Cylindrical Test Section Research Articles

Experimental investigation and modelling of hydrodynamics and heat transfer in flow boiling in normal and microgravity conditions

The development of long-term space thermal management systems has informed research into the influence of gravity on boiling. This work explored the influence of gravity on the hydrodynamics and heat transfer of boiling flow. Experiments were carried out using two test loops each consisting of a 6 mmID transparent cylindrical test section. Upward (+1░g) and downward (−1░g) flow boiling experiments were carried out in the laboratory while microgravity (μg) experiments were carried out during a parabolic flight campaign. The results of flow visualisation showed significant influence of gravity on the flow patterns and the influence of gravity was generally limited to mass flux, G≤400kg/m2s and/or vapor quality, x≤0.35. In all three gravity conditions, the measured heat transfer coefficient was influenced by heat flux, mass flux and/or vapor quality. For liquid Reynolds number, Relo≤2000(G≤150kg/m2s) and boiling number Bo<0.002 the measured heat transfer coefficient was highest in −1g flow and lowest in μg flow but becomes comparable at Bo>0.002. A correlation for predicting microgravity heat transfer coefficient was proposed in this work and the proposed correlation predicted 100 % of the μg data in the current work within ±20%, predicted nearly 100 % of the μg data of Ohta et al. (2013) within ±30% and around 85 % of the μg data of Narcy (2014) within -20 % to +50 %. A correlation for predicting the gravity dependent regime as it relates to heat transfer coefficient in +1g and μg flows was also proposed in this work. The proposed criterium correctly predicted over 85 % of the gravity-dependent heat transfer coefficient in the current work and the works of Lebon et al. (2019), Narcy (2014), Ohta et al. (2013).

Comparative investigation of convective heat transfer and pressure drop characteristics of MWCNT, Fe3O4, and MWCNT/Fe3O4 nanofluids

The present work discusses the experimental investigation of the convective heat transfer (CHT) and pressure drop characteristics of a hybrid nanofluid (HNF). The multi-walled carbon nanotube (MWCNT) NF, Fe3O4 NF, and MWCNT/Fe3O4 HNFs with 0.025 wt% to 0.2 wt% concentrations were prepared through ultrasonic dispersion. The CHT coefficient of the three NFs was investigated in a cylindrical test section at different Reynolds numbers (Re) and compared. As an outcome, the CHT coefficients of the MWCNT/Fe3O4 HNFs ranged from 1823.2 to 2030.5 W/m2·K, which is 6%–15.9% more than the base fluid. Furthermore, the MWCNT/Fe3O4 HNFs outperformed the MWCNT and Fe3O4 NFs individually during the CHT coefficient enhancement. At Re of 1000–1600, the increase in pressure drop of the MWCNT NF, Fe3O4 NF, and MWCNT/Fe3O4 HNFs varied from 62.4% to 91.7% as compared to water. Therefore, it is expected that the MWCNT/Fe3O4 HNF could be an effective heat transport medium.

Numerical Investigation of Mach 2.5 Axisymmetric Turbulent Shock Wave Boundary Layer Interactions

Shock wave boundary layer interactions are common to both supersonic and hypersonic inlet flows. Wall-resolved implicit large-eddy simulations of a canonical Mach 2.5 axisymmetric shock wave boundary layer interaction experiment at Glenn Research Center were carried out. A conical shock wave was generated with axisymmetric centerbodies with 16 deg half-angle cone. The centerbody radii were 9.2% and 14.7% of the test section diameter. The conical shock wave interacted with the turbulent boundary layer on the inside of the cylindrical test section. The experimental Reynolds number based on diameter was six million. For the simulations, the Reynolds number was reduced by a factor of 10 to lower the computational expense. The turbulent boundary layer separates for both centerbody radii and the separation is stronger for the larger centerbody radius. Frequency spectra of the spanwise-averaged wall-pressure coefficient reveal low-frequency content at Strouhal numbers based on separation length between 0.02 and 0.05 in the vicinity of the separation shock and mid-frequency content between 0.1 and 0.2 downstream of separation. A proper orthogonal decomposition captures spanwise coherent structures with a Strouhal number of 0.03–0.04 over the interaction region and streamwise coherent structures inside and downstream of the interaction with a Strouhal number of 0.1–0.4.

