• Federal Highway Administration >
• Publications >
• Research Publications >
• 09044 >
• Longdescr.Cfm
• Integrity of Infrastructure Materials and Structures

Publication Number: FHWA-HRT-09-044
Date: October 2009

Back | Table of Contents

Integrity of Infrastructure Materials and Structures—508 Captions

Figure 1. Photo. 2304 SS bar after bending. The photo shows a 0.5-m length of #5 (16-mm-diameter) 2304 SS bar bent into approximately a 30-degree angle. The bar is flat and shown on a horizontal brown surface, and the photo is taken from above. The bar is a shiny metallic material with a lugged surface typical of reinforcing bars for concrete. This is the type of sample used for the stress corrosion cracking test.

Figure 2. Illustration. A bent specimen in the restrained position. The diagram shows a specimen forming a U-shape in the restrained position with the open end of the specimen facing up and the closed end facing down. The location of the maximum tensile stress is shown at the bottom outside curve of the U, and it is marked in blue. The location of the C-clamp vice is shown 20 cm from the maximum tensile stress location on the left leg of the specimen (marked in green). The location of the strain gauge is on the left leg on the outside surface of the specimen 15 cm from the maximum tensile stress location (marked in red).

Figure 3. Photo. Test tank with cover. The test tank is semitransparent white plastic and shaped like a rectangular box with a flat cover made out of the same material. The two ends of the specimen are protruding from the top of the cover. The left end appears shiny while the right end appears blue due to the blue tape. Lead wires from strain gauges appear in the foreground and run downward from the top edges of the tank on the right side. The test tank is approximately 610 mm long, 460 mm wide, and 610 mm deep. The lower quarter of the tank appears darker due to salt water filling that part of the tank.

Figure 4. Photo. Top view of two specimens with C-clamps in the test tank. The cover of the test tank is removed, and the photo is taken from an overhead position looking down into the tank. Clamps have channels welded to the clamping surfaces that secure the bar during compression and are shown holding the specimens in a compressed U-shape. The dark gray clamps are positioned above the liquid level and resting on support plates. Lead wires from the strain gauges on the specimens are red and white, and they are shown routed to the right side of the tank.

Figure 5. Photo. High temperature experimental arrangement. The photo is taken from the overhead position of an insulated bucket with a C-clamp. The insulated bucket is white plastic and sits within a larger white plastic bucket, and the C-clamp, which is dark gray, is on top of the bucket. The bucket installation appears around the perimeter of the inner bucket and is shiny metallic aluminized plastic sheet. Blocks of wood rest on the upper edge of the outer bucket and provide a platform for the C-clamp. The C-clamped U-shaped specimen is shown with the bent portion in the insulated bucket. Lead wires for heating and temperature control exit the insulated bucket. The white bucket lid is split and notched to allow reassembly after placement of the specimen to reduce evaporation.

Figure 6. Photo. Straight as-received 2304 SS bar with epoxy-mounted ends and an electrical lead. The photo, taken from above, shows a straight bar with a white label illustrating the bar type, 2304 SS. The length of the bar is 152 mm, and the size is #5 (16-mm diameter). The bar is metallic blue, and each end has an amber-colored epoxy cap cast on it. The right end of the bar shows a blue wire electrical lead that was attached directly to the bar end and sealed with the epoxy cap.

Figure 7. Illustration. Accelerated experimental arrangement. The diagram shows a potentiostat on the right, shown as a gray box, which is used to control the potential of eight specimens. The potentiostat is connected in parallel to specimens by wires, which are immersed in a tank with a reference electrode and a counter electrode. Connected between each specimen and the potentiostat is a resistor used to measure current. The reference electrode and counter electrode are also connected to the potentiostat. The immersion tank is shown as a blue rectangle. The immersion solution is synthetic pore solution (0.30N KOH plus 0.05N NaOH) to which chlorides is incrementally added. A data acquisition system, shown above the tank as a gray box, is used to record the individual currents of the specimens by measuring the potential across each resistor. Electrical connections are made at each end of the resistor to the data acquisition system.

Figure 8. Photo. Test system. A photo of the test system shows the potentiostat, the data acquisition system, and the test chamber—which is 0.6 m long, 0.6 m wide, and 0.3 m high. The potentiostat is a black box on the right side of the photo and is identified with a label and arrow. The data acquisition system is a black box resting inside a small white box with blue wires connected to the system. It is labeled and identified with an arrow and is located to the left of the potentiostat. The test chamber is white plastic with a white plastic lid and is identified with a label and arrow on the left side of the photo. The lid covers the test chamber, and the blue wire leads are shown exiting the chamber and connecting to the data acquisition system.

Figure 9. Illustration. Simulated deck slab specimen design. The diagram shows two images side by side. The image on the left is the side view of the simulated deck slab specimen design. It shows 15-cm-deep concrete as a patchy gray area with six bars as white circles placed in two rows of three bars each. The white box at the top of the slab represents a pond of sodium chloride contained in a white plastic trough. The upper row has 2.5 cm of concrete cover, and the lower row is located 2.5 cm from the bottom of the specimen. The top view, which is the second image on the right, shows the 30–Cm by 30–Cm slab specimen with three vertical bars shown in solid light gray color protruding from opposite ends of the specimen. In this view, the remaining three bars are hidden from view by the top three bars.

