Evaluation of Natural Convection Solar Dryer for Onion Slices Drying

Abstract

Experiments of onion slices solar drying in thin layer were conducted in Shambat. The objectives of the study were testing the performance of a natural convective solar dryer, determination of the drying characteristics of onion slices, simulation of the drying process of onion slices. Fresh onion were brought from a central market for carrying out the drying trials. A data logger with thermocouples were used for recording temperature at intervals of one hour at inlet of the solar collector, inside absorber out let of the solar collector and ambient air. Two hygrometers were used to record air relative humidities at intervals of one hour at inlet of the solar collector and out let of the solar collector. The drying process of onion was carried out for three successive days interrupted by two overnights. Results indicate that, the solar dryer heated the drying air satisfactorily and this was depicted by the maximum attainable temperature difference between inside absorber and ambient temperature which was 38oC and the maximum attainable relative humidity difference between inlet and out let of the solar collector air relative humidities which was 20%. Onion drying took place in the falling rate period i.e onion initial moisture content was less than the critical moisture content. The effective drying process for onion slices occurred in the first day while equilibrium state was attained during the second and third days of the drying process. Two drying models namely, Lewis and Page, were used as a tester to simulate onion slices drying process consequently two mathematical models were written in turbo Pascal for window. ix Results showed that, the two tested drying models predicted onion slices drying moisture contents accurately, however page model gave close agreement between measured and predicted data for validating the two drying models tested. Statistical analysis showed that, average model error (AME) average absolute difference (AAD) and standard error of estimate (SEE), were 0.120, 0.230 and 0.250, respectively for Lewis model while for Page model, the parameters are 0.112, 0.130 and 0.150, respectively.