Skip to contentU.S. Department of Transportation/Federal Highway Administration
Pavement | Bridge
FHWA > Pavement : Bridge > ASR > HIF-09-004 > Appendix E Determination of Internal Damage and Expansion to Date in ASR-Affected Concrete

Report on the Diagnosis, Prognosis, and Mitigation of Alkali-Silica Reaction (ASR) in Transportation Structures

Appendix E Determination of Internal Damage and Expansion to Date in ASR-Affected Concrete

The Modified Stiffness Damage Test (SDT)

E.1 Introductory Remarks

The Stiffness Damage Test (SDT) was originally proposed by Chrisp et al. (1989, 1993) and adopted by the Institution of Structural Engineers (ISE 1992). It is based on the cyclic uniaxial compressive loading of concrete core samples (5 cycles) between 0 and 5.5 MPa (800 psi). The reduction in the Young's elastic modulus, the energy dissipated during the load-unload cycles, which corresponds to the surface area of the hysteresis loops, and the accumulated plastic strain after these cycles, are associated with the closure of the existing cracks and to a slip mechanism, and thus represent a measure of the damage in the specimen (microcracking) in the direction of the applied stress.

The capacity of the test to estimate the expansion attained to date by ASR-affected concrete has been recently investigated in detail (Smaoui et al. 2004a, 2004b, 2004c, Bérubé et al. 2005a). The test was found to be useful in that respect, particularly when applying a maximum load of 10 MPa (1450 psi) during the 5 load/unload cycles (rather than 5.5 MPa (800 psi) in the original method) and using one or both of the two following parameters: (1), the energy dissipated during the first cycle (hysteresis loop 9 see Figure E1), and (2), the accumulated plastic strain after the 5 load/unload cycles. These two parameters progressively increase along with the internal microcracking and the consequent bulk expansion of the concrete under study (Figure E1 and E2). However, as the first parameter was found to be more closely correlated to ASR expansion this is the parameter that is selected in the following procedure. This modified SDT can thus supply useful information about the internal damage (microcracking) of the concrete under study and the expansion attained to date due to ASR (or to any other expansive mechanism causing cracking). Moreover, when the number of cores available for mechanical testing is limited, the elastic modulus can be determined in the modified SDT and the compressive strength can be estimated by reloading the cores up to failure after the 5 load/unload cycles. However, despite the good correlation obtained, between the energy dissipated during the first load/unload cycle and the expansion due to ASR and this, for each type of reactive aggregate investigated up to now in the laboratory, the results are aggregate-dependent (Figure E3); this thus requires the use or the establishment of a calibration curve suitable to the particular reactive aggregate present in the concrete under study or corresponding to a similar type (see Section E.2.9).

Figure E1. Results from the modified SDT at various expansion levels for concrete cylinders made with a highly reactive siliceous limestone coarse aggregate from Québec (Canada) (After Smaoui et al. 2004a, 2004c).

This figure shows five graphs, depicting the stress versus strain relationship for a concrete sample. The last graph illustrates the area between the curves as a hysteresis loop and is shown in red.

Figure E2. Results from the modified SDT for a concrete cylinder made with an extremely-reactive sand from Texas at an expansion level of 0.392 percent. The energy dissipated during the first load/unload cycle (area of the first hysteresis), in Joule/m3, is obtained by subtracting the area (in red) under the first unload curve from the area (in blue) under the first load curve. Third-degree polynomial relationships can be obtained for these curves from which the above calculation can be easily made. (After Bérubé et al. 2005b).

This figure shows two graphs; the first graph depicts the stress versus strain relationship for a concrete specimen. The distance between two curves is shown and is further illustrated in the second graph.

Figure E3. Linear relationships between the energy dissipated during the first load/unload cycle of the modified SDT and the expansion of concrete cylinders made with different coarse (Québec, New Mexico, Potsdam, Limeridge) or fine (Texas) aggregates susceptible to ASR. (After Smaoui et al. 2004a, 2004c).

This graph shows the relationship between percent expansion and the dissipated energy during a stiffness damage test for five different concrete specimens.

E.2 Test Procedure

E.2.1 Core diameter

The test procedure is essentially recommended for cores of 100 mm (4-in) in diameter.7

E.2.2 Storage conditions before testing

The cores must be tested in the same moisture conditions that prevail in nature. Therefore, they must be tightly sealed in plastic sheets immediately after coring in order to prevent any volumetric change due to drying (e.g., unsealed cores stored at ambient temperature) or to moisture uptake (e.g., unsealed cores stored in a wet room).

