Chargement
Informations sur le livre
Brittle Matrix Composites 10
Actions du livre
Commencer à lireDescription
The subjects of the symposia are on composite materials behaving as brittle, normal and special conditions of exploitation. Brittle matrix composites are applied in various domains and the series of symposia are closely related to their applications in civil engineering. In the last decades their importance is increasing along with their variety and the use of most advanced methods of testing. Papers include concretes, fibre concretes and ceramics, particularly their composition, microstructure and fracture processes. Various new and advanced engineering problems are presented in the papers.
Actions du livre
Commencer à lireInformations sur le livre
Brittle Matrix Composites 10
Description
The subjects of the symposia are on composite materials behaving as brittle, normal and special conditions of exploitation. Brittle matrix composites are applied in various domains and the series of symposia are closely related to their applications in civil engineering. In the last decades their importance is increasing along with their variety and the use of most advanced methods of testing. Papers include concretes, fibre concretes and ceramics, particularly their composition, microstructure and fracture processes. Various new and advanced engineering problems are presented in the papers.
Aperçu du livre
Brittle Matrix Composites 10
Sobczak
Keynote Paper
Effects of mechanical properties of ASR damaged concrete on structural design
Hans W. Reinhardt¹ and Oliver Mielich², ¹Department of Construction Materials, University of Stuttgart; ²Materialprüfungsanstalt Universität Stuttgart (Otto-Graf-Institute) Pfaffenwaldring 4, 70569 Stuttgart, Germany. E-mail address: reinhardt@iwb.uni-stuttgart.de
ABSTRACT
The paper reports on an experimental investigation of concrete with three slow reacting aggregates (greywacke, quartz-porphyrite, crushed gravel). The specimens were stored in a fog room at 40 °C for 560 days. The compressive strength, the tensile strength, the static modulus of elasticity, and the dynamic modulus of elasticity were measured at certain intervals. The results show how these mechanical properties are affected by the alkali-silica reaction (ASR). It turned out that the dynamic modulus of elasticity cannot be used for the prediction of deterioration due to ASR. However, when the compressive and tensile strengths and the static modulus of elasticity have been determined a re-evaluation of an ASR affected structure can be performed.
Keywords
Aggregate
alkali-silica reaction
concrete
durability
mechanical properties
INTRODUCTION
There are many concrete structures which suffer from alkali-silica reaction (ASR) due to the reaction of certain types of aggregate. The aggregates can be fast reacting such as opal, chalcedony, flint, chert, or they can be slow reacting such as greywacke, quartzite, granite, quartz-porphyrite, rhyolite, and others. The reaction starts with sufficient supply of alkalis in aqueous environment. Fast reacting aggregates have been investigated very thoroughly whereas slow reacting aggregates need more investigations. The treacherous feature of slow reacting aggregates is that the damage may manifest itself only after 15 to 20 years and that the testing of aggregates is sometimes ambiguous. That is why the research on this subject has been carried out. Frequently, the expansion behavior is of main interest. However, as soon as existing structures have to be evaluated more information on mechanical properties is necessary. The design engineer should know the strength of the deteriorated material and the modulus of elasticity, sometimes also the creep behavior. A literature review has been presented in [1].
However, the mineralogical composition of rocks varies strongly from place to place. That is why three slow reacting aggregates from Germany have been investigated. In this study the expansion, the tensile and compressive strength, and the modulus of elasticity have been measured during the exposure in a fog room at 40 °C for 560 days. The results will be reported and discussed with respect to structural design.
TESTING PROGRAM
Overview
The testing program considered three concrete mixes with one binder and three types of aggregates which were supposed to be slow reacting with respect to ASR reaction. The slow reacting aggregates have been selected because this type of aggregates causes severe problems in Germany in the last years [2] and not much is known about the ASR degradation. Exposure tests were carried out for 560 days. The expansion was measured as well as the dynamic and static modulus of elasticity, the compressive strength and the tensile strength. The specimens were inspected visually with regard to the extent of cracking.
