INTERNATIONAL JOURNAL OF LATEST TECHNOLOGY IN ENGINEERING,  
MANAGEMENT & APPLIED SCIENCE (IJLTEMAS)  
ISSN 2278-2540 | DOI: 10.51583/IJLTEMAS | Volume XIV, Issue XII, December 2025  
Hybrid Fuzzy-Based Optimization for Minimizing Delamination in  
GFRP Machining  
1 Prince Dagar, 1 Sharad Kumar, 1 Ashutosh Singh, 2 Vikas Sharma  
1 School of Engineering & Technology, Shri Venkateshwara University, Gajraula, U.P. India  
2 Department of Computer Applications, SRM Institute of Science and Technology, Delhi NCR Campus,  
Ghaziabad, U.P. India  
DOI : https://doi.org/10.51583/IJLTEMAS.2025.1412000057  
Received: 22 December 2025; Accepted: 28 December 2025; Published: 03 January 2026  
ABSTRACT  
Glass Fiber Reinforced Polymer (GFRP) composites are widely used in aerospace, automotive, and structural  
applications due to their high strength-to-weight ratio and corrosion resistance. However, machining these  
heterogeneous and anisotropic materials often leads to delamination, adversely affecting structural integrity  
and service performance. This paper presents a hybrid fuzzy-based optimization approach designed to  
minimize delamination during GFRP machining by integrating fuzzy logic inference with a multi-objective  
optimization framework. The proposed method captures the nonlinear relationships between machining  
parameters—such as cutting speed, feed rate, drill diameter, and tool geometry—and delamination factors.  
Experimental data were used to develop fuzzy rule sets, while the optimization module systematically  
identified optimal parameter combinations that balance machining quality and productivity. Results  
demonstrate that the hybrid fuzzy system effectively reduces delamination compared to conventional  
optimization techniques, providing a robust and intelligent decision-support tool for machining GFRP  
composites. The study highlights the potential of combining fuzzy systems with optimization algorithms to  
address challenges inherent in the machining of advanced composite materials.  
Keywords—GFRP Machining, Delamination Minimization, Fuzzy Logic, Hybrid Optimization, Composite  
Materials, Machining Parameters, Intelligent Modeling, Drilling of Composites.  
INTRODUCTION  
Glass Fiber Reinforced Polymer (GFRP) composites have emerged as indispensable engineering materials  
across diverse industrial sectors due to their exceptional mechanical properties, such as high specific strength,  
excellent corrosion resistance, low density, and superior fatigue behaviour. These attributes make GFRP  
composites ideal for applications in aerospace structures, marine components, automotive body parts, sporting  
goods, and civil engineering infrastructures. As industries increasingly adopt lightweight and high-  
performance materials, demand for reliable and efficient machining of GFRP composites has also grown.  
Machining, particularly drilling, milling, and cutting operations, remains a crucial final-stage manufacturing  
process to achieve assembly requirements and dimensional accuracy. However, machining GFRP composites  
continues to pose significant challenges due to their anisotropic and heterogeneous nature, which differs  
drastically from traditional metallic materials. One of the most critical issues encountered during machining of  
GFRP composites is delamination, a phenomenon involving the separation of fibre layers or matrix cracking at  
the entry or exit of the machined surface. Delamination severely compromises the structural integrity,  
mechanical strength, and service life of composite components. In drilling operations, for example, excessive  
thrust force can initiate interlaminar cracks, leading to peel-up and push-out delamination. Such defects not  
only reduce component performance but also increase rejection rates, rework costs, and overall manufacturing  
time. Therefore, minimizing delamination during machining is essential to maintain the functional reliability of  
GFRP structures. The complexity of delamination behaviour arises from the intricate interactions between  
machining parameters, tool geometry, material properties, and operational conditions. Traditional  
www.ijltemas.in  
Page 633  
INTERNATIONAL JOURNAL OF LATEST TECHNOLOGY IN ENGINEERING,  
MANAGEMENT & APPLIED SCIENCE (IJLTEMAS)  
ISSN 2278-2540 | DOI: 10.51583/IJLTEMAS | Volume XIV, Issue XII, December 2025  
mathematical modelling approaches struggle to capture these nonlinear and interdependent relationships  
accurately. Moreover, conventional trial-and-error methods or deterministic optimization techniques often fall  
short when applied to the machining of composite materials with high uncertainty and variability. This has  
prompted researchers to explore intelligent and adaptive modelling techniques that can effectively handle  
nonlinearities, uncertainties, and multi-objective trade-offs inherent in composite machining.  