Figure 10. Photo. SDS specimens reinforced with 2304 SS under test. The photo shows three slabs straddling two wooden boards at a height of approximately 1 m, which allows air circulation around the entire specimen. The slabs are painted white except for the bar ends. Ponding chambers are white plastic and caulked to the tops of the specimens. The wooden boards are supported by gray cinder blocks on both ends.

Figure 11. Graph. Accelerated corrosion test data. In the graph, current density and pH are plotted on the vertical axis with the scale ranging from zero to 60 in increments of 10. The units are microamperes per cm square for current density with units of pH sharing the same scale. Time is plotted on the x-axis in hours ranging from zero to 2,000 hours in increments of 500 hours. The pH data are shown as a dashed gray horizontal line. Chloride content  in weight (wt) percent is plotted using the right vertical axis with the scale ranging from zero to 16 wt percent in increments of 2 wt percent. The chloride data are shown as a solid black line increasing from zero on the left to 14 in the upper right portion of the graph. The 10 2304 specimens are plotted separately using combinations of black, white, and gray symbols and black or gray lines. The initial current densities range from zero to 15 and decrease with time. The current densities stabilize and fall to nearly zero over the first 100 hours. The chloride content increases, starting at 100 hours, from zero to approximately 14 wt percent. The individual specimens show sharp increases in current density starting at approximately 300 hours. All specimens showed sharp increases by 1,700 hours.

Figure 12. Graph. Cumulative distribution plot of CT for 2304 SS from accelerated testing. In the graph, the chloride concentration in weight (wt) percent is plotted on the x-axis ranging from zero to 10 wt percent in increments of 1 wt percent. Percentage probability is plotted on the y-axis ranging from 1 to 90 percent as 1, 2, 5, and 10 percent, then in increments of 10 to 90 percent. The critical chloride concentration for initiation of corrosion, C subscript uppercase T, is approximately 5 percent at a probability of 10 percent assuming a normal probability distribution. The percent probability increases with increasing concentration from 10 percent to over 80 percent probability at approximately 9 percent chloride concentration. A straight line running upward to the right is drawn through the data points.

Figure 13. Graph. Potential data for the 2304 SS-reinforced concrete specimens. In the graph, potential in millivolts versus saturated calomel electrode is plotted on the y-axis ranging from -400 to zero millivolts in increments of 50 millivolts. Exposure time in days is plotted on the x-axis ranging from zero to 1,000 days in increments of 200 days. The data are plotted for three specimens using three colors connected by lines of the same colors as the points. Specimen 2304-1 is shown in dark blue, specimen 2304-2 is shown in red, and specimen 2304-3 is shown in green. Over 929 days, the potentials generally ranged from -150 to -200 mV versus SCE (saturated calomel electrode) except for specimen 2304-1 (blue), which exhibited excursions as low as -356 mV versus SCE.

Figure 14. Graph. Macrocell current data versus time for 2304 SS-reinforced concrete specimens. In the graph, current, measured in microamperes, is plotted on the y-axis ranging from zero to 0.20 microamperes in increments of 0.05 microamperes. Exposure time in days is plotted on the x-axis ranging from zero to 1,000 days in increments of 200 days. The data are plotted for three specimens using three colors connected by lines of the same colors as the points. Specimen 2304-1 is shown in dark blue, specimen 2304-2 is shown in red, and specimen 2304-3 is shown in green. The currents were zero except for one data point that corresponded to the low potential excursion for specimen 2304-1 shown in figure 13. This correspondence of potential (400 days, -356 mV) and current (400 days, 0.1 microamperes) suggests that this specimen experienced activation followed by repassivation.

Figure 15. Graph. Concrete chloride concentration profiles determined from 10 cores. In the graph, chloride concentration measured in weight (wt) percent is plotted on the y-axis ranging from zero to 7.00 wt percent in increments of 1.00 wt percent. Distance in m is plotted on the x-axis ranging from 0.00E+00 m to 7.00E-02 m in increments of 0.01E-02 m. The concrete contained no reinforcing bar but was subjected to the same exposure conditions as the reinforced concrete specimens. The cores were obtained at different times. Slices parallel to the top surface were taken and ground to powder. The results of the Florida Department of Transportation wet chemistry method chloride analysis are shown. These data are used for Fick's law calculations. Combinations of symbol shapes and line colors are used to plot the results. Generally, the results showed high concentrations at zero distance decreasing to zero with increasing distance. Core specimens obtained at longer times showed higher chloride concentration at zero distance decreasing to zero chloride concentration at greater distances.

