Development of a cyclone rice dryer

摘要

This thesis evaluates the suitability of a cyclone dryer for the drying of paddy (rice grain).udThe design aim is to reduce the moisture content of paddy from a fresh harvest level ofud33% dry-basis to a more manageable level of 22% of dry basis, and to replace alludclassical methods of drying. The cyclone dryer consists of a cylindrical tower containinguda series of inverted conical baffles with central orifices that divide the tower intoudchambers. The moist particulate matter is fed into a stream of hot, dry air which entersudtangentially at the base of the tower, creating a rotating flow within the dryer. The centraludvortex and through-flow jet transport the particles upwards from chamber to chamberuduntil discharged tangentially at the top. Recirculation of flow within the chambersudlengthens the particle residence timeudSingle-phase numerical calculations with the commercial RANS-based ComputationaludFluid Dynamics (CFD) code CFX 5.7 are used for flow field and pressure dropudpredictions. Experimental observations in a small scale laboratory model are used forudvalidation. Useful descriptions of the axial and tangential velocity distributions areudobtained, and the pressure drop across the cyclone dryer chamber is predicted to aboutud20% accuracy.udParticle residence time in the laboratory model cyclone dryer is measured by the pulseudtracer stimulus response technique. Observations using paddy grain and spherical silicaudgel particles show the mean residence time to vary quadratically with particleudconcentration. The residence time distribution (RTD) is explained well by a tank-in-seriesudmodel. Numerical predictions of particle residence time obtained from one-way coupledudparticle transport modelling without particle dispersion using a Lagrangian/Eulerianudapproach produce RTDs differing significantly from the experimental observations.udHowever, the trends of mean residence time variations in response to changes in inlet airudvelocity and number of cyclone chambers are correctly predicted. udSingle-pass drying tests with paddy grain demonstrate maximum moisture reductions ofud2.6-6.5% dry-basis obtained at inlet air temperatures of 82-89 °C and paddy grain feedudrate of 0.03 kg/s. The specific energy consumption (SPEC) varies between 7.5-20.5udMJ/kg of water evaporated, depending on the initial moisture content of paddy grain.udCompared with fluidised bed and spouted bed paddy dryers employing 50-70% exhaustudair recycling, the cyclone dryer gives a lower moisture reduction and a higher SPEC. Thisudindicates that practical application of the model-scale cyclone dryer would require multipassudoperation with exhaust air recycling to be sufficiently economic.udMulti-pass laboratory tests show three-pass drying in a four-chamber cyclone dryer withudinlet air temperature of about 80°C and 0.03 kg/s paddy grain feed rate to reduce theudmoisture content of paddy grain by about 11% dry-basis, with an SPEC of 13 MJ/kg ofudwater evaporated. Recycling about 90% of the air would give a 70-75% reduction inudSPEC compared to non-recycled air operation, and 3.5-4 MJ/kg water evaporated. This isudcomparable to fluidised bed paddy dryer operation for similar initial moisture content ofudthe paddy grain.udNumerical simulations of silica gel particle drying based on two-way coupled particleudtransport modelling with a Lagrangian/Eulerian approach are also reported. Theudsimulations consistently overpredict the moisture and heat transfer observedudexperimentally using silica gel particle 3.25 mm in average diameter. As the meanudparticle residence time was underpredicted , it is concluded that the water evaporationudmodel used here gives a much higher moisture transfer rate than that observedudexperimentally.udComputational studies of the increase in residence time with geometric scale-up of theuddryer indicate that a commercial scale unit small enough to be field portable couldudachieve the desired moisture reduction of paddy grain with single pass operation.