Volume 13, Issue 4 - Special Issue: Multiphase Flow in Industrial Plant, Guest Editors: Renzo Di Felice, Marco Fossa, Mario Misale, Annalisa Marchitto

In the present work, we formulate a simplistic two-fluid model for bubbly steam-water flow existing between fuel pins in nuclear fuel assemblies. Numerical simulations are performed in periodic 2D domains of varying sizes. The appearance of a non-uniform volume fraction field in the form of meso-scales is investigated and shown to be varying with the bubble loading and the domain size, as well as with the numerical algorithm employed. These findings highlight the difficulties involved in interpreting the occurrence of instabilities in two-fluid simulations of gas-liquid flows, where physical and unphysical instabilities are prone to be confounded. The results obtained in this work therefore contribute to a rigorous foundation in on-going efforts to derive a consistent meso-scale formulation of the traditional two-fluid model for multiphase flows in nuclear reactors.

The limiting aeration condition necessary to produce the collapse of established pipes (ratholes) is evaluated in a powder bed stored in a flat-bottomed aerated silo. The main parameters tested are the powder properties and the width of the aeration pad at the silo bottom. The measured quantities are the aeration rate and the corresponding size of the pipe. The gas pressure gradient close to the pipe is estimated from the numerical solution of the air flow field in the bulk solids assuming the Darcy flow regime. The powder permeability necessary for this calculation is estimated by fluidization experiments. The hoop stress evaluated according to Johanson (2004) is in good agreement with the unconfined yield strength of the powder, provided that the contribution of aeration to powder consolidation during the pipe formation is properly accounted for.

Multi Stage Flash (MSF) distillation plants are widely used in saline water desalination. In order to enhance MSF, it is important to increase an evaporation rate in the flashing stage. A spray flash method, in which superheated water jets are injected through nozzles into a depressurized environment to increase the gas/liquid interface area, is a promising technique to make the increase of evaporation rate, which leads directly to the reduction of energy consumption and cost of the MSF plant. In this paper, the introduction of microbubbles into the spray jet as the nucleation sites to increase the evaporation rate of the spray flash is proposed. The spray flash behaviours with/without microbubbles at outside/inside of the nozzle-inside were observed by means of a high speed camera to investigate the mechanism of enhancement of spray flash due to microbubbles. Moreover, the number densities of droplets and bubble volume increase were obtained from visualized images in order to discuss quantitatively on the effects of introduction of microbubbles.

Single gas bubbles have been injected into an stirred liquid tank and the eventual breakup process of the bubbles was examined through high-speed imaging. Experimental observations of the breakup probability, breakup time, number of daughter bubbles and daughter bubble size distribution have been collected. The occurrence of non-equal-size breakup events dominated over equal-size breakup events. The frequency of binary and multiple breakup events was about equal. The largest uncertainty is associated with the determination of the breakup time because the bubbles take continuously altering deformed shapes already from the point of injection into the tank. The present experimental data do not support the standard model assumption regarding the number of daughter particles. The active breakup zone in the stirred tank was in the large velocity field close to the radial impeller. It is not evident whether the breakup events are due to time average shear or pressures and velocity fluctuations.

Two-phase flow is widely encountered in minichannels heat exchangers such as air-cooled condensers and evaporators for automotive, compact devices for electronic cooling and aluminum condenser for air-conditioning applications. In the present work, frictional pressure drop during adiabatic liquid-vapor flow is experimentally investigated inside a single 0.96 mm diameter minichannel. Tests have been run with three mixtures of R32/R1234ze(E) (23/77%, 50/50% and 75/25% by mass composition) at mass flux ranging between 200 and 600 kg m −2 s −1 . Since pressure drop has a strong influence on the two-phase heat transfer, it is crucial to have reliable pressure drop prediction methods for two-phase heat transfer modeling and optimization. Therefore, with the aim of extending its validity range, a model to calculate the frictional pressure gradient during two-phase flow in small diameter channels is tested against the present two-phase pressure drop database. An assessment is also done with two low-GWP refrigerants: the halogenated olefin R1234ze(E) and the hydrocarbon R290. The present model accounts for the effect of internal surface roughness as a function of the liquid-only Reynolds number.

Hydrodynamic slugs in pipelines are usually analysed by using steady-state flow simulation software as a first approximation and then modelled in transient simulation software in order to get more accurate values. In this work, the ILF approach for hydrodynamic analysis is presented by using an empirical and a mechanistic method for steady-state simulations and by using a transient approach. For both, steady-state and transient simulations commercial software is applied. Furthermore, a new pseudo-transient approach as an alternative to the Lagrangian slug tracking scheme that is currently tested is presented. A set of operational hydrodynamic slug examples illustrates the use and accuracy of both commercial simulation programs and the newly developed pseudo-transient approach. Finally, a set of considerations for the hydrodynamic slug analysis in pipelines is proposed as a guide for accurate estimation of the slug properties.

