Accelerated Chloride Threshold Level Test 16 

The accelerated chloride threshold (ACT) test was developed to determine the critical chloride threshold of steel samples embedded in mortar in a relatively short time compared to the standard corrosion test methods (Trejo and Miller 2003). This test accelerates the transfer of chloride ions to the steel surface using an electrical field. Chloride ions migrate under the effect of an electrical field to the steel surface instead of slowly diffusing into concrete due to concentration differences.

17

Steel reinforcement samples are embedded in 3 x 6 in (75 x 150 mm) mortar cylinders that have attached chloride solution reservoirs at the top. Chloride reservoirs are filled with 3.5 percent by weight sodium chloride solution. Figure II-1 shows the layout of ACT test setup. An electrical field is generated by applying a potential difference of 20V between an anode at the reinforcement level and a cathode in the chloride reservoir. Negatively charged chlorides migrate to the anode, and when they reach the chloride threshold level, they initiate corrosion of the steel reinforcement. Corrosion of steel reinforcement is monitored through polarization resistance method. A counter electrode and Haber-Luggin probe, which contains a reference electrode, are embedded in the mortar above the steel reinforcement to run the polarization resistance tests. The anode, cathode, and counter electrodes are fabricated from Ni-Cr mesh.

Figure II-1 ACT Test Layout.

An electrical field is applied in intervals of 6 hours, and polarization resistance of reinforcement is measured at the end of a wait period of 42 hours after each application of the electric field. Testing is stopped when initiation of corrosion is detected. The amount of

18

chlorides at the steel reinforcement level is then measured to determine the chloride threshold value.

Castellote et al. (2002) also developed a similar chloride threshold measurement method that uses an electrical field to accelerate the transfer of chloride ions. This method uses mortar cubes with an embedded steel reinforcement, and it monitors corrosion initiation through polarization resistance. It should be noted that in this test method the anode and the reference electrode are not embedded in the mortar. The anode is underneath the sample, and the cathode and the reference electrodes are placed in the chloride reservoir. Castellote et al. recommends the use of 10 to 13 V potential difference to drive the chlorides and uses a 1 M Sodium Chloride solution. Trejo and Pillai (2003) evaluated the use of 1, 5, 10, 20, and 40 V potential differences for the electrical field and determined that up to 10 V the chloride profile development in the mortar did not change significantly.

The use of an electrical field to accelerate the transport of chlorides into the mortar or concrete also causes polarization of the steel reinforcement that is being tested for chloride threshold. In the literature it is reported that chloride threshold is independent from the potential of the reinforcement for potential values greater than -200 ±50 mV vs. SCE (Alonso et al. 2002) and that it linearly increases with decreasing potential for potential values less than - 200 ±50 mV vs. SCE (Izquierdo et al. 2004). The ACT test minimizes the polarization of reinforcement by placing the anode at the same level as the reinforcement and by connecting the anode to the ground terminal of the power source (Trejo and Pillai 2003). Castellote et al. (2002) reported that, although applied, the electrical field polarized the reinforcement being tested in direct proportion to the applied voltage, the reinforcement potential returned to its original value soon after the power supply was switched off. They also reported that because of this reversibility of potential the electrical field can be switched off when a drop in the potential is observed and that this way the chloride threshold could be measured in a quasi-natural state. Because the ACT test also measures the polarization resistance when the electrical field is switched off, the same claim about the quasi-natural state of reinforcement can be made for the ACT test.

