Comparative Corrosion Testing Cui Krauss

Comparative Corrosion Testing and Analysis of
MMFX 2 Rebars for Reinforced Concrete Applications

Northbrook, Illinois
Final Report
19 February 2008
WJE No. 2003.0707.0
Prepared for:
MMFX Technologies Corporation
Prepared by:
Wiss, Janney, Elstner Associates, Inc.
Comparative Corrosion Testing and Analysis of

Fushuang Cui
Project Engineer
Paul Krauss
Project Manager
Final Report
19 February 2008
WJE No. 2003.0707.0
Prepared for:
MMFX Technologies Corporation
2 Corporate Park, Suite 102
Irvine, CA 92606
Prepared by:
Wiss, Janney, Elstner Associates, Inc.
330 Pfingsten Road
Northbrook, Illinois 60062
847.272.7400 tel | 847.291.5189 fax
TABLE OF CONTENTS
INTRODUCTION ........................................................................................................................................ 1
MODIFIED ASTM G109 Test ..................................................................................................................... 1
Background ............................................................................................................................................ 1
Test Procedure ........................................................................................................................................ 1
Autopsy and Weight Loss Measurement ................................................................................................ 2
Results and Discussion ........................................................................................................................... 2
Phase I: 20-week exposure............................................................................................................... 2
Phase II: Continuation of G109 Testing .......................................................................................... 3
20-WEEK MODIFIED SOUTHERN EXPOSURE TEST ......................................................................... 13
Background .......................................................................................................................................... 13
Test Procedure ...................................................................................................................................... 13
Results and Discussion ......................................................................................................................... 14
DISCUSSION ............................................................................................................................................. 20
SUMMARY AND CONCLUSIONS ......................................................................................................... 24
Comparative Corrosion Testing and Analysis of MMFX 2 Rebars for

Reinforced Concrete Applications
INTRODUCTION
Wiss, Janney, Elstner Associates, Inc. (WJE) was contacted by MMFX Technologies to conduct a
laboratory study of the corrosion resistance of its MMFX 2 reinforcing bars. The objectives of this study
were to compare the corrosion performance of MMFX 2 bars to A615 carbon steel bars and Type 304
stainless bars in accelerated laboratory tests.
Two laboratory tests, ASTM G109 and time-to-corrosioni (also known as Southern Exposure) tests were
included in this study. These two methods have been widely used to evaluate the corrosion resistance of
reinforcing bars and are generally accepted by Federal and state DOT agencies. For comparison purposes,
ASTM A615 carbon steel reinforcement (black bar) and Type 304 stainless steel reinforcement were
tested under identical exposure conditions. Some MMFX 2 bar samples had mill scale and others did not.
In order to meet the original project requirements of MMFX of completing the test program in five
months, the standard test methods were slightly modified for this study. The ASTM G109 test, however,
was extended as no bars started to corrode at the end of the initial 20 weeks (5 months) of testing.
MODIFIED ASTM G109 TEST

Background
Originally, the ASTM G109 Standard Test Method for Determining Effects of Chemical Admixtures on
Corrosion of Embedded Steel Reinforcement in Concrete Exposed to Chloride Environments test method
was developed to evaluate the effectiveness of admixed corrosion inhibitors in two-mat laboratory
concrete specimens exposed to a salt solution.ii The ASTM G109-99 test method employs a moderately
good quality concrete, a monthly wet/dry cycle using mild salt exposure (3% NaCl solution) and a 100ohm resistor inserted between top and bottom mats for macro-cell current measurement. These test
parameters can result in slow corrosion initiation at the top mat bar and often the test duration must be
extended to over one year. The presence of a 100-ohm resistor also introduces a slight obstruction to the
macro-cell current flow and our previous laboratory data showed about a 13 percent reduction in
measured current compared to samples having a smaller 1-ohm resistor. The moderate exposure
conditions and slow corrosion response of the test specimens is believed by some to be more
representative of field conditions compared to other more accelerated tests.
Test Procedure
For this test program, we employed a modified G109 test specimen by replacing the 100-ohm resistor
with a direct (banana plug) electrical connection. This connection was periodically disconnected in order
to measure the macro-cell current directly using a zero resistance ammeter (ZRA). This provided a
continuous and resistance-free connection between the top and bottom bars during the test. A schematic
Comparative Correlative Corrosion Testing and Analysis of

19 February 2008
Page 2
of the test specimen is shown in Figure 1 and photographs of the general testing are shown in Appendix
A. The remaining test procedure was the same as the ASTM G-109 test method, that is, uncracked
concrete test blocks, exposure to 50 5 percent relative humidity (RH), and a 2-week wetting and 2-week
drying cycle with 3% salt solution.
Examined in this study were four types of rebar materials, which included MMFX 2 (with and without
mill scale), ASTM A615 with mill scale (control), and Type 304 stainless steel. The four different rebar
types were embedded in four specimens each. All bottom mat bars were black A615 rebars. The opencircuit potential (corrosion potential of top bar only, OCP), short-circuit potential (coupled potential of the
top and bottom bars, SCP), and macro-cell current data were collected at regular intervals. The test states
that corrosion is sufficient for visual evaluations when a macrocell threshold current of 10 A is reached.
Prior to placing the bars in concrete, the initial weight of each bar was measured using a digital balance to
allow for a weight loss measurements at the end of the testing. Macro-cell current was measured by
insertion of a zero resistance ammeter (ZRA) between rebar mats. A saturated calomel reference
electrode (SCE) was used to measure the half-cell potentials.
Autopsy and Weight Loss Measurement