Figure 16. Chart. Standard SAE J2334 cyclic test with five cycles/week. The diagram depicts a flow chart with four boxes placed vertically on top of each other filled with text. The cycle starts at the top with 6 hours of wet stage at 100 percent relative humidity (RH) and 50 degrees Celsius. The text reads, “Wet stage: RH=100%; T=50 °C (6 hours).” The box points to the one below it with an arrow, indicating the next stage, which is salt application by soaking for 15 minutes. The text reads, “Salt application at ambient conditions (1 Soak = 15 minutes).” Another arrow points to the box below it, which is the last portion of the 24-hour cycle. It consists of a dry stage at 50 percent relative humidity and 60 degrees Celsius. The text reads, “Dry stage: RH=50%; T=60 °C (17 h 45 min).” To the left of the box, there is an arrow pointing left with the text “Repeat daily for 5 days.” That arrow connects to an arrow pointing upward, which then connects with a third arrow pointing back to the first box, which completes the cycle. Below the third text box, there is a fourth box. The arrow pointing down from the third box to the fourth box has the text “Weekends only.” On weekends only, the dry stage continues during Saturday and Sunday, resuming the cycle on Monday with the 6-hour wet stage. The text in the fourth box reads, “Dry stage: RH=50%; T=60 °C (48 hours).” To the left of the textbox, there is an arrow pointing left, which then connects with the arrow pointing upward from the third box, which points back to the first box and completes the cycle.

Figure 17. Photo. CARON environmental chamber for cyclic SAE J2334 tests. The environmental chamber appears similar to a large gray refrigerator (2 m tall by 0.9 m wide by 0.6 m deep) with wire shelving and a glass door. The control unit with temperature and humidity displays is at the top of the chamber. The control unit manages the temperature and humidity during 24-hour cycles.

Figure 18. Photo. Specimens on holder rack and in soak tank. The figure provides two views of the specimens on a holder rack. The rack in the upper photo is fabricated from white polyvinyl chloride (PVC) pipe with angled slots for holding the specimens. Specimens are shown in the rack and appear as dark rusted metal angled 15 degrees from the lower left to upper right. The rack is shown as it is placed on the shelf in the environmental chamber. The soak tank (lower photo) is large enough to accommodate two PVC racks and deep enough to completely submerge the metal specimens. The soak tank is shown with one PVC rack and specimens immersed in soak solution.

Figure 19. Graph. Example of X-ray powder diffraction spectrum. Counting rate or intensity is plotted on the y-axis ranging from zero to 250 in increments of 50, and angles of 2 theta are plotted on the x-axis ranging from 10 to 70 in increments of 10. The example shows a powder diffraction pattern for iron oxide powder, which contains both hematite and maghemite. The individual compound is identified by peaks occurring at specific angles and intensities. A magenta colored line shows the counting rate from 10 degrees increasing from left to right. Sharp increases in rate identified with hematite (blue symbols) and maghemite (red symbols) are shown at the top of each peak.

Figure 20. Illustration. Atmospheric corrosion sensor (Model FAU2). A cross section of the corrosion sensor is shown. The copper- or gold-plated brass screw (dark orange color, head end up) serves as the cathode. The components are assembled onto the screw from the bottom starting with a nylon (green color) washer, followed by the steel anode (brown color) that has urethane (light brown color) coating the sides and bottom. Assembly is continued from the bottom with a shoulder washer (green color) followed by a nylon nut (black color). The cathode and anode are drilled and tapped for the copper lead wires (black curving line). The copper leads are screwed into the threaded holes using silver metal-filled epoxy for good electrical conduction. The lead attachment points are sealed with caulk (blue color) to electrically insulate the conductive surfaces.

Figure 21. Illustration. Anode detail for atmospheric corrosion sensor. The top and side views of the steel anode are shown. The anode is in the form of a thin disk labeled in the diagram as 33.02 mm with a hole in the center labeled 10.16 mm. The side view of the disc shows bevels at edges to allow uniform coating. The bevels are approximately 45-degree cuts on the edges to help round off the sharp edges. The top view shows diameters for the through-hole, the insulating washer area (light brown color), the inner and outer active area (reddish-brown color) diameters, and the coated area (light blue color).

Figure 22. Photo. Data logger incorporating a ZRA. The figure shows an overhead photo of a data logger (black colored plastic box with “018” written on the right side in white) with the cover removed. The circuit board, which appears green, and the battery compartment are observed with connectors for the sensor (red- and black-paired wires) on the right and the data cable connector (black colored end cap) on the left end of data logger.

Figure 23. Graph. Sensor output for 0.7-inch active anode steel washer diameter for one SAE J2334 cycle in the standard solution. Current in microamperes is plotted on the y-axis ranging from zero to 70 microamperes in increments of 10 microamperes. Time in hours is plotted on the x-axis ranging from 1 to 20 hours in increments of 1 hour. During the dry portion of the cycle, the soaked sensor current increased from zero to 65 microamperes between the fifth and sixth hours followed by decreasing currents to zero microamperes. The wet cycle began at the 10th hour, which caused increasing currents as the sensor became wetted again. After soaking the sensor again at the 13th hour, the sensor current increased to 60 microamperes between the 16th and 17th hours during the wet portion of the cycle followed by decreasing currents to zero microamperes when the cycle entered the dry stage.