Specimens and exposure condition
Beams with the dimensions of 100 mm x 100 mm x 500 mm (volume/surface ratio = 23 mm) are used for the determination of the expansion and of the dynamic elastic modulus with the resonance frequency method (Grindo Sonic device). Cubes of 300 mm edge length are cast for the visual inspection of crack formation. Cylinders with 300 mm height and 150 mm diameter (volume/surface ratio = 30 mm) are tested in compression for the modulus of elasticity and in splitting to get the tensile strength. Cubes with 150 mm edge length (volume/surface ratio = 25 mm) serve for the determination of the compressive strength. All specimen shapes and sizes follow the German guideline [3].
The specimens were stored after casting at 20 °C and 95 % RH for one day. Thereafter they were placed in a fog room at (40 ± 2.0)°C for 560 days. Measurements of the expansion and the dynamic modulus of elasticity were taken at 2, 7, 28, 35, 70, 140, 280, and 560 days. At the same dates, the specimens were visually inspected. Compressive strength, tensile strength, and the static modulus of elasticity were determined at 35, 70, 140, 280, and 560 days.
Materials
The crushed aggregates were greywacke (GW), quartz-porphyrite (QP) and crushed gravel (from boulders) from the Upper Rhine valley (UR). The composition of the aggregates has been described elsewhere [4,5]. The binder was a Portland cement CEM I 32.5 R [6] with 1.51 % K2O and 0.25 % Na2O which makes the Na2O equivalent 1.24 % by mass.
The three concrete mixes are the same except the type of aggregate. The cement content is 400 kg/m³, the water-cement ratio amounts to 0.45, and the grading curve follows almost the Fuller curve with a maximum grain size of 16 mm according to DIN 1045-2 [7]. 30 % by vol. of the aggregates consist of non-reactive sand up to 2 mm grain size while 40 % with grain size 2-8 mm and 30 % of grain size 8-16 mm stem from the reactive rocks.
RESULTS
Expansion and dynamic modulus of elasticity
The mean expansion and the dynamic modulus of elasticity of three beams are plotted in the following graphs as function of exposure time. Fig. Fig.1 shows the measurements on the concrete with greywacke aggregate.
Fig.1 Expansion and dynamic modulus of elasticity of greywacke concrete
The arrow marks the time when a crack with ≥ 0.2 mm width appeared at the edge of the 300 mm cube. This happened at 280 days storage. The expansion is small and reaches 0.2 mm/m after 560 days. There are two criteria for alkali sensitive aggregates according to DAfStb Guideline [3]:one is the expansion of the beam which should not be more than 0.6 mm/m after 9 months, the other is the edge cracking of the cube. The first is fulfilled, the second is not. That means the greywacke has to be considered reactive. The lines of the dynamic modulus of elasticity show a steep increase during the first 28 days which is the result of the continuing hydration. Thereafter, only a small increase can be seen. The curves are rather smooth and do not show any discontinuities.
Fig. 2 shows an analogous graph for the concretes made of quartz-porphyrite aggregates. The expansion of the concrete reaches the critical mark of 0.6 mm/m after 280 days and increases up to 0.75 mm/m after 560 days. It shows also a crack ≥ 0.2 mm after 224 days.
Fig. 2 Expansion and dynamic modulus of elasticity of quartz-porphyrite concrete
Fig. 3 gives the results for the crushed gravel concretes. The expansion curve of the concrete shows a slow increase up to 140 days, then a steep increase up to 224 days and after it stays constant. Cracking occurs after 168 days. The dynamic modulus of elasticity of the concrete shows some reverse sign of the expansion curve since it decays when the expansion increases strongly.
Fig. 3 Expansion and dynamic modulus of elasticity of crushed gravel concrete
Strength and static modulus of elasticity
The compressive strength is determined on 150 mm cubes, whereas the tensile strength is gained on cylinders by splitting and the static modulus of elasticity is measured by compression of cylinders. All values are the mean of three specimens. Fig. 4 shows the results of the concretes with greywacke aggregates. The values of the tensile strength and the static modulus of elasticity are normalized with respect to the averaged measured value after 35 days of fog room storage at 40 °C (fctm,35cd and ECm,35d, resp.). The compressive strength is plotted in absolute numbers in order not to mix up the lines. It can be seen that the modulus of elasticity of the concrete stays constant up to 140 days and then drops to 50 % after 560 days.
