Testing of Randolph Street Bridge

In 1986, the deck of the bridge that carries Randolph Street over I-35E in St. Paul, Minnesota, was rehabilitated. The rehabilitation process included the application of a Low Slump Dense Concrete (LSDC) overlay. Cortec MCI-2000 was added to the LSDC at 1 lb/yd³ (0.6 kg/m³) for the two westbound traffic lanes. Prior to overlay, the deck was milled to a depth of 0.5 in (13 mm) and the areas of unsound concrete were removed. The cavities from the removal of the unsound concrete were filled with the overlay concrete.

Corrosion assessments were conducted on the eastbound (control) and westbound (MCI) travel lanes of the structure by Virginia Tech researchers on two occasions, June 1991 and August 1992. The assessments included visual inspection, delamination survey, cover-depth survey, chloride contents as a function of depth, corrosion potentials, and estimates of corrosion current densities (i_corr). In November 2000, technicians from Cortec Corporation and American Engineering Testing returned to the bridge and took new measurements. These included Gecor 6 readings and copper/copper sulfate half-cell potentials. A new chloride analysis was also taken at various depths. SHRP-S-658 contains all information from the 1991 and 1992 study. This report will contain data obtained in 2000.

Powdered concrete samples for chloride analysis were taken at mean depths of 1.0, 2.0, and 3.0 in (25, 51, and 76 mm) from 3 different locations on each side of the bridge.

A diagram of these locations is included at the end of this report. Samples were taken using a rotary impact type drill with a _" sized bit. Three-gram samples that passed through a #20 sieve were obtained from each depth. The powder was then mixed with 20 ml of digestion solution for a total of 3 minutes and then 80 ml of stabilizing solution was added. A calibrated electrode coupled to an Orion Model 720-pH/ISE meter was then immersed in the solution, and the chloride-ion concentration was recorded. This method was consistent with the AASHTO: T260 procedure. The standard deviation for this chloride test was determined by testing the six pulverized concrete Quality Assurance (QA) samples of known chloride content. Each QA sample was tested five times.

Half-Cell Potentials

ASTM C 876 corrosion half-cell potentials were measured for both the eastbound and westbound travel lanes with a copper/copper sulfate electrode (CSE) in June 1991, August 1992, and November of 2000, five, six and fourteen years, respectively after construction; and with Gecor 6 in November of 2000. According to ASTM: C876, the results can be interpreted as follows:

Corrosion Potential

Corrosion Current Measurements

Corrosion current density (i_corr) estimates were taken with the Gecor 6 device in November 2000. These same estimates were taken in June 1991 and August 1992 using a 3LP device. The i_corr measurement using the 3LP device is proportional to the corrosion rate through Faraday’s Law. The Gecor 6 measures the corrosion rate of steel in concrete by "polarization resistance" or "linear polarization" techniques. This is a non-destructive technique that works by applying a small current to the rebar and measuring the change in the half-cell potential. Then the polarization resistance, R_p, (the change in potential measured by Gecor 6), is divided by the applied current. The Gecor 6 obtains the corrosion rate, i_corr, from the polarization resistance, R_p, by means of the "Stearn and Geary" relationship:

i_corr = B/R_p, where B = 26 mV.

The criteria for estimating the reinforcement’s condition in relation to the measured value of the rate of corrosion have been defined as¹:

R = resistance by the "iR drop" from a pulse between the sensor counter-electrode and the rebar network

D = counter-electrode diameter of the sensor

The value of the concrete’s resistance is used as an additional parameter for the interpretation of the rate of corrosion. According to Andrade², the following interpretation of the results is possible:

Concrete Resistivity

Carbonation

Carbonation of concrete is a process by which carbon dioxide from the air penetrates the concrete and reacts with the hydroxides, such as calcium hydroxide, to form carbonates. This process increases shrinkage on drying (promoting crack development) and reduces the alkalinity of the concrete. High alkalinity is needed to protect embedded rebar from corrosion; consequently, concrete should be resistant to carbonation to prevent steel corrosion³. The carbonation of powdered concrete samples taken from the bridge was determined by using phenolphthalein and pH measurements.

Gas Chromatography

A method was generated for the detection of MCI in the concrete overlay. GC analyses were performed on powdered samples taken from several locations on both the control and MCI-treated sides to determine whether or not MCI was present in the concrete.