Figure 24. Graph. Sensor output for 0.8-inch active anode steel washer diameter for one SAE J2334 cycle soaked with standard solution. Current in microamperes is plotted on the y-axis ranging from zero to 100 microamperes in increments of 20 microamperes. Time in hours is plotted on the x-axis ranging from 1 to 20 hours in increments of 1 hour. During the dry portion of the cycle, the soaked sensor current increased from zero to 65 microamperes between the fifth and sixth hours followed by decreasing currents to zero microamperes. The wet cycle began at the 10th hour, which caused increasing currents as the sensor became wetted again. After soaking the sensor again at the 13th hour, the sensor current increased to 95 microamperes between the 16th and 17th hours during the wet portion of the cycle followed by decreasing currents to zero microamperes when the cycle entered the dry stage. The flat shape of the curve between the 15th and 16th hours indicates that the current has exceeded the range of the data logger.

Figure 25. Photo. Bottom side of a sensor mounted on its holder experiencing under-paint corrosion. The photo is taken in the upward direction of a sensor mounted on one arm of its cross-shaped holder. The holder is white plastic with a threaded hole into which the sensor screw is threaded. The photo shows reddish rust stains and paint peeling in areas on the sensor that were intended to be protected from corrosion by the paint.

Figure 26. Photo. Four sensors set up on the cross-shaped holding rack. The photo was taken from the side in a slightly elevated position downward toward the sensor holding rack. The holding rack is cross-shaped and fabricated out of clear acrylic plastic. The photo shows four sensors of the final design using urethane protective coating (white) and either copper or gold cathodes (screw heads). Directly beneath the screw heads, a nylon washer can be observed. Between the nylon washer and the white protective coating is the grey metallic anode. These sensors are mounted with the flat disc in the horizontal position. White lead wires are attached to the tops and bottoms of the sensors and run vertically upward and exit the chamber to connect with the data loggers.

Figure 27. Photo. Components and fabrication of the cable sensor. The figure is comprised of three photos with arrows to the right of the first two photos pointing to the next photo indicating a sequence of cable sensor assembly. The photos show sensor components and a connection sequence for the first part of the fabrication. The first photo on the left shows three components including a steel wire anode, a resistor, and a thin copper wire. The middle photo shows solder connections of the red lead wire to the thin copper wire, the opposite end of the copper wire to the resistor, the opposite end of the resistor to the steel wire anode, and finally the opposite end of the steel wire anode to the black lead wire. The right side photo shows protective epoxy coating of the solder connections and black semitransparent cylinders illustrating protection of the epoxied solder connections with heat shrink tubing. The fabrication process is continued in the next figure (figure 28).

Figure 28. Photo. Completion of the cable sensor. The figure is comprised of four photos with arrows to the right of the first three photos pointing to the next photo. The photos show from left to right (1) the addition of the inner fiberglass sleeve (shown as a semitransparent gold colored tube), (2) the wrapping of the copper cathode around the sleeve (shown as a white fabric in the adjacent photo), (3) the installation of a second outer white fabric sleeve over the wrapped copper wire followed by heat shrink tubing at the free and lead wire ends, and (4) the sealing of both ends of the sensor with marine sealant (white coating).

Figure 29. Graph. The first test of the cable corrosion sensor, wetted, dried out, rewetted, and redried. Current in microamperes is plotted on the y-axis ranging from zero to 100 microamperes in increments of 20 microamperes. Time in hours is plotted on the x-axis ranging from zero to 600 hours in increments of 100 hours. When the sensor is wet, the current approaches 100 microamperes as shown in the first few hours on the graph. When the sensor is drying, the current decreases to nearly zero microamperes. At 100 hours, the sensor is rewetted, and the current increases to nearly 100 microamperes. When subjected to drying conditions a second time at approximately 180 hours, the current decreases to zero microamperes.

Figure 30. Illustration. Cable specimen showing arrangement of strands and sensors. The upper portion of the diagram shows in cross section an arrangement of 3 sensors (green circles) in a vertical column with 12 steel rods (blue circles) surrounding the sensors in a bundle. The whole bundle is wrapped with polyethylene film around the perimeter. The middle portion of the diagram shows a loose bundle of sensors and steel rods in a side view in which there is a tie in the middle, which leaves the ends open to air and evaporation/drying conditions. The polyethylene film wrapping around the bundle is shown as a gray rectangle space, and the tie is shown as a black loop. The bottom portion of the diagram is similar to the loose bundle, except there are additional ties at the ends securing the plastic wrap in a tight bundle.

Figure 31. Graph. A606 corrosion as a function of concentration of NaCl. In the graph, corrosion penetration in mm is plotted on the y-axis ranging from zero to 0.18 mm in increments of 0.02 mm. Sodium chloride (NaCl) concentration in weight (wt) percent is plotted on the x-axis ranging from zero to 6 wt percent in increments of 1 wt percent. There are eight sets of experiments for NaCl concentrations of zero to 5 percent with the combinations including either one or two soaks, 15 or 30 cycles, and pH 6 or 8. The corrosion values increase from zero to 5 percent NaCl. The corrosion values also increase with lower pH, higher number of soaks, and higher number of cycles. The highest corrosion value was observed for pH 6, two soaks, 30 cycles, and 5 percent NaCl. The graph shows that corrosion penetration increases rapidly up to 2 percent NaCl and then more gradually up to 5 percent NaCl. These increases occur in two sets corresponding to 15 cycles and 30 cycles. The 15-cycle set increases to 0.06 mm. The 30–Cycle set increases to 0.17 mm. The legend for these sets of data is as follows: dark blue solid diamond for pH 6 with one soak and 15 cycles, magenta solid square for pH 6 with one soak and 30 cycles, orange solid triangle for pH 6 with two soaks and 15 cycles, cyan cross for pH 6 with two soaks and 30 cycles, purple cross for pH 8 with one soak and 15 cycles, brown solid diamond for pH 8 with one soak in 30 cycles, teal vertical bar for pH 8 with two soaks and 15 cycles, and blue horizontal bar for pH 8 with two soaks and 15 cycles.