At the end of the initial 20-week testing (Phase I), none of the specimens exhibited a threshold current of
10 A. For each type of specimens, only the specimen with the highest macro-cell current was autopsied
while further testing of the remaining specimens continued. During the extended testing (Phase II), the
specimens were autopsied as soon as a 10 A threshold current was exceeded. The bar condition was
examined and an 1/8 in. slice of concrete immediately above the top bar was sampled for acid-soluble
chloride analysis according to ASTM C1152 Standard Test Method for Acid-Soluble Chloride in Mortar
and Concrete. The bar imprint was ground off the concrete sample before testing to minimize
interference from iron oxide residue to the chloride analysis.
The weight loss of the bars was determined to estimate the rate of corrosion for the extracted reinforcing
bars. Prior to weight measurements, the corrosion products formed on the bars were removed with a wire
brush and then the bars were chemically cleaned according to ASTM G1, "Standard Practice for
Preparing, Cleaning, and Evaluating Corrosion Test Specimens". The final weight of individual bars was
measured and the weight loss of each bar was determined by subtracting the final weight of the bar from
the initial value.
Results and Discussion

Phase I: 20-week exposure
The first 20-week exposure results of the modified G-109 testing are presented in Figures 2 to 4. The data
indicates that corrosion activity had not yet begun in these specimens. The OCP readings generally
remained high and larger (more positive) than -0.1 V. This is well above the usual ~-0.28 V (-0.350 V vs
CSE) upper limit for active corrosion in reinforced concrete. The SCP is also high and well above levels
that normally indicate the presence of active corrosion.
The macro-cell current data, shown in Figure 3, supports the findings of the potential measurements. The
low levels of current combined with stable values indicate that the onset of corrosion had not begun. The
onset of corrosion will always be accompanied by a significant increase in the measured current.

19 February 2008
Page 3
Due to the lack of corrosion activity, only one specimen of each bar type was removed from the testing
regime for autopsy after 20 weeks. The bars with the highest macrocell readings of each group were
selected (specimens BM-3, S-3, M-2 and MM-4). The bars were cleaned and weighed. The weight loss
data is shown in Figure 4. The black A615 bar showed some slight weight loss while the pristine
appearance and the lack of any significant weight loss confirm that corrosion had not yet begun in the
other test samples. Concrete samples were taken from each of the autopsied specimens, and Table 1
shows the chloride concentration at the bar level.
Table 1 - Chloride Contents at 20 weeks
Sample
Chloride, wt% (lbs/cy)
Black bar
0.067 (2.6)
MMFX without scale
0.087 (3.3)
MMFX with scale
0.068 (2.6)
Stainless steel
0.115 (4.4)
Phase II: Continuation of G109 Testing

The testing of the remaining modified G109 specimens (three for each bar type) was continued after the
initial twenty weeks of testing. As shown in Figure 5, the specimens were subjected to normal ponding
cycles of 2-week wetting and 2-week drying, except for three periods when the samples were maintained
at the 50% RH but were not ponded.
Figure 5 shows the open-circuit potential profiles of the specimens. A significant drop in the potential
profile and corresponding increase in the macrocell corrosion current indicates initiation of corrosion
activity of the top bar. As shown, the black bar specimens (BM) started to corrode on day 434, 664, and
758. MMFX bars with mill scale (MM) started to corrode on day 1169, 1266, and 1352. Two MMFX bars
without mill scale started to corrode on day 1310 and 1575, while one specimen (M1) remained passive
through 1625-days (although the M1 bar had a trace of corrosion product when autopsied). All stainless
steel 304 specimens remained passive through the total 1625-day test period.
Figure 6 shows the macrocell current measured between the top bar and the bottom bars of each
specimen. Corresponding to the potential drop events described above, macrocell currents higher than 10
A were registered confirming the onset of corrosion activity.
Table 2 shows the time of activation (corrosion initiation) for specimens reinforced with black bars and
MMFX bars (with and without mill scale). As noted, one specimen with MMFX without mill scale (M-1)
remained passive through the 1625-day test period, but showed a trace of corrosion on the bar when
autopsied. All three specimens with stainless steel 304 bars remained passive throughout the testing
period. The table also shows the chloride concentrations at the top bar level after the test exposure was
terminated. Since the exposure of the black bar control specimens was continued for additional time after
corrosion initiation, the chloride concentrations obtained after exposure termination were higher than the
actual chloride concentrations at which corrosion started. Therefore, the chloride threshold for the black
bar was estimated based on the well-known solution to Ficks Law, using 1 in. concrete cover, a corrosion
initiation time of 434 days (BM-2), and a surface chloride concentration of 0.75 wt% (29 lbs/cu. yd.).
This surface chloride concentration was estimated from a regression analysis of a detailed chloride profile

19 February 2008
Page 4
of Specimen MM1 (shown in Table 3). The lowest level of chloride for each test condition at corrosion
initiation was conservatively assumed as the chloride threshold for MMFX reinforcement.
Table 2 - Bar-top chloride contents determined after exposure termination
Type
A 615 Black
MMFX-2
without
mill scale
MMFX-2
with
mill scale
Stainless
Steel 304
Sample
Estimated
corrosion
initiation time
(days)
Total
exposure time
(days)
BM 1
758
924
BM 2
434
924
BM 4
664
939
M1
1625
1625
M2
1575
1625
M3
1310
1310
MM 1
1266
1266
MM 2
1169
1169
MM 3
1352
1352
S1
--
1625
S2
--
1625
S4
--
1625
Chloride,
wt%
(lbs/cy)
0.218
(8.4)
0.213
(8.2)
0.209
(8.0)
0.450*
(17.3)
0.405
(15.6)
0.328
(12.6)
0.405
(15.6)
0.297
(11.4)
0.405
(15.6)
0.555*
(21.4)
Not
tested*
0.539*
(20.7)
Estimated
chloride
threshold, wt%
(lbs/cy)
Threshold ratio
0.10 (3.8)**
0.33 (12.7)
3.3
0.30 (11.5)
3.0
>0.54 (>20.8)
>5.4
* Did not reach chloride current threshold values

**Ficks law estimate
Table 3 Chloride profile of Specimen MM-1 after 1266 days of exposure

Depth (in.)
Chloride, wt% (lbs/CY)
1/8 - 1/4
0.640 (24.6)
7/8 - 1
0.405 (15.6)
2 - 2 1/8
0.002 (0.08)