Fig. 4 Normalized tensile strength and static modulus of elasticity of greywacke concrete on the left ordinate, compressive strength at the right ordinate
The tensile strength increases a little up to 140 days and then decays to 67 %. The compressive strength of concrete is reduced by 5 MPa after 560 days.
Fig. 5 shows the results with quartz-porphyrite concrete. Tensile strength and modulus of elasticity of the concrete decrease by 32 and 54 %, resp., after 560 days. The compressive strength of the concrete is reduced by 2 MPa after 560 days.
Fig. 5 Normalized tensile strength and static modulus of elasticity of quartz-porphyrite concrete on the left ordinate, compressive strength at the right ordinate
Finally, the results of the crushed gravel concrete are presented in Fig. 6. The decay of the modulus of elasticity of the concrete is very obvious during the first 140 days. After that time it stays constant. The same applies for the tensile strength. The compressive strength of the concrete is reduced by 3 MPa.
Fig. 6 Normalized tensile strength and static modulus of elasticity of crushed gravel concrete on the left ordinate, compressive strength at the right ordinate
DISCUSSION
Compressive strength
As Figs. 4 to 6 indicate the compressive strength is the least affected by the ASR. The slow reacting rocks become cracked due to reaction of the alkalis with the rock minerals. Greywacke showed in the pristine state veins of pyrite and calcite which were dissolved during the exposure. Veins in quartz-porphyrite were originally filled with chlorite and hematite which was also dissolved during the exposure [4]. The dissolution creates open cracks in the grains which are compressed at compressive loading, however, the compressive strength does not suffer much from such a degradation process. The strength loss of the greywacke concrete with amounts to about 7 % after 560 days, the one of quartz-porphyrite concrete is 2 %, and the one of crushed gravel is 4 %.
Tensile strength
When the grains have got cracks due to the dissolution of cementing substances an important component of the concrete is weakened. Opposite to compressive loading tensile loading causes an opening and propagation of the crack. Figs. 4 to 6 show the relation between tensile strength and exposure time. The concrete with greywacke aggregate lost 33 % of tensile strength, with quartz-porphyrite 32 %, and with crushed gravel 21 %. Comparing the strength losses at compressive and tensile loading the loss at the latter loading is 3 to 5 times higher. The observation that the tensile strength is much more affected by ASR than the compressive strength has been made at structures in service several times before [8], however, the observations were not made on slow reacting aggregates.
Modulus of elasticity
The modulus of elasticity has been measured dynamically and statically. Figs. Fig.1 to 3 show the dynamic modulus of elasticity as function of exposure time whereas Figs. 4 to 6 contain the static modulus of elasticity in normalized form. The dynamic modulus of greywacke concrete seems not to be affected by the exposure, also quartz-porphyrite concrete seems unaffected. The dynamic modulus of crushed gravel concrete shows a decay of 20 % in the course of exposure but recovers later again. This result means that the dynamic modulus of elasticity is not very sensitive to the alkali reaction and is therefore not a good indicator for ASR. The reason for this fact is the type of loading which is, from a mechanical point of view, negligibly low and the mass and geometry of the concrete specimen plays an important role in the test. One can also assume that ettringite crystals grow into the cracks and increase the stiffness but, to a lesser amount, the strength.
In contrast to the measurement of the dynamic modulus of elasticity, the static modulus of elasticity is determined from the stress-strain diagram between 5 and 30 % of the compressive strength, i.e. at a much higher stress level. Cracks in the aggregate grains are compressed and therefore the static modulus could decrease with the exposure time since the dissolution process is progressing. Figs. 4 to 6 show how the static modulus develops during 560 days of storage in the fog room. With CEM I as a binder, the static modulus of greywacke concrete diminishes by 51 %, of quartz-porphyrite concrete by 54 %, and of crushed gravel concrete by 50 %. The comparison between the two moduli of elasticity is made by dividing the static modulus by the dynamic modulus for the exposure time between 35 and 560 days. The results are plotted in Fig. 7 for the concretes tested.
