Electromagnetic Acoustic Damage Imaging Using Compressive Sensing and Total Focusing Method

Research output: Contribution to journalArticlepeer-review

Abstract

The total focusing method (TFM) is widely used in ultrasonic imaging [by means of piezoelectric transducers (PZTs)] due to its inherent nature of high precision. However, for the electromagnetic acoustic transducer (EMAT), the imaging precision and efficiency are mainly limited by its large dimension and the substantial amount of data generated by full matrix capture (FMC) signals in TFM. To solve this problem, a compressive sensing (CS)-virtual element signal reproducing (VESR)-TFM method is proposed by innovatively combining the CS and VESR techniques for high-resolution array imaging of EMATs. This method improves the image quality by regenerating virtual element signals from adjacent real element signals. Meanwhile, the amount of data stored and transmitted by FMC is reduced by optimizing the CS parameters. Based on the lab-built EMAT system, experiments were carried out on the specimens with hole and crack damages only with the eight EMATs. Experimental results showed that the location error of hole damage is 6.08 mm, while the location error of crack damage is 4.12 mm and the angle error of crack damage is 1.04°. The CS-VESR-TFM method not only reduces the amount of data acquisition but also improves the imaging quality. It provides an effective way to reduce the complexity of hardware acquisition system and realize high-precision imaging based on sparse array EMATs.

Original languageEnglish
Article number9522309
JournalIEEE Transactions on Instrumentation and Measurement
Volume74
DOIs
StatePublished - 2025

Keywords

  • Compressive sensing (CS)
  • electromagnetic acoustic transducer (EMAT)
  • full matrix capture (FMC)
  • signal reproduction
  • total focusing method (TFM)

Fingerprint

Dive into the research topics of 'Electromagnetic Acoustic Damage Imaging Using Compressive Sensing and Total Focusing Method'. Together they form a unique fingerprint.

Cite this