Design and Development of a Semi-Automatic Robusta Coffee Bean Sorting Machine with Optimal Operating Conditions Determined by Response Surface Methodology

Abstract

Hin Chang Si Coffee community enterprise, Khon Kaen, Thailand faced a production bottleneck in coffee bean sorting (3.04 kg/h vs. target 100 kg/day). A semi-automatic Robusta coffee bean sorting machine was designed and developed comprising seven subsystems: electrical control, power transmission, bean storage, suction conveying, air-blowing impurity removal, a three-tier rotating sieve (apertures 7.5, 6.5, and 5.2 mm per TAS 5700-2561), and discharge. FEM verified structural integrity (σmax=33.73 MPa, FoS=18). Optimal conditions were determined via RSM-BBD with three factors: suction time X₁(10-20 s), air-blowing velocity X₂ (6.08-10.09 m/s), and drum speed X₃ (10-20 rpm). The quadratic model (R²=0.9989) revealed X₃ as dominant (β₃=-4.092, p<0.001), consistent with Residence Time theory; X₂ as secondary (β₂=-0.751, p<0.001); X₁ non-significant (p=1.000). The global optimum at X₁ =15 s, X₂ =6.08 m/s and X₃=10 rpm yielded η=99.35%, throughput 38.49 kg/h (12.66× baseline, +1,166.12%), contamination ≤4%, conforming to TAS 5700-2561.