Fig. 7 Ratio of static to dynamic modulus of elasticity for the concretes tested as function of exposure time
When the measurements started the ratio of the two moduli amounts to about 0.90 for the concretes. These values are in accordance with literature findings [9]. With increasing exposure time, the ratio decreases substantially. For the concretes the ratios come to 0.42 for greywacke concrete, to 0.39 for quartz-porphyrite concrete and to 0.45 for crushed gravel from the Upper Rhine valley. The numbers reveal the sensitivity of this property with respect to the alkali reaction. It is suggested to use either the static modulus of elasticity as indicator for ASR degradation or the ratio between static and dynamic modulus. This result is in accordance with the Stiffness Degradation Test [10,11] and agrees well with the deliberations in the Canadian Guide [12]. It shows also that the proposal from Siebel&Dahms [13] to use the dynamic modulus of elasticity as indicator for the diagnosis of ASR affected concrete is questionable.
EFFECTS OF RESULTS ON STRUCTURAL DESIGN
When an experimental assessment of ASR affected concrete has been carried out in which compressive strength, tensile strength, and static modulus of elasticity has been determined the structural designer can use the results and reanalyze the structure. However, when only compressive strength and dynamic modulus of elasticity have been measured there may be a great discrepancy between the analysis and the true bearing capacity. In a conventional analysis, all mechanical properties are taken as a function of compressive strength. These are the tensile strength, the bond strength, and the modulus of elasticity. It has been shown in the previous sections that compressive strength is least affected by ASR and that the other properties are much more affected. That means that considering compressive strength as the standard property only can lead to a large error as pointed out in [14]. Two examples will be given.
Almost every structure is not only loaded in bending but also in shear. And the shear strength of structures depends, especially of those structures which do not have a shear reinforcement like many slabs, on the tensile strength of concrete. The mean tensile strength of normal strength concrete is calculated as
(1)
withfck the characteristic value of the compressive strength [15]. If the compressive strength is assumed too high the tensile strength will follow. But it has been shown in the previous sections that that would lead to wrong results.
When the deflection of a structure is essential for the structural behavior, such as of cantilever beams, or flat roofs, the static modulus of elasticity has to be known for the analysis. Once again, the mean value of the modulus is a function of the compressive strength according to [15]
(2)
Table 1 shows the measured and calculated mechanical properties derived from the mean compressive strength. It can be seen that the calculated tensile strength is always larger than the measured tensile strength by 27 % for GW concrete, 17 % for QP concrete and 3 % for UR concrete. The modulus of elasticity is overestimated by 63 %, 76 % and 67 %, resp., for the three concretes. When only compressive strength is known the actual modulus of elasticity is greatly overestimated. It has also been shown that the determination of the dynamic modulus of elasticity would also lead to erroneous results.
Table 1
Measured and calculated mechanical properties
¹acc. to Eq. (1),
²acc. to Eq. (2),
³fcm = 0.8 fcm,cube fck = fcm - 8
The examples show that a reevaluation of an ASR affected structure is only realistically feasible if tensile strength, compressive strength and static modulus of elasticity of the relevant concrete are known.
CONCLUSIONS
The investigations with three types of slow reacting aggregates and Portland cement have brought up some new facts about alkali-silica reaction which have been found during 560 days of exposure in the fog room:
a) Compressive strength is much less affected by ASR than tensile strength and static modulus of elasticity.
b) Tensile strength and static modulus of elasticity can decay by about 50 % with Portland cement with an alkali content of 1.24 % by mass Na2Oequ. during 560 days of exposure.
c) The dynamic modulus of elasticity has not characterized the detrimental action of ASR. It is questionable whether it can be used as a design property of concrete which suffers from ASR.
d) The prediction of tensile strength and static modulus of elasticity from measured compressive strength leads to erroneous results.
ACKNOWLEDGEMENTS
The reported investigations have been financially supported by the German Science Community (DFG) which is gratefully acknowledged by the authors.
REFERENCES
1. Reinhardt, H.W., Mielich, O. Mechanical properties of ASR damaged concrete with CEM I cement and phonolite rock powder, submitted for publication.
