چكيده به لاتين
Liquid-liquid extraction (LLE) used in many industries, such as petroleum, food, hydrometallurgy, and chemical industries. LLE processes include two main steps: mixing and separation. One of the mixing methods, without mechanical agitation, is using an eductor. The main objective of the present work, the hydrodynamic and mass transfer aspects of the use of a novel eductor-based contacting device for liquid–liquid extraction (LLE) system has been evaluated by computational fluid dynamics (CFD) modeling, dimensionless numbers and cure fitting, and response surface methodology (RSM). Educator performance was studied by investigating the effects of jet velocity, the throat to nozzle area ratio (At/An), the column to nozzle diameter ratio (Dc/Dn), the projection ratio (Ltn/Dt), and two phases flow ratio (Qc/Qj) on the suction ratio (Rs), the mixing efficiency (ηm), and the mixing energy efficiency (ηe), sauter mean diameter (SMD), drop size distribution (DSD), and mass transfer coefficient. The high surface tension water/acetic acid/toluene LLE system was applied in the experiments. CFD results were validated by droplet rise velocity and dispersed phase holdup of water/toluene system; errors were 20.7% and 15.4%, respectively. Results show the lower jet velocities corresponds to higher Rs and ηe, but lower ηm and it is low than 2 m/s is recommended. By decreasing At/An ratio, Rs and ηm decreases, while at the higher At/An, Rs and ηe are lower and the value around 100 can be selected. The more Dc/Dn has better ηm and ηe, while Rs reduces insignificantly; The low Ltn/Dt corresponds to lower ηm and higher ηe, while Rs reduces insignificantly, and less than 50 is more appropriate. The optimum Ltn/Dt approximately is between 1 to 2; By increasing Qc/Qj, ηm and ηe decreases, while Rs reduces with small value, and the lower continuous phase rate causes the better venturi mixing performance. Experimental results show that for different nozzle diameters, from 1 to 3 mm, the same maximum jet length appears. Velocity range of formation was much wider for smaller nozzle diameters. Existence of venturi in front of nozzle reduces the SMD and produces narrower and symmetric DSD. Decrease of throat to nozzle area ratio and projection ratio decrease SMD and produce narrow and more symmetric DSD. Increase of two phases flow rate ratio increases the SMD trivially, and does not effect on DSD. A power law model was fitted to predict SMD. N-T model could predict DSD appropriately, while normal and log-normal models don’t fit well with the experimental data. The two parameters of N-T model were correlated, one with SMD and the other one with operational and geometrical parameters of equipment. Also, a power law model was fitted to predict mass transfer coefficient (Kc) in eductor LLE. The droplet to nuzzle diameter ratio is the most important parameter in predicting of Kc. By reducing of droplet size, Kc is increasing. Kc in disperse to continues phase is higher than continues to disperse phase. The optimum conditions for higher Kc in eductor LLE is Dn=1.567, Dt=29.659, Lnt=15.852, and RQ=2.7933. The eductor LLE device can be used economically and gives some advantages such as less corrosion and sealing problems, less contact time, no requirement for moving parts, provision of high interfacial area and higher overall mass transfer coefficients, reduced contactor volume and etc.