Figure 32. Graph. SAE1010 corrosion as a function of concentration of NaCl. In this graph, corrosion penetration in mm is plotted on the y-axis ranging from zero to 0.2 mm in increments of 0.02 mm. Sodium chloride (NaCl) concentration measured in weight (wt) percent is plotted on the x-axis ranging from zero to 6 wt percent in increments of 1 wt percent. There are eight sets of experiments for NaCl concentrations of zero to 5 percent with combinations of either one or two soaks, 15 or 30 cycles, and pH 6 or 8. The corrosion values increase from 0 to 5 percent NaCl. The corrosion values also increase with lower pH, higher number of soaks, and higher number of cycles. The graph shows that corrosion penetration increases rapidly up to 2 percent NaCl and then more gradually up to 5 percent NaCl. These increases occur in two sets corresponding to 15 cycles and 30 cycles. The 15-cycle set increases to 0.07 mm. The 30–Cycle set increases to 0.18 mm. The legend for these sets of data is as follows: dark blue solid diamond for pH 6 with one soak and 15 cycles, magenta solid square for pH 6 with one soak and 30 cycles, orange solid triangle for pH 6 with two soaks and 15 cycles, cyan horizontal bar for pH 6 with two soaks and 30 cycles, purple horizontal bar for pH 8 with one soak and 15 cycles, brown solid diamond for pH 8 with one soak in 30 cycles, teal vertical bar for pH 8 with two soaks and 15 cycles, and blue horizontal bar for pH 8 with two soaks and 15 cycles. The corrosion values are slightly higher than those observed for A606 specimens (compare to figure 31).

Figure 33. Graph. Relative corrosion between A606 and SAE1010. In this graph, percentage difference is plotted on the y-axis ranging from -30 to 50 percent in increments of 10 percent. The horizontal axis shows four sets in groups of two—the first group on the left is for one dip, and the second group on the right is for two dips. The differences in relative corrosion between A606 and SAE1010 are shown as percentage differences for 0.5 to 5 percent sodium chloride (NaCl). Positive differences indicate better performance or lower corrosion rate of A606 versus SAE1010. The largest positive differences occurred for 0.5 percent NaCl. The legend for these sets of data is as follows: solid black for 0.5 percent NaCl at pH 6, solid gray for 0.5 percent NaCl at pH 8, downward left to right diagonal pattern for 2 percent NaCl at pH 6, upward left to right diagonal pattern for 2 percent NaCl at pH 8, vertical bar pattern for 5 percent NaCl at pH 6, and horizontal bar pattern for 5 percent NaCl at pH 8.

Figure 34. Graph. A606 corrosion as a function of chloride concentration during a one soak/cycle exposure. Corrosion penetration in mm is plotted on the y-axis ranging from zero to 0.18 mm in increments of 0.02 mm. Sodium chloride (NaCl) concentration in weight (wt) percent is plotted on the x-axis ranging from zero to 5 wt percent in increments of 0.5 wt percent. The trend for these plots shows an increasing corrosion penetration with increasing NaCl concentration. The four plots are in groups of two—one group is for 30 cycles that shows corrosion penetration up to 0.16 mm, and the other group is for 15 cycles that shows corrosion penetration up to 0.06 mm. The legend for the sets of data is as follows: solid black circle for pH 6 and 15 cycles, solid black diamond for pH 6 and 30 cycles, solid gray triangle for pH 8 and 15 cycles, and solid gray square for pH 8 and 30 cycles. Higher rates are observed for pH 6 versus pH 8 and for 30 cycles versus 15 cycles. The rates also increase with higher NaCl concentrations.

Figure 35. Graph. A606 corrosion as a function of chloride concentration during a two soak/cycle exposure. Corrosion penetration in mm is plotted on the y-axis ranging from zero to 0.18 mm in increments of 0.02 mm. Sodium chloride (NaCl) concentration in weight (wt) percent is plotted on the x-axis ranging from zero to 5 wt percent in increments of 1 wt percent. The trend for these plots shows an increasing corrosion penetration with increasing NaCl concentration. The four plots are in groups of two—one group is for 30 cycles that shows corrosion penetration up to 0.17 mm, and the other group is for 15 cycles that shows corrosion penetration up to 0.06 mm. The legend for the sets of data is as follows: solid black circle for pH 6 and 15 cycles, solid black diamond for pH 6 and 30 cycles, solid gray triangle for pH 8 and 15 cycles, and solid gray square for pH 8 and 30 cycles. Higher rates are observed for pH 6 versus pH 8 and for 30 cycles versus 15 cycles. The rates also increase with higher NaCl concentrations. The rates are slightly higher for two soak/cycle exposure versus one soak/cycle exposure (compare to figure 34).