2. Stark, J. Alkali-silicareaction (in German). Finger Institute, Weimar: Bulletin of F.A; 2008.
3. Preventive measures against alkali-silica reaction in concrete (in German). Berlin: DAfStb Guideline; 2007.
4. Reinhardt, H.W., Mielich, O. A fracture mechanics approach to the crack formation in alkali sensitive grains. Cement and Concrete Res.. 2011; 41(3):255–262.
5. Mielich, O., Contribution to damage mechanisms in concrete with slow reactive alkali sensitive aggregates (in German), DAfStb Bulletin 583, Berlin. 2010.
6. EN 197-1 "Cement – Part 1: Composition, specifications and conformity criteria for common cements. CEN; 2000.
7. , DIN 1045-2: Concrete, reinforced and prestressed concrete structures – Part 2: Concrete – Specification, properties, production and conformity (in German). 2001. [DIN].
8. Siemes, T., Visser, J., Low tensile strength in older concrete structures with alkali-silica reaction. 11th ICAAR. 2000:1029–1038. [Québec].
9. Neville, A.M. Properties of concrete, 4th ed. Burnt Mill: Longman Group; 1995.
10. Chrisp, T.M., Wood, J.G.M., Norris, P., Towards quantification of microstructural damage in AAR deteriorated concrete. Int. Conf. on Recent Developments in the Fracture of Concrete and Rocks. Cardiff, 1989:419–427.
11. Chrisp, T.M., Waldron, P., Wood, J.G.M. Development of a non-destructive test to quantify damage in deteriorated concrete. Mag. Concrete Res.. 1993; 45:247–256.
12. Canadian A864-00 Guide to the Evaluation and Management of Concrete Structures Affected by Alkali-Aggregate Reaction
, CSA International, Toronto Feb. 2000, Reaffirmed 2005
13. Siebel, E., Dahms, J. Assessment of structures with respect to a damaging alkali-silica reaction (in German). Beton. 1997; 47(9):533–537.
14. den Uijl, J.A., Kaptijn, N., Walraven, J.C., Shear resistance of flat slab bridges affected by ASR. 11th ICAAR. 2000:1129–1138. [Québec].
15. EN 1992-1-1:2011-01, Eurocode 2 Design of concrete structures
Influence of alkali silica reaction on the chemistry of pore solutions in mortars with and without lithium ions
T. Kim and J. Olekkim312@purdue.edu, Purdue University, School of Civil Engineering, 550 Stadium Mall Drive, West Lafayette, Indiana-47907, USA,. E-mail address: olek@purdue.edu
This paper presents the results of the investigation of the chemistry of pore solutions of mortars containing both reactive and non-reactive aggregate. The effects of lithium ions (Li+) on chemical compositions of the pore solutions were also explored. In order to accelerate the A SR, all experiments were performed at 55 °C.The compositions of the pore solution were measured at short intervals for the period of up to 120 days. The results showed clear difference between the composition of the pore solution of the mortar with non-reactive aggregate (control mortar) and the mortar containing reactive aggregate. The concentrations of Na+, K+ and OH− ions in the reactive systems without Li+ ions continuously decreased until they reached certain threshold level at which point they stabilized. The time of the stabilization of chemistry of pore solution in the reactive systems coincided with the time of the ultimate expansion of mortar bars of the same composition.When LiNO3 was added as a source of Li+ ions, the consumption of Na+ and K+ ions was significantly reduced. However, the concentrations of Li+ and OH− ions in the pore solution decreased over time.
Keywords
Alkali-silica reaction
pore solution
lithium ions
INTRODUCTION AND RESEARCH SIGNIFICANCE
Alkali-silica reaction (ASR) is the chemical reaction that occurs between alkali cations and hydroxyl ions in the pore solution of hydrated cement paste and certain reactive silica phases present in the aggregates used in concrete. The product of that reaction, called ASR gel, can expand by imbibing water and thus lead to the expansion and cracking of concrete structures.
