the behavior of droplets on fibers. The force and fiber number relationship were analyzed when the droplets were separated from the fiber junctions. Two detachment modes were observed during separation. Furthermore, the drag force satisfied the additivity theorem by increasing the number of fibers under the same detachment mode. In addition, the minimum Reynolds number prediction model was established under two separation forms, which had good prediction compared to the experimental data.

Numerical simulation of forced convection of ferro-nanofluid in a U-shaped tube subjected to a magnetic field

In this paper, the forced convection of forced convection of ferro-nanofluid in a U-shaped tubed subjected to a magnetic field is investigated. Modeling is performed in three sizes of the bending radius. The Reynolds number and Hartmann number ranges are 600 ≤ Re ≤ 1200 and 2 ≤ Ha ≤ 12, respectively. The ferro-nanofluid is composed of water-based Fe3O4 particles that vary in volume fraction from 0.01 to 0.03. The simulation is carried out under a 3-D model, incompressible, laminar and steady-state by using the finite volume method. The results show that at a constant bending radius, as the Hartmann number increases, the current density increases. In addition, in all cases, as the Hartmann number increases, the coefficient of friction increases, and the presence of a magnetic field reduces the velocity of the fluid flow. Also, the maximum and minimum friction coefficients are related to the minimum Reynolds number and the maximum Reynolds number, respectively. The results also show that the effect of flow velocity on the heat transfer rate is much more significant than the intensity of the magnetic field and increasing the bending radius of the tube leads to a further reduction of fluid energy and the fluid receives less heat flux from the wall.

Investigation of thermal behavior and performance of different microchannels: A case study for traditional and Manifold Microchannels

In this study, numerical investigation of forced convection of non-Newtonian nanofluid into Traditional Microchannel (TMC) and Manifold Microchannel (MMC) containing porous media is investigated. The bottom of the microchannel wall is kept at a constant temperature and the other walls are heat insulated. The inside of the microchannel is filled with porous media. This study addresses a new sample of microchannels (MMC and TMC) filled with porous material in a three-dimensional flow. The results show that increasing the Reynolds number and volume fraction of nanoparticles (φ) and decreasing the porosity, lead to heat transfer improvement. At low Reynolds numbers, there is no relative advantage over TMC or MMC. Finally, it is concluded that to achieve the acceptable performance evaluation criterion PEC (PEC> 1) at porosity ε= 0.9, thermal conductivity ratio (γ) must be γ <0.9. When thermal conductivity of porous foam is considered low, PEC variations and heat transfer rates in the traditional microchannel are uniform. For TMC and MMC in the maximum Reynolds number, the percentage of increase in heat transfer at ε= 0.9 compared to the absence of porous material is 13.5% and 26.26%, respectively. For MMC and TMC in minimum permeability and maximum Reynolds number, the percentage of reduction of heat transfer rate from n = 0.5 to n = 1.5 is equal to 0.86% and 0.37%, respectively.

Impact of Partial Slip on Double Diffusion Convection of Sisko Nanofluids in Asymmetric Channel with Peristaltic Propulsion and Inclined Magnetic Field.

The current article discusses the outcomes of the double diffusion convection of peristaltic transport in Sisko nanofluids along an asymmetric channel having an inclined magnetic field. Consideration is given to the Sisko fluid model, which can forecast both Newtonian and non-Newtonian fluid properties. Lubricating greases are the best examples of Sisko fluids. Experimental research shows that most realistic fluids, including human blood, paint, dirt, and other substances, correspond to Sisko’s proposed definition of viscosity. Mathematical modelling is considered to explain the flow behavior. The simpler non-linear PEDs are deduced by using an elongated wavelength and a minimal Reynolds number. The expression is also numerically calculated. The impacts of the physical variables on the quantities of flow are plotted graphically as well as numerically. The results reveal that there is a remarkable increase in the concentration, temperature, and nanoparticle fraction with the rise in the Dufour and thermophoresis variables.

