Noncontact ultrasonic materials identification based on improved frequency responses.

Area of Science:

• Materials Science
• Acoustics
• Signal Processing

Background:

• Accurate and cost-effective materials identification is crucial across industries.
• Existing noncontact methods often face limitations in accuracy or complexity.
• Developing novel techniques for rapid and reliable material characterization is an ongoing challenge.

Purpose of the Study:

• To propose a novel noncontact ultrasonic materials identification method.
• To enhance identification accuracy using advanced signal processing techniques.
• To establish a low-cost and broadly applicable solution for object identification.

Main Methods:

• Utilizing off-the-shelf ultrasonic probes for noncontact signal acquisition.
• Applying frequency domain filtering (bandpass, notch) and proposing improved frequency responses (IFR) for initial signal processing.
• Employing empirical mode decomposition (EMD) for time domain filtering, extracting information entropy from intrinsic mode functions (IMFs) as features.
• Developing material-specific characteristic templates by fitting probability density functions to information entropy values from repeated tests.

Main Results:

• Achieved 100% identification accuracy for four identical-sized aluminum alloy plates.
• Demonstrated the effectiveness of the combined frequency and time domain filtering approach.
• Validated the proposed method's feasibility for accurate materials identification.

Conclusions:

• The proposed noncontact ultrasonic method offers high identification accuracy with simple, low-cost equipment.
• The technique, integrating IFR signals and information entropy derived from EMD, provides a robust approach to material characterization.
• This method has significant potential for diverse object identification applications due to its efficiency and cost-effectiveness.