The ASR process can be divided into two main steps: (a) the chemical reaction to form ASR gel and (b) expansion of gels induced by taking-up water. Although fundamental mechanisms of ASR have been researched for more than 70 years (since the pioneering work of Stanton [1]), they are still not fully understood. This is particularly true for the second step, which remains controversial despite the fact that several different mechanisms have been proposed over the years (Hansen [2], Mcgowan and Vivian [3], Powers and Steinour [4], Chatterji, et al. [5], Rodrigues, et al. [6] and Garcia-Diaz, et al. [7]). With respect to the first step (reaction to form ASR gels), it is generally agreed that two distinct processes are involved: (a) breakdown of reactive silica during the attack by the OH− ions and (b) formation of ASR gels (Powers and Steinour [4] and Glasser and Kataoka [8]).
When reactive forms of silica are exposed to the highly alkaline solution, the hydroxyl ions (OH ) in the solution attack the siloxane groups ( ≡ Si − O−Si ≡ ) in the aggregate as illustrated by Eq. (1) to form the silanol groups. The neutral silanol groups (≡ Si − OH) further react with OH− ions to neutralize the basic pore solution as shown in Eq. (2).
(1)
(2)
The continuous attack of the silanol groups by OH− ions present in the pore solution results in formation of dissolved silica ions (mainly H3SiO4− and H2SiO4² −). Afterwards, these dissolved silica ions combine with potassium (K+), sodium (Na+) and calcium (Ca² +) ions, and produce the ASR gel.
In general, the chemical reactions involved in ASR can be treated as acid-base reactions. The acid reactant is silica in the aggregate and the basic reactants are potassium and sodium hydroxides present in the pore solution] The reaction medium is water and the reaction product is the calcium potassium (or sodium) silicate hydrate. Therefore, rigorous observation of the changes in the chemistry of the pore solution in systems undergoing ASR can provide useful information regarding the potential ASR expansion or the degree (extent) of ASR over time.
The current paper describes changes in the chemistry of pore solutions as a function of time in reactive or non-reactive aggregate mortars. In addition, the effects of lithium ions (Li+) on the chemical composition of the pore solutions were also investigated. The beneficial effect of Li+ ions on ASR expansion was first reported by McCoy and Caldwell [9] and has been studied for more than 60 years. Specifically, the effects of following parameters on the chemistry of pore solution are described: (a) length of curing time (up to 120 days), (b) presence of reactive or non-reactive aggregates and (c) presence of lithium ions in the pore solution.
MATERIALS
High-alkali Type I portland cement (1.04% Na2Oeq) was used in the study. The composition of the cement is summarized in Table 1. One non-reactive (NR) aggregate (ASTM C 778 standard Ottawa graded sand) and one reactive (R) aggregate (lobe sand) were used to prepare mortar specimens. The specific gravities of the Ottawa sand and Jobe sand were, respectively 2.65 and 2.55 and their absorptions were, respectively 0.0% and 1.0%. The reagent grade lithium nitrate (LiNO3) was used as the source of lithium (Li+) ions.
Table 1
Oxide and compound compositions of cement
EXPERIMENTAL METHODS
The experiments were performed on two types of mortar specimens: (a) cylinders (50 mm in diameter and 63 mm in height) for analyses of pore solutions and (b) prisms (25 mm × 25 mm × 285 mm) for expansion measurements.
The cylindrical specimens were prepared with both reactive (R) aggregate and non-reactive (NR) aggregate. These specimens were cured (100% RH and at the temperature of 55 °C) for period of time raging from 1 to 120 days. The information related to composition of the mortars is given in Table 2. The water-to-cement ratio used was 0.55 and the total aggregate content of the mortars was also held constant (at 50% the No. 30 (0.595 mm) sieve and retained in the No. 40 (0.425 mm) sieve and (b) passing the No. 40 (0.425 mm) sieve and retained in the No. 50 (0.297 mm) sieve were used. The quantity of each aggregate was the same (25% by volume of the mixture).