19 February 2008
Page 5
Using the G109 test procedure and as shown in Table 2, the chloride thresholds for black bars, MMFX
bars with mill scale and without mill scale are approximately 0.10, 0.30, and 0.33 wt% (or 3.8, 11.5, and
12.7 lbs/cy), respectively. The removal of mill scale appears to have only a minor beneficial effect on the
corrosion resistance of MMFX bars.
Chloride threshold values determined from accelerated laboratory test programs of this kind may
overestimate the actual threshold for corrosion in bridge decks, since deck structures often accumulate
chlorides more slowly, only see intermittent deicer applications, and are often at low temperatures. It is
common to use the ratio of the threshold values when modeling and comparing the service life of bar
alternatives. While varying with many factors such as cementitious material contents, temperature and
exposure conditions, a value of 0.03 wt% (1.2 lbs/cu. yd.) is commonly used as the chloride threshold for
modeling service life of black bars in bridge decks.iii Based on findings from this program, it is
conservatively projected that the field chloride threshold of MMFX bars is three times that of the black
A615 steel or 0.09 wt% (3.5 lbs/cu. yd.). Stainless 304 bars resisted chloride levels over five times the
black bars without corrosion. It is conservatively projected that the chloride threshold of the stainless 304
bars is greater or much greater than 0.16 wt% (6.2 lbs/cu. yd.).
Table 4 summarizes the weight loss data of autopsied specimens. Black bar specimens were not
immediately autopsied after the first sign of corrosion activation, while MMFX bar specimens were
autopsied immediately after observation of corrosion activation. Consequently, weight loss for black bars
is generally higher than MMFX bars.
Table 4 - Weight loss of autopsied top bars
Sample
BM1
BM2
BM4
BM3*
M3
M2*
MM1
MM2
MM3
MM4*
S3*
Before exposure
(g)
314.8
317.4
311.6
317.6
319.0
339.8
334.9
344.2
317.7
319.6
291.7
After exposure
(g)
314.3
313.0
310.9
317.0
318.7
339.8
334.5
344.0
317.7
319.5
291.8
Weight loss (g)

0.5
4.4
0.7
0.6
0.3
0
0.4
0.2
0
0.1
0
Weight loss
(wt%)
0.16
1.39
0.22
0.19
0.09
0
0.12
0.06
0
0.03
0
*20-week Phase 1 results

The testing was stopped after 1625 days, and all remaining samples were autopsied. A trace of corrosion
was found on sample M-1 during the autopsy, even though the corrosion current remained well below 10
A. Active corrosion product was noted on sample M-4, as expected based on the high measured
corrosion current. None of the stainless steel bars showed any signs of corrosion, and the bars appeared
pristine. Photos 1-6 show the typical condition of the test bars after exposure.

19 February 2008
Page 6
Ponding well
with lid
3% NaCl
solution
2-layer heat
shrink tube
c*
ZRA
6.0 in.
c*
6.0 in.
4.5 in.
8.0 in.
*Clear cover = 1.0 in.
11.0 in.
[End View]
15.0 in.
[Side View]
Figure 1. Schematic of modified G-109 test specimen.

Change of mean open-circuit potential of G-109 top mat bars with time
A615 with mill scale
Stainless Steel
MMFX without mill scale
MMFX with mill scale
0.050
Potential (V vs. SCE)
0.000
-0.050
-0.100
-0.150
-0.200
-0.250
-7
21
35
49
63
77
91
105
119
133
Elapsed time (day)
Figure 2. Mean OCP data of G-109 test specimens for the first 20 weeks.
147

19 February 2008
Page 7
Change of mean macro-cell current for G-109 specimens with time

Stainless Steel
1.2
1.0
Current (uA)
0.8
0.6
0.4
0.2
0.0
-0.2
-7
21
35
49
63
77
91
105
119
133
147
Elapsed time (day)
Figure 3. Mean macro-cell current data of G-109 test specimens for the first 20 weeks.
Mean Final Percent Weight loss G-109
1.600
1.400
Percent Weight Loss (%)
1.200
1.000
0.800
0.600
0.400
0.189
0.200
-0.034
0.000
2
Stainless Steel
3
MMFX without Mill Scale
0.031
0.000
1
A615 with Mill Scale
4
MMFX with Mill Scale
-0.200
Bar Type
Figure 4. Mean percent weight loss of G-109 test specimens after 20 weeks of exposure.

19 February 2008
Page 8
146
251
477
924
1520 1625
0.1
BM1
Open Circuit Potential Vsce
BM2
BM4
-0.1
S-1
S-2
-0.2
S-4
Test condition
ponding cycle
-0.3
M-1
M-3
dry
M-4
-0.4
MM1
MM2
MM3
Corrosion initiation time (days)

BM-1 BM-2 BM-4 MM-1 MM-2 MM-3 M-3 M-4
758
434
664
1266 1169 1352 1310 1570
-0.5
M-1 and all SS specimens remained passive

-0.6
0
200
400
600
800
1000
Elapsed Time (days)
1200
1400
1600
1800
Figure 5. Open circuit potential G-109 test specimens. (BM: black bar with mill scale;
S: stainless steel; M: MMFX 2 without mill scale; MM: MMFX 2 with mill scale)
0
146
251
477
924
1520
1625
120
Corrosion initiation time (days)
BM-1 BM-2 BM-4 MM-1 MM-2 MM-3 M-3 M-4
758
434
664
1266 1169 1352 1310 1575
BM1
M-1 and all SS 304 specimens remained passive
BM4
BM2
Macrocell current ( A)
100
Test condition
ponding cycle
80
S-1
S-2
dry
S-4
60
M-1
M-3
M-4
40
MM-1
MM-2
20
MM-3
0
0
200
400
600
800
1000
1200
1400
1600
1800
Elapsed Time (days)
Figure.6. Mat-to-mat macro-cell current;(BM: black bar with mill scale; S: stainless
steel; M: MMFX 2 without mill scale; MM: MMFX 2 with mill scale

19 February 2008
Page 9
Photo 1. Black bar before cleaning: BM2 has extensive corrosion; moderate rust on BM1 and
BM4.
Photo 2. Black bars after cleaning: superficial attack on BM1 and BMI; significant attack on
BM2.
Photo 3. Sample MM2 corroded at a rib, and the corrosion product is black and liquid after
1169 days.

