U.S. Department of Transportation
Federal Highway Administration
1200 New Jersey Avenue, SE
Washington, DC 20590

Skip to content
Facebook iconYouTube iconTwitter iconFlickr iconLinkedInInstagram

Federal Highway Administration Research and Technology
Coordinating, Developing, and Delivering Highway Transportation Innovations

Public Roads
This magazine is an archived publication and may contain dated technical, contact, and link information.
Public Roads Home | Past Issues | Subscriptions | Article Reprints | Guidelines for Authors: Public Roads Magazine | Sign Up for E-Version of Public Roads | Search Public Roads
| Current Issue |
Back to Publication List        
Publication Number:  FHWA-HRT-21-001    Date:  Autumn 2020
Publication Number: FHWA-HRT-21-001
Issue No: Vol. 84 No. 3
Date: Autumn 2020


Detecting Bridge Corrosion with a Robotic Magnetic-based NDE System

by Hoda Azari and Sadegh Shams

FHWA is researching a promising nondestructive evaluation method to uncover corrosion in the Nation's prestressed concrete bridge girders

Two men in hardhats inspect the collapsed girder of a Pennsylvania overpass.
A destroyed overpass and the highway below.
In 2005, a catastrophic failure of a prestressed concrete girder caused a bridge to collapse onto I‐70 in Pennsylvania. Nondestructive evaluation (NDE) techniques such as magnetic flux leakage detection can help prevent events like this.

Concrete deterioration and steel corrosion are major concerns to bridge owners and engineers. The corrosion of steel reinforcement in concrete construction impairs the durability and longevity of prestressed concrete girders. Although steel reinforcing compensates for the weakness in tensile strength of the concrete, it is the leading cause of concrete deterioration.

Prestressed concrete is less permeable and has a higher alkalinity than normal concrete. However, steel reinforcement can be corroded in the case of poorly detailed or constructed systems, or when the environment is more severe than expected. The corrosion of the steel strands decreases the ultimate strength and ductility of strand and leads to fracture, and may cause premature failure of concrete structures. Concrete cracks form over prestressed steel strands, permitting water and de-icing chemicals to penetrate to the steel and accelerate corrosion and create delamination and spalling.

There are more than 617,000 bridges in the United States, most of which are constructed of steel, conventional reinforced concrete, or prestressed concrete.

In 2002, the Federal Highway Administration, in partnership with NACE International released a benchmark study, Corrosion Costs and Preventive Strategies in the United States (FHWA-RD-01-156), on costs associated with metallic corrosion in a wide range of industries. The report estimated the annual direct costs for replacement and maintenance of bridges in poor condition to be $8.3 billion, but the indirect costs of corrosion incurred by users and owners increase exponentially. The report estimated the indirect costs to the user, such as traffic delays and lost productivity, to be as high as 10 times that of direct corrosion costs. In 2013, NACE International estimated the annual direct cost of corrosion for highway bridges to be $13.6 billion.

Catastrophic collapses have led to a reevaluation of the condition of many bridge structures, which results in bridges being posted for weight restrictions. For instance, on December 27, 2005, the SR1014 Lake View Drive Bridge in Washington, PA, collapsed onto Interstate 70 when one of the prestressed concrete girders failed from dead load. The forensic evaluation of the bridge revealed heavy spalling and corrosion of the strands on the bottom flange of the failed box beam member. Therefore, detecting corrosion in bridges is critical to effective maintenance and repair.

"As such, some NDE data are becoming essential for more effective and economical management of bridges, and concrete bridge decks in particular," says Dr. Joey Hartmann, director of FHWA's Office of Bridges and Structures. "To make more informed decisions addressing safety, reliability, and maintenance of bridge structures, owners have increasingly turned to NDE over the past 15 years to support bridge inspections."

Magnetic-Based NDE Implementation and Technology Plan

Corrosion is a progressive process, and, if left unrepaired, leads to steel section loss or fracture resulting in lowered member capacity. Timely inspections of bridge structures—coupled with appropriate interventions and mitigation strategies—significantly reduce the dollar impact of corrosion by documenting the corrosion and allowing for rehabilitation of the structures to increase the life of these valuable assets. By developing new and improved evaluation techniques and testing devices, bridge engineers can more accurately determine inspection intervals, prioritize the most effective strategies for corrosion prevention and mitigation, and, ultimately, increase the service life of the Nation's bridges.

