INTRODUCTION AND MOTIVATION

• The overall rating for safety of the bridges in the United States is C+ and one in nine of the nation’s bridges are rated as structurally deficient (2013 ASCE Report Card).
• Structural health monitoring (SHM) is becoming a viable tool to collect real-time quantitative data from in-service structures concerning condition and performance.
• When carbon nanotubes are properly dispersed into the polymer matrix of a composite they form electrically percolating networks that can act as in situ strain and damage sensors.

STRUCTURAL HEALTH MONITORING

Concept of a Structural Health Monitoring (SHM) System

CARBON NANOTUBE-BASED SENSING

Why Carbon Nanotube (CNT)-Based Sensing?
• Extremely high sensitivity to strain
• Reliable performance (redundancy through network)
• Real-time and distributed sensing capability
• SHM and NDT (non-destructive testing) applications
Mechanism of CNT-based sensing:
• Electrically percolating networks
• Piezoresistive composite sensors
• In situ electrical resistance measurements

SELECTION OF CARRIERS FOR CNT-BASED SENSORS

Nonwoven fabrics, which are plane sheets consisting of randomly distributed and oriented short fibers, were used as the CNT carriers. Main advantages:
• Mechanical and electrical properties of nonwoven fabrics are isotropic.
• Enable repetitive electrical resistance response
Description of some considered fiber architectures:

SHM APPLICATION: SURFACE STRAIN SENSING

TENSILE STRAIN SENSING

COMPRESSIVE STRAIN SENSING

NDT APPLICATION: 2-D DAMAGE EVALUATION BY ELECTRICAL IMPEDENCE TOMOGRAPHY (EIT)

CANDIDATE EIT CURRENT INJECTION / DISTRIBUTION PATTERNS

LABORATORY STUDY

• Initial area of damage evaluation: 3.875 x 3.875 in. plane
• Evaluation sequence: Two-stage post-damage evaluation
• Damage type: manually simulated by removing sensing areas

TWO-STAGE DAMAGE EVALUATION

• Initial measurements of electric potentials between electrodes to determine the initial (undamaged) electrical conductivity distribution in the sensing patch;
• Applying the first damage location of a 0.75 in. dia. hole;
• NDT measurement 1: Injecting the electrical current and collecting the electric potentials (post-damage data set 1);
• Imposing the second damage location of 0.5 in. by 0.5 in. square hole ;
• NDT measurement 2: Collecting the post-damage data set 2;
• Processing data set and reconstructing the EIT mappings.

FINITE ELEMENT MODEL

FE modeling was used first to establish the electrical properties (i.e. conductivity) of the 2-D sensing patch corresponding to its initial state and then to conduct the reconstruction of spatial electrical conductivity for different damage stages.
• Calculating initial conductivity for each discrete element;
• Determining potential distribution at each node corresponding to different current injections;
• Conductivity mapping: seeking the optimal spatial conductivity distribution by minimizing the difference between predicted boundary electrode potentials with actual experimental boundary voltage measures.

EIT CONDUCTIVITY MAPPING BY NEIGHBORING METHOD

CONCLUSIONS

• Real-time sensing capability and excellent strain sensitivity of non-woven CNT-based sensing layer verified and compared to discrete strain sensors;
• Linear resistance response of CNT-based sensing layer correlated to elastic member behavior;
• Initiation of plastic member deformation captured by CNT-based sensing layer;
• CNT-based 2D damage sensing/evaluation tool established and demonstrated.

ACKNOWLEDGEMENTS

This research is funded by the National Science Foundation (NSF), CMMI Grant No.1234830 and the program director is Dr. Kishor Mehta.