- Email: [email protected]
Behavior of the reinforced concrete at cryogenic temperatures
Cryogenics 47 (2007) 517–525 www.elsevier.com/locate/cryogenics
Behavior of the reinforced concrete at cryogenic temperatures Lahlou Dahmani *, Amar Khenane, Salah Kaci Department of Civil Engineering, Faculty of Engineering Construction, University of Tizi-Ouzou, Tizi-Ouzou, Algeria Received 11 October 2006; received in revised form 23 July 2007; accepted 24 July 2007
Abstract Concrete has successfully been used at extremely low temperature ( 160 °C) for the storage of liquefied natural gas. Such use will induce a cycle of cooling every time a container is filled. In addition, lack of control, as for example a leak may produce a sudden temperature change. Some knowledge of concrete properties under these conditions is therefore necessary for successful design and operation. The first and basic cause of the failure of concrete is repeated freezing (thawing) of moisture contained in the pores, microcracks, and cavities of the concrete. On transition to ice, water existing in the free state in cracks increases in volume, expanding the recess in which freezing occurs. A reduction in strength below the initial value is to be expected and further cycle of freezing and thawing have a further marked effect. The main objective of this paper is to describe the principal reasons for the reduction in strength and structural damage (frost damage) of concrete following repeated freeze–thaw cycles. Some experimental work was carried out at the Institute of Cryogenics, University of Southampton, UK, to determine what happens to water in concrete during the freezing transition. Ó 2007 Elsevier Ltd. All rights reserved. Keywords: Cryogenic; Concrete; Properties; Steel; Strength; Water; Frost
Re´sume´ Le be´ton a e´te´ employe´ avec succe`s a` la tempe´rature cryoge´nique ( 160 °C) pour le stockage du gaz naturel lique´fie´. Une telle utilisation induira un cycle de refroidissement chaque fois qu’un re´cipient est rempli. En outre, le manque de controˆle, comme par exemple dans le cas de fuite, peut produire un changement brusque de tempe´rature. La connaissance des proprie´te´s du be´ton et des armatures dans ces conditions est donc ne´cessaire pour la conception et la bonne re´ussite de l’ope´ration. Une re´duction de la re´sistance du be´ton au dessous de la valeur initiale doit eˆtre pre´vues quand il est soumis au cycle de refroidissement re´chauffement. L’objectif principal de cet article est de de´crire les principales raisons de la re´duction de la re´sistance et des dommages structuraux (dommages du au gel) du be´ton durant le cycle de refroidissement et re´chauffement. Quelques travaux expe´rimentaux ont e´te´ mene´s a` bien a` l’Institut de la Cryoge´nie, Universite´ de Southampton, UK, pour de´terminer ce qui arrive a` l’eau dans le be´ton pendant la transition de conge´lation. Ó 2007 Elsevier Ltd. All rights reserved. Motscle´s: Cryoge´nie; Be´ton; Proprie´te´s; Armatures; Re´sistance; Eau; Gel
1. Introduction
*
Corresponding author. E-mail address: [email protected] (L. Dahmani).
0011-2275/$ - see front matter Ó 2007 Elsevier Ltd. All rights reserved. doi:10.1016/j.cryogenics.2007.07.001
Generally speaking, the mechanical properties of concrete improve when cooled down to 165 °C [1,3,6,7,14, 19,20,34,35]. The improvements in performances are thought to be due to the freezing of water inside the
518
L. Dahmani et al. / Cryogenics 47 (2007) 517–525
Nomenclature K W kJ m mm kg K °F °C N
thermal conductivity watt (energy) kilo joule meter millimeter kilograms kelvin fahrenheit degree celsius Newton
concrete: Ice fills and seals a fraction of pores and microcracks. This is consistent with the evidence that for ovendried concrete the variations in mechanical properties due to cooling are only marginal, while for saturated concrete the compressive strength, for example, may show a threefold increase, or more, when cooled down to 170 °C. However, concrete performances, although better at low temperature, become clearly poorer after the concrete returns to room temperature. Damage appears that manifests itself as a degradation of strength and stiffness, the more severe the larger the water content of the concrete. The complexity of the thermomechanic behavior of concrete at low temperatures seems to be due to water freezing inside the pores [16,22]. The expansion of water on freezing induces microcracking damage which does not manifest itself until the ice melts again. In order to understand who is responsible for the reduction in strength and structural damage (frost damage) of concrete under thermal cycling, an experimental work conducted at the Institute of Cryogenics, University of Southampton, UK [31–34], showed that water in concrete can be divided between weakly and strongly bound water, and that removal of weakly bound water is the way forward for making cryogenic and frost proof concrete without any need for air entrainment techniques. Before discussing this breakthrough in the next section, some knowledge of concrete properties under cryogenic temperature is therefore necessary. Most of the available information on cryogenic concrete can be found in the proceedings of two recent international conferences [6,7], a FIP state of the art report [12] and references therein [20]. In this section, some mechanical properties at low temperature of concrete and steel reinforcement will be briefly summarized.
