Advanced Corrosion Detection Combining Chemical Odor and Magnetic Flux Measurements

Project Information
Link to Latest Report: Coming Soon

Background:
It has been reported that of the over 600,000 bridges in the United States, 7.5% are structurally deficient. This is often a result of some form of corrosion whether external or internal. Common advances to monitor corrosion include magnetic flux and acoustic emission measurements. However, chemical reactions that occur during metal corrosion may generate changes in volatile organic compounds (VOCs) above and around the corroded material. Over the past 25 years the Furton group has studied the VOCs detected by canines and developed training aids for applications ranging from agricultural applications, medical applications (including COVID-19) and forensic applications including mass storage devices and we are currently studying fentanyl odors and canines ability to detect fentanyl. Although there are no reports of applying odor analysis and canine detection for the application proposed a preliminary study on dogs ability to detect corrosion under insulation in oil and gas processing plants was conducted by Schoon et al. suggesting that the application merits further attention.

Objective:
This proposal combines the non-destructive test Magnetic Flux Leakage Method (MFL) with odor analysis by SPME-GC/MS and canine olfaction. The combined methods will be optimized utilizing AI/ML to make better decisions and accurately locate the source of significant corrosion.) with odor analysis by SPME-GC/MS and canine olfaction. The combined methods will be optimized utilizing AI/ML to make better decisions and accurately locate the source of significant corrosion.

Scope:

  • Odor of areas above and within corroded test specimens will be collected on sorbent materials utilizing a miniature VacPen trace chemical collection device with the same areas being analyzed by MFL.
  • The VOC odors will be collected onto a sorption disk within the VacPen including utilizing an optimized sorption material that will then be analyzed via headspace solid phase microextraction (SPME) coupled with gas chromatography mass spectrometry (GC/MS) or direct desorption into a portable GC/MS.
  • Canines will be trained to detect odor on the sorbents and differentiate the odor of corroded and non-corroded test specimens and at specific corrosion levels. Initially one canine will be trained to demonstrate the proof of concept and ideal training method followed by 4 additional canines in order to have statistically significant publishable data.
  • Canines’ accuracy and specificity for detecting corrosion from structures will be assessed and compared to magnetic flux leakage measurements and the combined measurements will be optimized using AI/ML. This task will be accomplished by placement of steel strands with known levels of corrosion inside empty ducts of the full scale segmental concrete bridge that is located at FIU engineering Center, shown below. Various tests will be carried out to develop data and achieve the project objectives.

Research Team:
Principal Investigator: Kenneth Furton
Co-Principal Investigator: Atorod Azizinamini
Research Assistant:  TBD