DEMONSTRATION OF THE ELECTROCHEMICAL FATIGUE SENSOR SYSTEM AT THE TRANSPORTATION TECHNOLOGY CENTER FACILITY

Transcription

1 DEMONSTRATION OF THE ELECTROCHEMICAL FATIGUE SENSOR SYSTEM AT THE TRANSPORTATION TECHNOLOGY CENTER FACILITY Brent M. Phares, PE, Ph.D. Material Technologies, Inc. 7 NW Boulder Brook Dr. Ankeny, IA 53 bphares@iastate.edu KEYWORDS: Fatigue, crack, nondestructive, NDE, inspection, testing Material Technologies (MATECH) demonstrated its Electrochemical Fatigue Sensor (EFS) system at the Transportation Technology Center (TTC) in Pueblo, CO in May. During testing at TTC, the EFS system was installed on several notable railroad assets to determine their fatigue crack condition. Of principal importance are the installations on various structural components of the Facility for Accelerated Service Testing (FAST) steel bridge, a rail joint bar, and a section of standard rail. At fifteen locations on the FAST bridge the EFS was installed at visually identifiable cracks with installations at two additional locations that exhibited similar fatigue sensitive details. It is interesting that testing of cracks in the FAST steel bridge by TTC using the acoustic emission (AE) and ultrasonic testing (UT) technologies indicated no discernable crack growth. Using the EFS system, MATECH determined that, at five of the fifteen locations tested, cracks were continuing to actively grow including those that AE and UT indicated had no discernable growth. In the case of the rail joint bar, though no growing cracks were found, testing of the rail joint bar illustrated that the EFS system is well suited for inspecting these components for the presence of actively growing cracks. Likewise, testing of the standard rail section with a known crack revealed that EFS is well suited for active fatigue crack detection in rails as well. In early March of 7, MATECH visited the FAST Facility again in order to visually inspect the previously tested cracks and to collect additional data on many of the same cracks. Based on a visual inspection, it was found that all of the cracks that were inspected with EFS in May of 7 and found to not be growing had, indeed, not grown after more than 8, load cycles. Additionally, all of the 5 cracks that were found to be growing in had visually grown. One crack which was indicated by EFS to be growing in May of showed signs of crack growth based on historical pictures of the crack, but upon a re-inspection with EFS in March 7, indicated no current crack growth. This paper summarizes the EFS system including the fundamentals principles that embody its operation. The testing completed at TTC is also discussed with detailed information on the test procedures and results. INTRODUCTION In early May of Material Technologies, Inc. (MATECH), working with the Association of American Railroads (AAR) and the Federal Railroad Association (FRA), demonstrated its Electrochemical Fatigue Sensor (EFS) system at the Transportation Technology Center (TTC) in Pueblo, CO. TTC is a unique, world-class research, development, and demonstration facility that serves to meet the research needs of AAR member railroads via its unique accelerated testing capabilities. During testing at TTC, the EFS system was installed on various components at the Facility for Accelerated Service Testing (FAST) including a steel bridge, a rail joint bar, and a standard section of rail. At fifteen locations on the FAST steel bridge the EFS was installed at visually identifiable cracks. Testing of cracks in the FAST steel bridge by TTC staff using acoustic emission (AE) and ultrasonic testing (UT) technologies indicated no discernable crack growth. Using the EFS system, MATECH determined in May that, at five of the fifteen locations tested, cracks were actively growing, including those that AE and UT indicated had no discernable growth.