Experimental and numerical studies on the two-dimensional flow characteristics in the radially stratified porous bed

The experiment and numerical simulations are both conducted in the present study to better understand the two-dimensional flow characteristics in the radially stratified porous bed. Spherical particles of two different sizes are packed in the left part and right part of a cylindrical test section separately to form a two-layer bed with the configuration of radial stratification. The variations of pressure drops in each part of the stratified bed are measured when water flows up through the packed bed. Meanwhile, the numerical simulation is also carried out to investigate the flow field in the stratified bed, especially the flow characteristics around the interface of two parts. The results indicated that the pressure drops in the two layers of radial stratified bed are almost equal. When the fluids flow up through the radially stratified porous layers, the lateral flowing from the low permeability layer to the high permeability layer leads to a decrease of pressure drop in the low permeability layer and an increase of pressure drop in the high permeability layer. Most of the lateral flow occurs in the initial part of the test section. Besides, the lateral and vertical pressure gradient can be well predicted by Ergun equation respectively.

Time-resolved measurements of the local wall heat flux in turbulent Rayleigh–Bénard convection

We report direct and time-resolved measurements of the local wall heat flux in turbulent Rayleigh–Bénard convection in air. The measurements have been performed in a cylindrical test section with a diameter of 7.1 m and a diameter-to-height ratio of 8. They cover Rayleigh numbers 8.5×105<Ra<2.6×109, while the Prandtl number was fixed at Pr=0.7. In order to measure the local wall heat flux, we use heat flux plates that have been flush mounted into the hot/cold surfaces of the convection experiment. The results of our measurements show that the local wall heat flux in turbulent Rayleigh–Bénard convection strongly fluctuates. With increasing Rayleigh number, the variance of the fluctuations goes down with Ra−0.5 and the probability of large excursions from the mean decreases. The probability density of the fluctuations can be well-described by a distribution according to the Generalized Extreme Value Theory. We also analysed the typical time scales, and we found that the power of the fluctuations becomes more weighted towards higher frequencies, if the Rayleigh number increases.

Estimation of dynamic load factors for elastic cylinders under dynamic internal pressure

This paper discusses the design of a cylindrical test section subjected to a dynamic internal pressure for severe accident experiments performed at CEA. A commonly used method consists in computing the static equivalent response of the structure and then applying DLF coefficients. In this paper, Dynamic Load Factor (DLF) coefficients are obtained for cylinders. Based on a cylinder subjected to a stepped internal pressure, a set of dynamic equations is set up using the membrane theory with a modal approach (bending moments are neglected with this theory). As already commonly established, it is found that radial and axial displacements are coupled, resulting in a Multi-Degree Of Freedom (MDOF) model with coupling. The maximum DLF of the cylinder is therefore determined for both radial and axial displacements. It is found that the axial DLF reaches a maximum value when there is a specific ratio between the radius and the length (parameter \(\pi R/L\) equal to 1.0), while the radial DLF does not depend on the geometry. Results are compared to Finite Elements dynamic simulations. On the whole, the dynamic results are validated for long cylinders (i.e. \(\pi R/L\le 0.1\)). However, differences in the axial displacement tend to increase with an increase in the ratio \(R/L\) (shorter cylinders). The regular value of 2.0 established for the Single Degree Of Freedom model and for both axial and radial directions is exceeded when the radial and axial modes are coupled. The difference can be significantly higher (DLF \(\ge\) 3.9).