E.2.3 Specimen preparation

The modified SDT primarily consists in measuring the deformation of concrete cores when submitted to cyclic uniaxial compressive loading. The test specimens are thus prepared as for the standard uniaxial compression test ASTM C 39 ("Standard test method for compressive strength of cylindrical concrete specimens"), i.e., sawed and capped or ground to obtain two end faces that are perfectly parallel and perpendicular to the core axis. A great care must be given to satisfy the above requirement since small volumetric changes have to be precisely measured. The length of cores to test, 100 mm (4-in) in diameter, shall be 200 ± 5 mm (8 ± 0.2 in).

E.2.4 Testing machine

Any type of testing machine capable of imposing load/unload cycles at the rate and of the magnitude prescribed in Section E.2.7 may be used, provided the data acquisition system allows the simultaneous measurement at any time of the stress applied and the longitudinal strain. The machine shall also conform to all other requirements of Practices E 4 ("Standard practices for load verification of testing machines") and Test Method ASTM C 39 ("Standard test method for compressive strength of cylindrical concrete specimens").

E.2.5 Measuring device

For measuring the longitudinal deformations during the test, an unbonded sensing device shall be used, which conform to Section 4.2 of standard ASTM C 469 ("Standard test method for static modulus of elasticity and Poisson's ratio of concrete in compression"). In addition, this device shall allow measurements at the nearest micrometer along the two diametrically opposite gage lines, since small volumetric changes have to be precisely measured.

E.2.6 Number of test specimens

A minimum of three cores, 100 mm (4 in) in diameter by 200 ± 5 mm (8 ± 0.2 in) in length, should be tested, for each concrete under study.

E.2.7 Load/unload cycles

The test involves 5 cycles of loading up to 10 MPa (1450 psi) and unloading down to 0 MPa.8 The rate of loading and unloading is 0.1 N/mm2/s. The stress applied and the longitudinal deformations along each of the two gage lines are recorded by the data acquisition system for each increment of 0.5 N/mm2during loading and unloading.

When the 5 cycles are completed, the compressive strength can be obtained by reloading the specimen tested up to failure.

E.2.8 Handling and interpretation of test results

Stress-strain graph -A graph of the stress applied as a function of the longitudinal strain (average of the two diametrically opposite measurements) is drawn (Figure E1 and E2).

Elastic modulus -The Young's elastic modulus can be estimated from the slope of the linear relationship corresponding to the first loading (Figure E1 and E2). This parameter is of great interest in the diagnosis of AAR and for determining the progress of the reaction (provided tests are periodically conducted over time on core samples).

Energy dissipated during the first load/unload cycle -The energy dissipated during the first load/unload cycle (area of the first hysteresis), in Joule/m3, is obtained by subtracting the area (in red in Figure E2) under the first unloading curve from the area (in blue in Figure E2) under the first loading curve. Third-order polynomial relationships can be obtained for these two curves from which the energy dissipated can be easily calculated. In general, these relationships present correlation coefficients of 0.999 or more (Figure E2).

Expansion attained to date -The expansion attained to date by the ASR-affected concrete under test is ideally estimated from an empirical linear relationship between the energy dissipated during the first load/unload cycle and the expansion due to AAR, established in the laboratory for concrete cylinders made with the reactive aggregate present in the concrete under study or a similar type of reactive aggregate (see Section E.2.9). As mentioned earlier (Section E.1), the results of the test are greatly aggregate dependent (Figure E3). In absence of the above information, a relationship already obtained for quite similar reactive aggregates can be used. The results presented in Figure E3 have been obtained for cylinders made with 4 different types of reactive aggregates from the U.S.A. and Canada (Smaoui et al. 2004a, 2004c). Each corresponding relationship presents a correlation coefficient (R2) of 0.92 or more (Table E1) but is clearly different from one aggregate to another. In the absence of any more relevant information, the expansion to date could be estimated using the linear relationship in Table E1 corresponding to the reactive aggregate the most similar to the type present in the cores subjected to the modified SDT. It is important to recall that these relationships apply to concrete specimens of 100 mm (4 in) in diameter, only.

Table E1. Relationships obtained in the SDT for different types of reactive aggregates. (After Smaoui et al. 2004a, 2004c).
Reactive aggregate type State (Country) Relationship R2
Siliceous limestone Quebec (Canada) Energy = 0.00003614 + 0.5871 x expansion 0.98 (6 results)
Natural gravel rich in volcanic particles New Mexico (U.S.A.) Energy = 0.0003741 + 0.1900 x expansion 0.99 (5 results)
Quartzitic sandstone (Potsdam Group) Quebec (Canada) Energy = 0.0003494 + 0.9702 x expansion 1.00 (2 results)
Natural sand rich in volcanic particles Texas (U.S.A.) Energy = 0.0004577 + 0.001427 x expansion 0.92 (5 results)
E.2.9 Establishment of the calibration curve

