부산대학교 도서관

This paper adopts an artificial intelligence (AI) based fuzzy logic inference system to investigate the drilling parameters influence on 12 wt.% seashells (SS), 1 wt.% multi-walled carbon nanotubes (MWCNTs) fortified in Nylon66 matrix (NSM). For specimen preparation, extruders and injection moulding equipment are used. Drilling studies are performed on the CNC vertical machining centre considering rotational speed (RS), feed rate (FR) and helix angle (HA) as inputs and surface roughness (SR) and material removal rate (MRR) as outputs. The experimental matrix is designed using Taguchi’s L 9 orthogonal array, and the outputs are optimized using multi-objective grey relational analysis (GRA). A fuzzy inference system (FIS) is used to remove the fuzziness in the system to improve the optimal conditions. Results show that MRR increases with higher RS and FR whereas SR lowers with higher RS but increases with higher FR. Optimal HA (30°) produces better SR and higher MRR, whereas lower HA produces less amount of MRR and poor finish. The hybrid grey-fuzzy approach produces better results than the grey analysis with a higher R 2 value and lower error %.