Reflected shock-initiated ignition probed via simultaneous lateral and endwall high-speed imaging with a transparent, cylindrical test-section

A TRAnsparent cylindrical (TRAC) shock tube test section has been developed to explore the ignition structure behind reflected shockwaves via simultaneous lateral and endwall high-speed imaging using three fuel/oxidizer mixtures: 0.1% n-C7H16/1.10% O2/4.14% N2/94.66% Ar, 0.1% iC8H18/1.25% O2/4.69% N2/93.96% Ar, and 5% CH4/10% O2/85% CO2. During mild ignition, the first exothermic centers were located outside of the observable test section. It is shown that with increasing temperature the far-wall ignition events move closer to the endwall, until a strong ignition is achieved. The temperature gradients that promote mild ignition were quantified and were found to have a slight dependence on fuel with the RMS temperature gradients being 0.80% for n-heptane and 1.07% for isooctane. The impact of temperature gradients on metrics used to improve kinetic models, such as species time histories, is noteworthy. From this temperature gradient, an experimental correlation is provided to estimate the transition temperature from mild to strong ignition for a mixture. The effect of diagnostic location in non-homogeneous reactions is quantified in reference to ignition delay. The unique and insightful perspective on the ignition process under heavy shock bifurcation is also discussed.

Investigations on two-phase flow resistances and its model modifications in a packed bed

During severe core melting accidents of light water reactors with failures of all cooling systems, the molten core fuel (corium) would meet and interact with residual coolant water (FCI), then break up and fragment into a porous debris bed. Therefore, the debris bed coolability depending mainly on the flow friction laws, will be in great significance to nuclear reactor safety. Motivated by reducing the uncertainties in debris coolability assessment, this paper reports an experimental study on two-phase flow resistance in porous packed beds, and a modified model is proposed to predict the pressure drops of two-phase flow through packed beds with coarse particles. The tests are performed on the DEBECO-LT facility which is constructed to investigate the friction laws of adiabatic single and two-phase flow in a particulate bed. The spherical particles of 1.5 mm, 2 mm, 3 mm, 4 mm, 6 mm and 8 mm in diameter are packed separately in the cylindrical test section with the inner diameter of 120 mm and the height of 600 mm. The quality of experimentation and instrumentation can be ensured by the good agreement between the measured pressure drops of single phase flow tests and the predictions of Ergun equation. Then air-water co-current two-phase flow tests are conducted to investigate the flow resistance characteristics, providing the basic data for verifying, analyzing and modifying the existing models. The results show that: 1) when two-phase fluids flow upward through the porous beds packed with the smaller particles (e.g. 1.5 mm and 2 mm), the flow resistance increases gradually with the fluid flowrate, and the predictions of Reed model are more comparable with the measured pressure drops. While for the beds with larger size parties (e.g. 4 mm, 6 mm, and 8 mm), the flow resistances of two phase flow show a down-up tendency and hardly be predicted well by previous models; 2) it is believed that the interfacial drag plays non-ignorable role to the flow resistances for the beds packed with larger size particles, and should be considered specifically in the analysis models; 3) A modified model is proposed to predict the two-phase flow resistance in the packed bed with coarse particles. Compared with the previous models, the predictions of modified model show a favorable agreement with the experimental data under different conditions.

Flow resistances characteristics in a particulate bed with the configurations of uniform mixture and stratification

During severe core melting accidents of light water reactors with failures of all cooling systems, the molten core fuel (corium) would meet and interact with residual coolant water (FCI), then break up and fragment into a porous debris bed. Therefore, the debris bed coolability depending mainly on the flow friction laws, will be in great significance to nuclear reactor safety. This paper is concerned with reducing uncertainty in quantification of debris coolability in a severe accident of light water reactors (LWRs), and the experimental results on the flow resistance characteristics of homogeneous and stratified particulate beds are reported here. The objective is to get an idea of how the particles bed characteristics (such as the stratified information and its hierarchical arrangement) affect its flow resistance which is crucially important to debris bed coolability analysis. Three types of beds are packed in a cylindrical test section with the inner diameter of 120mm and the height of 600mm. Type-1 bed is a homogeneous bed packed with single size spheres. Type-2 bed is also a homogeneous bed with uniformly mixing by two sizes of spheres. Type-3 bed is the axially stratified bed which is composed of two sizes of glass spheres same as that in the Type-2 bed. Both single and two phase flow tests are carried out, the pressure drops and its flow resistance characteristics are measured and recorded during the tests. The results show that for gas-water co-current flow through a homogeneous bed, the predictions of Reed model are more comparable with the measured pressure drops. For a bed packed with uniform mixture of particles, the measured pressure drops are close to the predictions of Ergun equation with area mean diameter at low flowrate (e.g. Rep<7), but the length mean diameter should be considered as increasing of the Reynolds number of fluid. Compared with the homogeneous bed with the same particles, the stratified bed will generate a lower flow resistance, and consequently result in a higher dry-out heat flux under boiling conditions.

