Mathematical modeling and multiresponse optimization to reduce surface roughness and adhesion in Al 5083 H116 alloys used in ammunition propulsion actuators

Purpose This study aims to obtain the lowest surface roughness (Ra) and drill bit adhesion values (AV) depending on the change in control factors (cutting speed-Vc, feed rate-f and drill bit-D) during drilling of the Al 5083 H116 alloy. Low roughness values increase the fatigue strength of the final part and affect tribological properties such as lubrication and friction. In the machining of ductile materials, the AV increases the Ra value and negatively affects the tool life. Design/methodology/approach Drilling tests were conducted using Taguchi L16 orthogonal array. The experimental measurement findings for Ra and AV were adjusted utilizing the Grey Relational Analysis (GRA), the Response Surface Method (RSM) and Artificial Neural Networks (ANN) to generate prediction values. SEM detected drill-tip adhesions and chip morphology and they were analyzed by EDX. Findings Ra and AV increased as the f increased. Vc affects AV; 86.04% f on Ra and 54.71% Vc on AV were the most effective control parameters. After optimizing Ra and AV using GRA, the f is the most effective control factor. Vc: 120 m/min, f: 0.025 mm/rev and D2 were optimal. ANN predicted with Ra 99.6% and AV 99.8% accurately. Mathematical models are obtained with RSM. The increase in f increased AV, which had a negative effect on Ra, whereas the increase in Vc decreased the adhesion tendency. With the D1 drill bit with the highest flute length, a relatively lower Ra was measured, as it facilitates chip evacuation. In addition, the high correlations of the mathematical models obtained indicate that the models can be used safely. Originality/value The novelty of this study is to determine the optimum drilling parameters with GRA and ANN for drilling the necessary holes for the assembly of ammunition wing propulsion systems, especially those produced with Al 5083 H116 alloy, with rivets and bolts.