2 In early March of 7, MATECH re-visited the FAST Facility in order to visually inspect the cracks tested the previous year and to collect additional data on many of the same cracks for comparison purposes. Based on a visual inspection, it was found that all of the cracks that were inspected with EFS in May of 7 and found not to be growing were, indeed, not growing after more than 8, load cycles. Additionally, all of the 5 cracks that were found to be growing during the last inspection had all visually grown. One crack which was indicated by EFS to be growing in showed some signs of crack growth (based on historical pictures of the crack) but a re-inspection with EFS in 7 indicated no current crack growth. THE EFS SYSTEM The EFS system is a nondestructive fatigue crack inspection system. Specifically, EFS is a tool for determining if a previously unidentified actively growing fatigue crack is present in the inspection area or if known fatigue cracks are actively growing. During an EFS inspection an EFS sensor is applied to each location of interest and crack detection occurs for areas under, or in the immediate vicinity of, the sensor. The EFS system consists of an electrolyte filled sensor, a potentiostat that applies a constant polarizing voltage between the structure and the sensor, and data collection and analysis software. The EFS system works on fundamental electrochemical principles. During testing, the inspection area is anodically polarized to create a protective, passive film on the area of interest. The polarizing voltage produces a DC base current in the electrochemical cell. When the structure being inspected undergoes a cyclic stress, the current flowing within the cell fluctuates in a complex relation to the variation of the mechanical stress. As a result, an AC current is superimposed on the base DC current. Dependent upon the structural material, the loading conditions, as well as the state of the fatigue damage in the structure, the transient current within the cell provides information on the status of the fatigue damage. During cyclic loading, the fatigue cracking process causes micro-plasticity and strain localization on a very fine scale. The interaction of the cyclic slip and the passivating process causes temporary and repeated alterations to the passive films. These alterations, including both dissolution and repassivating processes, give rise to transient currents. The resulting EFS transient currents are complex and result from cyclic changes in the electrical double layer at the interface of the metal and the EFS electrolyte. The transient currents generally possess the same frequency as that of the mechanical stress, but also have a complex phase relationship. In addition, the disruption of the surface oxide film by the cyclic slip causes an additional component of the transient current which has double the frequency of the elastic current. This occurs because plasticity effects occur during both the tensile and compressive portions of the cycle. As fatigue damage develops, the cracks induce localized plasticity at different parts of the fatigue cycle from those in which the background micro-plasticity occurs and in locations where cracks have not yet formed. The crack-induced plasticity, thus, introduces higher harmonic components into the transient EFS current. It is the analysis and calibration of these various current components that allows determination of whether a growing crack is present. The EFS technique offers several distinct advantages over other methods of nondestructive evaluation because of the inherent ability to detect active fatigue crack growth and because of the very small detectable crack size. Differential EFS uses two sensors, one as the reference (R) and one as the crack measurement (CM) sensor. The two sensors are both installed near the location of interest. The CM sensor is specifically located over the area to be inspected while the R sensor is located near the CM sensor but in a location where a crack is not probable. Using signal processing, the two signals are compared to determine if a crack is present. The more complex loadings present, the more advanced the resulting EFS data analysis needs to be to extract and decipher the crack data signatures. Several proprietary techniques have been developed and are contained within the EFS system software. The EFS hardware system consists of three major components: the EFS sensor, the EFS electrolyte, and the EFS Potentiostat data link (PDL). The following sections briefly discuss each of these components.

3 EFS Sensor The basic EFS sensor consists of several parts as shown in Figure. Each sensor has a convenient contact adhesive on one side for attachment to the structure. The open area in the middle of the sensor holds the subsequently described EFS electrolyte. The sensor is filled with electrolyte through the lower filler tube while air escapes out of the upper bleeder tube. The EFS sensor electrode is sandwiched between the upper and lower sensor sections. When the sensor is filled with electrolyte, the electrode is completely covered. Depending on the area to be tested, EFS sensors can be custom-made to fit virtually any three-dimensional geometric requirements (including size, shape, orientation, etc.). Figure Drawing of an EFS sensor. EFS Electrolyte The EFS electrolyte is a proprietary, water-based solution that has been tested on multiple materials including aluminum, titanium, copper, and steel. The EFS electrolyte has been found, in all cases, to be benign to metals. The U.S. Air Force, in fact, has conducted extended fatigue tests with EFS installed on steel, aluminum, and titanium components and has found that EFS did not cause premature failure or impact the fatigue life. The EFS electrolyte is chemically inert and environmentally safe. EFS PDL The battery-powered, wireless EFS PDL, shown in Figure, provides all of the features necessary to collect data in the field. The PDL is compact, lightweight, and provides separately isolated channels for the R sensor and the CM sensor. The PDL features onboard A/D conversion, data collection to a removable MMC card, wireless data streaming, and an easy to use wireless setup for bias, gain, and sample rate. The potentiostat portion of the EFS PDL is an electronic module that controls the voltage difference between a working electrode and a reference electrode. In the case of use during an EFS inspection, both electrodes are contained within the previously mentioned electrochemical cell. The variable controlled by the potentiostat is the cell potential and the measured variable is the cell current. The EFS potentiostat has been custom-designed to not only control the voltage but to also measure the current flow between the working and reference electrodes. During inspection, the working electrode is the structure and the reference electrode is the previously mentioned EFS sensor electrode which is sandwiched within the EFS sensor.