r compressive strength e strain su ultimate bond strength Fr ultimate strength F0.2% elastic limit H enthalpy dh/dT = C heat capacity NMR nuclear magnetic resonance a alpha peak b beta peak
At temperatures ranging from the freezing point of water down to about 200 °C, the strength of concrete is markedly higher than at room temperature. The compressive strength may be as high as two to three times the strength at room temperature when the concrete is moist while being chilled, but the compressive strength of airdry concrete increase very much less. The difference in the increase in strength between wet and dry concretes is related to the formation of ice in the hydrated cement past.
Fig. 1. Variation of compressive strength with temperature [2].
2. Properties of concrete at low temperature Many researches [1,3,6,7,19–21,29,34] are involved in the study and behavior of the various physical parameters (conductivity, permeability, thermal strain, modulus of elasticity, tensile, compressive strength, etc.) at low temperatures (going until towards 170 °C).
Fig. 2. Variation of tensile strength with temperature [2].
L. Dahmani et al. / Cryogenics 47 (2007) 517–525
The relation between compressive strength and temperature, both for moist and air dry lightweight aggregate concrete is shown in Fig. 1, the corresponding data for splitting – tensile strength are shown in Fig. 2. The data in Figs. 1 and 2 refer to lightweight aggregate concrete which, for cryogenic purpose, has the advantage of good insulating properties [2]. However in normal weight concrete (Fig. 3), the increase in strength at low temperature is greater than is the case with lightweight aggregate concrete [20]. The modulus of elasticity of moist concrete is shown in Fig. 4. It increases steadily with the decrease in temperature
150 Monfore & Lentz (wet)
Compressive strength (N/mm 2)
125
Tognon Okada & Iguro
100 Monfore & Lentz (50% RH)
75
50
25
0 -200
Monfore & Lentz (dry)
-175
-150
-125
-100
-75
-50
-25
0
25
519
down to 190 °C. At that temperature, the modulus of elasticity is about 1.75 times the modulus at room temperature. For air dry concrete the corresponding value is about 1.65 [2]. The three principal thermal properties of concrete are thermal conductivity, specific heat and coefficient of thermal expansion or contraction. The thermal conductivity of the concrete depends on several parameters of which most significant are the moisture content. Experimental research [1,3,6,7,19] generally concluded that the thermal conductivity of the concrete increase with low temperature (Fig. 5). The specific heat exhibits a similar dependence; it varies according to the moisture content and temperature [20]. Several authors [6,7,29,30] report that, as concrete is cooled below 0 °C, it exhibits a gradual contraction with temperature. However according to Marshal, within the range from 20 °C to 70 °C [20] it expands before resuming its contraction. The amount of expansion and temperature range depends on moisture content. The incident met between 20 °C and 70 °C shows that in this space, the saturated concrete, instead of contracting with cooling, dilates because the much smaller pores, which until 20 °C contained only super fused water, are filled little by little with ice. Towards 70 °C the finest pores become frozen, and starting from this temperature the variation takes again its normal course. Fig. 6 shows the variation of coefficient of thermal expansion with temperature. The phase of dilation does not appear on the dried concrete.
Temperature ºC
Fig. 3. Variation of compressive strength with temperature [20].
3. Force–deformation relation Fig. 7 shows an evolution very marked towards a purely brittle elastic behavior for the saturated concretes with 170 °C, and hardly less sensitive for the concretes preserved under the normal conditions (to 20 °C and 65% HR) [29,30]. 4. Reinforced steel
Fig. 4. Influence of temperature on modulus of elasticity [19].
Current reinforced concrete steels have a notable increase in their ultimate resistance Fr and their elastic limit F0.2% at the low temperature (Fig. 8). The modulus of elasticity increases by 10% while the thermal dilation coefficient
Fig. 5. Evolution of conductivity for various percentage of moisture [21].
