Optimazation Of Operational Parameters And Numerical Simulation On The Flow Field In A Turbo Air Classifier

Material classification is seen as the most basic technology of powdertechnology. Turbo air classifier is one of the most common extension devices,the research of turbo air classifier mainly concentrated on improving the partof structure, not do enough study on the theory and the internal flow field.Classification indexes affect each other, so only find a relatively optimalsolution can enable high efficient production. Airflow movementcharacteristics affect the trajectories of particles, and then affect theclassification performances. In this paper, orthogonal experimental method isapplied. The model of objective function was setted up by Matlab to find therelatively optimal solution. Smooth and steady flow field was established withFluent software through studying the inner field of a turbo air classifier.Eventually, using material experiment to validate the results of the simulation.Using the Fluent discrete phase model （DPM） to simulation the trajectory ofsingle particle. Through setting up the model of multi-objective programmingoptimization for classification performance, The model was established formaterial experiment. The optimal solution could be obtained under theexperimental conditions. The optimal solution: the cut size was21μm, theaccuracy was0.6, the classification efficiency of Newton was61%, and finepowder production rate was57%.It is easy to produce turbulent pulse in the annular region when thespeed of rotor cage is too large or too small. When the rotor speed was toosmall, negative pressure and swirl were formed at the surface area of rotorblades. When the rotor speed was too large, negative pressure and anti-swirlwere formed at the surface area of rotor blades. When the wind speed was8m·s-1, the rotating speed was600-800r·min-1, the accuracy of turbo airclassifier was high. Through this simulation, we can find the correspondingreasonable range of the rotor cage speeds under different air inlet velocity.The change of the number of rotor cage blade could affect the flow fieldin the blades channel. When the radius of the cage was same, there was abetter number of rotor cage blades, making the flow field stability. Theanti-swirl between the blades could be effectively weakened when bowedrotor cage blade was adopted, and the flow field was more uniform.The position of the material fed point had effect on the trajectories ofparticle in a turbo air classifier, but had little effect on whether could it enterinto the area between neighboring blades of the rotor or into coarse powder. The trajectories were different when different size particles were fed at thesame position. Smaller diameter particles moved into the annular zone, largerdiameter particles rotated in the annular region, and eventually were gatheredinto the coarse powder-collecting cone at the base. Simulation results showedthat: When the air inlet velocity was8m·s^-1, rotor cage rotary speed was800r·min^-1, the cut size of talcum powder material could be calculated as30–40μm, the cut size of quartz sand material could be calculated as40–50μm;When rotor cage rotary speed was1200r min-1, the cut size of talcum powdermaterial could be calculated as20–30μm, the cut size of quartz sand materialcould be calculated as30–40μm. The simulation method provides a newmethod to determine the cut size of a turbo air classifier, as well as provides areference method to study the cut size of various types of classifier.