A wide view of a partially collapsed overpass.
The collapsed girder of a Pennsylvania overpass.
In 2005, a catastrophic failure of a prestressed concrete girder caused a bridge to collapse onto I‐70 in Pennsylvania. Nondestructive evaluation (NDE) techniques such as magnetic flux leakage detection can help prevent events like this.

Since the late 1990s, FHWA has conducted several projects on the NDE of steel corrosion embedded in concrete bridges. The magnetic-based methods have been investigated through various projects seeking more efficient and cost-effective corrosion inspections of prestressed concrete girders. Studies have demonstrated that magnetic flux leakage (MFL) systems are an effective NDE means used in the transportation and energy industries to detect corrosion such as loss of general uniform thickness, stress corrosion cracking, and concentrated pitting.

MFL inspection relies on the fact that the gradient of magnetic scalar potential in a magnetized metallic object rises in approaching any loss or added material. The amplitude of the MFL is generally proportional to the magnetization level: the gradient can be significant when the ferromagnetic materials are locally magnetized to full or near saturation. For the purposes of examining prestressed concrete girders, the magnetizer travels the length of the prestressing strands to detect any flux leakage caused by section loss or gain. In other words, the MFL system, in the same manner with leakage from section loss at corrosion, identifies lateral reinforcements (stirrups) from added section at strand-stirrup interfaces.

A close up of the circuit board for the magnetic flux leakage system.
The magnetic flux leakage device includes 64 sensors placed on its circuit board.

The magnetic sources can be either permanent or electromagnets. However, permanent magnets have more versatile applications in the field because of their simpler hardware requirements and speedy operation. Sensors are located between two permanent magnet blocks and can pass near the defects to measure the resulting flux leakage field. Sensors can be mounted axially and normally with respect to the direction of the magnetic field, producing a voltage output that changes proportionally with the magnetic field. As concrete is essentially a nonmagnetic material with a relative magnetic permeability of unity, it has negligible influence on the magnetic measurements.

FHWA's NDE Laboratory developed a more effective and field worthy magnetic-based NDE system in 2019 for the detection of steel corrosion in AASHTO-type prestressed concrete girders. The new MFL system is designed by taking computational and analytical measurements using the finite element multiphysics method to size and lay out magnets and determine rare-earth metal strength, select the magnetic field detection sensor, and predict test outputs under corrosion sizes and different concrete steel placements. Sixty-four such sensors are laid out on eight printed circuit boards in 32 pairs along a straight line with a 0.25-inch (0.64-centimeter) spacing. Each pair of sensors represents one channel of data for the normal magnetic field and one channel for the axial field.

Virginia's MFL Research

Corrosion of reinforcement in concrete bridges is a major concern for the Virginia Department of Transportation (VDOT). This is particularly the case for prestressed concrete because of the sensitivity of corrosion-induced cross-section losses on the strands under large tensile stresses.

"VDOT had found severely corroded prestressed strands in some bridge beams during inspections," says Soundar Balakumaran, the associate director at the Virginia Transportation Research Council (VTRC), which is part of VDOT. "This motivated VDOT and VTRC to explore technologies that are sensitive to cross-section losses in steel strands before the corrosion propagates to an advanced stage."

The research conducted at VTRC showed MFL as the technology with the most potential. VDOT built a rough prototype MFL device for laboratory testing and found it to be successful in identifying small cross-section losses in steel strands. Since a commercial device with this technology is not available at this time, VDOT has been looking to collaborate on the development of a production-level practical device for bridge inspection.

"FHWA's extensive testing reaffirmed the effectiveness of MFL technology for VDOT," says Balakumaran.

The Robotic MFL

In the latest version developed by FHWA's NDE Laboratory, the MFL system is integrated into a robotic rover that travels the girder length using eight articulating arms engaged with the bottom flange on an AASHTO-type prestressed concrete girder. The robotic rover crawls along the length of the girders' bottom flange while an X-Y positioning carriage supports an NDE device. The rover is capable of navigating diaphragms such that four arm pairs move around obstructions sequentially. The new rover follows a modular scheme that allows different evaluation devices (payloads), other than the MFL, to be used for testing and evaluation of the AASHTO-type prestressed concrete girders.

