MODELING THE DRYING CHARACTERISTICS OF FERMENTED LOCUST BEANS

Abstract

The production of good quality fermented locust beans is essential, having known its rate of spoilage. Different methods have been used over time and this has probably affected the quality of stored fermented locust beans. This project is meant to guide the producer in selecting the best drying method and model to be used for processing of fermented locust bean. Three different drying methods were used: the mechanical (convective) dryer, indirect (cabinet) solar dryer and sun drying. The mechanical (convective) drying experiments were conducted for four drying air temperatures at 45°C, 55°C, 65°C and 75°C at air velocity of 0.9 m/s. During the drying process, the weight loss was measured at 30 min intervals for all the drying methods. The result was used to determine the moisture content, moisture ratio and the drying rate. The drying data were fitted with thirteen published thin layer drying models. The best model was investigated by comparing the determination of coefficient (R2), reduced chi-square (χ2) and root mean square error (RMSE) between the experimental and predicted values. The organoleptic properties and the proximate analysis were determined at the end of the experiments. Result from this work shows that at the mechanical drying method, the highest air temperature of 75°C gave the least drying time. The moisture content and the drying rate decreases with the drying time for all the air temperatures. . The constant and falling rate periods were observed. The drying characteristics also vary with the drying methods. The mechanical drying method gave the least drying time compared to sun and solar drying methods. The highest drying rate was observed at the mechanical drying method. Comparison between the three drying methods showed that mechanically dried condiments at air temperature of 55°C gave the best quality for both the organoleptic test and the proximate analysis. The Modified Henderson. and Pabis model gave the highest value of R2 (0.997), the lowest χ2 (0.000239) and RMSE (8.53471E-05), highest value of R2 (0.997), the lowest χ2 (0.000187) and RMSE (5.84742E-05), highest value of R2(0.995), the lowest χ2 (0.000506) and RMSE (0.000158), highest value of R2(0.994), the lowest χ2 (0.000996) and RMSE (0.000226) for 45°C, 55°C, 65°C and 75°C experiment respectively. The Modified Henderson. and Pabis model was also found best to described the drying curves for the drying methods . It gave the highest value of R2 (0.987), the lowest χ2 (0.001068) and RMSE (0.000388) and highest value of R2 (0.99), the lowest χ2 (0.000988) and RMSE (0.000353) at the sun and solar experiment respectively. The results showed that the Modified Henderson. and Pabis model was found best to describe all the drying experiment of fermented locust beans. The Artificial Neural Network model was successful in predicting the experimental drying kinetics of fermented locust beans. Fick’s second law was used to calculate the Effective moisture diffusivity. The moisture diffusion coefficient varied between 2×10−10 and 8×10−10 m2/s for the given temperature range and corresponding activation energy was 1.099KJ/mol.