L. Dahmani et al. / Cryogenics 47 (2007) 517–525 Coefficient of thermal expansion, µm/mºC
520
10 9 8
Dry concrete
7 6 5 4 3
Wet concrete
2 1 0 -180
-160
-140
-120
-100
-80
-60
-40
-20
0
20
Temperature ºC
Fig. 6. Variation of coefficient of thermal expansion with temperature [20].
Fig. 9. Prestressed concrete steel [12].
is maintained constant between 70 °C and 165 °C around 10 5/°C [8,9]. One however observes on these steels a marked loss of ductility at low temperature. The improvement of the properties of the steels at cryogenic temperatures is obtained starting from addition, in small quantity of titanium, or of niobium or aluminium (KRYBAR). That makes it possible to preserve ductility even at very low temperature. 5. Prestressing steels
Fig. 7. Stress–strain diagram of Portland concrete [26].
As for the case of current reinforced concrete steels, the steels used in prestressing have a notable increase in their ultimate resistance Fr and their elastic limit F0.2% at low temperature (Fig. 9). The studies showed that [8,10,12,28] for a decrease in temperature until 195 °C, the modulus of elasticity increases by 3–10% compared to the ambient temperature. The losses of tension by relaxation of the steels stretched after 10 h of test were negligible towards 100 °C. 6. Reinforced and prestressed concrete
Fig. 8. Reinforced concrete steel [9].
An important aspect when the structural behavior of reinforced and prestressed concrete is that of thermal expansion, since internal stresses may be built up due to different thermal expansion of concrete and steel [5,9,12]. In fact, the thermal expansion of concrete must be determined down to 165 °C for every concrete used in cryogenic conditions. Experiments show that the thermal expansion of concrete at low temperatures can differ considerably from that of steel, especially for saturated concrete. This is, again, due to the formation of ice that
L. Dahmani et al. / Cryogenics 47 (2007) 517–525
521
the various water states, and their effect on concrete properties during cooling through the freezing transition temperature range. 9. Nuclear magnetic resonance NMR studies
Fig. 10. Variation of ultimate bond strength with temperature for different curing conditions [37].
makes concrete to expand upon cooling – instead of contracting – over certain temperature intervals.
Since its discovery in 1946 nuclear magnetic resonance (NMR) [4,36] has evolved from a scientific curiosity into one of the most powerful spectroscopic techniques. Today it is routinely used as an analytical tool in chemical and bio-chemical research but also in physics, material science and even in geochemistry. It provides information about the physical state of matter. By studying hydrogen nuclear magnetic resonance (1H NMR) signals of water in porous materials, it is possible to determine their effect on concrete properties during cooling through the freezing transition temperature range [29–32]. NMR studies can provide detailed information about the porous microstructure (porosity, pore-size distribution) of these materials [16,27]. The proton nuclear magnetic resonance absorption signal measured at various temperatures for small samples of concrete sealed in glass tube, have identified two type of water in concrete [31,32], alpha water which is tightly bound by chemical hydration, and beta water which is weakly bound by physical adsorption and capillary action. These two types are clearly revealed by NMR spectrum of two peaks, the alpha and beta peak respectively.
7. Bond strength A good bond between reinforcement and concrete is required so that the two materials can act together in the best way [37]. Fig. 10 shows the ultimate bond strength (su) plotted against temperature for different curing conditions. It appears that the test temperature and the moisture content of the concrete have a clear influence on su. 8. Behavior of water in the concrete during the freezing process Many researches [11,15,17,18,22,26] believed that the degree of saturation is a key parameter for the frost resistance of concrete. Due to its porous nature, it may hold considerable amounts of water that may cause frost damage. Cycles of freezing and thawing might result in progressive disruption of the cement paste due to the expansion of the absorbed water on freezing. The water in concrete can be in three forms [13,22– 25,31]:
10. NMR study in a saturated moist cured concrete sample with w/c ratio of 0.37 [31–33] 10.1. Effect of reducing temperature By reducing the temperature from 12 °C to 73 °C, the observed spectra for a saturated moist cured concrete sample with a water cement ratio of 0.37 is shown in Fig. 11, a change in the relative size of the alpha and beta peak is apparent. The beta peak diminishes in magnitude and totally disappears at temperature of 53 °C as the beta water freeze to ice.