Figure 36. Graph. SAE1010 corrosion as a function of chloride concentration during a one soak/cycle experiment. Corrosion penetration in mm is on the y-axis ranging from zero to 0.16 mm in increments of 0.02 mm. Sodium chloride (NaCl) concentration in weight (wt) percent is plotted on the x-axis ranging from zero to 5 wt percent in increments of 0.5 wt percent. The trend for these plots shows an increasing corrosion penetration with increasing NaCl concentration. The four plots are in groups of two—one group is for 30 cycles that shows corrosion penetration up to 0.15 mm, and the other group is for 15 cycles that shows corrosion penetration up to 0.06 mm. The legend for the sets of data is as follows: solid black circle for pH 6 and 15 cycles, solid black diamond for pH 6 and 30 cycles, solid gray triangle for pH 8 and 15 cycles, and solid gray square for pH 8 and 30 cycles. Higher rates are observed for pH 6 versus pH 8 and for 30 cycles versus 15 cycles. The rates also increase with higher NaCl concentrations. The rates are similar to those observed for A606 (compare to figure 34) except at 0.5 weight percent and 30 cycles, which shows higher rates.

Figure 37. Graph. SAE1010 corrosion as a function of chloride concentration during a two soak/cycle exposure. Corrosion penetration in mm is plotted on the y-axis ranging from zero to 0.2 mm in increments of 0.02. Sodium chloride (NaCl) concentration in weight (wt) percent is plotted on the x-axis ranging from zero to 5 wt percent in increments of 0.5 wt percent. The trend for these plots shows an increasing corrosion penetration with an increasing NaCl concentration. The four plots are in groups of two—one group is for 30 cycles that shows corrosion penetration up to 0.18 mm, and the other group is for 15 cycles that shows corrosion penetration up to 0.7 mm. The legend for the sets of data is as follows: solid black circle for pH 6 and 15 cycles, solid black diamond for pH 6 and 30 cycles, solid gray triangle for pH 8 and 15 cycles, and solid gray square for pH 8 and 30 cycles. Higher rates are observed for pH 6 versus pH 8 and for 30 cycles versus 15 cycles. The rates also increase with higher NaCl concentrations. The rates are slightly higher for two soak/cycle exposure versus one soak/cycle exposure (compare to figure 36).

Figure 38. Graph. Relative corrosion versus NaCl concentration during exposure to a one soak/cycle environment. The percentage difference in corrosion is shown on the y-axis ranging from -30 to 50 percent in increments of 10 percent. Sodium chloride (NaCl) concentration in weight (wt) percent is plotted on the x-axis ranging from 0.5 to 5 wt percent in increments of 0.5 wt percent. The legend for the sets of data is as follows: solid black circle for pH 6 and 15 cycles, solid black diamond for pH 6 and 30 cycles, solid gray triangle for pH 8 and 15 cycles, and solid gray square for pH 8 and 30 cycles. The trend for all sets of data shows large positive percentage differences at low NaCl concentrations decreasing to small negative differences at 2 wt percent NaCl followed by nearly no differences at 5 wt percent NaCl. Positive differences indicate better performance or lower corrosion rate of A606 versus SAE1010. The largest positive differences occurred for 0.5 wt percent NaCl.

Figure 39. Graph. A606 corrosion as a function of chloride concentration at pH 6. Corrosion penetration rate in mm is plotted on the y-axis ranging from zero to 0.18 mm in increments of 0.02 mm. Sodium chloride (NaCl) concentration in weight (wt) percent is plotted on the x-axis ranging from zero to 5 wt percent in increments of 0.5 wt percent. The four plots are in groups of two—one group is for 30 cycles that shows penetration up to 0.17 mm, and the other group is for 15 cycles that shows corrosion penetration up to 0.06 mm. The legend for the sets of data is as follows: solid black circle for one soak and 15 cycles, solid black diamond for one soak and 30 cycles, solid gray triangle for two soaks and 15 cycles, and solid gray square for two soaks and 30 cycles. The plot compares one soak versus two soaks. For both 15- and 30–Cycle exposures, two soaks resulted in higher corrosion rates, especially at high concentrations.

Figure 40. Graph. A606 corrosion as a function of chloride concentration at pH 8. Corrosion penetration rate in mm is plotted on the y-axis ranging from zero to 0.18 mm in increments of 0.02 mm. Sodium chloride (NaCl) concentration in weight (wt) percent is plotted on the x-axis ranging from zero to 5 wt percent in increments of 0.5 wt percent. The four plots are in groups of two—one group is for 30 cycles that shows penetration up to 0.16 mm, and the other group is for 15 cycles that shows corrosion penetration up to 0.05 mm. The legend for the sets of data is as follows: solid black circle for one soak and 15 cycles, solid black diamond for one soak and 30 cycles, solid gray triangle for two soaks and 15 cycles, and solid gray square for two soaks and 30 cycles. The plot compares one soak versus two soaks. For both 15- and 30–Cycle exposures, two soaks resulted in higher corrosion rates, especially at high concentrations.

