A Novel Inspection Technique for Electronic Components Using Thermography (NITECT).

Area of Science:

• Materials Science and Engineering
• Electrical Engineering
• Non-Destructive Testing

Background:

• Unverified and counterfeit electronic components present a significant global risk, potentially causing critical system failures and compromising high-hazard assets.
• Existing inspection methods for electronic components are often costly or time-consuming, hindering large-scale screening efforts.

Purpose of the Study:

• To propose and validate a pulsed thermography-based screening technique, augmented by digital twin methodology, as a complementary solution for rapid inspection of electronic components.
• To address the limitations of current inspection techniques by offering a faster and potentially more cost-effective alternative for detecting unverified components.

Main Methods:

• Development of a Finite Element Method (FEM)-based simulation unit to model internal component structures and thermal behavior under heat exposure, informed by X-ray data.
• Integration of a physical inspection unit to validate and refine the digital twin simulation, incorporating real-world uncertainties from equipment and samples.
• Application of Principle Component Analysis (PCA) for feature extraction, followed by machine learning classifiers for quantitative classification of component authenticity.

Main Results:

• The developed digital twin methodology accurately simulates the thermal response of electronic components with varying physical properties.
• The integrated physical inspection unit successfully verified and improved the simulation's fidelity by accounting for experimental uncertainties.
• Evaluation on 17 chips from diverse sources demonstrated the technique's effectiveness in distinguishing between verified and unverified components.

Conclusions:

• The proposed pulsed thermography technique, combined with digital twin modeling, offers a promising and effective solution for the rapid screening of unverified electronic components.
• This approach can significantly enhance the reliability and safety of electronic systems by providing a scalable method for counterfeit detection.