Report on the Diagnosis, Prognosis, and Mitigation of Alkali-Silica Reaction (ASR) in Transportation Structures
Appendix I Evaluation of the Potential for Further Expansion due to ASR (Prognosis)
Procedure for Laboratory Assessment of the Current Rate of ASR Expansion in Field Concrete
I.1 Introductory Remarks
The potential for further expansion due to ASR is a critical factor in the process of selecting the most appropriate measures of mitigation/remediation. ASR and associated expansion and deterioration processes will continue in a concrete member as long as the reactive mineral phases within the aggregate particles are not completely consumed, and the two other essential conditions are still satisfied, i.e., high humidity and high alkali concentration in the concrete pore solution. Expansions are also function of temperature and restraint. In-situ monitoring of concrete deformation and movement is the best way to assess the current and future expansion due to ASR in existing concrete members or structures; however, reliable results usually take over three years to obtain, as the ASR contribution needs to be isolated from seasonal variations in temperature and humidity. Laboratory tests on cores generally lead to fairly reliable estimates within 6 to 12 months, and are commonly used to assess the potential for further ASR expansion.
In the procedure proposed by Bérubé et al. (2002) and described hereafter, the risk for further expansion and damage in a concrete member or structure due to ASR, is estimated from the following parameters (Table I1 and Figure I1):
- The "residual concrete expansivity", determined in the laboratory by testing core samples in humid air at >95 percent R.H. and 38°C (100°F) (see Appendix F).
- The "residual aggregate reactivity", determined in the laboratory by testing core samples in 1N NaOH solution at 38°C (100°F) (see Appendix F) or, even better, when the reactive aggregates belong to the coarse aggregates, by extracting these aggregates from cores (see Appendix G) and testing them through the Concrete Prism Test ASTM C 1293.
- The amount of "water-soluble alkalis" in the concrete, which is estimated using the hot-water extraction method described in Appendix H.
- Reasonable estimates of the humidity, temperature, and restraints (reinforcement, prestressing, postensioning, confinement, loading,...) in the field.
The individual risk indices corresponding to each of the above parameters are then combined to estimate the potential/current rate of ASR expansion in the field of the concrete member under study, either already affected by ASR or not.
Table I1 - Classification of the various coefficients proposed to estimate the current rate of ASR expansion in concrete members in service either already affected by ASR or not.
|percent exp./yr1||Residual exp.1||RCE1|
|0.003 to 0.005||very low||1|
|0.005 to 0.01||low||2|
|0.01 to 0.015||moderate||4|
|0.015 to 0.02||moderate||6|
|0.02 to 0.025||high||9|
|0.025 to 0.03||high||12|
|> 0.03||very high||16|
|percent exp. at 1 yr1||Reactivity1||RAR1|
|0.04 to 0.08||low||1|
|0.08 to 0.12||moderate||2|
|0.12 to 0.20||high||3|
|> 0.20||very high||4|
|kg/m3(lb/yd3) Na2Oeq||Alkali content||WSA|
|< 1.0 (0.6)||very low||0|
|1.0(0.6) to 1.5(0.9)||low||1|
|1.5(0.9) to 2.0(1.2)||moderate||2|
|2.0(1.2) to 2.5(1.5)||high||3|
|> 2.5(1.5)||very high||4|
|Internal humidity||Humidity risk||HUM|
|< 80 percent RH||very low||0|
|80-85 percent RH||low||0.25|
|85-90 percent RH||moderate||0.5|
|90-95 percent RH||high||0.75|
|95-100 percent RH||very high||1|
1 Values are considered underestimated if the concrete cores were abnormally fractured or porous compared to the overall concrete member under study (RCE and RAR) or quite impermeable to the alkaline solution (RAR).
Figure I1. Laboratory assessment of the current rate of ASR expansion in concrete members in service either already affected by ASR or not.