Figure 42. Graph. SAE1010 corrosion as a function of chloride concentration at pH 8. Corrosion penetration rate in mm is plotted on the y-axis ranging from zero to 0.2 mm in increments of 0.02 mm. Sodium chloride (NaCl) concentration in weight (wt) percent is plotted on the x-axis ranging from zero to 5 wt percent in increments of 0.5 wt percent. The four plots are in groups of two—one group is for 30 cycles that shows penetration up to 0.18 mm, and the other group is for 15 cycles that shows corrosion penetration up to 0.06 mm. The legend for the sets of data is as follows: solid black circle for one soak and 15 cycles, solid black diamond for one soak and 30 cycles, solid gray triangle for two soaks and 15 cycles, and solid gray square for two soaks and 30 cycles. The plot compares one soak versus two soaks. For both 15- and 30–Cycle exposures, two soaks resulted in higher corrosion rates, especially at high concentrations.

Figure 43. Graph. Corrosion of A606 and SAE1010 versus chloride concentration for the second test set. Corrosion penetration in mm is plotted on the y-axis ranging from zero to 0.14 mm in increments of 0.02 mm. Sodium chloride (NaCl) concentration in weight (wt) percent is plotted on the x-axis ranging from zero to 2 wt percent in increments of 0.5 wt percent. The four plots are in groups of two—one group is for 30 cycles that shows corrosion penetration up to 0.14 mm, and the other group is for 15 cycles that shows corrosion penetration up to 0.05 mm. The legend for the sets of data is as follows: solid black diamond for A606 and 15 cycles, solid black squares for A606 and 30 cycles, solid gray triangle for SAE1010 and 15 cycles, and solid gray square for SAE1010 and 30 cycles. For this concentration range at equal cycles, A606 exhibits lower corrosion rates than SAE1010, especially for the 30–Cycle test. This difference is expected for weathering steel versus carbon steel.

Figure 44. Graph. Relative corrosion versus chloride concentration. The percent difference is plotted on the y-axis ranging from zero to 45 percent in increments of 5 percent. Sodium chloride (NaCl) concentration in weight (wt) percent is plotted on the x-axis ranging from zero to 1.8 wt percent in increments of 0.2 wt percent. The general trends for the two plots are increasing percentage difference up to 0.7 wt percent NaCl followed by stable or decreasing percentage difference. The legend for the sets of data is as follows: solid gray diamond for 15 cycles and solid black square for 30 cycles. Positive differences indicate better performance or lower corrosion rate of A606 versus SAE1010. The most significant differences occur at 0.5 to 0.7 wt percent NaCl.

Figure 45. Graph. Output for a Cu-A606 atmospheric corrosion sensor using soaking solution 3-1. Sensor output in microamperes is plotted on the y-axis ranging from zero to 90 microamperes in increments of 10 microamperes. Time is plotted on the x-axis in hours ranging from zero to 24 hours in increments of 1 hour. Wet stage occurred from the 6th to 12th hours, and soaking occurred at the 11th hour (indicated on the graph by a solid circle symbol at coordinates (11, 90). Up to the beginning of the wet stage, the sensor output (current) was zero microamperes. During the wet stage, the current increased to almost 50 microamperes. At the end of the wet stage (11th hour), the sensor was soaked in salt solution, and the current increased additionally to nearly 80 microamperes until sufficient drying of the salt solution caused the current to decrease to zero at the 15th hour and remained zero microamperes to the 24th hour.

Figure 46. Graph. Output for Cu-A606 sensor during a 15-cycle test. Current in microamperes is plotted on the y-axis ranging from zero to 120 microamperes in increments of 20 microamperes. Time in hours is plotted on the x-axis ranging from zero to 400 hours in increments of 20 hours. The D and W indicate approximate dry and wet periods. During a dry period (D), the current output is zero microamperes. During wet periods (W), the currents increase to 40 microamperes or more. There are 15 wet periods in this data set with corresponding increases in current. At approximately 300 hours, there is a long dry period that is 48 hours longer than the normal dry period, which corresponds to a weekend when no soaks were performed. The three current increases after the dry weekend were noticeably smaller than most of the previous current increases. The output is shown for sensor C-W, which is the copper-A606 sensor using solution 3-1 (0.2 percent NaCl plus 0.075 percent CaCl2) as the soaking solution.

Figure 47. Graph. Corrosion of A606 versus sodium chloride concentration. Corrosion penetration in mm is plotted on the y-axis ranging from zero to 0.14 mm in increments of 0.02 mm. Sodium chloride (NaCl) concentration in weight (wt) percent is plotted on the x-axis ranging from zero to 6 wt percent in increments of 1 wt percent. The legend for these data sets is as follows: solid black squares for 30 cycles and solid gray diamonds for 15 cycles. There are 10 data points for each line. Data from test sets 1 and 2 (15 and 30 cycles) are plotted together and show an inflection at 2 percent NaCl. The corrosion penetration increases with increasing NaCl concentrations. The inflection is a slight decrease with increasing NaCl concentrations. This inflection suggests that seven cycles per week (test set 2, for 0.2 to 1.7 percent) produces higher corrosion rates than five cycles per week (test set 1, for 2 and 5 percent) attributed to the longer dry periods for the seven cycles per week test.

