Opportunities and Challenges for Predicting the Service Status of SLM Metal Parts Under Big Data and Artificial Intelligence

Abstract

• Selective Laser Melting (SLM) Technology: Crucial for aerospace, medical equipment, automotive, defense industries.
• Issues: Uncontrollable service states, nonlinear and complex operating conditions.
• Approach: Combine ultrasonic nonlinear responses and big data/AI to evaluate SLM parts' service status.

Introduction

• SLM: Promising additive manufacturing technology for dense, complex parts.
• Challenges: Microscopic defects (inclusions, pores, cracks) in SLM parts.
• Importance: Early damage detection, accurate service status assessment for safety and economy.

Research Progress and Development Trend

Nonlinear Ultrasound Theory

• Classical Nonlinearity and Contact Acoustic Nonlinearity: Classical models exist but lack complete lifecycle characterization.
• Numerical Simulation: Helps clarify correlation between damage evolution and ultrasonic response.

Data-Driven Approaches

Machine Learning Approaches

• Strengths: Deep data extraction, complex structural data fitting.
• Weaknesses: Lack of physical meaning, hyper-parameter selection challenges, deterministic predictions.

Statistical Data-Driven Approaches

• Strengths: Quantify prediction uncertainty, probabilistic models.
• Weaknesses: Limited big data processing, multivariate complexity.

Combination Approaches

• Objective: Integrate strengths, compensate weaknesses. Combine machine learning and statistical models.

Mechanism Model Combined with Data

• Approach: Use big data to improve mechanism models, but challenges in model accuracy due to complexity.

Problem Analysis and Solution Exploration

• SLM Process Complexity: Leads to micro defects and service state evolution.
• Chimeric Model: Combine classical theories with big data insights for better service status assessment.
• Challenges: Integration of failure mechanisms, data mining, model parameter optimization.

Summary and Prospect

• Ultrasonic Testing: Key in revealing mesoscopic damage and service state evolution.
• Chimeric Model Approach: Scientific method to ensure SLM parts' safety under complex conditions.
• Future Research: Focus on improving model accuracy, enhancing interpretability, real-time data application.

Conclusion

• Scientific and Operable: Proposal promises to solve uncontrollable service state issues.
• Integration of AI and manufacturing: Essential for advancing intelligent manufacturing and strategic industries.

References: Cited works summarize recent advancements and contributions in SLM technology and related fields.

Funding and Conflicts of Interest: Noted funding source and declaration of no conflicts of interest by the authors.

This summary outlines the critical advancements and challenges in predicting the service status of SLM metal parts, emphasizing the integration of big data, AI, and ultrasonic technologies.