The Effect Of Reinforcement On The Risk Of Cracking in Hardening High Strength Concrete

Transcription

1 The Effect Of Reinforcement On The Risk Of Cracking in Hardening High Strength Concrete M Sule K van Breugel Delft University of Technology The Netherlands Summary: High strength concrete (HSC) undergoes volume changes during hardening. These changes are generated by thermal effects and autogenous shrinkage and can lead to cracking if they are restrained. An engineering tool in order to control cracking is reinforcement. There are indications that the presence of reinforcement postpones the occurrence of major cracks at early age. This matter has now been investigated with the help of a Temperature Stress Testing Machine (TSTM). The reinforcement percentage (0%, 0.75%, 1.34%, 3.01%) and reinforcement configuration (1 bar and 4 bars) have been varied in a HSC with a water to cement ratio The results show that in specimens with four reinforcement bars the formation of microcracks increases the strain capacity of a concrete element before major cracks occur. Specimens with one reinforcement bar crack as sudden as plain concrete specimens. The influence of reinforcement percentage and configuration on the development of self-induced stresses is discussed. Furthermore the question is dealt with how the gain in strain capacity can be used for judging the risk of cracking. In this respect a strain criterion for designing durable HSC seems to be more suitable than a stress criterion and is formulated in this paper. Keywords: High strength concrete, reinforcement, Temperature Stress Testing Machine, early-age cracking, cracking criterion 1 INTRODUCTION High strength concrete (HSC) has been developed for carrying higher loads than normal strength concrete. This is realized by applying lower water-cement ratios and admixtures. A low water-cement ratio leads to a denser microstructure, which makes HSC an ideal material for durable structures. However, HSC undergoes additional volume changes during hardening, so called autogenous shrinkage. Under restrained conditions this makes HSC even more prone to cracking in the early phase. With reinforcement this cracking can be controlled. In order to enhance our understanding of the influence and effect of reinforcement on early-age cracking, an experimental research project has been conducted. Unreinforced and reinforced concrete specimens were tested in a Temperature Stress Testing Machine (TSTM). In this test device the concrete hardened under sealed conditions and the deformations were 100% restrained. Consequently stresses were generated in the specimen, which could lead to cracking when the tensile strength (or strain capacity) of the concrete was exceeded. The results presented here refer to a HSC that was cured semi-adiabatically. Test variables were the reinforcement percentages and configurations. In parallel with the TSTM test the free deformation of plain and reinforced concrete was measured. From these measurements the effect of the reinforcement on the free deformation could be deduced. Furthermore, the development of the modulus of elasticity and the cube compressive strength were monitored. The results indicated that a substantial gain in strain before the occurrence of major cracks can be achieved by applying reinforcement. In order to allow for this gain in strain capacity a stress-strain diagram reinforced concrete specimen in uniaxial tension is suggested. With the help of the proposed stress-strain relationship a strain criterion could be formulated for judging the risk of reinforced concrete at early age. 9DBMC-2002 Paper 150 Page 1

2 2 EXPERIMENTS 2.1 Programme The stress development and the cracking behaviour of a reinforced concrete element was tested in a Temperature Stress Testing Machine (TSTM) with full restraint of the deformations. The free deformations of plain concrete and reinforced concrete were measured in dummies. Table 1 shows the concrete composition of the tested HSC (w/c ratio = 0.33). Specimens of each experiment were cast of the same batch of concrete and cured semi-adiabatically under sealed conditions. Water Table 1. Concrete composition per m 3 CEM III/ B 42.5 LH HS CEM I 52.5 R Silicol SL (50/50 slurry) Lignosulphonate Naphtalene sulphonate Gravel 4-16 mm Sand 0 4 mm kg kg kg 50.0 kg 0.9 kg 9.5 kg kg kg The TSTM specimens were tested with three different percentages of reinforcement, 0.75%, 1.34% and 3.02%, respectively, and without reinforcement. Each reinforcement percentage (ω) was realised with two different configurations of reinforcement bars (rebars): 4 6 or 1 12, 4 8 or 1 16, and, 4 12 or 1 25, respectively (Table 2). code Table 2. List of experiments reinforcement configuration reinforcement percentage (ω) A - - 4B % 1B % 4C % 1C % 4D % 1D % 2.2 Test equipment The TSTM used for these experiments was used in combination with two dummies and cubes (Lura et al. 2001). One dummy was used to create reference conditions for the temperature development during hardening. In the second dummy the effect of reinforcement on the free deformation was tested. The cube compressive strength and the modulus of elasticity were tested at different early ages in order to monitor the development of concrete properties. Strain gages measured the self-induced stresses in reinforcement bars with a diameter of 16 mm and 25 mm The Autogenous Deformation Testing Machine (Dummy) During hardening, the concrete temperature and the load-independent deformation were measured in an unrestrained specimen: the dummy (ADTM). The temperature was measured with thermocouples inserted in the concrete at different positions immediately after casting. So the temperature measurements could be started within approximately 30 minutes after mixing. The load-independent deformations were measured with LVDT s. They measured the length changes of the dummies over a measuring length of 750 mm. This was done with the help of small steel bars perpendicular to the measuring direction. The bars were embedded in the specimen and passed the mould through holes (comparable to Figure 2). 9DBMC-2002 Paper 150 Page 2