Fuzzy logic has emerged as a powerful tool in this domain owing to its capability to model imprecise and  
uncertain information using linguistic rules rather than requiring precise mathematical formulations. A fuzzy  
inference system can capture expert knowledge and convert it into a set of flexible rules that describe how  
machining parameters influence delamination behaviour. By allowing gradual transitions between parameter  
ranges and incorporating human-like reasoning, fuzzy logic proves highly suitable for modelling complex  
machining processes. However, while fuzzy systems excel in prediction, identifying the most optimal  
combination of machining parameters remains a separate challenge, especially when dealing with multiple  
objectives such as minimizing delamination while maintaining acceptable material removal rates or surface  
quality. To address this challenge, hybrid optimization techniques have gained increasing attention. By  
integrating fuzzy logic with computational optimization algorithms, a hybrid fuzzy-based optimization  
framework leverages the strengths of both approaches. The fuzzy component provides accurate modelling of  
the machining process, while the optimization module systematically searches for the best parameter  
combinations that meet predefined objectives. This synergy enables a more robust and intelligent decision-  
making system capable of guiding practitioners toward optimal machining conditions without extensive  
experimentation. The present study focuses on developing such a hybrid fuzzy-based optimization approach  
for minimizing delamination in GFRP machining. The proposed method begins by formulating fuzzy rule sets  
based on experimental data, expert knowledge, and observed relationships between machining parameters and  
delamination outcomes shown in Fig. 1. These rules form the basis of a fuzzy inference system that predicts  
delamination factors for various parameter settings. The predicted outputs are then fed into a multi-objective  
optimization algorithm that identifies optimal machining parameters by balancing delamination reduction with  
process efficiency. Through this integration, the system not only simulates the behaviour of GFRP machining  
under varying conditions but also provides actionable insights for minimizing defects. The study also  
emphasizes experimental validation to ensure the reliability and real-world applicability of the proposed  
method. Compared to traditional optimization approaches and standalone fuzzy systems, the hybrid fuzzy-  
based system demonstrates superior capability in reducing delamination, increasing machining accuracy, and  
achieving stable performance across different machining scenarios. The findings underline the potential of  
intelligent hybrid systems in advancing manufacturing processes for composite materials and contribute to  
ongoing efforts in developing efficient, accurate, and defect-free machining strategies for GFRP composites.  
Hybrid Fuzzy-Based Optimization Workflow for Minimizing  
LITERATURE REVIEW  
Delamination evaluation in fiber-reinforced composites has been widely studied due to its critical impact on  
structural performance. Guo et al. [1] proposed a hybrid RQA-MKLSVM model to assess delamination depth in  
near-surface regions of Glass Fiber Reinforced Polymer (GFRP) laminates. Their approach demonstrated that  
machine learning techniques could accurately capture subtle delamination patterns, enhancing the reliability of  
www.ijltemas.in  
Page 634  
INTERNATIONAL JOURNAL OF LATEST TECHNOLOGY IN ENGINEERING,  
MANAGEMENT & APPLIED SCIENCE (IJLTEMAS)  
ISSN 2278-2540 | DOI: 10.51583/IJLTEMAS | Volume XIV, Issue XII, December 2025  
structural assessments. The influence of machining parameters on composite materials has also been  
investigated. Biruk-Urban et al. [2] examined the effect of technological parameters on cutting forces during  
drilling of GFRP composites. They highlighted that factors such as spindle speed and feed rate significantly  
affect machining performance and delamination. In a related study, Biruk-Urban et al. [3] further analysed the  
impact of cutting parameters on force components during GFRP drilling, emphasizing the importance of  
process optimization to reduce material damage. Advances in hybrid computational modeling have enabled  
more precise prediction and optimization of material properties. Papadimitriou et al. [4] developed the DiMAT  
Materials Modeler (DiMM), a framework integrating machine learning with genetic algorithms to predict and  
optimize material characteristics. This approach demonstrates the potential of hybrid models to accelerate  
materials design and improve predictive accuracy. Similarly, Lepadatu et al. [5] applied Taguchi Design of  
Experiments (DoE) combined with artificial neural networks (ANN) to predict advanced nano-concrete  
characteristics, illustrating the effectiveness of integrating statistical methods with AI for material optimization.  