4 Figure The EFS PDL. TESTING DETAILS The initial series of testing at was completed on the FAST steel bridge and the two rail components during the week of May 8 th - th,. The FAST bridge consists of two open-deck, all welded steel deck plate girder spans. The first fatigue cracks were found a few weeks after the installation of the bridge in November 997. Over twenty cracks of various sizes, locations, and configurations currently exist in the bridge. The two rail components tested as part of this work consisted of a rail joint bar and a standard section of rail known to have a crack that was thought to be growing. Testing was conducted during normal operation of the FAST test train. This operation consisted of the operation of a 5-car, loaded unit train operating at approximately mph. At no time did testing impact normal train operation. A re-inspection on the FAST steel bridge occurred March th,7. During this re-inspection, a visual inspection of all fifteen locations inspected in May was conducted. In addition to the visual inspection, seven locations were tested using EFS. Five of the locations were the same as locations tested the previous year. Testing was conducted during normal operation of the FAST test train. Sensor Locations FAST Steel Bridge EFS sensors were located at the locations shown in the plan view of the bridge given in Figure 3. In this figure the inspection locations are designated as MT through MT9 and are so referenced herein. At all locations, the two EFS sensors (i.e., the CM and R sensors) were installed close to one another with the CM sensor located directly over the specific area of interest. Rail Joint Bar One of several rail joint bars installed at FAST was inspected with the EFS system in. In this case, the two EFS sensors were installed on the mid section of the rail joint bar. There were no visibly identifiable cracks at either EFS sensor or at any other location on the rail joint bar. Standard Track Section The EFS system was used to inspect a known, and thought to be actively growing, crack emanating from a hole in a standard rail section in. The hole was a remnant of a previously removed bolted repair. The CM sensor was installed over the crack in order to determine if the crack was actively growing with the R sensor located as close to the CM sensor as physically possible.

5 Sensor Installation In all cases, sensor installation required a clean and bare, but not polished, metal surface. A power wire wheel and grit sandpaper were used to remove the paint and any dirt from the inspection areas. After paint removal, the metal surface was cleaned with acetone to remove any residual paint and debris. The sensors were then installed by simply removing the self-adhesive backing and firmly pressing them in place. Following sensor placement, the sensors were filled with the EFS electrolyte by injecting the electrolyte through the filler tube. Once full, the filler and bleeder tubes were sealed to prevent leakage. Once the sensors were installed, the PDLs were placed (held by magnets) and connected to the sensors. Wireless communication was then registered and the bias, gain, and data acquisition rate was set for each potentiostat. Typically, a bias of. volts, a gain of, and a data acquisition rate of Hz were used for the testing described herein. Figures and 5 show typical sensor and PDL installations. Data Collection and Analysis Data were wirelessly streamed from the potentiostats directly to a laptop computer for data archiving. The EFS data were collected for each of the EFS sensors while the FAST train operating at approximately mph passed over the test locations. Multiple data sets were collected for each of the locations to illustrate system repeatability. RESULTS After data collection, the data were examined and analyzed using the custom EFS system software to determine crack growth activity. The software consists of frequency and time domain based algorithms used to analyze and parse the data. Multiple data sets from each location were examined in order to determine specific results. A visual inspection of all the EFS inspection locations was conducted in 7. The growth of the cracks was determined from historical pictures and the locations of the crack relative to the position of the sensors from. The EFS crack growth indications are in good agreement with the visual inspection completed in 7 (Table ). The specific data with associated photographs of the inspection locations for several locations are given in the following pages. At location MT there is a crack emanating from a stiffener welded to the web of girder near the mid span of the longer span (see Figure 3). Figure shows a photograph of the inspection location, the raw EFS data, and the associated FFT of the time-based data. Crack growth is indicated by the multiple harmonics of the first peak (. Hz) and the overall relative magnitudes of the CM sensor peaks as compared to the R sensor peaks. For example, the CM sensor s nd - 5 th harmonic peaks are all times greater than the R sensor peaks. The combination of the existence of multiple harmonics and the differences in the magnitudes of the R sensor and CM sensor data are sufficient to indicate that this crack is growing. Cracks at locations MT and MT 9 are back-to-back cracks of a through crack growing vertically from a girder bottom flange. The May EFS data for these two locations indicate that both of these cracks were growing. The crack growth was verified with a visual inspection in March 7. The EFS inspection in March 7 also indicates continued crack growth by the large transients in the time signal upon loading and unloading as well as clear harmonics at. Hz, 3. Hz, and.8 Hz (Figure 7.) In May, the EFS inspection of location MT indicated that there was crack growth at this fracture critical location. By examining the photographs from both inspections ( and 7), it can be seen that the two CM sensor s are in different positions. The conclusion is that since the sensors are located just over the visible crack tip and the sensor positions are different for the two inspections, then the crack grew sometime over the past 9 months. However, the EFS inspection in March 7 indicates that at that time the crack was no longer growing. This can be seen in the FFT of the time signal (see Figure 8), where the CM and R data from this location are very similar both in the time domain and frequency domain, indicating that the crack is not growing.

