ORIGINAL PAPER Artificial Intelligence System Approach for Optimization of Drilling Parameters of Glass-Carbon Fiber/Polymer Composites U. Hari Babu 1 & N. Vijaya Sai 2 & Ranjeet Kumar Sahu 3 Received: 6 April 2020 /Accepted: 4 August 2020 # Springer Nature B.V. 2020 Abstract In recent times, the study on machining characteristics of combined (hybrid) fiber polymer composites has drawn a remarkable research attention because of its emerging industrial applications. The present study focuses on the drilling of hybrid glass-carbon fiber reinforced (GCFR) epoxy composites fabricated using hand layup technique. The machining characteristics were consid- ered in the drilling of GCFR composites which include thrust force, torque, delamination factor and surface roughness. The influence of the drilling process parameters such as spindle speed, drill diameter and feed rate on the characteristics are studied. To avoid the confounding effect of the individual optimized characteristics, an artificial intelligence system i.e. fuzzy inference system approach is adopted. The fuzzy inference system transformed all the performance characteristics of drilled hybrid composites into a multi response performance index (MPI) and optimized the MPI at the common factor level setting. The optimal combination of process parameters for minimum thrust force, torque, delamination factor and surface roughness found to be: speed 3000 RPM, drill diameter 5 mm and the feed rate 50 mm/min. The analysis of variance results show that drill diameter is the most significant parameter followed by feed rate and speed. Further, a theoretical model was proposed for the estimation of MPI and found that an average absolute error of 14.8% with respect to the experimental MPI data is obtained. Keywords GCFR composites . Drilling parameters . Delamination factor . Surface quality . Artificial intelligence system 1 Introduction Over the last decades, various researchers studied the formu- lation of individual fiber incorporated polymer composites and their optimum machining characteristics for various in- dustrial applications. Amidst different fiber incorporated com- posite materials, respective glass and carbon fiber reinforced polymer (GFRP) and (CFRP) composites have been gaining popularity because of their high specific strength property, etc. and interdisciplinary applications. The research findings on the parametric optimization of drilled GFRP and CFRP com- posites and their influence on the machining characteristics are as follows. Kumar et al. [1] drilled polyester matrix based composites reinforced with carbon fiber at 45° and 90° orien- tation using a carbide drill bit coated with titanium aluminium nitride. The machining responses such as torque, delamination factor at inlet and outlet and thrust force were optimized using harmony search algorithm. The optimal parametric combina- tion of the drilled composites found to be spindle speed: 1019 rpm and 1088 rpm, feed: 350 mm/min and 349.5 mm/ min, and diameter of drill: 5 mm (both cases) for 45° and 90° carbon fiber oriented composites, respectively. Anand et al. [2] optimized the thrust force, delamination and torque on the drilled hybrid vinyl ester composites using grey relational Taguchi method. The parameters for favorable hole quality were found to be drill diameter - 4 mm, spindle speed - 2700 rpm and feed rate - 40 mm/min. Latha and Senthilkumar [3] optimized the responses of drilled GFRP composites and found that the feed rate and diameter of drill significantly effect the surface roughness and delamination followed by spindle speed. Xu et al. [4] experimentally studied the drilling of carbon/ epoxy composites and multilayered composite-Ti6Al4V stacks using WC twist drill. The thrust force and drilling * U. Hari Babu upputuriharibabu@gmail.com 1 Department of Mechanical Engineering, Acharya Nagarjuna University, Guntur, India 2 Department of Mechanical Engineering, V.R. Siddhartha Engineering College, Vijayawada, India 3 Department of Mechanical Engineering, National Institute of Technology Karnataka, Surathkal, India Silicon https://doi.org/10.1007/s12633-020-00637-5