---

A practical method for improving the COP of an air-cooled chiller is pre-cooling the entering air of its condenser via a water mist system. This article studies a water mist system with hollow-cone spray nozzles and investigates the effects of water flow rate, water droplet diameter and the number of spray nozzles on system performance. Simulations were run by software FLUENT applying Eulerian-Lagrangian method. Solution grid independency was obtained and it was validated with experimental data. According to the results, in a constant air flow rate of 8.3 (kg/s), with increasing the water flow rate from 0.05 to 0.4 (kg/s), percent increase of COP increases from 3 to about 14, but the percentage of evaporated water decreases from 12.13 to 7.62 (however the value of evaporated water increases). Besides, decreasing the water droplets’ diameter from 200 to 50 micrometer, results in percent increase of COP from 4 to 24. Due to less water evaporation in higher flow rates, the number of spray nozzles was raised in a constant total flow rate that according to the results, increasing the number of nozzles improves the system performance. Also with other simulations it was observed that increasing the number of nozzles is more effective in higher flow rates and less drop diameters. Finally by the case study, it was demonstrated using sufficient number of nozzles, it is possible to achieve higher COPs in lower flow rates and therefore in addition to energy consumption decline, the water consumption could be lowered.

---

The poor formability of Mg results in crack and failure in workpiece material during rolling process and limits its applications in different industries. Numerical modeling of the process can guarantee that the required product properties are met with a minimum production cost. The numerical modeling of the rolling processes requires the coupling of several models including different physical phenomena such as the deformation of the workpiece with its thermo-metallurgical evolution and the thermal evolution of the rolls with its mechanical deformation. On the other hand, in finite element modeling of the rolling process, the meshes of the workpiece are often highly distorted. The high distortion in meshes decreases the confidence in the predicted results. Many formulations based on the viscoelasticity behavior of workpiece material are encountered in the literature to model the rolling process, each with their pros and cons. This present work introduces the Coupled Eulerian-Lagrangian (CEL) formulation, in which the workpiece is divided into three regions (unrolled, in rolling deformation, rolled) to simulate material flow during the process. The comparison of the results with the literature shows that the temperature and strain fields are well predicted by the proposed model. All of the simulations have been done in the two-dimensional mode with ABAQUS/Explicit software.

---

Condensation phenomena in steam flow, can cause droplets with different sizes to form. For an exact prediction of two phase flow behavior, it is necessary to consider the effects of all droplets with different sizes on the steam. In this paper, nucleation equation in an Eulerian–Lagrangian framework has been used to analyze one–dimensional flow of wet steam in a supersonic convergent–divergent nozzle. Polydispersed and monodispersed radius methods for modeling the formed droplets are compared. In polydispersed method, all the formed droplets in the spontaneous condensation zone, are retained in the calculations, without being merged with other groups; but in monodispersed method, all groups are merged, and only one group with averaged radius is retained in the calculations. The polydispersed method has an advantage and can predict the complete droplet spectra. Results of comparing the two methods with experimental data indicates that the predicted radius in the polydispersed method, in every four investigated cases, is closer to experimental data, than that of monodispersed method.

---

Density currents flow due to the density difference between the current and surrounding environment. An important category of density currents is called turbidity currents, which density difference created as a result of suspended solid particle presence in fluid. In the present study, it is tried to use both Eulerian-Eulerian and Eulerian-Lagrangian methods, to take advantage of each one. In this way, the larger particle that have a more effective role in sedimentation mechanism due to the more falling velocity are calculated as Lagrangian and smaller particles by the Eulerian method. In order to obtain a criterion for particle assortment, seven currents with different particle sizes in the Eulerian-Eulerian model have been numerically simulated in a simple channel and it is compared with no particle case, and also the Eulerian-Eulerian method has been verified with experimental results and identified when the particle sizes is less than 12 micron, the sedimentation process is not appreciable, and the presence effect of these kind of particle can be ignored. Therefore, the Eulerian-Eulerian method is a suitable method for this case. The Eulerian-Lagrangian method validation has been performed with experimental results. Finally, the current inside the channel with a spectrum of particle dimensions is simulated and described the results by the proposed method (the combination of two methods). To perform numerical simulations, the development of open-source OpenFOAM codes has been used to take into account the effect of particle. Due to the current’s turbulence, a Large Eddy Simulation method has been used for turbulent modeling.

---

in this paper, conjugate heat transfer in wavy microchannels filled with nanofluid is studied numerically. Homogeneous single-phase models underestimate the experimental results. Then, nanofluid simulated by two-phase model using an Eulerian-Lagrangian approach. Nanofluids are water-Cu or water-Al2O3 suspensions with a particle diameter of 100-150nm and a volume fraction of up to 2%. The three-dimensional governing equations including continuity, Navier-Stokes and energy equations are solved by the well-known SIMPLE method. The governing equations for particles are solved by a 4th order Runge-Kutta algorithm. due to the 3-D governing equation four equations includinf velocity components and energy should be solved for all particles. the computer program has been written in parallel processing method (MPI). Then a super computer with several CPU,s should be used. In one phase model there some supposes, one of them is that the velocty and temperature of a particle is equal to the velocity and temperature of its surrounding fluid. But the main suppose is that the particle distribution is homogeneous. Results show that the main reason of difference between the results of Homogeneous single-phase models and two-phase model is non-homogeneous particle distribution in the domain.

---

In this study, the effects of impact of a projectile on a fuel tank are studied using the finite element method and compared with experimental method. Due to penetration of the bullet into the tank, large internal pressures from the fluid are imposed on the tank's walls which can damage it. The considered fluid structure interaction (FSI) problem is solved in an Eulerian-Lagrangian reference frame by using the LS-Dyna software. By comparing of the results obtained from the simulations and the experimental data, it can be seen that the LS-Dyna software is able to model the different phases of event accurately. In previous researches mostly the penetration and cavitation phases are investigated numerically. In this paper all phases namely penetration, cavitation, stresses applied to tank’s walls and bullet exit are investigated. The comparison between the Von Mises stress of walls in the fluid-filled tank and the empty one signifies 30 percent growth of the maximum Von Mises stress in the wall of the fluid-filled tank compared to the walls of the empty tank. Also in addition to what has been done in previous numerical works, the failure mode of fluid-filled tanks are determined numerically. The numerical results show that because fluid-filled tank walls are pre-stress due to the fluid shock waves, the failure mode of fluid-filled tank is quite different with the failure mode of the empty one.

---