(a) chemically bound water; (b) physically adsorbed water and (c) rewetting water. By using some experimental parameters [31–34] like nuclear magnetic resonance variation (NMR), enthalpy– temperature (or heat capacity) variation, and thermal strain–temperature measurement, we can resolve between
Fig. 11. Proton NMR in a saturated moist-cured concrete sample W/ C = 0.37. Effect of reducing temperature [31].
522
L. Dahmani et al. / Cryogenics 47 (2007) 517–525
Fig. 12. NMR in dried and rewetted concrete sample. Effect of cooling to 53 °C [31,32].
11. NMR study in a dried and rewetted concrete sample 11.1. Effect of cooling The saturated moist cured concrete sample now is dried to a constant weight and then rewetted in the glass tube and sealed. On cooling to 53 °C, the beta peak in Fig. 12 diminish with decreasing temperature, again as the beta water freeze to ice. 11.2. Effect of heating On heating to 120 °C, the beta peak first grows in relation to the alpha peak, and then diminishes with time. Fig. 13, indicates the relative magnitude of the alpha and beta line after 1/2 h, 1 h, and 10 h, respectively. 12. Concluding remarks It is concluded that beta-water, corresponding to the beta peak is at least part of the evaporable water which is more rapidly removed by heating to 120 °C. It should also be noted that on re-wetting the dried sample, the beta peak reappears hence the beta water. These examples indicate how NMR spectra provide a qualitative means for observing the behavior of different type of water in the concrete.
Fig. 13. NMR in dried and rewetted concrete sample. Effect of heating to 120 °C [31,32].
13. Enthalpy–temperature measurements The samples used to produce enthalpy–temperature plot down to 80 °C were circular cylinders 96 mm diameter 200 mm high [22,31,32]. The heat capacity measurements were carried out in a large calorimeter. All samples were moist cured. The enthalpy versus temperature (H–T) curve for dry (sample D), saturated (samples A, B, and C) and re-wetted (sample E) concrete samples are shown in Figs. 14 and 15, the respective differential enthalpy (heat capacity) plots are shown in Figs. 16 and 17. We note that the H–T plot for the dry sample (D) is a straight line and that the heat capacity is a constant. It is concluded that there is no change of phase tacks place in the water remaining in a dried sample. The normalized H–T curves for dry, saturated and rewetted concrete samples and water ice are shown in Fig. 18. It can be seen that the normalized enthalpy plots are closely similar for samples A, B, and C with different water/ cement ratio.
0
6
-20
5
-40
4
∂H/ ∂ T kj / kg. ºC
Enthalpy kj/kg
L. Dahmani et al. / Cryogenics 47 (2007) 517–525
-60
-80
Sample A Sample B Sample C Sample D
W/C = 0.30 W/C = 0.37 W/C = 0.40 Dried
Sample D Sample E
3
2
-50
-40
-30
-20
-10
0
0 -60
+10
-50
-40
0
-10
0
10
100
Normalised Enthalpy of water kj/kg.
-20
-40
Enthalpy kj/kg
-20
Fig. 17. Specific heat of rewetted concrete sample [32].
Fig. 14. Enthalpy of concrete samples A, B, C, and D [29].
-60
-80 Sample D Sample E
Dried Rewetted
-100
-50
-40
-30
-20
Temperature
-10
0
+10
0
Sample A Sample B Sample C Sample E
W/C = 0.30 W/C = 0.37 W/C = 0.40 Dried and rewetted
-100
-200
-300
-400
-500
-60
-50
-40
ºC
-30
-20
-10
0
10
Temperature ºC
Fig. 15. Enthalpy of concrete samples D and E [31].
Fig. 18. Normalised enthalpy of water in concrete samples [32].
The dried and re-wetted sample E shows a totally different behavior. The excess heat capacity is confined to a narrower temperature range from 0 °C to 10 °C (90% of the freezing is confined to a narrower temperature range from 0 °C to 10 °C). The remaining 10% of the freezing appears to take place between 35 °C and 45 °C.
2.0
∂ H/ ∂ T kj / kg .ºC
-30
Temperature ºC
Temperature ˚C
-120 -60
Dried Rewetted
1
-100
-120 -60
523
1.0 Sample A Sample B Sample C Sample D
-60
-50
-40
-30
14. Concluding remarks
W/C = 0.30 W/C = 0.37 W/C = 0.40 Dried
-20
-10
Temperature ºC Fig. 16. Specific heat of concrete samples [32].