A close up of the robotic rover installed on an onsite bridge girder.
During testing, FHWA researchers installed the robotic rover on an AASHTO-type IV concrete girder at FHWA's Turner-Fairbank Highway Research Center.

The newly designed MFL system consists of two permanent magnets to magnetize embedded strands and multiple sensors to detect normal and axial MFL. The magnet and sensor assembly moves linearly along the length of the girder. The direction of the applied magnetic field is aligned with the longitudinal strands embedded in the concrete girder. When the assembly reaches either the interface with stirrups or a corroded section of the bottom layer of strands, the magnetic field will be distorted.

The primary components of the robotic rover subsystem include a rigid structural frame, articulating arms with drive mechanism, proximity sensors, NDE device carriage, two encoder wheels, control modules, a power source, and data communication devices. Communication between the main computer on the device and the different onboard microprocessors and the data acquisition system occurs via two independent wireless gateways, one that is dedicated to the robotic rover and one to the MFL detection subsystem.

FHWA researchers validated the robotized MFL system in the laboratory on a mockup girder specimen. The team investigated the effectiveness of the MFL method under several impacting parameters using finite element multiphysics simulations. They found that MFL signals can be influenced by the distance between the magnetizer/sensor array and the object of interest. However, operators should be aware that signals from nearby secondary ferromagnetic objects such as stirrups, tie wires, and steel chairs may interfere with MFL signals.

The magnetic flux leakage system shown in the laboratory.
The MFL system comprises two permanent magnets, sensor arrays, data acquisition and communication, and a power source.

Researchers verified the real-world performance of the system by examining AASHTO-type prestressed concrete girders salvaged from an in-service bridge in Maryland. The girders were in service for almost 50 years before being salvaged from the bridge to be used for full-scale research at the Turner-Fairbank Highway Research Center.

The research team at FHWA's NDE Laboratory installed the robotic rover on the girder and established wireless communications with the integrated MFL system and ground controller. The developed software on the ground computer successfully controlled and monitored the scanning process. Using real-time data visualizations and further processing of recorded data revealed that the MFL system was able to disclose the location and extent of corrosion of prestressing strands and mapping stirrups.

Research Findings

The research done in FHWA's NDE Laboratory in collaboration with the University of Wisconsin‐Milwaukee has led to the development of a state-of-the-art magnetic-based corrosion detection system for use in prestressed concrete bridge girders. The system has benefitted from incorporating the latest technology, design, fabrication, and software engineering techniques. FHWA designed the MFL detection system using finite element multiphysics simulations to arrive at the optimum performance. The unique electronic circuits implemented in the data detection system enable the system to achieve extremely large dynamic range and excellent resolution, providing detection of corrosion-related cross section losses at less than 2 percent.

A computer screen shows graphs and data for a testing run of the magnetic flux leakage device.
The MFL software provides data visualization and analysis.

The robotic rover provides structural support and a platform for the payload so that it can be moved along the length of a test girder for magnetic scanning. The rover is a sophisticated computer-controlled robotic device designed and fabricated to navigate obstacles as it moves, making the system a more effective and field-worthy NDE tool. System software provides data visualization and analysis, including 3-D imaging of the measured magnetic field in two different orientations.

"FHWA's team subjected the MFL system to extensive laboratory testing and evaluation after its fabrication," says Dr. Jean Nehme, the team leader for FHWA's Long-Term Infrastructure Performance Team. "The testing and evaluation led to optimizing the system performance by enhancement of the components and the entire system."

Hoda Azari is the manager of the NDE Research Program and NDE Laboratory at FHWA's Turner-Fairbank Highway Research Center. She holds a Ph.D. in civil engineering from the University of Texas at El Paso.

Sadegh Shams is a contracted research engineer working in the NDE Laboratory at FHWA's Turner-Fairbank Highway Research Center. He holds a Ph.D. in civil engineering from the University of Wisconsin-Milwaukee.