I.2 Test Procedure
I.2.1 Residual concrete expansivity (Coefficient "RCE")
The "residual concrete expansivity" of the concrete under study is determined through expansion tests on cores in humid air at >95 percent R.H. and 38°C (100°F). The test procedure is described in the Appendix F, which also provides insights for the interpretation of the test results; the method also generally requires an examination of the cores before and after the expansion tests for petrographic signs of ASR. After the concrete cores have reached a mass equilibrium through a preconditioning period, the rate of expansion is calculated over a testing period of one year or more; the expansion rate is then used to determine the coefficient "RCE", which varies from 0 to 16 (Table I1).
An alternative method of estimating the residual concrete expansivity is based on the "residual aggregate reactivity" (coefficient "RAR", ranging from 0 to 4; Section I.3), which is combined with the "water-soluble alkali content" of the concrete (coefficient "WSA", ranging from 0 to 4; Section I.4). The product of the above two parameters varies from a minimum of 0 to a maximum of 16 (Table I1 and Figure I1).
I.2.2 Residual aggregate reactivity (coefficient "RAR")
The "residual aggregate reactivity" is based: (1), on expansion tests on cores stored in 1N NaOH at 38°C (100°F) (see Appendix F), and/or (2), even better when the reactive aggregates belong to the coarse aggregate fraction, on the expansion of concrete prisms incorporating the coarse aggregates extracted from the cores following the Concrete Prism Test ASTM C 1293 (see the Appendix G).
Expansion tests on cores - After the concrete cores have reached a mass equilibrium through a preconditioning period, the one-year expansion value thus obtained is used to determine the coefficient "RAR", which varies from 0 to 4 (Table I1). The interpretation of the results from expansion tests on cores in NaOH solution should also require the petrographic examination of the cores before and after the expansion tests, as discussed in the Appendix F.
Concrete Prism Tests ASTM C 1293 - The one-year expansion of concrete prisms incorporating aggregates extracted from the concrete cores can also be used for determining the coefficient "RAR" (Table I1). One must consider though that the concrete prisms made and tested according to ASTM C 1293 conditions may depart from the field concrete under study with regards to reactive aggregate content, aggregate size, air content, and/or water-to-cement ratio, all parameters which affect the expansion due to AAR (Bérubé et al. 2004).
I.2.3 Water-soluble alkali content (coefficient "WSA")
The test procedure for the measurement of the water-soluble alkali content of the concrete under study, in kilograms of Na2Oeq per cubic meter (or lbs/yd3), is described in the Appendix H along with some fundamental considerations on the subject. The result of the test is used for establishing the coefficient "WSA", which varies from 0 to 4 (Table I1).
I.2.4 Humidity conditions in service (coefficient "HUM")
It is generally well established that minimum external/ambient relative humidity conditions in the range of 80 to 85 percent are required for excessive expansion due to AAR to develop in concrete. However, according to a number of experimental studies, the less reactive the aggregate, the higher the critical humidity level required for significant ASR expansion to occur (Bérubé et al. 2002). Moreover, the higher the temperature, the lower this critical level (Olafsson 1987). Also, a clear distinction must be made between internal and external/ambient humidity, with different minimum levels for ASR (Table I1). In a particular study, a minimum level of "internal" humidity of 85 percent R.H. was necessary to sustain AAR in concrete specimens made with very reactive aggregates and high alkali contents, and stored at 38°C (100°F), which corresponded to external/ambient humidity conditions of less than 70 percent R.H. (Bérubé et al. 2002).
Internal humidity conditions in concrete can be periodically or continuously monitored using dedicated instrumentation (see Appendix D). On the other hand, the external/ambient humidity conditions may be continuously recorded by local or regional weather stations. The average yearly internal or external humidity conditions can be determined from such monitoring/ recording processes, otherwise reasonably estimated. The result is then used for establishing the coefficient "HUM", which varies from 0 to 1 (Table I1). This coefficient takes into account the fact that the humidity inside thick concrete members tend to remain higher than inside thin members, which makes the former less influenced by the external humidity conditions. It also takes into account the type of climate involved (i.e., desertic or not).
