OPTIMISATION OF THE PERFORMANCE OF A CLEANING-TYPE TEFF THRESHER USING RESPONSE SURFACE METHODOLOGY

Traditional teff threshing methods are labour intensive, inefficient and prone to considerable post-harvest losses. To address these limitations, a mechanical teff thresher was designed and developed. However, the initial prototype demonstrated suboptimal threshing performance and cleaning efficiency. This study aimed to optimise the performance of the developed teff thresher through the application of empirical modelling and response surface methodology. The optimisation focused on the threshing unit, particularly the drum length and diameter. The machine was fabricated from mild steel, angle iron, aluminium and round bar materials. Based on structural analysis, the total stress was 0.6776 MPa, the maximum shear stress was 0.00013242 MPa and the equivalent (Von Mises) stress was 16.126 kPa. Performance was evaluated at three drum speeds (1000, 1100, and 1200 rpm) and three feed rates (620, 660, and 700 kg/h), under a concave clearance of 0.03 m and a grain moisture content of 14%. A split-plot experimental design was employed, generating 27 observations that were analysed using the Design-Expert software. The results indicated that both drum speed and feed rate significantly influenced threshing performance. The maximum threshing capacity of 287.3 kg/h was achieved at a drum speed of 1200 rpm and a feed rate of 700 kg/h, representing an improvement from the baseline capacity of 187.5 kg/h. Increasing both drum speed and the feed rate within the studied range markedly enhanced the threshing efficiency and throughput of the machine. The optimised operating conditions are recommended to maximise the performance of the teff thresher.