Sedimentation behavior of suspensions in milliflow reactors

The increasing industrial interest in performing heterogeneous solid–liquid reactions in tubular reactors requires control over sedimentation and accurate data on the solid velocity in order to optimize residence time distribution. In this research, concentration profiles, velocity profiles and the minimum suspension speed of non-interacting solids in milliflow reactors are determined via image analysis and particle image velocimetry. First, the five different flow regimes are identified. Then, the minimum suspension speed, ensuring a homogeneous/pseudo-heterogeneous flow regime is observed to be between 0.7 and 1 m/s depending on liquid and solid properties. Reducing the liquid viscosity and density difference, decreases the minimum suspension speed. Based on the free settling velocity, the Reynolds number to avoid sedimenting of particles can be calculated. The identification of minimum suspension speed and Reynolds number were found to be the determining factors enabling broader application of heterogeneous solid–liquid reactions in milliflow reactors, in academic and industrial environments.

Stability of plane shear flows in a layer with rigid and stress-free boundary conditions

We study the stability of shear flows of an incompressible fluid contained in a horizontal layer. We consider rigid–rigid, rigid—stress-free and stress-free—stress-free boundary conditions. We study (and recall some known results) linear stability/instability of the basic Couette, Poiseuille and a laminar parabolic flow with the spectral analysis by using the Chebyshev collocation method. We then use an L2\\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$$L_2$$\\end{document}-energy with Lyapunov second method to obtain nonlinear critical Reynolds numbers, by solving a maximum problem arising from the Reynolds energy equation. We obtain this maximum (which gives the minimum Reynolds number) for streamwise perturbations Rec=Rey\\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$$\\mathrm{Re}_c={\ ext {Re}}^y$$\\end{document}. However, this contradicts a theorem which proves that streamwise perturbations are always stabilizing, Rey=+∞\\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$${\ ext {Re}}^y=+\\infty $$\\end{document}. We solve this contradiction with a conjecture and prove that the critical nonlinear Reynolds numbers are obtained for two-dimensional perturbations, the spanwise perturbations, Rec=Rex\\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$$\\mathrm{Re}_c={\ ext {Re}}^x$$\\end{document}, as Orr had supposed in the classic case of Couette flow between rigid planes.

Approximate approach for improving pressure attenuation accuracy during hydraulic transients

Abstract The quasi-steady friction model is generally adopted in water hammer simulation in pipe network systems, which cannot accurately reflect the attenuation of pressure, while the existing unsteady friction model is challenging to use in complex pipe network systems. In this study, a convenient method for treating the friction term is proposed based on the Moody diagram. The attenuation process of water hammer pressure can be accurately reflected by reading the relationship curve between Reynolds number and the Darcy friction factor in the pipeline transient process. Combined with the classical water hammer experiment and the long pipe valve closing experiment in our laboratory, the accuracy of this model is verified, and the influence of absolute roughness (e) and Reynolds number (Re) on the model was analyzed as well. The results show that the pressure attenuation using the Method of Characteristics (MOC) and the proposed friction model has a good agreement with the experimental data. The absolute roughness has little influence on the results in hydraulically smooth pipe, while the minimum Reynolds number has a significant influence. When selecting the minimum Reynolds number, 2% ∼ 5% of the initial flow rate is recommended for calculation.