Another issue with the use of an electrical field to accelerate chloride transfer is the effect of the electrical field on the pH of the mortar environment. In the literature many researchers reported that chloride threshold levels change proportionally with the hydroxyl ion concentration of the environment and proposed to state chloride threshold levels as chloride to

19

hydroxyl ratios (Glass and Buenfeld 1997). Trejo and Pillai (2003) reported that the pH of the environment around the reinforcement was decreasing with increasing magnitude of applied electrical field and with increasing time of application due to the oxidation of hydroxyl ions at the anode. Similar to chlorides, negatively charged hydroxyl ions are also attracted to the anode and if the rate of their oxidation is higher than their rate of transportation, the pH around the anode (same level as the reinforcement) decreases. On the contrary, Castellote et al. (1999; Castellote et al. 2002) recommended addition of HCl to the chloride solution to neutralize the environment around the reinforcement. They suggested that the extra hydroxyl ions generated at the cathode through reduction of water molecules were being attracted into the mortar causing an increase of the pH of the environment. Due to their higher transference numbers, hydroxyl ions were also slowing the penetration rate of chlorides.

ACT test was used to determine chloride threshold values of conventional and corrosion-resistant reinforcements embedded in mortars with a water-cement ratio of 0.5. Mean critical chloride threshold for ASTM A 615 and ASTM A 706 steels were 0.87 lb/yd3 _(0.52 kg/m3) and 0.34 lb/yd3 (0.20 kg/m3), respectively. The 95 percent confidence intervals of the means for the ASTM A 615 and A 706 steels were 0.51 to 1.20 lb/yd3 _{(0.3 to 0.71 kg/m}3_{) and} 0.25 to 0.40 lb/yd3 (0.15 to 0.24 kg/m3), respectively. The 95 percent confidence interval shows the range of values that includes the mean value with 95 percent probability and is a good indicator of variability of results, i.e., the bigger the interval the more variability. Mean critical chloride threshold values for microcomposite steel, stainless steel 316LN (SS 316LN), and stainless steel 304 (SS304) were 7.7 lb/yd3 _{(4.6 kg/m}3_{), 8.5 lb/yd}3 _{(5.0 kg/m}3_{), and 18.1 lb/yd}3 (10.8 kg/m3_{), respectively. Their 95 percent confidence intervals were 6.5 to 9 lb/yd}3 _{(3.8 to 5.3} kg/m3_{), 6.9 to 10.1 lb/yd}3 _{(4.1 to 6 kg/m}3_{), and 16 to 20.2 lb/yd}3 _{(9.5 to 12 kg/m}3_{), respectively.} Corrosion resistant steel exhibited higher chloride threshold values as expected but with a higher variability compared to the ASTM A 615 and A 706 steels. The duration of the ACT test for conventional steel samples was approximately 7 weeks and approximately 16 weeks for corrosion resistant steel types (Trejo and Pillai 2003; Trejo and Pillai 2004). Although the ACT test was used on different types of steel reinforcement, it was not evaluated for different mortar mixtures with or without corrosion inhibiting admixtures, and its results were not compared with long-term standard test results. This study aims to do both of these evaluations for the ACT method.

20

Castellote et al. (2002) used their chloride threshold determination method only with conventional corrugated steel rebar embedded in mortar with a water-cement ratio of 0.37, but they also tested some of their samples without the application of an electrical field, where chlorides diffused into the mortar similar to longer term standard corrosion tests. They reported that the results of the accelerated test method were similar to the results of the standard diffusion method. Chloride threshold for the accelerated test method was 0.152 percent by weight of sample and 0.227 percent by weight of sample for the diffusion method. Chloride thresholds expressed as chloride to hydroxyl ion ratio for the accelerated and diffusion tests were 2.0 and 1.5, respectively (Castellote et al. 2002). Viedma et al. (2006) also tested similar samples using an electrical field to accelerate the test and using only diffusion of chlorides into the mortar with a water-cement ratio of 0.45. The chloride threshold for the embedded conventional corrugated steel rebar was 2 and 1.07 percent by weight of cement for the diffusion and accelerated samples, respectively. Similar to Castellote et al.’s results, the chloride threshold determined by the diffusion method was higher compared to the result of the accelerated method. Initiation of corrosion for the diffusion samples took 432 days, and the coefficient of variation of the results was 50 percent. For the accelerated samples, initiation of corrosion took place at 1 to 6 days, and the coefficient of variation was only 23 percent.