0
-10
Some of the evaporable water is strongly physically bound and does not exhibit a latent heat of freezing. For saturate moist cured concrete samples A, B, and C, the freezing of the non-chemically bound water takes place in a similar manner over the whole temperature range between 0 and 60 °C.
524
L. Dahmani et al. / Cryogenics 47 (2007) 517–525
Sample C resoacked
-400
Sample C
-200
Saturated sample After 3 weeks in dry chamber
-400 warming
-200
0
Thermal strain μ strain
Thermal strain μ strain
Cooling
-200
-400
-600
0
-200
-400
-600
Cooling
-800
-1000 -80
-800
-60
-40
-20
0
+20
Temperature ºC
Fig. 19. Thermal strain of moist-cured concrete samples W/C = 0.40 [32].
For a dried and re-wetted concrete sample E, 90% of the freezing is confined to a narrower temperature range from 0 °C to 10 °C. The remaining 10% of the freezing appears to take place between 35 °C and 45 °C. 15. Strain-temperature measurements Fig. 19 shows the thermal strain behavior of a saturated moist cured concrete (sample C) with a water cement ratio of 0.4. There is a dilation on freezing for a saturated sample, however after drying for three weeks in a drying chamber, there is no dilation on freezing. Fig. 20 shows the thermal strain of the re-wetted concrete sample. The dilation and hysteretic behavior has returned. 16. General conclusion The behavior of concrete at low temperature is governed by its porosity. The changes at low temperatures are influenced primarily by the moisture content of the material. For moist concrete, the strength increases as the temperature is reduced. It is due to the formation of ice in the pores of the material. However, concrete performance, although better at low temperature, become clearly poorer after the concrete returns to room temperature. Damage appears that manifests itself as a degradation of strength and stiffness, the more severe the larger the water content of the concrete.
-80
warming
st
Saturated resoacked sample 1 cycle Saturated resoacked sample 2nd cycle Further resoack in water 1st cycle
-60
-40
-20
0
+20
Temperature ºC
Fig. 20. Thermal strain of rewetted concrete samples W/C = 0.40 after repeated cycles [32].
By using proton (hydrogen) nuclear magnetic resonance (NMR), heat capacity and thermal strain measurement on concrete samples, the following remarks can be made: – Only some of the physically bound water (beta water) change phase from liquid to solid with a latent heat of freezing. – The chemically bound water (alpha water) does not change state. – There is a sharp change of state for a rewetting water (beta water) just below °C. The beta water is present in a saturated moist-cured concrete or in a dried and rewetted concrete as a mixture of free water and capillary water which constitute a major portion of evaporable water. These studies have shown that beta water is almost exclusively responsible for the thermal strain hysterisis behavior and for cracking, etc. When beta water is absent,1 the cryogenic concrete has repeatable properties of strength, permeability, resistance to cracking, etc.
