Machinability and ANN based prediction of surface roughness for TiAlN and PCD coated end mill cutters on AA6061 hybrid composite
At a Glance
Section titled âAt a Glanceâ| Metadata | Details |
|---|---|
| Publication Date | 2025-08-26 |
| Journal | Scientific Reports |
| Authors | P. Haja Syeddu Masooth, V. Jayakumar, M. Kamatchi Hariharan, M. Satthiyaraju, M. Sathish Kumar |
| Institutions | SRM Institute of Science and Technology, Amrita Vishwa Vidyapeetham |
| Analysis | Full AI Review Included |
Executive Summary
Section titled âExecutive Summaryâ- Novel Material Machinability: The study successfully evaluated the machinability of a novel AA6061 hybrid composite (90% AA6061, 5% C, 5% ZrO2) fabricated via stir casting, focusing on surface roughness (Ra) and chip morphology.
- PCD Superiority: Polycrystalline Diamond (PCD) coated end mill cutters delivered the best surface finish (Ra ranging from 0.11 to 0.4 ”m) and exhibited the most stable cutting behavior, attributed to superior wear resistance and thermal conductivity.
- Tool Comparison: Uncoated carbide tools showed significant Ra fluctuations (0.7 to 7.8 ”m) and inconsistent performance, while TiAlN coated tools provided moderate, consistent Ra values (0.3 to 0.55 ”m).
- Chip Morphology: PCD tools produced the highest and most stable chip width-to-thickness (w/t) ratios (up to 39.11), indicating effective chip evacuation and minimal chip deformation compared to TiAlN and uncoated tools.
- ANN Predictive Accuracy: A machine learning-based Artificial Neural Network (ANN) model accurately predicted surface roughness for the coated tools, achieving high correlation coefficients (R2 up to 0.9838).
- Optimization Strategy: Taguchi optimization identified tool-specific optimal cutting parameters (Spindle Speed, Depth of Cut, Feed Rate) for minimizing Ra and maximizing w/t ratio, validating the need for tailored strategies for each coating type.
Technical Specifications
Section titled âTechnical Specificationsâ| Parameter | Value | Unit | Context |
|---|---|---|---|
| Workpiece Composition | 90% AA6061, 5% C, 5% ZrO2 | % weight | Hybrid Metal Matrix Composite |
| Fabrication Method | Stir Casting | N/A | Workpiece preparation |
| Tool Type Comparison | Uncoated Carbide (UCC), TiAlN, PCD | N/A | 8 mm diameter end mills |
| Coating Thickness | 2 | ”m | TiAlN and PCD coatings (PVD method) |
| Spindle Speed (SS) Range | 3000, 4000, 5000 | rpm | Input variable (3 levels) |
| Depth of Cut (DoC) Range | 0.5, 1, 1.5 | mm | Input variable (3 levels) |
| Feed Rate (FR) Range | 100, 200, 300 | mm/min | Input variable (3 levels) |
| Best Ra Range (PCD Tool) | 0.11 to 0.4 | ”m | Superior surface finish achieved |
| TiAlN Ra Range | 0.3 to 0.55 | ”m | Moderate, consistent performance |
| UCC Ra Range | 0.7 to 7.8 | ”m | Poor performance, high fluctuation |
| ANN R2 (TiAlN Ra Prediction) | 0.9838 | N/A | High accuracy prediction model |
| ANN R2 (PCD Ra Prediction) | 0.9699 | N/A | High accuracy prediction model |
| Optimal UCC Ra Setting | SS: 5000 rpm, DoC: 1.5 mm, FR: 100 mm/min | N/A | Taguchi optimization result |
| Optimal PCD Ra Setting | SS: 3000 rpm, DoC: 0.5 mm, FR: 300 mm/min | N/A | Taguchi optimization result |
Key Methodologies
Section titled âKey Methodologiesâ- Composite Fabrication: The AA6061 / C / ZrO2 hybrid composite was manufactured using the stir casting method, ensuring uniform dispersion of the 5% Graphite and 5% Zirconia reinforcements.
- Tool Coating: Three 8 mm carbide end mills were prepared: Uncoated Carbide (UCC), TiAlN coated, and Polycrystalline Diamond (PCD) coated. Coatings were applied using the Physical Vapor Deposition (PVD) method to a thickness of 2 ”m.
- Experimental Design: The experiment utilized a Taguchi L9 Orthogonal Array (OA) design to efficiently test the three input factors (Spindle Speed, Depth of Cut, Feed Rate) at three distinct levels each.
- Machining Setup: End milling operations were conducted on a CNC milling machine under dry conditions for a length of 150 mm per trial.
- Response Measurement:
- Surface Roughness (Ra) was measured using a Surfcom 1400G machine.
- Chip morphology (width-to-thickness ratio) was measured using an Optical Length Measurement (OLM) vision measuring system.
- Optimization Technique: Taguchi Signal-to-Noise (S/N) ratio analysis was applied to identify the optimal parameter combinations for minimizing Ra (smaller-is-better) and maximizing w/t ratio (larger-is-better).
- Predictive Modeling: An Artificial Neural Network (ANN) model was developed using a multilayer feed-forward architecture with Rectified Linear Unit (ReLU) activation functions. The model was trained using a back-propagation algorithm and Mean Squared Error (MSE) loss function.
- Model Validation: Model performance was assessed using the Regression coefficient (R2) and Mean Absolute Percentage Error (MAPE). Further statistical validation was performed using Analysis of Variance (ANOVA) and second-order polynomial regression equations (R2 > 0.95).
Commercial Applications
Section titled âCommercial Applicationsâ- High-Performance Automotive Components: Ideal for manufacturing parts requiring high wear resistance, thermal stability, and low friction, such as advanced brake rotors, piston crowns, and cylinder liners.
- Aerospace and Defense Structures: Suitable for lightweight, high-strength applications like structural panels and defense armor plating, where durability and high strength-to-weight ratio are paramount.
- Precision Machining and Tooling: The ANN model and optimized parameters provide a robust framework for process control and quality prediction in advanced manufacturing environments, particularly when machining hard Metal Matrix Composites (MMCs).
- Wear-Resistant Industrial Machinery: Applicable in systems requiring components with exceptional surface hardness and resistance to abrasive wear, leveraging the benefits of ZrO2 reinforcement and PCD coating.
- Robotics and Automation: Used in the production of precision-engineered parts for industrial automation systems where high surface quality and geometric tolerances are critical.