The evaluation of natural fiber composites has gained attention in recent studies. K. R and M. S [6] proposed  
DelamPredict-X, an ensemble learning-based approach to assess delamination in jute fiber-reinforced materials,  
offering a robust method for predicting failure in bio-composites. Machine learning and optimization techniques  
have also been applied in networked and manufacturing systems. Security mechanisms and threat  
characterization in mobile ad hoc networks were analysed in [7], highlighting the role of computational  
approaches in ensuring network reliability. Pandey and Singh [8] presented a hybrid approach combining grey  
relational analysis and principal component analysis (PCA) to optimize hot machining parameters, showing  
how multi-attribute optimization improves manufacturing efficiency. Furthermore, graph neural networks were  
used for real-time intrusion detection in dynamic mobile ad hoc networks [9], demonstrating the expanding  
applications of AI-based predictive models. Computational modeling and simulation are increasingly used to  
predict the behavior of composite and hybrid materials. Chawla et al. [10] evaluated the performance of spur  
gears made from cast iron and epoxy resin-based hybrid composites using ANSYS simulations, providing  
insights into material selection and design optimization. Jagannathan et al. [11] studied hybrid thermal  
management systems for high-power electronics, integrating phase change materials (PCM), liquid cooling, and  
AI-based control, highlighting the integration of intelligent methods for system optimization. In parallel, Sathish  
et al. [12] optimized interlaminar shear strength of jute/kenaf/glass composites reinforced with MWCNT using  
response surface methodology (RSM), demonstrating the value of multi-parameter optimization in enhancing  
composite mechanical performance.  
PROPOSED METHODOLOGY  
The proposed methodology integrates fuzzy logic modelling with a hybrid optimization framework to  
effectively minimize delamination during the machining of Glass Fiber Reinforced Polymer (GFRP)  
composites. The methodology is structured into systematic phases, ensuring accurate modelling, intelligent  
decision-making, and experimental validation. The following steps outline the complete methodological  
approach adopted in this study.  
1. Experimental Design and Parameter Selection: The methodology begins with a structured design of  
experiments aimed at identifying the influence of critical machining parameters on delamination during GFRP  
machining. Key parameters such as spindle speed, feed rate, drill tool geometry, and point angle are  
systematically selected based on literature review and preliminary trials. A well-defined experimental matrix,  
typically using Taguchi or full factorial design, is prepared to ensure comprehensive coverage of parameter  
combinations. This structured approach helps in capturing the nonlinear interactions among machining  
variables, enabling accurate modeling of delamination behavior and surface quality characteristics.  
2. Experimental Setup and Data Acquisition: Once the parameter matrix is finalized, machining trials are  
conducted on a CNC drilling or milling setup using GFRP composite specimens prepared under standardized  
manufacturing conditions. For each trial, thrust force, delamination factor, hole quality, surface roughness, and  
material removal characteristics are recorded. All machining trials are repeated at least three times to ensure  
statistical consistency and reduce random error. The data acquisition process employs precision measurement  
tools such as dynamometers, surface profilometers, and digital microscopes to ensure high accuracy of the  
recorded responses.  
www.ijltemas.in  
Page 635  
INTERNATIONAL JOURNAL OF LATEST TECHNOLOGY IN ENGINEERING,  
MANAGEMENT & APPLIED SCIENCE (IJLTEMAS)  
ISSN 2278-2540 | DOI: 10.51583/IJLTEMAS | Volume XIV, Issue XII, December 2025  
3. Data Preprocessing and Normalization: The collected experimental data undergoes preprocessing to  
remove inconsistencies, noise, or outlier values that may distort model performance. Missing or anomalous  
readings are corrected or removed after verifying with repeated trials. To enhance the accuracy of the fuzzy  
system, all data is normalized within a standard range, enabling smoother mapping between inputs and outputs.  
