A novel predictive model for abrasive waterjet deep hole drilling on AL7075 T6 using machine learning and...

Area of Science:

• Materials Science and Engineering
• Manufacturing Processes
• Computational Intelligence

Background:

• Abrasive Waterjet (AWJ) machining is a key non-traditional method for precision cutting of advanced aerospace alloys like AL7075 T6.
• Deep Hole Drilling (DHD) using AWJ presents challenges in achieving high geometrical precision and surface finish.
• Optimizing AWJ-DHD parameters is crucial for enhancing efficiency and quality in aerospace component manufacturing.

Purpose of the Study:

• To investigate the influence of process parameters on AL7075 T6 AWJ-based Deep Hole Drilling (AWJ-DHD) quality.
• To develop and validate machine learning models for predicting AWJ-DHD performance.
• To optimize AWJ-DHD settings using evolutionary algorithms for improved drilling efficiency and precision.

Main Methods:

• Full factorial design was employed for experimental data collection.
• Four machine learning models (Adaptive Boosted Regression, Extreme Gradient Boosting, Decision Tree, Random Forest) were developed for prediction.
• Three evolutionary algorithms (Moth-Flame Optimization, Differential Evolution, Sine Cosine Algorithm) were utilized for multi-response optimization.

Main Results:

• The Random Forest model demonstrated the lowest testing error across all responses.
• The Sine Cosine Algorithm (SCA) outperformed other optimization algorithms, identifying optimal parameters: 350 MPa water pressure, 1.5 mm standoff distance, and 300 g/min abrasive mass flow rate.
• Validation trials confirmed the model's accuracy, showing minimal percentage variations in predicted versus actual responses (kerf angle, kerf ratio, surface roughness, drilling rate).

Conclusions:

• The integrated framework of machine learning and evolutionary algorithms effectively models and optimizes AWJ-DHD for AL7075 T6.
• The optimized process significantly enhances drilling efficiency and hole quality, particularly by minimizing surface roughness.
• This approach provides a robust method for achieving high-precision deep hole drilling in aerospace applications.