I.2.5 Temperature conditions in service (coefficient "TEM")
The expansion tests described before are performed at a constant temperature of 38°C (100°F); however, concrete structures exposed outdoors can be subjected to fairly different average annual temperatures depending on their location across North America. On the other hand, the higher the temperature, the greater is usually the expansion rate due to ASR, while the ultimate expansion is not necessarily greater.
Temperature inside concrete can be periodically or continuously monitored using dedicated instrumentation (see Appendix D). On the other hand, external temperature may be continuously recorded by local or regional weather stations. The average yearly internal or external temperature can be determined based on the above approaches, otherwise reasonably estimated. The result is then used for establishing the coefficient "TEM", which varies from 0 to 1 (Table I1).
I.2.6 Restraints in service (coefficient "STR")
In general, internal (prestressing, reinforcement, etc.) and external (post-tensioning, loading, confinement, etc.) compressive stresses applied to concrete can significantly reduce expansion due to ASR, however not always the surface cracking. It is then necessary to apply a correction factor to the results obtained from expansion tests on cores or prisms (e.g. ASTM C 1293) that are free from restraints. The coefficient of correction "STR" is, however, a difficult parameter to estimate because of the relatively limited amount of relevant information generally available. Moreover, a same level of reinforcement or compressive stress, for instance, is more effective in the presence of a marginally reactive aggregate than with a highly-reactive one (Smaoui et al. 2007). The values proposed in Table I1 must then be considered with circumspection. They correspond to the median results found in a report by the Institution of Structural Engineers (ISE 1992).
When steel reinforcing bars are installed in a single plane in the concrete member (1D or 2D, depending if all bars are parallel or at right angle) or in several parallel planes (2D or 3D, again depending if all bars are parallel or at right angle), but without any anchorage between the different planes, the coefficient "STR" applies in the direction(s) of the reinforcing bars; however, this can result in expansions even higher in the unrestrained directions than when no confinement is applied (Smaoui et al. 2007). Similarly, in the cases of uniaxial (1D) or biaxial (2D) compressive stresses (prestressing, postensioning, loading, confinement,…), "STR" applies in the direction(s) of the stresses while, once again, expansions even higher than when no confinement is applied can develop in the unrestrained direction(s).
I.2.7 Current rate of AAR expansion in service (coefficient "CRE")
The coefficient "CRE" is proposed as an estimate of the current rate of expansion due to AAR in concrete members in service either already affected by ASR or not. This coefficient takes into account all of the above coefficients. As illustrated in Figure I1, it is obtained as follows:
CRE = (maximum [RCE] or [RAR x WSA]) x HUM x TEM x STR
The CRE ranges between 0 and 16. In accordance with Table I1, the current rate of ASR expansion in the concrete member under study can range from negligible to very high. This qualitative rating would be of greater interest/use if translated into an expected current rate of expansion (in %/yr). The information is very limited relating expansion results obtained in the laboratory with those observed in the field. However, after the above methodology had been applied to a number of existing structures in Québec (Canada), it appears realistic to use the same classification as for the above coefficient RCE (Table I1), which corresponds to Figure I2. In other words, the assumption is made that the expansion rate in service (in %/yr) will be somewhat similar to the one observed in the laboratory for cores tested in humid air at >95 percent R.H. and 38°C (100°F) when the following conditions are satisfied in service: (1), very high humidity conditions (i.e., HUM = 1); (2), temperature over 30°C (86°F) (i.e., TEM = 1), and (3), no reinforcement neither other restraints applied to concrete (i.e., STR = 1).
Figure I2. Approximation of the current rate of ASR expansion in concrete members in service either already affected by ASR or not, as a function of the coefficient CRE.
As discussed in the Appendix F, results from expansion tests on cores are considered underestimated if the concrete cores tested were abnormally fractured or porous compared to the overall concrete member under study (coefficients RCE and RAR) or quite impermeable to the alkaline solution (coefficient RAR); this would result in underestimated values for the overall coefficient CRE and the corresponding estimated rate of current expansion in the field.
