International Journal of Multiphase Flow最新文献

Wavy flow of straight rivulets down an inclined plate was studied experimentally and theoretically. Instantaneous three-dimensional shape of a rivulet was obtained experimentally using Brightness-Based Laser-Induced Fluorescence technique. The results were compared to the predictions by quasi-two-dimensional theoretical model. The model reproduces well the transition of wave pattern from soliton-like to sinusoid-like waves with increasing wave frequency, as well as the main wave characteristics such as shape, amplitude and velocity. The agreement gets worse for overall Reynolds number of 252, where the local thickness-based Reynolds number is large enough for turbulence to start. The discrepancy also grows with the rivulet width, since the region where turbulence develops is getting larger. In such conditions, the experimentally observed waves may acquire additional small-wave perturbations of their rear slopes and deep troughs across the slopes. Such perturbations are substantially three-dimensional and their prediction requires more complex theoretical models.

A single jet was studied in a rectangular fluidized bed by analyzing the pressure fluctuations data to characterize the jet hydrodynamics. The bed contained with glass beads of different mean sizes (450, 650, and 920 µm) as well as pharmaceutical pellets (920 µm). The analysis of the pressure fluctuation data was performed using the power spectral density function (PSDF) and discrete wavelet transforms technique. This study investigated how the flow regime changed from a jet in a fluidized bed to a spouted regime as the gas injection velocity was increased. Increasing the mean particle size led to an increase in the minimum spouting velocity. The minimum spouting velocity for 450, 650, and 920 μm glass bead particles and for 920 μm pharmaceutical pellets were determined to be 32.47 m s^-1, 38.66 m s^-1, 44.78 m s^-1, and 44.47 m s^-1, respectively. The study also examined how the dominant frequency of the various particles and the energy percentage of scales changed with increasing injection velocities. The ratio of energy percentages of meso-scale changes as the injection velocity increases and these changes were used to estimate the minimum spouting velocity. Wavelet sub-signals analysis revealed that the Daubechies 2 (DB2) wavelet was the most effective at capturing the characteristics of the jet. Based on the Shannon entropy of the approximate coefficients, the wavelet analysis showed that 11 levels of decomposition were required. By combining the wavelet analysis and PSDF, a more detailed analysis of the meso-scale was achieved. This study provided valuable insights into the behavior of jets and spouts in fluidized beds, which can greatly contribute to the optimization of a jet in fluidized beds and spout-fluidized beds by enhancing our understanding of jet behavior in such environments.

We report on experimental and numerical results of the impact of millimeter sized droplet with high-speed projectile. The impact velocity $V_{imp}$ ranges between 70 m/s and 245 m/s and is sufficiently high that the compressibility of the liquid becomes important. High-speed images reveal non-axisymmetric lamella spreading, hydrodynamic and secondary cavitation and the ejection of a thin jet from the distal side of the droplet. The spreading velocity is supersonic for the impact velocities tested. Secondary cavitation at the distal side of the droplet is found for $V_{imp}\gtrsim$ 120 m/s. The experiments are compared and analyzed further with a CFD model based on a compressible volume of fluid (VoF) method. Excellent agreement of the early droplet and lamella dynamics is obtained with the experimental data. Additionally, we provide quantitative data for the pressure loading and the shear stress on the projectile using spatio-temporal maps.

The thermal log-law ${\Theta }_{+}\left(y_{+}\right)=\beta +2.12log\left(y_{+}\right)$ is valid in flow boiling with a value of $\beta$ that evolves as the flow develops. Using a multiphase flow cross-literature database, this constant is shown to be ${\beta }_{OSV}=-7$ at the point of onset of significant void (OSV). This means that at the OSV the liquid is at saturation temperature up to $y_{+}\simeq 30$. The OSV predictions using this model have a similar mean average error as the Saha and Zuber 1974 correlation for one less fitted constant for channel, pipe and annular flows for pressure from 1 to 147 bar and Peclet numbers from $3.5\cdot 10^3$ to $4\cdot 10^5$. This model is used to build a heat flux partitioning (HFP) inspired from system-scale codes (Lahey 1978). It predicts the distribution of the heat flux between the liquid phase and the evaporation term when the total heat flux is known. It does not give information on the total heat flux as a function of wall temperature and cannot be used to draw a boiling curve. In imposed flux conditions, this partition provides more coherent flux distribution between the evaporation and liquid terms than Kurul and Podowski (1990) based approaches and improves void fraction predictions in high-subcooling regions on the DEBORA database (Garnier et al. 2001) and on experiments by Bartolomei and Chanturiya (1967) and Bartolomei et al. (1982). When the wall temperature is imposed, it must be coupled with an empirical boiling total heat flux correlation to replace a traditional HFP. The prediction of the total heat flux is then as good as that of the correlation.