19 February 2008
Page 10
Photo 4. Sample M1 showing small trace of corrosion on bar after 1625 days.

19 February 2008
Page 11
Photo 5. Sample M4 showing minor corrosion products that are black and liquid after 1625
days.

19 February 2008
Page 12
Photo 6. Stainless 304 bars samples showing no corrosion after 1625 days.

19 February 2008
Page 13
20-WEEK MODIFIED SOUTHERN EXPOSURE TEST

Background
The macro-cell current test slab program, as reported in the 1983 FHWA-RD-83-012 report entitled
Time to Corrosion of Reinforcing Steel in Concrete Slabs, Vol. 5 has become the most common method
used to evaluate the corrosion resistance of reinforcement in concrete. A majority of the recent corrosion
studies have used specimens similar to the macro-cell slab specimens. Landmark research projects
performed by WJE using these macro-cell test slabs include the 1987 FHWA-RD-86-193 report entitled
Protective Systems for New Prestressed and Substructure Concrete and the 5-year (1993-1998) FHWA
project (FHWA-RD-98-153) entitled Corrosion Evaluation of Epoxy-Coated, Metallic-Clad, and Solid
Metallic Reinforcing Bars in Concrete.
This accelerated test procedure has become a standard in the industry for testing high performance
reinforcing steels in concrete. The following wetting and drying cycles were employed for the FHWA
RD-98-153 study:
Three-day drying at 100 oF (38 oC) and 60-80 percent RH followed by four-day ponding
under a 15 percent NaCl solution at 60-80 oF (16-27 oC) for 12 weeks.
Continuous ponding under a 15 percent NaCl solution at 60-80 oF (16-27 oC) and 60-80
RH for another 12 weeks.
The cyclic ponding and continuous ponding cycles were repeated four times for a total
exposure period of 96 weeks.
The 15 percent salt solution was chosen to represent the high salt concentrations occurring on in-land
bridge structures from deicing salts. The long continuous ponding period was utilized to simulate a
sustained period of submersion or long periods of high concrete moisture common in marine structures in
winter months. The concrete slabs used in the FHWA study measured 12 x 12 x 7 in. and contained two
layers of 11.5-in. long and 5/8-in. diameter reinforcement. The top mat acts as a macro-anode and
contained either two straight or bent reinforcing bars, while the bottom mat was a macro-cathode that
contained four straight reinforcing bars. One top mat bar was connected to two bottom mat bars through
a 10 resistor. A clear cover of 1.0 in. was used in all specimens.
Test Procedure
WJE used a modified geometry of the Southern Exposure (time-to-corrosion) test slabs without altering
the test basics in order to accelerate the test to accommodate the desired 20-week test period. Figure 7
shows the schematic of the test slab and photographs of the general testing are shown in Appendix A.
First, the clear cover to the top mat bars was reduced from 1.0 in. to 0.5 in. in order to reduce the time-tocorrosion. Secondly, the number of rebars in the bottom mat was doubled, from two bottom bars
(cathode) to four bottom bars for each single top bar (see Figure 7). This change was introduced to
increase the corrosion activity of the top bar by increasing the cathodic area, where oxygen is reduced, of
the corrosion cell. Thirdly, the thickness of the test slab was reduced from 7 in. to 5 in. and consequently
the clear spacing between top and bottom mats was reduced from 3.75 in. to 2.25 in. This lowers the
concrete resistance between mats and improves the flow of macro-cell current. Lastly, the width of the
test slab was increased from 12 in. to 16 in. to accommodate a corrosion rate measurement probe (7.5 in.
diameter) over each of the top bars. A Gamry Electrochemical Testing Instrument was used to measure

19 February 2008
Page 14
the corrosion rates of the test bars. The electrochemical testing software (Framework) uses the principle
of linear polarization for a 3-electrode system.
Uncracked specimens were made containing each of the four reinforcing bar materials: A615 with mill
scale, Type 304 stainless steel, and MMFX 2 with and without mill scale. Three specimens of the A615
and stainless steel bar types and six specimens of each of the MMFX bar types were fabricated. For
weight loss measurement purposes, the initial weight of each bar was measured using a digital balance
prior to casting.
A multi-channel Hewlett Packard data logger was used to collect short-circuit potential (SCP) and macrocell current several times each day. Macro-cell measurements were made with the data logger without the
need of a circuit resistor. In order for the data logger to measure the SCP of a bar while the bar was
connected to the bottom mat, a pseudo-reference electrode was cast into the concrete next to each top bar.
This is shown in Figure 7. Baseline potentials of each pseudo-reference electrode were measured soon
after casting with respect to an external saturated calomel reference electrode (SCE) and periodic
calibration checks were made during the exposure testing. The pseudo-reference electrode baseline and
subsequent calibration data were used to convert the SCP data collected with the pseudo-reference
electrode to potential values with respect to the SCE. The SCP data provided information about the
electrochemical behavior of the bars in a coupled condition, indicating either anode control or cathode
control.
In addition, open-circuit potential (OCP) of the top bar only was periodically measured with respect to an
external SCE half-cell after disconnecting the circuit between top and bottom mats. The OCP data
provides useful information about the electrochemical state of the bars in a depolarized (freely corroding)
condition. Corrosion rate measurements were made using the Gamry Frameworks system at the
beginning and at the end of the testing.
Mat-to-mat AC resistance was measured periodically and can be an indicator of the concrete condition
and of physical deterioration around the bar/concrete interface. Normally as concrete ages, its AC
resistance steadily increases. However, when a rebar is severely corroding and the corrosion progresses
to cause microscopic cracks and delaminations at the bar interface, the AC resistance can change rapidly.
Autopsy of the test slabs was performed at the end of the 20-week testing. Powder samples were
collected by drilling along the top bar/concrete interface of selected slabs, and they were analyzed for
acid-soluble chloride according to ASTM C1152 Standard Test Method for Acid-Soluble Chloride in
Mortar and Concrete.
Weight loss was determined to estimate actual rate of corrosion (metal loss) for the extracted reinforcing
bars. Prior to weight measurements, the corrosion products formed on the bars were removed with a wire
brush, and then the bars were chemically cleaned according to the chemical method specified in ASTM
G-1, "Standard Practice for Preparing, Cleaning, and Evaluating Corrosion Test Specimens." The final
weight of individual bars was measured and weight loss of each bar was determined by subtracting the
final weight of the bar from the initial value. Photo documentation of the typical bar conditions as they
were removed from the slabs and after cleaning are shown in Appendix B.
Results and Discussion