Pressure losses and interfacial drag for two-phase flow in porous beds with coarse particles

Motivated by reducing the uncertainties in coolability assessment of debris bed with coarse particles formed during the severe accident of core melting in light water reactors, this paper reports the experimental investigations on the pressure losses for two phase flow through packed porous beds with coarse particles as well as the effects of the interfacial drag between fluid phases on debris coolability analysis. The experiments are carried out on the test facility of DEBECO-LT (DEbris BEd COolability-Low Temperature), which was designed to investigate single/two-phase flow in porous beds. The spherical particles of 4mm, 6mm and 8mm in diameter are packed in the cylindrical test section with the inner diameter of 120mm and the height of 600mm. The pressure losses are measured when fluids flow through the porous packed beds. The experimental data with predictable results by models are analyzed comparatively, and the influences of interfacial drag on two-phase flow pressure losses are discussed. The results show that, for the packed beds with coarse particles, the pressure losses curves show a descent-ascent tendency along with the fluid velocity, and this tendency will become more significant in packed bed with bigger size particles or under lower liquid velocity. The decrease for pressure losses in beds with coarse particles resulted from the effects of interfacial drag, and the rates of the interfacial drag to two-phase pressure losses are rising significantly in beds with greater size of particles, which indicates the interfacial drag plays an important and nonnegligible role in total pressure losses. Therefore, for the packed beds with coarse particles, the interfacial drag should be considered specifically and nonnegligible in debris coolability analysis models.

Role of gaseous film cooling injector orientation in duct flow: Experimental Study

Experimental investigations are carried out to analyze the effect of coolant injector configuration on overall film cooling performance in a cylindrical test section similar to a rocket combustion chamber. Three different injector orientations are investigated: (i) holes parallel to core gas flow, (ii) holes at a tangential angle to the core gas flow, and (iii) compound angle holes inclined to the wall both in the tangential and azimuthal direction. The objective is to provide a consistent set of measurements for cooling effectiveness within the same test facility with respect to different blowing ratios ranging from 1.69 to 3.9. The results suggest that coolant injection parallel to the axis should be used in situations where far field effectiveness is of concern. The tangential injection, in general, shows lower wall protection compared to other schemes. The results show that the compound angle configuration augments film cooling effectiveness in the near injection regimes. However comparison of the effectiveness values for the compound injectors suggests the existence of an optimum compound angle configuration. The circumferential wall temperature data represent systematic changes in the coolant flow behavior with the change from straight injection to compound angle injection.

Time-Resolved High-Speed Visualization and Analysis of Underwater Shock Wave Focusing Generated by a Magnetic Pulse Compression Unit