The LSDC (Low Slump Dense Concrete) has done an excellent job at reducing the corrosion rate of the steel on both the control and MCI treated sides by reducing the diffusion of chlorides into the concrete, as well as having a low rate of carbonation, and high alkalinity. The reduced chloride diffusion can be seen when looking at the data from American Engineering Testing. After fourteen years, the chloride concentration at three inches of depth (near the rebar) is 2.4 lb/yd³ on the control side and 1.3 lb/yd³ on the MCI treated side. Also, the concrete still has a very high pH (about 12), indicating that it alkaline after 14 years of service life. When concrete is at this high of a pH, we know that carbonation is low and also that it creates a passive environment for corrosion.

The conditions of the concrete on the bridge were wet and cool, taken after a period of cold weather and rain for several days. This was because the readings with the Gecor 6 are more accurate regarding the corrosion rate in wet conditions. While the LSDC has done an excellent job at protecting the reinforcing steel from corrosion on both sides of the bridge, a difference between readings taken on the control and MCI treated sides can still be seen:

2. The average copper/copper sulfate potential value for rebar embedded in the MCI treated overlay was 208.9 mV, 22% less than the average for the control overlay at 269.6 mV. Also, 12% of the readings taken on the control overlay indicated corrosion, while none did on the MCI treated side.
3. Rebar on the MCI 2000 treated side had very low corrosion currents, an average of 0.013 m A/cm², approximately 40% below readings taken on control side (average of 0.022 m A/cm²).
4. The LSDC resistivity was very high for both the MCI 2000 treated side and control sections on the bridge. However, the MCI-treated side had an average resistivity of 373.9 kW · cm, 10% higher than that of the control at 338.3 kW · cm.
5. The gas chromatography analysis indicated the presence of amines (indicating the presence of MCI) on the MCI-treated side, while there was no indication of amines on the control treated side (Figures 10 and 11).

For: Cortec Corporation

Date: January 25, 2001

cc: Boris Miksic

Anna Vignetti

Alla Furman

Art Ahlbrecht

Christophe Chandler

Dick Stehly

REFERENCES

1. Andrade, C., Alonso, M.C., Gonzalez, J.A., "An Initial Effort to Use the Corrosion Rate Measurements for Estimating Rebar Durability", Corrosion Rates of Steel in Concrete, ASTM STP 1065.
2. Andrade, C., Fullea, J., Alonso, C., "The Use of the Graph Corrosion Rate-Resistivity in the Measurement of Corrosion Current", International Workshop MESINA, Instituto Eduardo Torroja, Madrid, Spain, 1999.
3. Kosmatka, Steven and Panarese, William, Design and Control of Concrete Mixtures Thirteenth Edition, Portland Cement Association, Skokie, Illinois, 1994 rev. pp. 72.
4. Thomas, M.D.A. and Bentz, E.C., Life-365 Computer Program for Predicting the Service Life and Life-Cycle Costs of Reinforced Concrete Exposed to Chlorides. University of Toronto, Toronto, Canada, April 21, 2000, pp. 9.

Chloride-Ion Content (lbs/yd3), AASHTO: T260 Procedure C 1991 & 1992 Data: Virginia Tech; 2000 Data: American Engineering Testing Eastbound, LSDC Control Westbound, LSDC/Cortec MCI 2000 1991 1992 2000 1991 1992 2000 Depth (in) Mean _x Mean _x Mean _x Mean _x Mean _x Mean _x 0.5 10.7 2.4 13.7 3.4 NA NA 6.0 2.2 9.6 1.8 NA NA 1.0 4.7 2.3 5.4 2.9 17.2 1.8 1.1 1.3 3.3 1.5 11.7 0.9 1.5 2.2 1.8 3.0 1.8 NA NA 0.0 0.1 1.2 1.3 NA NA 2.0 2.7 0.9 4.0 2.0 6.2 0.6 0.0 0.0 1.0 1.1 1.6 0.5 2.5 2.3 1.4 3.0 2.0 NA NA 0.4 0.6 1.3 0.9 NA NA 3.0 1.4 1.6 1.9 1.8 2.4 0.4 1.0 0.9 2.5 1.0 1.3 0.2 3.5 0.3 0.4 NA NA NA NA 0.9 0.8 NA NA NA NA