Assessment of physical and aerodynamic properties of corn kernel (KSC 704)

AbstractPhysical and aerodynamic properties of agricultural products have been used to precision design of the machinery and different postharvest operations. In this study, some physical and aerodynamic properties of corn kernels were investigated as a function of size at three levels of small, medium, and large (8.41 mm, 9.60 mm, and 11.62 mm, respectively) and moisture content in the range of 10 to 25% (wet basis). According to results, with increasing the moisture content from 10 to 25%, the length; width; thickness; geometric mean diameter; mass; and volume of the small, medium, and large kernels linearly increased by 2.22%, 2.40%, and 1.68%; 3.76%, 1.82%, and 1.55%; 3.98%, 2.49%, and 2.67%; 3.29%, 2.23%, and 1.93%; 8.44%, 7.38%, and 8.12%; and 11.32%, 10.71%, and 10.21%, respectively. In this regard, the projection area of the specimens increased nonlinearly by 6.06%, 4.26%, and 3.26%, respectively; whereas, true density decreased nonlinearly by 2.65%, 2.91%, and 1.90%, respectively. The maximum terminal velocity (12.41 m/s) and Reynolds number (6,037.44) were observed in large kernels and the moisture content of 25%; whereas, the minimum terminal velocity (8.32 m/s) and Reynolds number (3,016.55) were observed in small kernels and the moisture content of 10%. Maximum and minimum drag coefficients (1.014 vs. 0.302) were observed in small kernels at the moisture of 10% and large kernels at the moisture of 25%, respectively. The polynomial model was chosen as a suitable model in order to foretell the terminal velocity, Reynolds number, and drag coefficient of kernels as a function of moisture content and size.Practical ApplicationsMechanical damages to seeds at harvesting, transporting, and threshing are major concerns for farmers and farming experts. All the processes of wind transfer, flotation, and seed removal from seed mixture and the stalks and husks depend on the behavior of the seeds in the wind flow. Therefore, it is crucial to determine the physical and aerodynamic properties of various crops to analyzing their behavior during transport, processing, and precise design of farm equipment and machines to minimize wastes.

Numerical analysis of heat transfer in peripheral air vaporizers used in cryogenic storage tanks

In this study, the numerical study of fluid flow and heat transfer in peripheral air vaporizers used in cryogenic tanks are studied. The vaporizers under consideration include two simple and non-simple peripheral air vaporizers. The fluid enters the vaporizer (in the liquid phase) which is connected to a cryogenic reservoir and, after heat exchanging, is converted to the gas phase, and exits the vaporizer. Inside the two vaporizers, porous foam is used to fill the entire cross-section of the channel. The k-ε realizable model is used for simulation. The Reynolds numbers and the surface temperature vary in the ranges of 2400≤Re≤3000 and 280 K≤Ts≤310 K, respectively. The results show that in all cases, the use of non-simple vaporizer versus simple vaporizer has more satisfactory results in increasing heat transfer. Also, the performance of vaporizers at low surface temperatures leads to a further increase in heat transfer. The presence of porous foam can be considered as an auxiliary factor in increasing heat transfer. In maximum and minimum Reynolds numbers, the percentage increase in convective heat transfer coefficient for surface temperature changes from 300 to 310 K is equal to 15 % and 17 % for the simple vaporizer and 30% and 20% for the non-simple vaporizer, respectively.

An iterative machine-learning framework for RANS turbulence modeling

Machine-learning (ML) techniques provide a new and encouraging perspective for constructing turbulence models for Reynolds-averaged Navier-Stokes (RANS) simulations. In this study, an iterative ML-RANS computational framework is proposed that combines an ML algorithm with transport equations of a conventional turbulence model. This framework maintains a consistent procedure for obtaining the input features of an ML model in both the training and predicting stages, ensuring a built-in reproducibility. The effective form of the closure term is discussed to determine suitable target variables for the ML algorithm, and the multi-valued problem of existing constitutive theory is studied to establish a proper regression system for ML algorithms. The developed ML model is trained under a cross-case strategy with data from turbulent channel flows at three Reynolds numbers, and a posteriori simulations of channel flows show that the framework is able to predict both the mean flow field and turbulent variables accurately. Interpolation tests for the channel flow show the proposed framework can reliably predict flow features that lie between the minimum and maximum Reynolds numbers associated with the training data. A further test of the ML model in a flow over periodic hills also demonstrates improved performance compared with the k-ω SST model, even though the model is only trained with planar channel flow data.