The results of the modified Southern Exposure testing are presented in Figures 8-14. All specimens,
except the stainless steel 304 bars, showed considerable corrosion activity during the testing period. The

19 February 2008
Page 15
OCP data presented in Figure 8 shows that the A615, the MMFX with mill scale, and MMFX without
mill scale bars are all in a similar potential range, with the MMFX bars only slightly lagging the A615
bars. The values indicated active corrosion after about 14 to 28 days and potentials remained below -350
mV after approximately week 8 (day 56). At about this same time, an increase in macro-cell current data
in Figure 10 was seen indicating the onset of corrosion.
The SCP and Macro-Cell current data (Figures 10-11) show that the MMFX bar types behaved similarly,
and their corrosion activities lie somewhere between that of the A615 and stainless steel. The MMFX
bars (both with and without mill scale) had macrocell corrosion currents of about 1/3 less than the A615
black bars with a final 20-week macro-cell current values around 0.27 mA, compared to approximately
0.44 mA for the A615 black bars. The stainless steel bars remained passive and close to 0.00 mA for the
duration of the testing. Due to an equipment or connection failure, the SCP and current data for the 18th
week were not collected for the MMFX samples and, as a result, there is a discontinuity in the data.
The mean corrosion rates, measured using the Gamry equipment, at the end of the 20-week test, as seen in
Figure 12, of both MMFX bar types was similar. The MMFX with mill scale specimens had a final mean
corrosion rate of 2.08 A/cm2 and the MMFX without mill scale specimens had a final mean corrosion
rate of 1.89 A/cm2. These values are approximately 66% and 70% lower, respectively, than the final
mean corrosion rate for the A615 black bar specimens of 6.27 A/cm2.
The measured weight loss of the test bars, shown in Figure 13, was determined after the autopsied bars
were cleaned, weighed, and compared to their initial weights. The MMFX with mill scale specimens had
a mean percent weight loss of 0.44 % and the MMFX without mill scale specimens had a mean percent
weight loss of 0.28 %. These values are much lower than the mean percent weight loss for the A615
specimens of 1.35%. The measured stainless 304 steel bar weight loss was zero.
Table 5 lists the estimated chloride content at bar level after the 20-week southern exposure. The chloride
contents at the bar level were high and ranged from 0.38 to 0.60 wt%. (14.7 to 23.0 lbs/cu. yd.) As
described, all A615 black bars and MMFX bars were corroding at the end of the 20-week test and none of
the stainless 304 bars were corroding.
Table 5 - Chloride Contents at bar level after 20-weeks SE

Sample
Acid-Soluble Chloride,
% by mass of sample
(lbs/cy)
Black bar No.1
0.408 (15.7)
Black bar No.2
0.513 (19.7)
MMFX without scale No.1
0.506 (19.5)
MMFX without scale No.2
0.453 (17.4)
MMFX with scale No.1
0.513 (19.7)
MMFX with scale No.2
0.383 (14.7)
Stainless steel No.1
0.597 (23.0)
Stainless steel No.2
0.546 (21.0)

19 February 2008
Page 16
The correlation of the measured final corrosion rate and the determined weight loss is presented in Figure
14. The general trend correlates well, with high corrosion rates corresponding to high weight loss
percentages. A linear function gives a fairly good fit, with an R2 value of 0.785.
The Southern Exposure test, as modified for this test program, was a very aggressive and accelerated test.
The MMFX 2 bars with and without mill scale performed better than the A615 bars but not nearly as well
as stainless 304 bars with respect to macro-cell corrosion, final corrosion rates, and weight loss. The data
does not indicate a significant difference in performance between the MMFX 2 bars with or without mill
scale.
Ponding well with lid
15% NaCl
solution
Pseudo-reference
electrode (Graphite)
Top mat
(Anode)
Data logger
Bottom mat
(Cathode)
A connecting bar
across bottom mat
11.0 in.
[Side View]
12.0 in.
2-layer heat
shrink tube
0.5 in.
Clear cover
[CrossSectional
View]
5.0 in.
1.0 in.
Clear cover
4.0 in.
8.0 in.
4.0 in.
16.0 in.
Figure 7. Schematic of modified southern exposure test specimen

19 February 2008
Page 17
Change of mean open-circuit potential of Southern Exposure top mat bars with time
Stainless Steel
0.100
Open-circuit potential (V vs. SCE)
0.000
-0.100
-0.200
-0.300
-0.400
-0.500
-14
14
28
42
56
70
84
98
112
126
140
Elapsed time (Day)
Figure 8. Mean open-circuit (OCP) half-cell potential data of Southern Exposure test
specimens
Change of mean AC resistance for Southern Exposure specimens with time

Stainless Steel
240
220
Resistance (ohm)
200
180
160
140
120
100
80
-14
14
28
42
56
70
84
98
112
126
Elapsed time (day)
Figure 9. Mean AC resistance data of Southern Exposure test specimens
140

19 February 2008
Page 18
Average macro-cell current

Stainless Steel
0.7
Macro-cell current (mA)
0.6
0.5
0.4
0.3
0.2
0.1
0
0
14
21
28
35
42
49
56
63
70
77
84
91
98 105 112 119 126 133 140 147 154
Elapsed time (day)
Figure 10. Mean macro-cell current data of Southern Exposure test specimens
Average short-circuit potential
Stainless Steel
0
0
14
21
28
35
42
49
56
63
70
77
84
91
98 105 112 119 126 133 140 147 154
-5
Short-circuit potential (mV)
-10
-15
-20
-25
-30
-35
-40
-45
Elapsed time (day)
Figure 11. Mean short-circuit (SCP) half-cell potential data of Southern Exposure test
specimens