1 A 10 cm cube sample [32] survive without damage under extreme thermal cycling by repeated dunking in liquid nitrogen.
L. Dahmani et al. / Cryogenics 47 (2007) 517–525
The solution to manufacturing reliable cryogenic concrete is therefore to remove beta water (by appropriate drying) after curing, and to stop beta water re-entering subsequently. The air entrainment is not necessary. References [1] Bamforth PB, Murray WT, Browne RD. The application of concrete at cryogenic temperature to LNG tank design. In: 2nd international confe´rence on cryogenic concrete, Amsterdam; 1983. [2] Berner D, Gerwick BC, Polivka M. Static and cyclic behavior of structural lightweight concrete at cryogenic temperature. In: Temperature effects on concrete. ASTM Sp Tech Publ No 858, Philadelphia (PA); 1983. p. 21–37. [3] Browne RD, Bameorth PB. In: Proceedings of the 1st international conference on cryogenic concrete, Newcastle, (March); 1981. [4] Callaghan PT. Principles of nuclear magnetic resonance microscopy. New York: Oxford Science Publications; 1991. [5] Canadian Prestressed Concrete Institute. Concrete sea structures in Artic regions. In: Proc calgary symposium. 1984. p. 109. [6] Cryogenic Concrete. In: Proceedings of the 1st international conference of the Concrete Society, March. Newcastle upon Tyne: Construction Press; 1981. p. 336. [7] Cryogenic Concrete. In: Proceedings of the 2nd international conference of the concrete societies of the Netherlands and UK, October. Amsterdam/London: Devon House; 1983. [8] Elices M, Rostasy FS, Faas WM, Wiedemann G. Cryogenic behaviour of materials for prestressed concrete. FIP State-of-theArt Report. Slough: Wexham Spring; 1982. p. 84. [9] Elices M, Corres H, Planas J. Behaviour at cryogenic temperatures of steel for concrete reinforcement. J ACI. Technical paper 83-40; 1986. p. 405–11. [10] Elices M. Cryogenic prestressed concrete: fracture aspects. Theoretical and applied fracture mechanics, vol. 7. North-Holland; 1987. p. 51–63. [11] Etzler FM, Conners JJ. Temperature dependence of the heat capacity of water in small pores. Langmuir 1990;6(7):1250–3. [12] Federation International De La Precontrainte. Cryogenic behaviour of materials for prestressed concrete, State of art report, FIP; 1982. [13] Geiker M, Thaulow N. Ingress of moisture due to freeze/thaw exposure. In: Lindmark S, editor. Frost resistance of building materials. Report TVBM-3072, Lund Institute of Technology, Div of Building Materials; 1996. [14] GotoY, Miura T. Mechanical properties of concrete at very low temperature. In: Proceedings of the 21st Japan congress on materials research; 1978. p. 157–9. [15] Hanaor A. Microcracking and permeability of concrete to liquide nitrogen. J Am Concr Inst 1985. [16] Hansen EW, Stocker M, Schmidt R. Low-temperature phase transition of water confined in mesopores probed by NMR. Influence on pore size distribution. J Phys Chem 1996;100:2195–200. [17] Hedenblad G. Moisture permeability of mature concrete, Cement mortar and cement paste. Lund Institute of Technology, Div of Building Materials, Report TVBM-1014; 1993.
525
[18] Marchand J, Pleau R, Gagne´ R. Deterioration of concrete due to freezing and thawing. In: Skalny Jan, Mindess Sidney, editors. Materials science of concrete IV; 1995. [19] Marechal JC. Variation in the modulus of elasticity and poisson’s ratio with temperature. In: International seminar on concrete for nuclear reactors. ACI SP-34, Detroit (MI); 1972. p. 495–503. [20] Marshal AL. Cryogenic concrete. Cryogenics. Butterworth & Co (Publishers) Ltd.; 1982. p. 555–65. [21] Menzes NC. Etude du comportement des structures en be´ton soumises a` des tempe´ratures cryoge´niques. The`se de IIIeme` cycle. Paris VI, Paris: Universite´ Pierre et Marie Curie; 1986. [22] Mohd Yusof KB. The effects of moisture on the thermal properties of concrete between 80 °C and 0 °C. PhD thesis, University of Southampton, UK; 1984. [23] Pel L, Kopinga K, Hazrati K. Water distribution and pore structure in concrete as determined by NMR. In: Proc 9, Feuchtetag, Weimar, September 17–18, 1997. p. 294–300. [24] Okada T, Iguro M. Bending behaviour of prestressed concrete beam under low tempe´rature. J Jpn Prestressed Concr Eng Assoc 1978. [25] Pel L. Moisture transport in porous building materials. Ph D thesis, Eindhoven University of Technology; 1995. [26] Pigeon M, Pleau R. Durability of concrete in cold climates. London: Chapman & Hall; 1995. [27] Prado PJ, Balcom BJ, Beyea SD, Bremner TW, Armstrong RL, Grattan-Bellew PE. Concrete freeze/thaw as studied by magnetic resonance imaging. Cement Concr Res 1998;28(2):261–70. [28] Rostasy FS. Prestressing steels and systems for storage tanks for refrigerated liquefied gases. In: FIP symposium, New Delhi; 1986. [29] Rostazy FS, Wiedemann G. Strength, deformation and thermal strains of concrete at cryogenic conditions. In: First international conference cryogenic concrete, Newcastle upon Tyne; 1981. [30] Rostazy FS, Wiedemann G. Stress–strain behaviour of concrete at extremely low temperature. Int J Cem Concr Res 1980;10:565–72. [31] Scurlock RG, Yusof Kamarudin M. Thermal behaviour of concretes between 80 °C and +20 °C. In: Proceedings of the 2nd international conference on cryogenic concrete, Amsterdam; 1983. [32] Scurlock RG, Mohd Yusof KB. Cryogenic and frostproof concrete made from conventional aggregates and Portland cement. Adv Cryo Eng 2000;45:1779. [33] Scurlock RG. Dewars low loss. Cryogenic and frostproof concrete, low loss dewars and tanks. Cryogenic Society of America. 2005; p. 35–7. [34] Scurlock RG. Cryogenics at the University of Southampton: A review. Institute of Cryogenics, University of Southampton, Southampton S09 5N H, UK, 19 February, 1988. p. 439–66. [35] Van Der Veen C. In: Delft TU, editors. Properties of concrete at very low temperatures. A survey of literature; 1987. [36] Wolter B, Dobmann G. Nuclear magnetic resonance as a tool for the characterisation of concrete in different stages of its development. In: Proceedings of the international symposium on non-destructive testing in civil engineering (NDT-CE), Berlin, Germany, vol. 1; 1995. p. 181. [37] Vandewalle L. Bond between a reinforcement bar and concrete at normal and cryogenic temperatures. J Mater Sci Lett 1989;8: 147–9.