The specimens required for the establishment of the calibration curve for a particular reactive aggregate correspond to concrete cylinders, 100 mm in diameter x 200 mm in length. These cylinders are made and tested in accordance with the Concrete Prism Test ASTM C 1293, except that different sets of specimens must be submitted to the SDT at different expansion levels (including a "zero" level). It is recommended that a minimum of 5 expansion levels (including the "zero" level) be considered using a constant increment between two successive levels. This increment should depend on the maximum expansion that the particular aggregate could generate, but it should never exceed 0.04 percent. For instance, if the maximum expansion is 0.10 percent, thus the recommended expansion levels are: 0.00 percent, 0.025 percent, 0.050 percent, 0.075 percent, and 0.10 percent (5 different expansion levels).9If the maximum expansion is 0.20 percent, thus the recommended expansion levels are: 0.00 percent, 0.04 percent, 0.08 percent, 0.12 percent, 0.16 percent, and 0.20 percent (maximum increment of 0.04 percent). Considering that a minimum of three specimens must be tested per SDT, i.e. for each expansion level, a minimum of 18 cylinders must then be made in the latter case for the establishment of the calibration curve. After the last expansion measurement, the specimens are immediately sealed with several layers of plastic wrap in order to prevent any moisture loss, stored for one day at room temperature, then subjected to the SDT following the above steps E.2.3 to E.2.8. When specimens cannot be tested immediately, they can be stored in a freezer in order to stop expansion. The specimens have to be thawed for 24 hours at room temperature, unwrapped, and then subjected to the SDT.

E.3 References

American Standards for Testing and Materials (ASTM), "Standard Test Method for Compressive Strength of Cylindrical Concrete Specimens," ASTM International, ASTM C39.

American Standards for Testing and Materials (ASTM), "Standard Test Method for Static Modulus of Elasticity and Poisson's Ratio of Concrete in Compression," ASTM International, ASTM C469.

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.

American Standards for Testing and Materials (ASTM), "Standard Practices for Force Verification of Testing Machines," ASTM International, ASTM E4.

Bérubé, M.E., Smaoui, N., Fournier, B., Bissonnette, B., and Durand, B., "Evaluation of the Expansion Attained to Date by ASR-affected Concrete - Part III: Application to Existing Structures," Canadian Journal of Civil Engineering, 32 : 463-479, 2005a.

Bérubé, M.A., Smaoui, N., Bissonnette, B., and Fournier, B., "Outil d'évaluation et de gestion des ouvrages d'art affectés de réactions alcalis-silice (RAS)," Études et Recherches .en Transport, Ministère des Transports du Québec, 140 p., 2005b

Chrisp, T.M., Waldron, P., and Wood, J.G.M., "Development of a Non-destructive Test to Quantify Damage in Deteriorated Concrete," Magazine of Concrete Research, 45 (165): 247256, 1993.

Chrisp, T.M., Wood, J.G.M., and Norris, P., "Towards Quantification of Microstructural Damage in AAR Deteriorated Concrete," Proceedings of the International Conference on Recent

Developments on the Fracture of Concrete and Rocks, Elsevier Applied Science, London, U.K., pp. 419-427, 1989.

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.

Smaoui, N., Bérubé, M.E., Fournier, B., Bissonnette, B. and Durand, B., "Evaluation of the Expansion Attained to Date by ASR-Affected Concrete - Part I: Experimental Study," Canadian Journal of Civil Engineering, 31: 826-845, 2004a.

Smaoui, N., Fournier, B., Bérubé, M.E., Bissonnette, B. and Durand, B., 2004b, "Evaluation of the Expansion Attained to Date by ASR-Affected Concrete - Part II: Application to Non-Reinforced Concrete Specimens Exposed Outside," Canadian Journal of Civil Engineering, 31: 997-1011, 2004b.

Smaoui, N, Bérubé, M.A., Fournier, B., Bissonnette, B. and Durand, B., "Evaluation of Expansion Attained to Date by ASR-Affected Concrete," 12th International Conference on AAR, Beijing (China), pp. 1005-1015, 2004c.


7So far, testing has been carried out on 100-mm (4-in) cores only. Preliminary results (Smaoui et al. 2004a) indicated that the results are greatly affected by the core diameter and a study is presently underway with the objective of adjusting the method for cores of 75 and 150 mm (3 to 6 in) in diameter.

8Considering that the parameter used for assessing the expansion to date is the energy dissipated during the first load/unload cycle, thus neglecting the residual deformation after 5 cycles, only one cycle should be sufficient. However, the behavior of the cores tested during the other 4 cycles is useful information.

9It is obvious that exact expansion levels cannot be reached and that all specimens to be tested at the same expansion level will not present exactly the same expansion. The important point is that these specimens be close enough to the desired expansion level, while the results (i.e., exact expansion and SDT result) for each specimen tested are individually reported on the graph of expansion against SDT and used for calculating the calibration curve.

PDF files can be viewed with the Acrobat® Reader®
Updated: 06/14/2012