Figure 48. Graph. Corrosion (weight-loss) of A606 coupons and calculated mass-loss for A606 sensors versus sodium chloride concentration for 15-cycle exposure. Corrosion penetration in mm is plotted on the y-axis ranging from zero to 0.035 mm in increments of 0.005 mm. Sodium chloride (NaCl) concentration is plotted in weight (wt) percent on the x-axis ranging from zero to 1.4 wt percent in increments of 0.2 percent. There is a dashed box drawn on the graph, which identifies the over-range sensor data and is located at 0.015 mm with a corrosion penetration between 0.7 and 1.4 NaCl wt percent. The legend for these data sets is as follows: open black square for gold-weathering steel sensor, open gray square for copper-weathering steel sensor, and solid black square for weight loss coupons. The sensor data are scaled by two times. The trend for the weight loss coupons is increasing corrosion penetration with NaCl wt percent. The trend for both sensors shows an increasing corrosion penetration with an increasing NaCl wt percent up to 0.015-mm corrosion penetration followed by leveling off as indicated by the dashed box. The weight-loss data are compared to sensor data by converting to corrosion penetration using unit conversion and Faraday's law of electrolysis, respectively. The sensor data is expected to yield lower values than the weight-loss data yet follow the trend for the lower concentrations. The higher concentrations did not follow the trend due to data logger error.

Figure 49. Graph. Corrosion (weight-loss) of SAE1010 coupons and calculated mass-loss for SAE1010 sensors versus sodium chloride concentration for 15-cycle exposure. Corrosion penetration in mm is plotted on the y-axis ranging from zero to 0.05 mm in increments of 0.01 mm. Sodium chloride (NaCl) concentration is plotted in weight (wt) percent on the x-axis ranging from zero to 1.4 wt percent in increments of 0.2 wt percent. There is a dashed box drawn on the graph, which identifies the over-range sensor data and is located at 0.015 mm with a corrosion penetration between 0.7 and 1.4 NaCl wt percent. The legend for these data sets is as follows: open black square for gold-steel sensor, open gray square for copper-steel sensor, and solid black square for weight loss coupons. The sensor data are scaled by two times. The trend for the weight-loss coupons is increasing corrosion penetration with NaCl wt percent. The trend for both sensors shows an increasing corrosion penetration with an increasing NaCl wt percent up to 0.015-mm corrosion penetration followed by leveling off as indicated by the dashed box. The weight-loss data are compared to sensor data by converting to corrosion penetration using unit conversion and Faraday's law of electrolysis, respectively. The sensor data is expected to yield lower values than the weight loss data yet follow the trend for the lower concentrations. The higher concentrations did not follow the trend due to data logger error.

Figure 50. Graph. Response of cable sensor before and after dilute Harrison solution. In this graph, sensor response in microamperes is plotted on the y-axis ranging from zero to 100 microamperes in increments of 20 microamperes. Time in days is plotted on the x-axis ranging from zero to 40 days in increments of 10 days. The response remained zero microamperes until the sensor was wetted in Harrison solution on the 32d day. When wet, the sensor gave a large response (nearly 100 microamperes) followed by a decrease to zero microamperes and small responses (1 to 5 microamperes) on subsequent days during the wet cycle.

Figure 51. Graph. Response of cable sensor during constant 50-percent RH exposure after dilute Harrison solution soak. Sensor response in microamperes is plotted on the y-axis ranging from zero to 100 microamperes in increments of 10 microamperes. Time in days is plotted on the x-axis ranging from zero to 40 days in increments of 10 days. The response is zero microamperes until the sensor is wetted on the fist day. The sensor response increases gradually to nearly full scale as the sensor becomes fully wetted. The response gradually decreases to zero microamperes on the ninth day and remains at zero microamperes throughout the remaining days.

Figure 52. Graph. Response of cable sensor during constant 100-percent RH exposure after dilute Harrison solution soak. Sensor response in microamperes is plotted on the y-axis ranging from zero to 100 microamperes in increments of 10 microamperes. Time is plotted in days on the x-axis ranging from zero to 40 days in increments of 10 days. The response is zero microamperes until the sensor is wetted on the first day when the current increased to 40 microamperes then rapidly decreased to 10 to 20 microamperes, which was maintained to the 17th day. At the 17th day, the current increased to 50 microamperes and gradually decreased to 3 to 5 microamperes on the 35th day. The sensor response is moderate throughout the measurement period. The sensor does not completely dry out in the 100-percent relative humidity (RH) exposure condition and gives continuous (nonzero) current readings.

Figure 53. Graph. XRD pattern of steel rods in cable sensor bundle after single soak in dilute Harrison solution and exposure in cyclic chamber for 40 days. X-ray intensity is plotted on the y-axis ranging from zero to 160 in increments of 20. The angle 2 theta is shown on the x-axis ranging from 10 to 70 in increments of 10. The general trend of the data shows that intensity values vary rapidly by plus or minus 5 units as the angle 2 theta increases from 10 to 70. This general trend establishes a baseline for the detection of peaks, which are sharp transitions above the baseline. Peaks characteristic of six materials are identified with symbols above the peaks and include hematite (blue square), goethite (red triangle), lepidocrocite (cyan diamond), maghemite (red circle), iron (cyan asterisk), and butlerite (blue plus).