3 2.2.2 The Temperature Stress Testing Machine (TSTM) The TSTM is a horizontal steel frame in which hardening concrete specimens can be loaded in compression and in tension under various hardening conditions (Figure 2). Both load-controlled and deformation-controlled experiments can be performed under measured or prescribed thermal conditions. embedded steel bar reinforcement bar LVDT s insulation actuator LVDT s 750 load cell claws Figure 2. The Temperature Stress Testing Machine (top view) To perform experiments in tension, a dovetailed interlock is used between the concrete specimen and the frame. Two steel claws hold the dovetailed specimen. One of the claws is fixed to the frame, the other is placed on roller-bearings and can be moved with the hydraulic actuator. Loads are measured with a load cell Strain gages Strain gages measured the stresses generated in the reinforcement bar during the hardening phase. The gages and the connecting wires were placed in notches milled along the longitudinal welds. As the notches reduced the cross section considerably, particularly in case of small rebars, the strain was only measured in reinforcement bars with bar diameters of 16 mm and 25 mm. Combined strain gages were chosen and placed opposite to each other in a cross section eliminating the effect of possible bending of the reinforcement in the measurements. One gage of this combination measured the deformation in longitudinal (Fig. 3, gage 1 and 3, respectively) and the other in transverse direction (Fig. 3, gage 2 and 4, respectively). Two of these combined gages were placed opposite to each other in one cross section forming a full bridge. gages (1,2) + +ε 1 µ 4 Usig UBat gages (3,4) µ 2 3 +ε - Figure 3. Position of gages in the cross section of the reinforcement bar (left) and whole bridge formed by two combined gages (right) 3 RESULTS 3.1 Temperature development The heat evolution is an elegant parameter to which the development of concrete properties can be related (cube compressive strength, modulus of elasticity etc.). This heat evolution can easily be determined from the temperature curves during hydration with the help of an adequate maturity function (e.g. Arrhenius function). Thermocouples measured the temperature in all specimens of each experiment. In Fig. 4 the average temperature of all seven experiments is plotted. With a coefficient of thermal dilation α Τ, the thermal deformations ε T can be calculated with Eq. 1. 9DBMC-2002 Paper 150 Page 3

4 55 T max temperature [ C] mean value highest value lowest value ε T =α Τ T [1] α Τ coefficient of thermal dilation T temperature difference 15 t max Figure 4. Average temperature development in the plain dummies of all seven experiments with scatter 3.2 Development of cube compressive strength The cube compressive strength was tested after different ages (8h, 24h, 30h, 48h, 168h and 672h). The experimental results are shown in Fig. 5. In addition, the cube compressive strength is plotted that was calculated according to MC90 (Eq. 2 and 3). As can be seen a rather poor agreement exists between the values calculated with MC90 and the experimental results. Obviously MC90 is not very accurate in case of HSC at very early age cube compressive strength [MPa] experimental results MC f cm (t)=β cc (t)f cm [2] ßcc (t) = exp{s[1 28 1/ 2 ( ) ]} t [3] f cm mean compressive strength after 28 days f cm (t) mean compressive strength at the age of t days β cc coefficient depending on the age of concrete (t) s parameter, depending on the type of cement (0.2) Figure 5. Cube compressive strength tested in all experiments and calculated according to MC Development of modulus of elasticity The modulus of elasticity has been tested after 1 day, 3 days, 7 days and 28 days. As for the cube compressive strength, these results and the results found with the MC90 (Eq. 4 and 5) are shown in Fig. 6. Whereas the results of the cube compressive strength according to MC90 did not fit the experimental results, the tested modulus of elasticity could be well approximated with the help of Eqs. 4 and 5. 9DBMC-2002 Paper 150 Page 4