Paper reports on visualization and analysis of underwater shock waves generation, focusing, and propagation. Recently, shock waves have been attractive with rising attention in medical field. Extracorporeal shock wave lithotripter has been one of its very successful applications. However, the effects of shock waves on living tissue are not sufficiently understood, due to unknown parameters involving the shock wave interaction. For better understanding of affecting points, a precise and controllable shock wave source is required. In this paper, in order to study shock wave effects, pulsed electric discharges produced by magnetic pulse compression circuit were used to generate underwater shock waves. A half-ellipsoidal reflector was used for shock wave focusing. The whole sequences of the shock wave generation, propagation, focusing, and cavitation bubbles, which appeared after shock wave passage, were visualized by time-resolved high-speed shadowgraph method. The peak pressure of shock wave focusing at second focal point reached 180 MPa. A cylindrical test section was used to observe phenomena around the electrode from the pulse discharge to bubble collapse. The first cavitation bubble's collapse occurred at about 750 μs. In the experiment, the two shock waves, one generated by expansion of plasma and another generated by collapse of the cavitation bubble were studied. Using high-speed visualization and image analysis, the velocity and pressure of the two shock waves were evaluated. The peak pressure of shock wave generated by bubble collapse reached 157 MPa.

C204 球状燃料を有する高温ガス炉の燃料温度解析に及ぼす熱出力の影響(OS8 軽水炉・新型炉・原子力安全)

This paper indicates the predictive simulation of thermal-hydraulic analysis with packed spheres in high temperature reactor core. The pressure drop experiments were carried out in the air through cylindrical test section. The porosity simulation is calculated in a computer code of the particle method. The thermal-hydraulic computer code was developed, and it was identified as PEBTEMP. The predictive analysis was applied to the design condition of thermal power profile in the axial direction.

Influence of coolant injector configuration on film cooling effectiveness for gaseous and liquid film coolants

An experimental investigation is conducted to bring out the effects of coolant injector configuration on film cooling effectiveness, film cooled length and film uniformity associated with gaseous and liquid coolants. A series of measurements are performed using hot air as the core gas and gaseous nitrogen and water as the film coolants in a cylindrical test section simulating a thrust chamber. Straight and compound angle injection at two different configurations of 30°–10° and 45°–10° are investigated for the gaseous coolant. Tangential injection at 30° and compound angle injection at 30°–10° are examined for the liquid coolant. The analysis is based on measurements of the film-cooling effectiveness and film uniformity downstream of the injection location at different blowing ratios. Measured results showed that compound angle configuration leads to lower far-field effectiveness and shorter film length compared to tangential injection in the case of liquid film cooling. For similar injector configurations, effectiveness along the stream wise direction showed flat characteristics initially for the liquid coolant, while it was continuously dropping for the gaseous coolant. For liquid coolant, deviations in temperature around the circumference are very low near the injection point, but increases to higher values for regions away from the coolant injection locations. The study brings out the existance of an optimum gaseous film coolant injector configuration for which the effectiveness is maximum.

Experimental study of toroidal shock wave focusing in a compact vertical annular diaphragmless shock tube

The paper reports results of experiments regarding toroidal shock wave focusing in a vertical shock tube as a part of a series of converging shock wave studies. This compact vertical shock tube was designed to achieve a high degree of reproducibility with minimum shock formation distance by adopting a diaphragmless operating system. The shock tube was manufactured in the Institute of Fluid Science, Tohoku University. An aspheric lens shaped cylindrical test section was connected at the open end of the shock tube to visualize the diffraction and focusing of the toroidal shock wave released from the ring shaped shock tube opening. Pressure transducers were flush mounted on the shock tube’s test section to measure pressure histories at the converging test section. Double exposure holographic interferometry was employed to quantitatively visualize the shock waves. The whole sequence of toroidal shock wave diffraction, focusing, and its reflection from the symmetrical axis were successfully studied. The transition of reflected shock waves was observed.

Film Condensations on Horizontal and Slightly Inclined upward and downward Facing Plates

Condensation experiments were carried out at atmospheric pressure using water-cooled horizontal and slightly inclined flat plates, varying the air concentrations and the orientations (upward or downward facing). The plate was suspended in a cylindrical test section and exposed to a very slow flow of pure steam or mixtures of steam and air from underneath. For the slightly inclined cases, the test results showed good agreement with the existing studies and reproduced the typical inclined plate phenomena. For the horizontal cases, the downward facing plate showed twice as high heat transfer rates as the upward facing one. The air contained in the steam showed negligible effects for the upward facing plate but a systematic decrease in heat transfer for the downward facing one. The condensation heat transfer beneath the downward facing horizontal plate is influenced by the surface wetting characteristics.