19 February 2008
Page 19
Mean Final Corrosion Rate Data

10.000
Corrosion rate (uA/cm 2)
8.000
6.265
6.000
4.000
1.888
2.000
2.083
0.733
0.000
Stainless Steel
Bar Type
Figure 12. Mean corrosion rate (Icorr) data of Southern Exposure test specimens (20weeks).
Mean Final Percent Weight loss
1.6
1.4
1.352722289
1.2
0.8
0.6
0.436783561
0.4
0.283394724
0.2
-0.034408659
0
1
A615 with Mill Scale
2
Stainless Steel
3
MMFX without Mill Scale
4
MMFX with Mill Scale
-0.2
Bar Type
Figure 13. Mean percent weight loss of Southern Exposure test specimens (20-weeks)

19 February 2008
Page 20
Final Corrosion Rate vs. % Weight Loss for A615 and MMFX Test Specimens
10
Final Corrosion Rate ( A/cm2)
6
y = 5.0001x
R2 = 0.7851
0
0
0.2
0.4
0.6
0.8
1.2
1.4
1.6
Figure 14. Correlation of the final corrosion rate and percent weight loss for A615 and
MMFX samples
DISCUSSION
Carbon steel reinforcing bars are usually passive in new concrete because concrete provides a high pH
(approximately 13) medium and also acts as a physical barrier isolating the steel from aggressive
chemicals in the environment. However, steel will begin to corrode once those aggressive chemical
species (e.g., chloride ions) penetrate through the concrete cover and reach a certain concentration
(commonly called the chloride threshold, CT, in the case of chloride ions). The time required for chloride
to reach CT is called corrosion initiation time (t1), since it is assumed that the corrosion will initiate once
the threshold is exceeded. The corrosion products of steel in concrete occupy much more volume than the
original steel, and this introduces significant stresses in the surrounding concrete. As corrosion continues,
distress such as cracks or delamination in the concrete occurs and costly maintenance is required. The
time required for corrosion products to induce concrete distresses is called the corrosion propagation
time (t2), and the sum of t1 and t2 defines the total amount of time before corrosion-related damage is
apparent at a given location. Figure 15 illustrates this corrosion sequence. The service life of a structure is
defined as the time until the amount of damaged area exceeds some level of acceptability.
Corrosion of reinforced concrete is often caused by chloride ingress due to either usage of de-icing salts
or exposure to marine saltwater environments. Chloride diffusion in concrete, driven by a concentration
gradient, is described by Ficks Second Law of Diffusion:
dC
dt
d 2C
dx2
(1)

19 February 2008
Page 21
where C is the chloride concentration at a depth of x from the concrete surface at time t, and D is the
effective chloride diffusion coefficient.
If the surface chloride concentration Cs and D are assumed to be constants, the concentration C(x, t) at
depth of x and time t is given by:
C ( x, t ) Cs (Cs C0 ) erf (
x
)
D t
(2)
where erf() is the Gaussian error function, and C0 is the background or original chloride concentration in
concrete.
Reinforcing will start to corrode when the bar-level chloride concentration reaches or surpasses its
chloride threshold CT. The corrosion initiation time t1 can be calculated from:
CT
Cs
(C s
C 0 ) erf (
cc
)
D t1
(3)
where cc is the concrete cover.

In order to assess the service life extension offered by MMFX corrosion-resistant bars, a statistical model
developed by Sags et al.iv was adapted to estimate the service life of reinforced structures representative
of a marine environment (a pile exposed to splash of brackish water) and a northern bridge deck
environment (a deck exposed to significant deicing salts). In this model, concrete structures are virtually
divided into finite elements of equal size. Each element is randomly assigned with a unique value of
surface chloride concentration, concrete cover and diffusion coefficient based on assumed normal
distribution functions for these three parameters determined from data reported in the literature and
measured from actual structures. Using the Ficks Law solution, the model calculates t1, the time for the
chloride concentration at the steel to exceed the chloride threshold, for each element. The propagation
time, t2, is assumed to be a constant 3 years for both bar types. The MMFX bars may have a slightly
longer propagation time than A615 bars based on the SE tests results. The MMFX bars tested in the SE
test had macrocell corrosion currents about 1/3 less than the A615 bars and about 2/3 less weight loss.
Assuming that the corrosion of the MMFX bars occurs at 1/3 to 2/3 the rate of A615 bars, this would
potentially add an additional 1.5 to 9 years to the propagation time and service life of the MMFX bar
structures.
Combining t1 and t2, damage functions (percentage of area deteriorated versus time) for the MMFX and
black bar conditions can be generated and used to predict the service life of concrete structures. The
chloride threshold for MMFX and black bars are assumed to be 0.09 wt% and 0.03 wt% (3.5 and 1.2
lbs/cu. yd.), respectively, based on the values estimated from the modified G-109 tests performed in this
study.
Case 1 is for a marine concrete pile, with inputs shown in Table 6, which included actual data collected
from a structure along the Intracoastal Waterway in Georgiav.

19 February 2008
Page 22
Table 6. Input values for modeling a marine pile
Concrete cover (in.)

Chloride diffusion coefficient (in2/y)
Surface chloride concentration (wt%)
Background chloride (wt%)
Average
3.5
0.045
0.471 (18.1 lbs/CY)
0.002 (0.08 lbs/CY)
Standard deviation
0.3
0.025
0.148 (5.7 lbs/CY)
0
Figure 16 shows the projected damage functions for the pile reinforced with MMFX and black bars.
Assuming 10% damage as the end of service life, the model projected that MMFX bars would have a
service life of 43 years, while black bars have a service life of 24 years.
Case 2 is for a northern bridge deck using data, shown in Table 7, collected on decks in Iowa vi.
Table 7. Input values for modeling of a northern bridge deck
Concrete cover (in.)

