MODELING THE PERFORMANCE OF A BAOBAB SEED DECORTICATOR USING RESPONSE SURFACE METHODOLOGY

S. O. Ibrahim; F. A. Ola; T. A. Ishola; A. Saleh; B. A. Sanusi; F. B. Akande

S. O. Ibrahim Department of Agricultural Engineering, Ladoke Akintola University of Technology, Ogbomoso, Nigeria
F. A. Ola Department of Agricultural Engineering, Ladoke Akintola University of Technology, Ogbomoso, Nigeria
T. A. Ishola Department of Agricultural and Biosystem Engineering, University of Ilorin, Ilorin, Nigeria
A. Saleh Department of Agricultural and Bioresources Engineering, A. B. U. Zaria, Nigeria
B. A. Sanusi Department of Agricultural Engineering, Ladoke Akintola University of Technology, Ogbomoso, Nigeria
F. B. Akande Department of Agricultural Engineering, Ladoke Akintola University of Technology, Ogbomoso, Nigeria

Keywords: Baobab seed-Decorticator, Modeling, Crop and machine parameters, Decortication Efficiency, Percent clean kernel.

Abstract

Decortication of baobab seed manually after soaking or roasting is time consuming, labour-intensive and uneconomical. The demand for baobab kernel as a source of protein is increasing for use as supplement for animal protein ration, for edible oil consumption and for other industrial uses. To ameliorate this problem, a baobab seed decorticator has been developed. This study therefore focused on the effects of process parameters; moisture content of seed, speed of decorticator and concave-shaft clearance on percentage clean kernel (P_ck), percentage broken kernel (P_bk), percentage whole seed (P_ws) and decorticating efficiency (DE), using Response Surface Methodology (RSM). In designing the experiments, Box-Behnken Design (BBD) was selected. Analysis of Variance (ANOVA) was carried out to evaluate and select the appropriate dependent (P_ck P_bk, P_ws and DE) and independent variables using both the F and P-values calculated at 95% confidence level (a=0.05). Mathematical models relating the process parameters to the responses were developed. The developed models were validated by comparing the predicted and actual experimental values. The selected quadratic models were adequate for predicting the performance of the developed decorticator. It was observed that a decrease in percentage clean kernel recovery would result if concave-shaft clearance and moisture content were increased. Likewise, the concave-shaft clearance had the highest positive effect while moisture content had the highest negative effect on percentage of broken kernel. All the three factors (process parameters) had positive effects on decorticating efficiency.