5 modulus of elasticity [GPa] experimental results MC E ci (t) = β E (t) E ci [4] β E (t) = [β cc (t)] 1/2 [5] E ci modulus of elasticity after 28 days E ci (t) modulus of elasticity at the age of t days β E coefficient depending on the age of concrete (t) Figure 6. Development of modulus of elasticity from experiments and calculated according to MC Development of free deformation In all experiments the free deformation has been measured in a plain and a reinforced dummy specimen. As the coefficient of thermal dilation α T is an intrinsic material property of the tested HSC and the temperature development of all specimens was about the same (Fig. 4) the deformations measured in all seven plain specimens is also nearly the same (Fig. 7, left). The effect of reinforcement on the free deformation can be seen in Fig. 7, right. As a first approximation the free deformation ε c (t) of concrete can be calculated as follows ε c (t)=σ ε [6] ε= ε T (T)+ ε a (α,t) [7] strain [ ] in plain specimens mean value highest value strain [ ] in reinforced specimens mean value highest value lowest value ε c(144) lowest value ε c(144) Figure 7. Mean value of free deformation measured in all experiments with scatter on plain dummies (left) and reinforced dummies (right), HSC cured semi-adiabatically ε T (T) thermal strain increment depending on the temperature T ε a (α,t).autogenous strain increment depending on the temperature T and the degree of hydration α (Note: in fact autogenous deformations are also temperature dependent!) There are three main differences between the measurements in plain and in reinforced dummy specimens. Firstly, reinforced specimens expand more. Secondly, reinforced specimens pass the zero strain line later. Thirdly, measurements in reinforced specimens vary much more than in plain specimens (due to different reinforcement percentages and configurations). 9DBMC-2002 Paper 150 Page 5

6 As expected the total free deformation measured after 144 hours in reinforced specimens, i.c. ε c (144)=0.455*10-6, is smaller than in plain specimens (ε c (144)=0.468*10-6 ). On average, reinforcement restrains the free deformation of the tested HSC by 2.8 %. It thus appears that the presence of reinforcement does not to influence the free deformation of the samples very much. 3.5 Stress development in TSTM Table 3 shows the measured stress values for all experiments. It appears that the compressive stresses σ D,total (i.e. the force measured in the TSTM divided by specimen s cross section) at time t D when the compressive stress is maximal, increases with increasing reinforcement percentage (see also Figure 8, left). As a consequence the moment t 02 at which the stress development passes the zero-stress state in the cooling phase is postponed. In specimens with four reinforcement bars microcracks form before the first through-crack. This is indicated by a bending curve. In specimens with the smallest bar diameter (6 mm) the yield stress of the reinforcement was exceeded immediately after cracking of the concrete. Consequently this specimen failed as a plain specimen. In any case, specimens with one reinforcement bar cracked in the same sudden way as plain concrete specimens (t el t cr, Table 4). Table 3. measured stress values for all experiments code t 01 t D / σ D,total t 02 t el t cr / σ cr,total A / / B / / B / / C / / C / / D / / D / / 2.67 t [h], σ [MPa]refering to A total =15000 mm 2 stress [MPa] 5.0 strain [ ] 0.4 σ cr,total 2.5 A σ D,total 4D micro-cracking 0.0 ε micro t D t 02 1D t el A: plain 4D:4 12(3.01%) 1D:1 25(3.27%) t cr t 02 t el t cr Figure 8. Influence of reinforcement percentage and configuration on the stress development and the resulting gain in strain (De micro ) in specimen 4D (4 12) Stress-inducing strain The stress-inducing strain is given for all experiments in Tab. 4. These values have been measured in the plain dummy of each experiment and they are defined as in Eqs The strain gain ε micro, i.e. the strain measured from the moment that the stress-strain curve started to flatten until cracking, is highest in the TSTM specimen of experiment 1B. Despite this single result the common trend of the results on the strain gain ε micro is that ε micro is about ten to twenty times bigger in specimens reinforced with four rebars compared with specimens reinforced with one rebar. For both reinforcement configurations it can be said that the post-elastic strain gain increases with the reinforcement percentage. 9DBMC-2002 Paper 150 Page 6