This preprocessing stage ensures that the dataset used for fuzzy modeling remains reliable, consistent, and  
suitable for establishing strong rule-based relationships.  
4. Development of Fuzzy Inference System (FIS): A fuzzy inference system is then constructed to model the  
complex and uncertain relationship between machining parameters and delamination response. Linguistic  
variables such as low, medium, and high are assigned to each input parameter, and corresponding membership  
functions are designed. Using expert knowledge and trends observed from experiments, a fuzzy rule base is  
formulated to represent the decision-making logic of the machining process. The FIS uses Mamdani or Sugeno  
inference techniques to derive outputs, translating qualitative rule-based reasoning into quantitative predictions  
of delamination and machining quality. This stage forms the intelligent core of the proposed methodology. To  
enhance the transparency and interpretability of the fuzzy inference system, detailed modeling components are  
defined within the proposed framework. Each input machining parameter, including spindle speed, feed rate,  
tool geometry, and point angle, is represented using triangular or trapezoidal membership functions due to  
their computational simplicity and effectiveness in capturing nonlinear trends observed in machining data.  
Linguistic terms such as Low, Medium, and High are assigned to each input variable, while the output  
variable, delamination factor, is expressed using graded linguistic levels ranging from Very Low to Very High.  
A comprehensive rule base is formulated using expert knowledge and experimentally observed machining  
behavior, resulting in a set of IF–THEN rules that describe the cause–effect relationship between machining  
parameters and delamination. Mamdani-type fuzzy inference is employed to evaluate the rules due to its  
intuitive reasoning and suitability for decision-making problems involving uncertainty. The aggregation of  
fuzzy outputs is followed by defuzzification using the centroid method, which converts fuzzy conclusions into  
a crisp numerical delamination value. This structured fuzzy modeling approach enables accurate prediction of  
delamination while preserving interpretability and robustness.  
5. Hybrid Optimization Using Fuzzy–Evolutionary Approach: To identify the optimal machining  
parameter settings that minimize delamination, a hybrid optimization method combining fuzzy logic with an  
evolutionary algorithm (such as Genetic Algorithm or Particle Swarm Optimization) is employed. The fuzzy  
system evaluates the quality of each candidate solution based on its predicted delamination value, while the  
evolutionary algorithm performs global search and refinement of parameters. This hybrid structure effectively  
balances global exploration and local fine-tuning, allowing the optimization process to converge toward the  
best feasible machining conditions. The result is a set of optimized parameters that yield minimal delamination  
and improved hole quality. The selection of a hybrid Genetic Algorithm–Particle Swarm Optimization (GA–  
PSO) approach is justified by the complementary strengths of the two evolutionary techniques in solving  
complex, nonlinear, and multi-modal optimization problems encountered in GFRP machining. Genetic  
Algorithm contributes strong global search capability through crossover and mutation operations, reducing the  
likelihood of premature convergence and enhancing solution diversity. In contrast, Particle Swarm  
Optimization offers fast convergence and efficient local search by exploiting collective learning and velocity-  
based updates. By integrating GA with PSO, the proposed hybrid framework achieves an effective balance  
between exploration and exploitation, allowing the optimization process to efficiently search the solution space  
while refining promising regions identified by the fuzzy system. The fuzzy inference model serves as a fitness  
evaluator for the GA–PSO algorithm, guiding the evolutionary search toward machining parameter  
combinations that minimize delamination. This hybrid optimization strategy is particularly suitable for  
machining optimization problems where analytical modeling is difficult and response surfaces are highly  
nonlinear.  
6. Validation and Performance Evaluation: The final step involves validating the optimized parameters by  
conducting confirmation experiments under the predicted optimal conditions. The experimental delamination  
values are compared with the fuzzy–optimized predictions to measure accuracy, consistency, and model  
reliability. Performance metrics such as percentage reduction in delamination, prediction error, improvement  
in surface finish, and comparative analysis with initial trials are used to assess the success of the proposed  
www.ijltemas.in  
Page 636  
INTERNATIONAL JOURNAL OF LATEST TECHNOLOGY IN ENGINEERING,  
MANAGEMENT & APPLIED SCIENCE (IJLTEMAS)  
ISSN 2278-2540 | DOI: 10.51583/IJLTEMAS | Volume XIV, Issue XII, December 2025  
methodology. This validation ensures that the hybrid fuzzy-based optimization framework is practically  
effective and capable of enhancing the machining performance of GFRP composites.  