Behavior of the reinforced concrete at cryogenic temperatures Lahlou Dahmani *, Amar Khenane, Salah Kaci Department of Civil Engineering, Faculty of Engineering Construction, University of Tizi-Ouzou, Tizi-Ouzou, Algeria Received 11 October 2006; received in revised form 23 July 2007; accepted 24 July 2007
Abstract Concrete has successfully been used at extremely low temperature ( 160 °C) for the storage of liquefied natural gas. Such use will induce a cycle of cooling every time a container is filled. In addition, lack of control, as for example a leak may produce a sudden temperature change. Some knowledge of concrete properties under these conditions is therefore necessary for successful design and operation. The first and basic cause of the failure of concrete is repeated freezing (thawing) of moisture contained in the pores, microcracks, and cavities of the concrete. On transition to ice, water existing in the free state in cracks increases in volume, expanding the recess in which freezing occurs. A reduction in strength below the initial value is to be expected and further cycle of freezing and thawing have a further marked effect. The main objective of this paper is to describe the principal reasons for the reduction in strength and structural damage (frost damage) of concrete following repeated freeze–thaw cycles. Some experimental work was carried out at the Institute of Cryogenics, University of Southampton, UK, to determine what happens to water in concrete during the freezing transition. Ó 2007 Elsevier Ltd. All rights reserved. Keywords: Cryogenic; Concrete; Properties; Steel; Strength; Water; Frost
Re´sume´ Le be´ton a e´te´ employe´ avec succe`s a` la tempe´rature cryoge´nique ( 160 °C) pour le stockage du gaz naturel lique´fie´. Une telle utilisation induira un cycle de refroidissement chaque fois qu’un re´cipient est rempli. En outre, le manque de controˆle, comme par exemple dans le cas de fuite, peut produire un changement brusque de tempe´rature. La connaissance des proprie´te´s du be´ton et des armatures dans ces conditions est donc ne´cessaire pour la conception et la bonne re´ussite de l’ope´ration. Une re´duction de la re´sistance du be´ton au dessous de la valeur initiale doit eˆtre pre´vues quand il est soumis au cycle de refroidissement re´chauffement. L’objectif principal de cet article est de de´crire les principales raisons de la re´duction de la re´sistance et des dommages structuraux (dommages du au gel) du be´ton durant le cycle de refroidissement et re´chauffement. Quelques travaux expe´rimentaux ont e´te´ mene´s a` bien a` l’Institut de la Cryoge´nie, Universite´ de Southampton, UK, pour de´terminer ce qui arrive a` l’eau dans le be´ton pendant la transition de conge´lation. Ó 2007 Elsevier Ltd. All rights reserved. Motscle´s: Cryoge´nie; Be´ton; Proprie´te´s; Armatures; Re´sistance; Eau; Gel
1. Introduction
*
Corresponding author. E-mail address: [email protected] (L. Dahmani).