6 TTCI Bridge, Pueblo CO May 8-, MT R NORTH GIRDER MT LONG SPAN 5 - Abutment to Center Pier -9 MT 3 SOUTH GIRDER MT 7 MT 8 MT 5 MT MID SPAN - -9 MT MT MT 9 MT 8 SHORT SPAN 55-5 Abutment to Center Pier NORTH GIRDER -5 MT - MID SPAN SOUTH GIRDER -5 MT MT 3 MT 5 TTCI Bridge, Pueblo CO March, 7 NORTH GIRDER MT a. Test Locations MT 7 LONG SPAN 5 - Abutment to Center Pier -9 SOUTH GIRDER MID SPAN - -9 MT MT MT 9 MT 8 SHORT SPAN 55-5 Abutment to Center Pier NORTH GIRDER -5 MT - MID SPAN SOUTH GIRDER -5 b. 7 Test Locations Figure 3 Top view of bridge and instrumentation layout.

7 Figure. Sensors installed at location MT 8. Figure 5. Potentiostat and sensors installed at location MT.

8 Table Tabulated results from the nineteen inspection locations. Inspection Location Visible Crack Crack Status EFS Measurement Crack Status 7 Visually Verified Crack Status 7 EFS Measurement MT Yes Not growing Not growing Not Growing MT Yes Not growing Not growing Not Inspected MT3 Yes Not growing Not growing Not Inspected MT Yes Not growing Not growing Not Inspected MT5 Yes Not growing Not growing Not Inspected MT Yes Not Inspected Growing Growing MT7 Yes Growing Growing Not Inspected MT8 Yes Not growing Not growing Not Inspected MT Yes Growing Growing Growing MT3 Yes Not Growing Not growing Not Inspected MT Yes Not Growing Not growing Not Inspected MT5 No No growing crack No growing crack Not Inspected MT Yes Growing Growing Not Growing MTR No No Growing Crack No growing crack Not Inspected MT8 Yes Growing Growing Growing MT7 Yes Not Inspected Not Growing Not Growing MT9 Yes Growing Growing Not Inspected Flange Weld Crack Yes Not Inspected Unknown Not Growing Near MT9 Rail Joint Bar No No Growing Crack Not Verified Not Inspected Rail Yes Not Growing Not Verified Not Inspected

9 x -7 Location MT Crack in Stiffener - Girder Near Midspan (Mar 7) EFS Signal (amps) x -9 Location MT Crack in Stiffener - Girder Near Midspan (Mar 7) Figure. Results from location MT.

10 x -7 Location MT Vertical Crack in Web (May ) x -7 Location MT Vertical Crack (March 7) EFS Signal (amps) EFS Signal (amps) x -8 Location MT Vertical Crack in Web (May ) 7 x -8 Location MT Vertical Crack (March 7) Figure 7. Results from location MT.

11 3 x -7 Location MT Girder - Cracked Flange (May ) x -7.5 Location MT Girder - Cracked Flange (March 7) EFS Signal (amps) - EFS Signal (amps) x -8 7 Location MT Girder - Cracked Flange (May ). x -8. Location MT Girder -Cracked Flange (March 7) Figure 8. Results from location MT. The crack at location MT8 is a long crack emanating out of a web stiffener near the vertically growing crack at location MT and MT9. The FFT of the time signal for location MT8 indicates that there is active crack growth. The 7 EFS measurement data indicate that the crack is much more active than the previous year as indicated by the general magnitudes of the harmonics at. Hz, 3., Hz, and.8 Hz. (Figure 9).

12 EFS Signal (amps).5 x -7 MT8 Long Crack Initiating from Below the Web Stiffener (May ) x -8 MT8 Long Crack Initiating from Below the Web Stiffener (May ) 7 EFS Signal (amps) x -7 MT8 Long Crack Initiating from Below the Web Stiffener (March 7) x -8 MT 8 Long Crack Initiating from Below the Web Stiffener (March 7) Figure 9. Results from location MT8. CONCLUSIONS EFS was successfully implemented and used on several existing cracks both on the FAST bridge, on a section of standard rail, and on a rail joint bar. The collected data were sufficient to show that five of the inspected cracks on the FAST steel bridge, some of which were thought to no longer be growing, were growing in. In the case of the rail joint bar no growing cracks were found. However, testing on the rail joint bar illustrated that the EFS technology is well suited for inspecting these components. Testing of the known crack in the standard rail indicated that the crack was, in fact, no longer growing. Follow-up testing in 7 confirmed the conclusions of the testing. The demonstration of EFS at FAST clearly showed that the EFS system is capable of determining if actively growing cracks are present in a variety of railroad assets. By their repetitive and essentially regular nature, railroad fatigue loadings are well suited to using the EFS inspection procedure. Further, the fact that the EFS sensors are easily customized makes the system easy to use on assets with a variety of geometries, loadings, and configurations. Implementation of the EFS system can result in improved safety and operational efficiency. Additionally, as has been shown with previous

13 implementations on highway bridges, using the information resulting from an EFS inspection can yield significant cost savings from knowing which problems need to be fixed and when they need to be fixed.