The silicon tracker of the ATLAS experiment at CERN Large Hadron Collider will operate around –15°C to minimize the effects of radiation damage. The present cooling system is based on a conventional evaporative circuit, removing around 60 kW of heat dissipated by the silicon sensors and their local electronics. The compressors in the present circuit have proved less reliable than originally hoped, and will be replaced with a thermosiphon. The working principle of the thermosiphon uses gravity to circulate the coolant without any mechanical components (compressors or pumps) in the primary coolant circuit. The fluorocarbon coolant will be condensed at a temperature and pressure lower than those in the on-detector evaporators, but at a higher altitude, taking advantage of the 92 m height difference between the underground experiment and the services located on the surface. An extensive campaign of tests, detailed in this paper, was performed using two small-scale thermosiphon systems. These tests confirmed the design specifications of the full-scale plant and demonstrated operation over the temperature range required for ATLAS. During the testing phase the system has demonstrated unattended long-term stable running over a period of several weeks. The commissioning of the full scale thermosiphon is ongoing, with full operation planned for late 2015.

This work faces the challenge of cutting the specific energy demand in the CO 2 capture process based on Ca-looping technology. The use of high-temperature sorbents allows an efficient integration of the excess heat flows. Up to now, several investigations studied the Ca-looping integration with external systems such as a steam cycle. In this research, a further step is done by comparing technological solutions for the internal heat integration with the aim of reducing the energy needs. Particles preheating before entering the regeneration reactor appears as an opportunity for energy saving since solids have to be heated up around 250–300°C from one reactor to another. Two different internal heat integration possibilities making use of a particle separation device and a mixing valve are presented and compared. The former consists of the inclusion of a cyclonic preheater. This configuration presents the a priori advantage of a more developed technology since it is widely used in the cement industry but the drawback of a worse gas–solid heat exchange. Although there is a lack of practical experience regarding the use of a single seal valve to feed two reactors, this configuration presents a promising prospective related to the excellent heat exchange features of the solid flows. The aim is to obtain comparative results by means of implementing advanced thermochemical models, in order to make progress on the development of less energy-intensive configurations of the calcium looping.

The air-water two phase critical flows through a safety relief valve commonly used in the refrigeration industry is examined with particular emphasis on the prediction of the critical mass flowrates using CFD based approaches. The expansion of the gas through the valve and the associated acceleration is coupled to the liquid phase and results in changes to the velocity slip with the possibility of influencing the choking conditions and the magnitude of the critical mass flows. These conditions are poorly reported in the literature for safety valves. This paper presents a study where the ability of established two phase multi-dimensional modelling approaches to predict such conditions are investigated. Comparison with the simplified mixture model will show that this model tends to underestimate mass flowrates for medium to high liquid mass fraction. However, the two fluid model can adequately account for the thermal and mechanical non equilibrium for these complex flow conditions with the use of simplified droplet sizing rules.

Slug flow is one of the most frequently occurring multiphase flow patterns in industrial processes. A deep knowledge of its fundamentals is necessary to accurately model not only the fluid flow but also reaction and heat and mass transfer in several operation units. A numerical study is reported on the dynamics of slug flow, under laminar regime, in vertical columns of stagnant and co-current Newtonian and non-Newtonian liquids (shear-thickening and shear-thinning). A CFD package (Ansys FLUENT) with the VOF methodology was applied to simulate the flow of individual and pairs of consecutive Taylor bubbles. The behaviour of the most relevant hydrodynamic features with the approach of the trailing bubble towards the leading one is addressed, with particular emphasis to the role of the liquid rheology and flow configuration. The main results presented are the velocity ratio curves between consecutive bubbles, the variation of the bubbles shape, and the axial velocity and viscosity fields in the surrounding liquid. This bubble-bubble interaction data can be a keystone to improve and complement continuous slug flow simulators used for very long columns.

Uneven distribution of phases in plate heat exchangers is a cause of reduction in both thermal and fluid-dynamic performances. With respect to two-phase flows, phase separation in manifolds with several outlets is a complex phenomenon and no general rules are available for predicting the phase distribution at header–channel junctions. The design of compact heat exchangers and their distributors is still based on empirical approaches and both experimentation and numerical analyses are needed for defining the best geometries able to reduce the mass flow rate non-uniformities in parallel channels. In this paper, a series of CFD simulations are carried out to infer the effects of a protrusion fitting (inside the header) on the single-phase distribution in parallel upward vertical channels fed by a common horizontal distributor. The numerical results are compared with both experimental single-phase and two-phase (liquid/gas) experimental data. The effects of the operating conditions are investigated and general conclusions on the differences and analogies between single-phase and two-phase flows in the present problem are discussed.