7 ε el = ε c (t el )- ε c (t 02 ) [8] ε cr = ε c (t cr )- ε c (t 02 ) [9] ε micro = ε cr - ε el [10] Table 4. measured strain values for all experiments code ω ε el ε cr ε micro η cr A B B C C D D reinforcement percentage ω[%], ε [ ], η cr [-] Stresses in concrete and in reinforcement In the previous section the effect of reinforcement on the stress and strain development in composite specimens is shown. Before cracking the stress in the concrete is fully dependent on the imposed strain, irrespective of the amount of reinforcement. Figure 9 shows the stresses in the concrete, the steel and the composite cross section of specimen 1D prior and after cracking. After the first crack occurred the stress in the concrete at the location of the crack is about zero. The further performance of the specimen is determined by the bond properties between steel and concrete stress [MPa] concrete TSTM steel (x1/30) s total Atotal s s As s c = [11] A c t 01 t D 24 t 02 t cr Figure 9. Stress development in TSTM specimen, in the cast-in reinforcement bar of 25 mm diameter and in the concrete cross section of test 1D The stresses in the rebar (σ s ) were calculated from the measured strain (see strain gages). In order to fit into the figure they were plotted with a factor 1/30. The strain gage happened to be where the specimen cracked for the first time. The concrete stresses in the concrete cross section (A c =A total -A s ) were calculated with Eq ADDITIONAL STRAIN CAPACITY PRIOR TO MAJOR CRACKS From the experimental results it could be seen that the strain capacity of specimens reinforced with 4 rebars is considerably increased caused by the formation of microcracks. It seems to be appropriate, therefore, to formulate a strain criterion in order to judge the risk of cracking at early age. Different researchers proposed diagrams for the stress-strain behaviour of early age concrete under tension. In normal strength concrete (NSC) a non-linearity can be found at approximately 40% of the failure load (Larson 2000). Hedlund (1996) found a 9DBMC-2002 Paper 150 Page 7

8 higher level for high performance concrete. In order to account for the non-linear stress-strain behaviour at high tensile stresses Jonasson (1994) and Hedlund (1996) developed the stress-strain relation shown in Figure 10, left. The MC90 handles for 28 days old concrete the diagram shown in Figure 10, right. σ 1 σ ct f ct,f ctm tensile strength f ct 1 αα ct 1 1 ε ε m 0 f ctm 0.9f ctm E ci ε ct α ct relative stress level above which non-linear stress-strain behaviour is present [-] ε m material strain ε 0 fictitious strain linearly related to the tensile strength, f ct /E c Figure 10. Non-linear stress-strain behaviour under tension according to Jonasson (1994) and Hedlund (1996), left and stress-strain diagram for uniaxial tension according to MC 90, right In our experiments we did not find any non-linearity in plain concrete specimens. One reason might be the high stiffness of the tested concrete mixture. Another reason might be the long measuring length (Hordijk 1991) resulting in snap back behaviour. However, in experiments on specimens with 4 reinforcement bars it was observed that the stress-strain curve became flatter at high stress levels. This flattening could be explained by the formation of microcracks. These microcracks can result in an additional strain ε micro. The values of ε micro can be taken from Tab. 4. According to Lokhorst (1998) plain concrete specimens crack at 0.75 times the actual tensile splitting strength (σ cr =f ct,sp (t cr )) at early age. This was taken into account when adjusting the stress-strain diagram of MC90 for a stress-strain diagram for early-age reinforced concrete (Figure 11). In this diagram the value of ε micro depends on the reinforcement ratio and configuration of the rebars in the cross section (see next section). σ ct 0.75f ct,sp E(t) ci ε el ε cr ε ct ε micro Figure 11. Stress-strain diagram for early age reinforced concrete under tension The reason why more microcracks form in specimens with four reinforcement bars is illustrated in Fig. 12. Cracks always start in the weakest part of the specimen. The surface zones, especially the corners of the specimen, are likely to be the weakest parts of the cross section. From there the crack grows to the centre of the specimen. In specimens with four reinforcement bars a crack growing from the corner will meet a reinforcement bar in an early stage of crack growth. Thanks to the reinforcement bar the stresses are redistributed and more microcracks are generated before the specimen really cracks (major cracks). The sum of the microcracks widths Σw micro constitute the additional strain capacity ε micro of the reinforced concrete specimen. Σwmicro w w cr Figure 12. Cracking in specimens with 4 rebars and 1 rebar (longitudinal and cross section) 9DBMC-2002 Paper 150 Page 8