Controlling the form of strong converging shocks by means of disturbances

The influence of artificial disturbances on the behavior of strong converging cylindrical shocks is investigated experimentally and numerically. Ring-shaped shocks, generated in an annular cross sectional shock tube are transformed to converging cylindrical shocks in a thin cylindrical test section, mounted at the rear end of the shock tube. The converging cylindrical shocks are perturbed by small cylinders placed at different locations and in various patterns in the test section. Their influence on the shock convergence and reflection process is investigated. It is found that disturbances arranged in a symmetrical pattern will produce a symmetrical deformation of the converging shockfront. For example, a square formation produces a square-like shock and an octagon formation a shock with an octagonal front. This introduces an alternative way of tailoring the form of a converging shock, instead of using a specific form of a reflector boundary. The influence of disturbances arranged in non-symmetric patterns on the shape of the shockfront is also investigated.

The effect of corium composition and interaction vessel geometry on the prototypic steam explosion

This paper discusses the results of steam explosion experiments using molten material consisting of UO 2 and ZrO 2 mixture, which is called corium, to simulate a prototypic steam explosion in a nuclear reactor during a postulated severe accident. About 5–10 kg of molten material with enough superheat was poured into a pool of water in a test section at room temperature to simulate ex-vessel steam explosion in the reactor situation. Most of the experiments were externally triggered. The purpose of the experiments was to investigate the effect of material composition and average void fraction on the strength of a prototypic steam explosion, which were highlighted as major unresolved issues. The experiments were performed using two kinds of mixtures, one, corium A, at 70:30 weight percent composition of UO 2 and ZrO 2, close to eutectic composition, and the other, corium B, at 80:20 weight percent. Also, two kinds of cylindrical test sections having a different diameter were used. It turned out that corium A was likely to produce an energetic steam explosion, while corium B seldom led to an energetic steam explosion. The existence of mush phase for the non-eutectic mixture is suggested to be the reason for the difference. Comparative cross sectional views of the corium particles by scanning electron microscope supported the proposed argument. The tests performed with a narrow test section seldom led to an energetic steam explosion for both materials. An increase in average void is suggested to be the reason for the non-explosive behavior, which is consistent with the physical models employed in the current steam explosion computer codes.

Film condensations of flowing mixtures of steam and air on an inclined flat plate

A programme of condensation experiments was carried out at atmospheric pressure using a water-cooled flat plate varying the air concentrations, the plate inclination and its orientation (upward or downward facing). Rates of heat transfer have been measured on a single face of the condensing plate suspended in a cylindrical test section as steam and mixtures of steam and air flowed over it. The rate of heat transfer decreased as the angle of the plate to the horizontal was reduced and as the concentration of air was increased. A notable observation was that comparison of results for the upward and downward facing cases showed that the heat transfer rates with pure steam are higher for an upward facing plate than for a downward facing one. However, with air present in the steam, this trend is reversed. The effects of plate orientation, mixture flows and buoyancy are discussed.

Experimental comparison of film-wise and drop-wise condensations of steam on vertical flat plates with the presence of air

This paper presents experimental results comparing film-wise and drop-wise condensations. The condensing plates were developed to promote film-wise or drop-wise condensation respectively. Rates of heat transfer have been measured on a single face of water-cooled flat plates suspended vertically in a cylindrical test section as steam and mixtures of steam and air flowed over it. In the pure steam cases, the drop-wise condensations showed much higher heat transfer rates than film-wise condensations, which showed good agreements with the Nusselt theory of natural convection condensation. However in the steam and air mixture cases, as expected, both modes of condensations fell in similar range of heat transfer rates. Due to the difference in the condensate flows, the drop-wise condensation showed even lower heat transfer rates than film-wise condensation with the presence of air.