An inverted liquid bottle with a small neck diameter empties through periodic admission of air bubbles at the neck, followed by liquid discharge, a process termed ‘glugging’. In contrast, a large Taylor bubble rises to the air–water interface in the bottle for large neck diameters, followed by an instant interface collapse and a chaotic liquid discharge. We numerically find that a spiral large-scale rotating structure due to churning motion develops in the liquid during time evolution in an emptying bottle, though an initial swirl is not imposed. The induced structure is strong for a large neck bottle, and the circulation strength is maximum near the neck region. The spiral structure’s strength decreases for small neck diameters, and a pure oscillatory ‘glugging’ mode is preserved. The high circulation strength near the neck region for the large neck bottle causes liquid to accelerate and is the reason for the existing empirical models on liquid discharge to deviate from experimental observations.

This study experimentally investigates the spatiotemporal evolution and associated local instabilities of a drop subjected to a weak shock wave. The front and side views of the drop are captured to understand its three-dimensional evolution and breakup. The interaction of shock causes the windward side of the drop to compress and generate a surface wave over it. Its temporal amplification is found to be governed by Kelvin–Helmholtz instability. The core of the deformed drop expands in a stream-wise direction, forming a Rayleigh–Taylor instability-driven bag structure. Consistent pressure gradients across the bag cause its continuous elongation until the pressure gradient overcomes the surface tension. This continuous elongation leads the sheet to undergo kinematic thinning, which causes the sheet to destabilize and nucleate the hole. This hole recedes and gathers liquid from upstream to thicken its interface, called the bag rim. The accelerating receding motion of the bag rim triggers Rayleigh–Taylor instability, and the corrugation that forms over it grows into ligaments and destabilizes to shed droplets through end pinching and ligament merging. Additionally, the accelerating rim undergoes radial expansion, with its further destabilization governed by the coupled effect of Rayleigh–Taylor and Rayleigh–Plateau instabilities, as well as the collision of the receding bag rim. This leads to the formation of corrugations, which grow into ligaments and further destabilize to shed drops via end pinching. Nonlinear effects dominate ligament dynamics and increase with the Weber number. The asymmetric ejection of the daughter drop from the rim causes it to evolve into a bag, undergoing tertiary breakup.

In this study, we investigate a novel type of static thermal instability that occurs in two-phase flow systems. The instability is theoretically and experimentally confirmed to occur for flow boiling micro-channel scenarios and flow condensing scenarios. The theory presented is independent of the evaporator geometry as well as flow boiling/condensing situations. A stability criteria is derived using a Reynolds Transport approach applied to the evaporator of a two-phase pumped loop (TPPL) system. The theory is then validated experimentally using a TPPL containing a parallel micro-channel evaporator. The instability is confirmed with pressure, temperature, and void fraction measurements acquired at the exit of the evaporator. The findings reveal that static thermal instabilities can arise when simultaneous heat addition and reduction in system saturation temperature occurs (or vice versa). The implications of the thermal instability result in dramatic changes in evaporator heat flux, as well as a flow transition from stratified laminar flow to vigorous turbulent flow with high void fractions. By identifying the additional instability mechanisms, this work contributes to enhancing system reliability and predictability of TPPL systems.

Heavy and light particles are commonly found in many natural phenomena and industrial processes, such as suspensions of bubbles, dust, and droplets in incompressible turbulent flows. Based on a recent machine learning approach using a diffusion model that successfully generated single tracer trajectories in three-dimensional turbulence and passed most statistical benchmarks across time scales, we extend this model to include heavy and light particles. Given the particle type – tracer, light, or heavy – the model can generate synthetic, realistic trajectories with correct fat-tail distributions for acceleration, anomalous power laws, and scale dependent local slope properties. This work paves the way for future exploration of the use of diffusion models to produce high-quality synthetic datasets for different flow configurations, potentially allowing interpolation between different setups and adaptation to new conditions.

The analysis of bubbly two-phase flows is challenging due to their turbulent nature and the need for intrusive phase-detection probes. However, accurately characterizing these flows is crucial for safely designing critical infrastructure such as dams and their appurtenant structures. The combination of dual-tip intrusive phase-detection probes with advanced signal processing algorithms enables the assessment of pseudo-instantaneous 1-D velocity time series; for which the limitations are not fully fathomed. In this investigation, we theoretically define four major sources of error, which we quantify using synthetically generated turbulent time series, coupled with the simulated response of a phase-detection probe. Based on the analysis of 10¹⁰ simulated bubble trajectories, our findings show that typical high-velocity flows in hydraulic structures hold up to 15% error in the mean velocity estimations and up to 35% error in the turbulence intensity estimations for the most critical conditions, typically occurring in the proximity of the wall. Based on thousands of simulations, our study provides a novel data-driven tool for the estimation of these baseline errors (bias and uncertainties) in real-word phase-detection probe measurements of bubbly flows (air concentrations $c<40\text{\%}$).