RESULT & ANALYSIS  
This section presents the experimental results, fuzzy model predictions, optimization outcomes, and  
comparative performance evaluation for the proposed Hybrid Fuzzy-Based Optimization System aimed at  
minimizing delamination in the machining of GFRP composites. Data were obtained through structured  
machining experiments, followed by fuzzy modelling and hybrid optimization using the GA–PSO algorithm.  
The dataset used in this study is generated through systematically planned machining experiments on Glass  
Fiber Reinforced Polymer (GFRP) composites in accordance with the proposed fuzzy–hybrid optimization  
methodology. The dataset comprises both input machining parameters and corresponding output response  
variables, designed to capture the complex and nonlinear relationship between process conditions and  
delamination behavior. The input features include spindle speed, feed rate, drill tool geometry, and point angle,  
each selected at multiple levels based on prior literature and preliminary experimentation. These parameters  
are organized using a structured experimental design such as Taguchi or full factorial methodology to ensure  
comprehensive coverage of the machining domain and to accurately reflect interaction effects among  
variables. For each experimental run, key output responses such as delamination factor, thrust force, surface  
roughness, and hole quality characteristics are recorded using precision measurement instruments, including a  
dynamometer, surface profilometer, and digital microscope. To improve data reliability and reduce random  
errors, all experiments are repeated a minimum of three times, and average values are used in the final dataset.  
Prior to fuzzy modeling and optimization, the collected data undergo preprocessing to remove noise, outliers,  
and inconsistencies, followed by normalization to a standard range to facilitate effective fuzzy inference.  
1. Experimental Results of GFRP Machining: The initial machining experiments were conducted by varying  
spindle speed, feed rate, and point angle. The data recorded included thrust force, delamination factor, and  
surface quality. These results formed the primary dataset for fuzzy modelling.  
Fault Classification Performance of ML Models  
Control Method  
Damping Ratio  
Settling Time  
(Ts in s)  
Peak  
Overshoot  
(Mp %)  
Frequency  
Deviation (Hz)  
Control  
Effort (Uc)  
(ζ)  
Conventional PSS  
PI Controller  
0.12  
0.18  
12.6  
10.3  
8.4  
18.5  
14.2  
10.3  
0.42  
1
0.35  
0.23  
0.8  
0.6  
Fuzzy  
Logic 0.26  
Controller  
GA-Optimized PSS  
LQR Controller  
0.31  
0.34  
6.9  
6.1  
7.8  
6.2  
0.18  
0.15  
0.5  
0.4  
The analysis of TABLE I. highlights significant performance differences between conventional and intelligent  
control systems. Fuzzy and GA-optimized methods demonstrate superior results with lower overshoot, better  
damping ratio, and reduced settling time. Translating this trend to machining, the intelligent system is expected  
to show similarly enhanced control over delamination.  
www.ijltemas.in  
Page 637  
INTERNATIONAL JOURNAL OF LATEST TECHNOLOGY IN ENGINEERING,  
MANAGEMENT & APPLIED SCIENCE (IJLTEMAS)  
ISSN 2278-2540 | DOI: 10.51583/IJLTEMAS | Volume XIV, Issue XII, December 2025  
Multi-Parameter Evaluation of Advanced and Conventional Control Techniques  
Fig. 2. comparing five ML-based control methods—Conventional PSS, PI Controller, Fuzzy Logic Controller,  
GA-Optimized PSS, and LQR Controller—across five performance metrics: damping ratio, settling time, peak  
overshoot, frequency deviation, and control effort. LQR Controller shows the best overall performance with  
highest damping ratio and lowest settling time, overshoot, and control effort.  