Table 2
Composition of mortars, initial lithium dosages and storage temperature
NR: Nonreactive aggregate (Ottawa sand)
R: Reactive aggregate (Jobe sand)
100% Li: [Li]/[Na + K] = 0.74
⁎Due to overall low level of initial concentrations of all ions, the molality (m) and molarity (M) of the solutions were assumed to be approximately equal
For all mortar specimens containing Li+ ions, the lithium nitrate (LiNO3) admixture was introduced by dissolving it in the mix water. The amount of LiNO3 was selected to provide 0.74 molar ratio of lithium to alkali, which is traditionally considered to be a sufficient for the effective control of the ASR expansion (Feng, et al. [10], Thomas, et al. [11] and Millard and Kurtis [12]). In the present paper, to simplify the naming scheme for individual series of mortars, the amount of LiNOa needed to ensure 0.74 molar ratio of lithium to alkali is referred to as 100% dosage of LiNO3
(see note under Table 2).
All mortars were prepared and mixed in accordance with the provisions of ASTM C 305. When preparing cylindrical specimens, all mortars were cast into ointment jars (50 mm in diameter and 63 mm in length), which were then capped, sealed tightly and placed in the moist room where they remained for 24 hours. After 24 hours, the jars (with mortars still inside) were removed from the moist room and placed in the air tight container, the interior of which was kept in a saturated condition during the test. The saturated condition was maintained by placing a thin (10 mm thick) layer of water at the bottom of the container. This container was stored in the convection oven (maintained at 55 °C) for up to 120 days.
The cylindrical mortar samples were removed from the jars after the prescribed curing period, weighted (to determine their initial weight) and placed in the pore solution extraction device described elsewhere (Barney back Jr and Diamond [13]). After the pore solution extraction was completed, the remnants of the specimens were removed from the device, broken up into several pieces, flushed with acetone (to stop the hydration) and dried at 105 °C to a constant weight. The difference between the initial weight of the sample and its dried weight was then calculated and assumed to represent the amount of evaporable water in the sample. This calculated amount of the evaporable water was used to adjust the measured concentration of ions in the pore solution using the Eq. (3). This adjustment allowed for direct comparison of the concentrations of ions from samples at various degrees of hydration.
(3)
where, CAdjsuted and CMeasured are, respectively, the adjusted concentration (mol/L) and the measured concentration (mol/L); weva and winitial are, respectively, the amount of evaporable water (g) and the amount of the initial mix water (g) in the mortar specimens.
The pore solutions were chemically analyzed for Na+, K+ and Li+ ions using the atomic absorption/emission spectrophotometer (Varian® SpectrAA − 20). The concentrations of OH− ions in the solutions were determined by titration with a standardized hydrochloric acid using methyl red indicator.
The expansion tests were performed using the previously mentioned prismatic bars. Only the reactive aggregate (Jobe sand) was used to prepare the bars. Immediately after casting the molds, containing fresh mortar were placed in a 100% RH curing room and stored there (at 23 °C) for 24 hours. After 24 hours of curing, the mortar prisms were removed from the molds and placed horizontally (on the support rods) in the air-tight container. Shallow layer of water (about 10 mm thick) was placed at the bottom of container to maintain a saturated condition. The containers were then placed in the convection oven maintained at 55 °C. The length changes were monitored daily for a period of up to 120 days.
RESULTS AND ANALYSIS
Evaporable water contents of mortar specimens
As expected, the evaporable water contents (calculated as the fraction of the initial amount of mix water (weva / winitial) in the mortar specimens) tend to decrease over time (Table 3) due to the removal of the free water by continuous hydration. In the NR mortars, the biggest drop in evaporable water (from 1 to 0.748 or about 25.2%) was observed after just 1 day of hydration. Similar 1 day drop (from 1 to 0.762 or about 23.8%) was also observed in mortars with the reactive (R) aggregate. The subsequent reduction in the evaporable water content was more gradual, reaching (after 100 days) values of about 0.64 to 0.65 for mortars without lithium admixture and about 0.64 to 0.67 for corresponding mortars with lithium admixture.
Table 3
Water content and chemistry of pore solution in mortar specimens (all concentrations were adjusted using
Vous avez atteint la fin de cet aperçu. Inscrivez-vous pour en lire plus !
Page 1 sur 1