9 5 CRACK CRITERION In literature (Rostasy 2000, Larson 2000) different criterions are used in order to quantify the risk of cracking in early age concrete. In plain concrete a stress criterion is mostly applied, relating the stress at the moment of cracking to the tensile splitting strength at the same moment (e.g. Lokhorst 1998). Our results show that due to reinforcement microcracks can form which enhance the strain capacity of concrete before major cracks occur. Therefore a strain enhancement factor, η cr, is introduced in order to indicate the proneness to cracking of reinforced specimens compared with plain concrete specimens (Eq. 12), viz.: ε cr,re inf orced η cr = [12] ε cr,plain ε cr,reinforced and ε cr,plain are the stress-inducing strains until the moment that major cracks are formed in a reinforced specimen and a plain specimen, respectively. In this study the value of ε cr,plain is to be taken from Table 4, i.e. ε cr,plain = ε el = (specimen A). The strain enhancement factor depends on the reinforcement percentage and configuration and has been calculated for all experiments in Table 4. The results show that the strain enhancement factor increases with increasing reinforcement percentage. Even more important appear to be the reinforcement configuration: four reinforcement bars are much more effective than only one bar. It is noticed that the value of ε cr,plain = is based on only one test. This might be the reason why the strain enhancement factor η cr for the specimen with only one rebar are below one because of a relatively high value of ε cr,plain. 6 DISCUSSION The effect of reinforcement on the risk of cracking has been investigated on seven single experiments. In order to verify trends more experiments are to be performed. Nevertheless the following trends are discussed below not considering the lowest reinforcement percentage (series B). 6.1 Restraining free deformation The experimental results on the dummies show that reinforcement restrains the free deformation. Assuming that the coefficient of thermal dilation of steel and concrete is about the same, the thermal shrinkage is not restrained. Therefore reinforcement has only an effect on autogenous shrinkage. This was also found by Sule et al. (2000a) in isothermal tests on different temperatures. Due to their bigger specific surface and their distribution in the cross section, four rebars restrain the autogenous shrinkage more effectively than one rebar. 6.2 Postponing the moment of stress free state t 02 The results summarized in Table 3 show that in semi-adiabatic tests the compressive stress (σ D,total= F TSTM /A total ) increases with increasing reinforcement percentage. When separating concrete and steel stresses (as in Fig. 9) it becomes obvious that in reinforced specimens only the steel stress increases while the concrete stress stays nearly the same as in plain specimens. However the compressive stresses σ D,total increase with the reinforcement percentage and the moment t 02 at which the composite specimen passes the stress-free state is postponed (t 02 : transition from compression to tension). Emborg & Bernander (1994) consider the mean temperature T(t 02 ) of the cross section at t 02 as a crucial parameter when calculating thermal stresses. The later the zero stress is passed, the smaller is the remaining imposed strain in the cooling phase that has still to be compared with the ultimate failure strain for judging the risk of cracking. So, the later the zero stress is passed, i.e. the lower the zero stress temperature, the lower the risk of cracking. In the case of reinforced specimens, however, postponing the moment when the stress-free state is passed does not decrease the risk of cracking. The stresses measured in the TSTM refer to the composite cross section. This means that when the stresses measured in the TSTM specimen passes the stress-free state (t 02 ) the concrete is already under tension while the reinforcement bar is under compression (Fig. 9). In this case the overall zero stress is not a good indicator of the remaining strain capacity of the concrete. 6.3 Increasing the strain capacity With increasing reinforcement percentage the strain capacity prior to major cracks increases. This increase is only very small for specimens with one reinforcement bar as they show about the same sudden cracking behaviour as plain specimens. Specimens with four reinforcement bars, however, show a substantial increase in strain capacity before the first through-crack. This is supposed to be due to microcracking. The increase in strain capacity can reach values up to (4D), which is 44% of the concrete tensile strain of plain concrete at an age of 28 days (Fig. 10 right). 9DBMC-2002 Paper 150 Page 9