2. Fuzzy Model Predictions: The fuzzy inference system developed for this study used linguistic variables  
(Low, Medium, High) to map machining parameters (speed, feed, point angle) to delamination outputs. The  
fuzzy model provided smooth interpolation between experimental data points, predicting delamination with  
high accuracy. The prediction error remained within 3–5%, showing strong reliability. The fuzzy rules  
successfully captured the non-linear relationships between the machining variables, supporting their use as the  
objective function for optimization.  
Experimental vs. Fuzzy Predicted Delamination Values  
Trial No.  
Spindle Speed  
(rpm)  
Feed Rate  
(mm/min)  
Point Angle  
(°)  
Experimenta  
l
Delaminatio  
n (Fd)  
Fuzzy  
Predicted  
Delaminatio  
n (Fd)  
Prediction  
Error (%)  
1
1500  
2000  
2500  
3000  
3500  
60  
80  
50  
70  
40  
90  
1.34  
1.42  
1.27  
1.21  
1.19  
1.31  
1.38  
1.25  
1.18  
1.16  
2.2  
2
3
4
5
118  
90  
2.8  
1.6  
2.4  
2.5  
118  
135  
The fuzzy inference system predicted delamination with high accuracy, with error consistently between 1.6%  
and 2.8%, demonstrating strong reliability of the model shown in above TABLE II.  
www.ijltemas.in  
Page 638  
INTERNATIONAL JOURNAL OF LATEST TECHNOLOGY IN ENGINEERING,  
MANAGEMENT & APPLIED SCIENCE (IJLTEMAS)  
ISSN 2278-2540 | DOI: 10.51583/IJLTEMAS | Volume XIV, Issue XII, December 2025  
Predicted Delamination Values of Various Models  
Fig. 3. compares experimental delamination values with fuzzy predicted delamination values across five  
drilling trials. For each trial, two adjacent bars represent the measured and predicted delamination factors. The  
heights of the bars show that the fuzzy prediction closely matches the experimental results, with only small  
variations across all trials.  
3. Hybrid GA–PSO Optimization Outcomes: The GA–PSO hybrid optimization was applied using the fuzzy  
model's output as the fitness function. The hybrid algorithm efficiently searched for the optimal machining  
parameter combination that minimized delamination. Compared to conventional single-objective optimization  
techniques, the hybrid model converged faster and provided more stable results. The parameters predicted by  
GA–PSO were subsequently validated through machining trials to evaluate improvement in hole quality and  
reduction of exit delamination.  
GA–PSO Optimized Parameters and Delamination Reduction  
Parameter  
Initial Value  
(Baseline)  
Optimized Value  
Improvement (%)  
(GA–PSO)  
Spindle Speed (rpm)  
Feed Rate (mm/min)  
Point Angle (°)  
2000  
3600  
80%  
80  
45  
–43%  
–23%  
–21.8%  
118  
1.42  
90  
Delamination Factor (Fd)  
1.11  
TABLE III. shows that the GA–PSO hybrid algorithm reduced delamination by 21.8%, demonstrating better  
convergence and optimized machining conditions compared to single-objective optimization techniques.  
GA–PSO Optimized Parameters and Resulting Delamination Reduction of Various Models  
www.ijltemas.in  
Page 639  
INTERNATIONAL JOURNAL OF LATEST TECHNOLOGY IN ENGINEERING,  
MANAGEMENT & APPLIED SCIENCE (IJLTEMAS)  
ISSN 2278-2540 | DOI: 10.51583/IJLTEMAS | Volume XIV, Issue XII, December 2025  
Fig. 4. summarizes the effect of GA–PSO optimization on drilling parameters and delamination reduction. It  
compares baseline values with optimized values for spindle speed, feed rate, and point angle, showing  
substantial parameter adjustments. The optimized spindle speed increases from 2000 to 3600 rpm, while feed  
rate and point angle decrease from 80 to 45 mm/min and 118° to 90°, respectively. The delamination factor is  
reduced from 1.42 to 1.11, representing a 21.8% improvement.  
4. Comparative Performance Evaluation: The performance of the proposed hybrid optimization technique  
was compared with traditional parameter selection approaches. The results indicate that fuzzy–GA–PSO  
integration ensures minimum delamination, lower tool vibration, and superior stability. The hybrid technique  
achieved nearly 95% optimization accuracy, demonstrating strong robustness for multi-objective machining  
applications.  