Chloride diffusion coefficient (in2/y)
Surface chloride concentration (wt%)
Background chloride (wt%)
Average
2.75
0.05
0.350 (13.5 lbs/CY)
0
Standard deviation
0.45
0.038
0.090 (3.5 lbs/CY)
0
Figure 17 shows the projected damage functions for a northern bridge deck reinforced with MMFX and
black bars. Assuming 10% damage as the end of service life, the model projected that MMFX bars would
have a service life of 27 years, while black bars have a service life of 14 years.
The modeling predicts that the MMFX bars will extend the service life of the marine pile by 1.8 times and
the northern bridge deck by 1.9 times. Accounting for the reduced corrosion rates measured for the
MMFX bars compared to A615 bars after corrosion initiates (during the propagation period), this service
life extension could be over two times that of the A615 steel.

19 February 2008
Page 23
Deterioration
Propagation
Activation
Initiation
Time
Chloride buildup but C < CT
Passive Steel
C > CT
Active steel
Figure 15. Corrosion sequence (Tuutti 1982)
100
MMFX
Black Bar
90
80
Damage (%)
70
60
50
40
30
20
24
10
43
0
0
10
20
30
40
50
60
70
Service life (years)
Figure 16. Projected damage function for a marine concrete pile reinforced with
MMFX and black bars.

19 February 2008
Page 24
100
MMFX
Black Bar
90
80
Damage (%)
70
60
50
40
30
20
14
10
27
0
0
10
20
30
40
50
60
70
Service life (years)
Figure 17. Projected damage function for a northern bridge deck reinforced with
MMFX and black bars
SUMMARY AND CONCLUSIONS

Two test methods, Modified ASTM G-109 and 20-Week Modified Southern Exposure tests, were used in
this project to investigate the corrosion performance of MMFX 2 deformed reinforcing bars with and
without mill scale compared to black ASTM A615 and stainless steel 304 bars. The following
conclusions can be drawn from the tests:
MMFX 2 steel bars have higher chloride thresholds than A615 bars, but less than Type 304 stainless steel.
The ASTM G-109 test program suggested that the chloride threshold of MMFX bars is about three times
of that of black bars. The stainless 304 bars did not corrode during the test program and resisted chloride
levels over five times that of the black bars. The removal of mill scale was found to have only slightly
increased the corrosion resistance of MMFX bars in the G-109 test.
The modified Southern Exposure (SE) test was very aggressive and corrosion initiation of the MMFX
bars occurred shortly after that of the black A615 bars. However, the macro-cell corrosion rate of the
MMFX bars remained about 1/3 less than that of the A615 black bars. The final weight loss of the
MMFX bars was 20% and 30% of the A615 black bars for the MMFX without and with mill scales,
respectively. Stainless 304 bars did not corrode during the 20-week SE test, even at chloride levels of
0.54% by wt. (21.0 lbs/cy). Corrosion rates measured by the linear polarization technique and weight loss
analysis of the bars correlated well.
Modeling analyses of a marine pile and a northern bridge deck exposed to deicers showed that the use of
MMFX bars in lieu of black bars may extend the structure service life by about 1.8 times, assuming a
chloride threshold for the MMFX bars of three times that of the A615 black bars. Slower corrosion rates
for the MMFX bars during the propagation period could increase the service life further, to about twice
that of A615 bars.
References
i
FHWA-RD-83-012, Time to Corrosion of Reinforcing Steel in Concrete Slabs, Vol. 5, FHWA, 1983.
ASTM Standard G109, 1999a (reaffirmed 2005), Standard Test Method for Determining Effects of Chemical
Admixtures on Corrosion of Embedded Steel Reinforcement in Concrete Exposed to Chloride Environments, Annual
Book of ASTM Standards, Vol. 03.02, ASTM International, West Conshohocken, PA.
iii
McDonald, D.B., Pfeifer, D.W., and Sherman, M.R. Corrosion Evaluation of Epoxy-coated, Metallic-clad and
Solid Metallic Reinforcing Bars in Concrete. FHWA-RD-98-153, McLean, VA: FHWA, 1998
iv
Sags, A.A. "Modeling the Effects of Corrosion on the Lifetime of Extended Reinforced Concrete Structures."
Corrosion, October 2003: 854-866.
v
Fushuang Cui, John S. Lawler and Paul D. Krauss, Corrosion Performance of Epoxy-Coated Reinforcing Bars in
a Marine Bridge Substructure CORROSION 2007, Paper No. 07286, Houston, TX: NACE International, 2007 .
vi
Krauss, Paul and Lee, S.K., Service Life Extension of Northern Bridge Decks Containing Epoxy-Coated
Reinforcing Bars, Concrete Reinforcing Steel Institute, Schaumburg, IL. 2003.
ii
APPENDIX A
TESTING, SETUP, AND SPECIMEN PHOTOGRAPHS
Photo A1. Concrete molds for ASTM G-109 test specimens.
Photo A2. ASTM G-109 test specimens just before

putting them in a fog curing chamber.
Photo A3. ASTM G-109 test specimens

in the RH and temperature controlled room.
Photo A4. ASTM G-109 test specimens in the first drying cycle.
Photo A5. Southern Exposure test specimens just

before putting them in a fog curing chamber.
Photo A6. Collection of baseline data for Southern Exposure test specimen.
Photo A7. The first wetting cycle of Southern Exposure test specimens.
Photo A8. Data acquisition systems for Southern Exposure testing.
Photo A9. Close-up of wiring arrangement for a