10 6.4 Judging the increase of strain capacity Judging the increase of strain capacity with help of the strain enhancement factor (Eq. 12) reveals that specimens reinforced with one rebar perform less advantageous than plain specimens. In specimens with four rebars the strain capacity was found to increase substantially, i.e. by up to 48%. Although the strain gain ε micro is bigger in 4D (higher reinforcement percentage) than in 4C the total strain ε cr at the moment of cracking is lower. The reason is that the formation of microcracks start very early and consequently the value of ε el is relatively small. 6.5 Discussing microcracks in view of durability Figure 12 illustrates how microcracks form in specimens with four rebars, increasing the strain capacity of these specimens. In view of durability it is of primary importance to know if these microcracks impair the quality of the concrete cover. Via a permeable surface, water can reach the interconnected network of pores, microcracks and macrocracks. As water is the medium that plays a very important role in deterioration processes, such as sulphate attack, alkali-silica expansion and frost action, the high quality of the surface is indispensable in order to provide reliable long-term performance, especially when exposed to aggressive environments. Little information is available about the effect of microcracks in real sructures on the durability. It is still a matter of spectaculation to which extent these microcracks may close due to self healing. Sometimes surface cracks are compressed later on in the hydration process. Sule and van Breugel (2000b) investigated how these compressed cracks performed in a TSTM without the presence of water. They found that only very small cracks close again so that the specimen regains the strength to bear loads applied after a period of compression. Based on theoretical and experimental research Edvardsen (1999) investigated the effect of crack healing in structures subjected to water-pressure loads. She proposes permissible crack widths, which can be expected to obtain an almost total self- healing after a few weeks of water pressure exposure. Therefore it is important to control the crack width with reinforcement. 7 CONLUSION The experimental results (tests C and D) on reinforced TSTM specimens show that microcracks form before major cracks occur. As the formation of microcracks results in a gain in strain capacity, a strain enhancement factor has been formulated in order to judge the effect of reinforcement. It was found that specimens with four reinforcing bar performed better than plain specimens and specimen with only one bar. In specimens with 4 rebars the moment of the first through-crack could be postponed substantially: an increase in the strain capacity up to 48% has been found. In order to control the crack width in view of durability in early-age concrete further research is necessary. 8 ACKNOWLEDGMENTS The assistance of the laboratory technicians of Stevin laboratory is gratefully acknowledged. A special thanks goes to Mr A. van Rhijn for performing the experiments. This project can be realised thanks to the financial support of the Dutch Technology Foundation (STW) and the Ministry of Transport, Public Works and Water Management (RWS). 9 REFERENCES 1. CEB-FIP Model Code 1990, Comité Euro-International du Béton, Lausanne. 2. Edvardsen, C. 1999, Water permeability and autogenous healing of cracks in concrete, in ACI Material Journal, V96, No.4, July-August, pp Emborg, M. & Bernander, S. 1994, Thermal stresses computed by a method for manual calculations, in Crack Risk estimation. RILEM Symposium Thermal Cracking in Concrete at Early Ages, Proceedings 25. pp Hedlund, H. 1996, Stresses in high performance concrete due to temperature and moisture variations at early ages, Licentiate Thesis, Luleå, Sweden. 5. Hordijk, D.A. 1991, Local approach to fatigue of concrete, Doctoral Thesis, Delft, The Netherlands. 6. Jonasson, J.E. 1994, Modelling of temperature, moisture and stresss in young concrete, Doctoral Thesis, Luleå, Sweden. 7. Larson, M. 2000, Estimation of crack risk in early age concrete, Licentiate Thesis, Luleå, Sweden. 8. Lokhorst, S.J. 1998, Deformational behaviour of concrete influenced by hydration related changes of the microstructure, internal report, TU Delft. 9. Lura, P., Sule, M. & Breugel K. v Effect of water/cement ratio and curing temperature on early-age shrinkage and self-induced stresses of High Performance Concrete this proceedings. 10. Rostásy, F.S. & Krauβ M. 2000, Effects of thermomechanical properties of young concrete and their scatter on stress and cracking, International workshop on Control of cracking in early-age concrete, proceedings preprint, pp DBMC-2002 Paper 150 Page 10

11 11. Sule, M. & Breugel, K. v. 2000a, Cracking behaviour of reinforced concrete subjected to early-age shrinkage, international RILEM workshop on Shrinkage of concrete, Shrinkage 2000, proceedings, Paris, France, pp Sule, M. & Breugel, K. v. 2000b, Performance of compressed early temperature cracks in concrete cover in view of durability, 2nd international symposium on Cement and concrete technology in the 2000s, proceedings, Istanbul, Turkey, pp DBMC-2002 Paper 150 Page 11