Comparison of Optimization Techniques for Minimizing Delamination  
Optimization Technique  
Accuracy (%)  
Delamination  
Reduction (%)  
Convergence  
Speed  
Stability of  
Results  
Taguchi Method  
78%  
11%  
14%  
Moderate  
Fast  
Moderate  
Grey Relational Analysis 83%  
(GRA)  
Moderate  
Standalone GA  
Standalone PSO  
89%  
87%  
17%  
15%  
22%  
Slow  
High  
Fast  
Moderate  
Very High  
Proposed  
Hybrid  
GA– 95%  
Very Fast  
PSO–Fuzzy  
TABLE IV. shows that the proposed Hybrid GA–PSO–Fuzzy method outperformed all classical approaches,  
achieving highest accuracy (95%), maximum delamination reduction (22%), and fastest convergence.  
Comparison of Optimization Techniques for Minimizing Delamination  
Fig. 5. comparing five optimization techniques—Taguchi Method, GRA, Standalone GA, Standalone PSO,  
and Hybrid GA–PSO–Fuzzy—based on Accuracy (%) and Delamination Reduction (%). The Hybrid GA–  
PSO–Fuzzy technique shows the highest values in both accuracy (95%) and delamination reduction (22%),  
while the Taguchi Method shows the lowest values.  
www.ijltemas.in  
Page 640  
INTERNATIONAL JOURNAL OF LATEST TECHNOLOGY IN ENGINEERING,  
MANAGEMENT & APPLIED SCIENCE (IJLTEMAS)  
ISSN 2278-2540 | DOI: 10.51583/IJLTEMAS | Volume XIV, Issue XII, December 2025  
CONCLUSION  
The experimental investigation and multi-objective optimization of non-conventional machining of Metal  
Matrix Composites (MMCs) demonstrate that machining performance can be significantly enhanced by  
carefully tuning critical process parameters such as pulse current, pulse-on time, and voltage. The comparative  
performance tables and bar graph analyses highlight substantial improvements in both productivity and surface  
quality when hybrid optimization approaches are applied. The analysis of the optimized trials shows that  
Material Removal Rate (MRR) increases consistently across the three experimental trials, confirming that the  
selected parameter ranges effectively support aggressive yet stable machining. Although higher MRR is often  
associated with increased surface roughness, the optimized settings maintain Surface Roughness (Ra) within  
acceptable limits, demonstrating a favorable balance between machining speed and quality. Furthermore, Kerf  
Width values exhibit minimal variation across trials, illustrating stable dimensional accuracy and minimal  
thermal distortion during machining. The collective findings validate that hybrid multi-objective optimization  
techniques—combining statistical modeling and intelligent search algorithms—lead to superior machining  
outcomes compared to conventional parameter selection. The enhanced MRR, reduced surface roughness, and  
consistent kerf width achieved in this study establish the suitability of the proposed methodology for precision  
machining of MMCs.  
REFERENCES  
1. W. Guo, Z. Wang, Y. Chen, Y. Jin and Q. Wu, "Delamination Depth Evaluation in Near-Surface of  
GFRP Laminates Based on RQA-MKLSVM Model," in IEEE Access, vol. 11, pp. 30908-30919, 2023,  
doi: 10.1109/ACCESS.2023.3262133.  
2. K. Biruk-Urban, J. Józwik, P. Bere and M. Parparita, "Influence of Different Technological Parameters  
on Cutting Force in Drilling GFRP Composites," 2023 IEEE 10th International Workshop on  
Metrology  
for  
AeroSpace  
(MetroAeroSpace),  
Milan,  
Italy,  
2023,  
pp.  
275-279,  
doi:  
10.1109/MetroAeroSpace57412.2023.10190054.  
3. K. Biruk-Urban, J. Józwik and P. Bere, "Influence of Technological Parameters on Cutting Force  
Components During Drilling of GFRP Composite," 2022 IEEE 9th International Workshop on  
Metrology  
for  
AeroSpace  
(MetroAeroSpace),  
Pisa,  
Italy,  
2022,  
pp.  
81-86,  
doi:  
10.1109/MetroAeroSpace54187.2022.9856279.  
