Results and Discussion

The compressive 28-day strengths of the specimens are shown in Table 3. The PC50 mix with W/CM ratio of 0.5 had much lower strength than the other seven mixes with W/CM ratio of 0.4. Of those the silica fume concrete, SFC, had the highest strength of 49.1 Mpa.

Table 3. 28-day compressive

strength

for the eight

mixes

Mix

PC50

PC40

SR

SC

SFC

FAC

SSFC

FASF

W/CM ratio

0.5

0.4

0.4

0.4

0.4

0.4

0.4

0.4

Compressive Strength (MPa)

34.0

48.6

47.4

42.3

49.1

47.0

43.5

46.0

The splitting strengths were variable. The specimens exposed to H2S only gained some strength, about 1 kPa on average, between cycles 11 and 15, but lost about the same amount of strength between cycles 15 and 23. The specimens exposed to both H2S and sulphate gained similar amounts of tensile strength in the first part of the testing period but retained these strength until the end of testing, despite the loss of concrete volume to be discussed later.

The electrochemical potentials for the different mixes have been taken 19 times since the start of the experiment. They are plotted in Fig. 3 for the 8 mixes exposed to hydrogen sulphide and half submerged in sulphate solution. Fig. 4 presents the results for the 8 mixes exposed to hydrogen sulphide only. In examining the curves it should be noted that a potentials of -350 mV may be considered a 90% probability of active steel corrosion. All values are the mean of all available replicates.

In Fig. 3 the potential measurements of PC50 were consistently the highest. It crossed the -350 mV boundary at cycle 17. The values measured for the sulphate resistant concrete mix, SR were almost consistently the lowest, indicating excellent protection against steel corrosion.

They did not cross the -350 mV till near the end of testing. All other mixes performed in between; mix FASF crossed the -350 mV value at cycle 18, SC between cycles 18 and 19, mix SFC at about cycle 19 and SSFC at cycle 20.

The results in Fig. 4 indicate that the specimens exposed to H2S only experienced far less steel corrosion than did those exposed to both sulphates and H2S. Potentials were roughly one half. Indeed, none of the former reached the -350 mV value. But as before, the PC50 mix exhibited the highest values almost consistently. Mix SSFC shows the lowest values up to cycle 18 after which mix SR gave the lowest value. In all mixes except PC50 there appeared to be a distinct increase in the rate at cycle 16.

Results and Discussion

Fig. 3 Electrochemical potential of mixes exposed to sulphate and H2S

Results and Discussion

Fig. 4 Electrochemical potential of mixes exposed to H2S only

The degree of corrosion of the reinforcing steel was evaluated visually using numerical rating system of 1 to 3, 1 indicating no evidence of visible rust on the surface, and 3 indicating heavy corrosion. Table 4 presents the visual rating of steel surface corrosion of the 16 steel bars for the last set of specimens after the 23rd test cycle. The corrosion of the steel reinforcement was quite uniform in the specimens exposed to H2S only, regardless of the mix. All had medium amount of corrosion except mix FASF which had severe corrosion. There was considerable variability in the mixes exposed to H2S and sulphate. Mix SR had visibly the least amount of steel corrosion whereas in mixes PC50, PC40 and FAC the steel was heavily corroded. The visual corrosion ratings are in reasonable agreement with the electrochemical potential results.

Table 4. Visual rating of steel corrosion

Mix

H2S and sulphate

H2S only

PC50

3

2

PC40

3

2

SR

1

2

SC

2

2.5

SFC

2

2

FAC

3

2.5

SSFC

1.5

2

FASF

2

3

The volume loss experienced in the concrete because of corrosion is illustrated in Fig. 5. All concrete specimens exposed to sulphate solution and hydrogen sulphide gas showed higher volume loss than those exposed to hydrogen sulphide gas only. The volume loss was the greatest in mix PC40 exposed to both H2S and sulphate. The least amount of concrete lost to corrosion was in mix SR; SFC was second lowest.

Photographs of all specimens that experienced the full 23-cycle 3-year exposure to sulphate and/or sulphide are shown in Figs. 6a and 6b. Again, it is obvious from the photographs that all concrete specimens exposed to sulphate solution and hydrogen sulphide gas showed higher volume loss than those exposed to hydrogen sulphide gas only. And it is also obvious that most of the

Results and Discussion

Fig. 5. Loss of volume for all mixes after 23 cycles of testing

concrete loss is located above the sulphate solution level. As was discussed earlier, the evaporation of the water from the concrete surface that contains high amount of alkali sulphates causes crystallization of salts, which in turn generates a disruptive pressure (Skalny et al. 2002). This is in good agreement with field observations of the piers partially submerged in liquid manure in the barn that collapsed in 2001 as a result of corrosion, after only 12 years of service.

X-ray diffraction results were obtained from surface scrapings from mixes PC40, SR and SFC exposed to H2S only and those exposed to both H2S and sulphate after 23 cycles of exposure. The scrapings were taken above the level of the sulphate solution in the case of the latter three specimens. Scrapings from the surfaces of these six specimens contained gypsum, but no ettringite. Samples taken from the piers and beams over the manure pit of the barn that collapsed in 2001 as a result of corrosion, after only 12 years of service, showed very similar diffraction patterns; again gypsum was the predominant corrosion product. As gypsum is the expected corrosion product at sulphate concentrations over 8,000 ppm it is clear that very high concentrations of sulphate (over 8,000 ppm) do occur in parts of manure structures.

After 110 applications of H2SO4 over a period of approximately one year the PC50 experienced the greatest loss in mass of 5.2% of the original dry mass. The 5 specimens with a W/CM ratio of 0.4 had losses ranging from 4.0 to 4.7%. The least amount of mass loss was experienced by the sulphate resistant cement, SR. These preliminary results are not conclusive. What is evident from visual observation of the specimens that the limestone aggregate helps to reduce concrete deterioration from cement paste loss only and breakdown of the concrete structure. After a year of testing no pieces of coarse aggregate have become dislodged.