Southern Exposure test specimen.
Photo A10. Example of initial visible corrosion

product above bar locations in a Southern Exposure test specimen (BM2).
Photo A11. Example of initial visible corrosion product

above bar location in a Southern Exposure test specimen (MM1).
APPENDIX B
AUTOPIED REBAR PHOTOGRAPHS BEFORE AND AFTER CLEANING
SE TESTING AND ASTM G-109 TESTS
SE EXPOSURE TESTS
Top
Right
Left
Bottom
Right
Left
Specimen BM1 Autopsied Bars before Cleaning
Top
Right
Left
Bottom
Right
Left
Specimen BM1 Autopsied Bars After Cleaning
Top
SE EXPOSURE TESTS
Right
Left
Bottom
Right
Left
Top
Right
Left
Bottom
Right
Left
SE EXPOSURE TESTS
Top
Right
Left
Bottom
Right
Left
Top
Right
Left
Bottom
Right
Left
SE EXPOSURE TESTS
Top
Right
Left
Bottom
Right
Left
Specimen S1 Autopsied Bars before Cleaning
Top
Right
Left
Bottom
Right
Left
Specimen S1 Autopsied Bars After Cleaning
SE EXPOSURE TESTS
Top
Right
Left
Bottom
Right
Left
Top
Right
Left
Bottom
Right
Left
SE EXPOSURE TESTS
Top
Right
Left
Bottom
Right
Left
Top
Right
Left
Bottom
Right
Left
SE EXPOSURE TESTS
Top
Right
Left
Bottom
Right
Left
Specimen M1 Autopsied Bars before Cleaning
Top
Right
Left
Bottom
Right
Left
Specimen M1 Autopsied Bars After Cleaning
SE EXPOSURE TESTS
Top
Right
Left
Bottom
Right
Left
Top
Right
Left
Bottom
Right
Left
SE EXPOSURE TESTS
Top
Right
Left
Bottom
Right
Left
Top
Right
Left
Bottom
Right
Left
SE EXPOSURE TESTS
Top
Right
Left
Bottom
Right
Left
Top
Right
Left
Bottom
Right
Left
SE EXPOSURE TESTS
Top
Right
Left
Bottom
Right
Left
Top
Right
Left
Bottom
Right
Left

Top
SE EXPOSURE TESTS
Right
Left
Bottom
Right
Left
Top
Right
Left
Bottom
Right
Left
SE EXPOSURE TESTS
Top
Right
Left
Bottom
Right
Left
Specimen MM1 Autopsied Bars before Cleaning
Top
Right
Left
Bottom
Right
Left
Specimen MM1 Autopsied Bars After Cleaning
SE EXPOSURE TESTS
Top
Right
Left
Bottom
Right
Left
Top
Right
Left
Bottom
Right
Left
SE EXPOSURE TESTS
Top
Right
Left
Bottom
Right
Left
Top
Right
Left
Bottom
Right
Left
Top
SE EXPOSURE TESTS
Right
Left
Bottom
Right
Left
Top
Right
Left
Bottom
Right
Left

Top
Right
Left
SE EXPOSURE TESTS
Bottom
Right
Left
Top
Right
Left
Bottom
Right
Left
SE EXPOSURE TESTS
Top
Right
Left
Bottom
Right
Left
Top
Right
Left
Bottom
Right
Left
Specimen MM6Autopsied Bars After Cleaning
ASTM G109 TEST SAMPLES
Top
Bottom
Specimen BM3 G109 Autopsied Bars before Cleaning
Top
Bottom
Specimen BM3 G109 Autopsied Bars After Cleaning
Top
Bottom
Specimen S1 G109 Autopsied Bars before Cleaning
Top
Bottom
Specimen S1 G109 Autopsied Bars After Cleaning
Top
Bottom
Specimen M2 G109 Autopsied Bars before Cleaning
Top
Bottom
Specimen M2 G109 Autopsied Bars After Cleaning
Top
Bottom
Specimen MM4 G109 Autopsied Bars before Cleaning
Top
Bottom
Specimen M2 G109 Autopsied Bars After Cleaning

Comparative Corrosion Testing Cui Krauss

Uploaded by

Copyright:

Available Formats

Comparative Corrosion Testing Cui Krauss

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Comparative Corrosion Testing Cui Krauss

Uploaded by

Copyright:

Available Formats

Comparative Corrosion Testing and Analysis of

MMFX 2 Rebars for Reinforced Concrete Applications

Comparative Corrosion Testing and Analysis of

Comparative Corrosion Testing and Analysis of MMFX 2 Rebars for

MODIFIED ASTM G109 TEST

Comparative Correlative Corrosion Testing and Analysis of

Autopsy and Weight Loss Measurement

Results and Discussion

Comparative Correlative Corrosion Testing and Analysis of

Chloride, wt% (lbs/cy)

MMFX without scale

MMFX with scale

Phase II: Continuation of G109 Testing

Comparative Correlative Corrosion Testing and Analysis of

* Did not reach chloride current threshold values

Table 3 Chloride profile of Specimen MM-1 after 1266 days of exposure

Chloride, wt% (lbs/CY)

Comparative Correlative Corrosion Testing and Analysis of

Weight loss (g)

*20-week Phase 1 results

Comparative Correlative Corrosion Testing and Analysis of

*Clear cover = 1.0 in.

Figure 1. Schematic of modified G-109 test specimen.

MMFX without mill scale

MMFX with mill scale

Potential (V vs. SCE)

Elapsed time (day)

Comparative Correlative Corrosion Testing and Analysis of

Change of mean macro-cell current for G-109 specimens with time

MMFX without mill scale

MMFX with mill scale

Elapsed time (day)

Percent Weight Loss (%)

Comparative Correlative Corrosion Testing and Analysis of

Open Circuit Potential Vsce

Corrosion initiation time (days)

M-1 and all SS specimens remained passive

M-1 and all SS 304 specimens remained passive

Elapsed Time (days)

Comparative Correlative Corrosion Testing and Analysis of

Comparative Correlative Corrosion Testing and Analysis of

Comparative Correlative Corrosion Testing and Analysis of

Comparative Correlative Corrosion Testing and Analysis of

Comparative Correlative Corrosion Testing and Analysis of

20-WEEK MODIFIED SOUTHERN EXPOSURE TEST

Comparative Correlative Corrosion Testing and Analysis of

Results and Discussion

Comparative Correlative Corrosion Testing and Analysis of

Table 5 - Chloride Contents at bar level after 20-weeks SE

Black bar No.1

Black bar No.2

MMFX without scale No.1

MMFX without scale No.2

MMFX with scale No.1

MMFX with scale No.2

Stainless steel No.1

Stainless steel No.2

Comparative Correlative Corrosion Testing and Analysis of

Ponding well with lid

Figure 7. Schematic of modified southern exposure test specimen