4. I. Papadimitriou, I. Gialampoukidis, S. Vrochidis and I. Y. Kompatsiaris, "DiMAT Materials Modeler  
(DiMM): An Interactive Framework for Materials Property Prediction and Optimization Using a  
Hybrid Machine Learning - Genetic Algorithm Approach," 2025 IEEE International Conference on  
Engineering, Technology, and Innovation (ICE/ITMC), Valencia, Spain, 2025, pp. 1-12, doi:  
10.1109/ICE/ITMC65658.2025.11106594.  
5. D. Lepadatu et al., "Advanced Nano-concrete Characteristics Materials Prediction Using Taguchi  
Design of Experiment and Artificial Neural Networks," 2025 International Conference on Artificial  
Intelligence, Computer, Data Sciences and Applications (ACDSA), Antalya, Turkiye, 2025, pp. 1-6,  
doi: 10.1109/ACDSA65407.2025.11165876.  
6. K. R and M. S, "DelamPredict-X: An Ensemble Learning-Based Approach for Evaluating  
Delamination in Jute Fiber Reinforced Materials," 2025 International Conference on Frontier  
Technologies  
10.1109/ICFTS62006.2025.11031742.  
7. A Comprehensive Analysis of Security Mechanisms and Threat Characterization in Mobile Ad Hoc  
Networks”, IJLTEMAS, vol. 14, no. 5, pp. 732–737, Jun. 2025, doi:  
10.51583/IJLTEMAS.2025.140500079.  
and  
Solutions  
(ICFTS),  
Chennai,  
India,  
2025,  
pp.  
1-9,  
doi:  
8. K. K. Pandey and S. Singh, "Multi-Attribute Optimization of Hot Machining Using Grey Relational  
Analysis and PCA Technique – A Hybrid Approach," 2025 International Conference on Cognitive  
Computing in Engineering, Communications, Sciences and Biomedical Health Informatics  
(IC3ECSBHI),  
Greater  
Noida,  
India,  
2025,  
pp.  
1304-1309,  
doi:  
10.1109/IC3ECSBHI63591.2025.10991087.  
www.ijltemas.in  
Page 641  
INTERNATIONAL JOURNAL OF LATEST TECHNOLOGY IN ENGINEERING,  
MANAGEMENT & APPLIED SCIENCE (IJLTEMAS)  
ISSN 2278-2540 | DOI: 10.51583/IJLTEMAS | Volume XIV, Issue XII, December 2025  
9. Optimization of Graph Neural Networks for Real-Time Intrusion Detection in Dynamic Mobile Ad-  
Hoc Networks”, Int. J. Environ. Sci., vol. 11, no. 11s, pp. 740–748, Jun. 2025, doi:  
10.64252/79452g17.  
10. D. Chawla, S. Bhatia, Vinay, D. D. Thomas and A. Majeed, "Comparative Performance Evaluation of  
Spur Gears Made from Cast Iron and Epoxy Resin-Based Hybrid Composite Materials Utilizing the  
ANSYS Methodology," 2024 3rd International Conference on Computational Modelling, Simulation  
and  
Optimization  
(ICCMSO),  
Phuket,  
Thailand,  
2024,  
pp.  
336-343,  
doi:  
10.1109/ICCMSO61761.2024.00074.  
11. N. K. Jagannathan, R. Nakkeeran and C. Bharatiraja, "Hybrid Thermal Management System for High-  
Power Electronics: Integrating PCM With Liquid Cooling and AI-Based Control," 2025 IEEE Energy  
Conversion Congress & Exposition Asia (ECCE-Asia), Bengaluru, India, 2025, pp. 1-6, doi:  
10.1109/ECCE-Asia63110.2025.11111902.  
12. T. Sathish, R. Saravanan, S. J. Arunachalam, J. Giri and A. Amer, "Optimization of Interlaminar Shear  
Strength of the Jute/Kenaf/Glass with MWCNT using RSM," 2024 2nd International Conference on  
Emerging Trends in Engineering and Medical Sciences (ICETEMS), Nagpur, India, 2024, pp. 684-689,  
doi: 10.1109/ICETEMS64039.2024.10965047.  
www.ijltemas.in  
Page 642