I.3 Concluding Remarks
The overall coefficient CRE takes into account many parameters which all affect, to some extent, the rates of ASR reaction and expansion. It takes a zero value when at least one of the three necessary conditions for ASR is not satisfied, i.e., when the aggregates are not reactive (i.e., RAR = 0), when the concrete alkali content is low (i.e., WSA = 0), or when the humidity conditions in service are low (i.e., HUM = 0). Also, it can predict the anisotropic expansion in concrete members whose different parts are exposed to different humidity, temperature and/or stress conditions. However, it does not predict for how long the expansion will continue in the concrete structure but just gives an indication of the current (or next future) rate of expansion due to AAR. It must be also emphasized that the coefficient CRE is mostly based on laboratory test results. As mentioned before, long-term in-situ monitoring remains the only reliable way to obtain relevant information on the current rates of expansion, which can then be extrapolated for the next 5 to 10 years.
Table I2 - Classification of the various coefficients proposed to estimate the current rate of ASR expansion in concrete members in service either already affected by ASR or not. (cont)
|External (ambient) humidity||Thin member (<0.5m)(0.55yd)||Thick member (>0.5m) (0.55yd)|
|Humidity risk||HUM||Humidity risk||HUM|
|Indoor||<70 percent RH||very low||0||low||0.5|
|70-80 percent RH||low||0.25||moderate||0.75|
|80-90 percent RH||moderate||0.5||high||1|
|90-95 percent RH||high||0.75||very high||1|
|95-100 percent RH or immersed||very high||1||very high||1|
|Outdoor in deserts||Not in contact with the ground||very low||0||moderate||0.5|
|In contact with the ground||low||0.25||high||0.75|
|Outdoor in other areas in North America||Not exposed to rain nor in contact with the ground||moderate||0.5||high||0.75|
|Not exposed to rain but in contact with the ground||high||0.75||very high||1|
|Exposed to rain, immersed or buried very||high||1||"||1|
|Annual avg. temp. (°C)(°F)||TEM|
|< 0 (32)||0.4|
|0(32) to 10(50)||0.55|
|10(50) to 20(68)||0.7|
|20(68) to 30(86)||0.85|
|percent of steel||STR|
|> 0 to 1||very low|
|1 to 2||low|
|2 to 6||moderate|
|>6 to 12||high|
|>12 to 16||very high|
American Standards for Testing and Materials (ASTM), "Standard Test Method for Determination of Length Change of Concrete Due to Alkali-Silica Reaction," ASTM International, ASTM C1293.
Bérubé, M.A., Frenette, J. and Rivest, M., "Laboratory Assessment of the Potential Rate of ASR Expansion of Field Concrete," Cement, Concrete, and Aggregates, 24 (1), 28-36, 2002.
Bérubé, M.A., Smaoui, N. and Côté, T., "Expansion Tests on Cores from ASR-Affected Structures," 12th International Conference on AAR, Beijing (China), pp. 821-832, 2004.
Institution of Structural Engineers (ISE), "Structural Effects of Alkali-Silica Reaction - Technical Guidance Appraisal of Existing Structures," Institution of Structural Engineers, 11 Upper Belgrave Street, London SW1X 8BH, 45 p., 1992.
Olafsson, H., "The Effect of Relative Humidity and Temperature on Alkali Expansion of Mortar Bars," Concrete Alkali-Aggregate Reaction, Noyes Publications, pp. 461-465 (Proc. 7th Int. Conf. on AAR in Concrete, Ottawa, Canada, July 1986), 1987.
Smaoui, N., Bissonnette, B., Bérubé, M.A., Fournier, B., and Durand, B., "Stresses Induced by ASR in Prototypes of Reinforced Concrete Columns Incorporating Various Types of Reactive Aggregates," Canadian Journal of Civil Engineering, submitted for publication, 2007.