0011-2275/$ - see front matter Ó 2007 Elsevier Ltd. All rights reserved. doi:10.1016/j.cryogenics.2007.07.001
Generally speaking, the mechanical properties of concrete improve when cooled down to 165 °C [1,3,6,7,14, 19,20,34,35]. The improvements in performances are thought to be due to the freezing of water inside the
518
L. Dahmani et al. / Cryogenics 47 (2007) 517–525
Nomenclature K W kJ m mm kg K °F °C N
thermal conductivity watt (energy) kilo joule meter millimeter kilograms kelvin fahrenheit degree celsius Newton
concrete: Ice fills and seals a fraction of pores and microcracks. This is consistent with the evidence that for ovendried concrete the variations in mechanical properties due to cooling are only marginal, while for saturated concrete the compressive strength, for example, may show a threefold increase, or more, when cooled down to 170 °C. However, concrete performances, although better at low temperature, become clearly poorer after the concrete returns to room temperature. Damage appears that manifests itself as a degradation of strength and stiffness, the more severe the larger the water content of the concrete. The complexity of the thermomechanic behavior of concrete at low temperatures seems to be due to water freezing inside the pores [16,22]. The expansion of water on freezing induces microcracking damage which does not manifest itself until the ice melts again. In order to understand who is responsible for the reduction in strength and structural damage (frost damage) of concrete under thermal cycling, an experimental work conducted at the Institute of Cryogenics, University of Southampton, UK [31–34], showed that water in concrete can be divided between weakly and strongly bound water, and that removal of weakly bound water is the way forward for making cryogenic and frost proof concrete without any need for air entrainment techniques. Before discussing this breakthrough in the next section, some knowledge of concrete properties under cryogenic temperature is therefore necessary. Most of the available information on cryogenic concrete can be found in the proceedings of two recent international conferences [6,7], a FIP state of the art report [12] and references therein [20]. In this section, some mechanical properties at low temperature of concrete and steel reinforcement will be briefly summarized.
r compressive strength e strain su ultimate bond strength Fr ultimate strength F0.2% elastic limit H enthalpy dh/dT = C heat capacity NMR nuclear magnetic resonance a alpha peak b beta peak
At temperatures ranging from the freezing point of water down to about 200 °C, the strength of concrete is markedly higher than at room temperature. The compressive strength may be as high as two to three times the strength at room temperature when the concrete is moist while being chilled, but the compressive strength of airdry concrete increase very much less. The difference in the increase in strength between wet and dry concretes is related to the formation of ice in the hydrated cement past.
Fig. 1. Variation of compressive strength with temperature [2].
2. Properties of concrete at low temperature Many researches [1,3,6,7,19–21,29,34] are involved in the study and behavior of the various physical parameters (conductivity, permeability, thermal strain, modulus of elasticity, tensile, compressive strength, etc.) at low temperatures (going until towards 170 °C).
Fig. 2. Variation of tensile strength with temperature [2].
L. Dahmani et al. / Cryogenics 47 (2007) 517–525
The relation between compressive strength and temperature, both for moist and air dry lightweight aggregate concrete is shown in Fig. 1, the corresponding data for splitting – tensile strength are shown in Fig. 2. The data in Figs. 1 and 2 refer to lightweight aggregate concrete which, for cryogenic purpose, has the advantage of good insulating properties [2]. However in normal weight concrete (Fig. 3), the increase in strength at low temperature is greater than is the case with lightweight aggregate concrete [20]. The modulus of elasticity of moist concrete is shown in Fig. 4. It increases steadily with the decrease in temperature
150 Monfore & Lentz (wet)
Compressive strength (N/mm 2)
125
Tognon Okada & Iguro
100 Monfore & Lentz (50% RH)
75
50
25
0 -200
Monfore & Lentz (dry)
-175
-150
-125
-100
-75
-50
-25
0
25
519
down to 190 °C. At that temperature, the modulus of elasticity is about 1.75 times the modulus at room temperature. For air dry concrete the corresponding value is about 1.65 [2]. The three principal thermal properties of concrete are thermal conductivity, specific heat and coefficient of thermal expansion or contraction. The thermal conductivity of the concrete depends on several parameters of which most significant are the moisture content. Experimental research [1,3,6,7,19] generally concluded that the thermal conductivity of the concrete increase with low temperature (Fig. 5). The specific heat exhibits a similar dependence; it varies according to the moisture content and temperature [20]. Several authors [6,7,29,30] report that, as concrete is cooled below 0 °C, it exhibits a gradual contraction with temperature. However according to Marshal, within the range from 20 °C to 70 °C [20] it expands before resuming its contraction. The amount of expansion and temperature range depends on moisture content. The incident met between 20 °C and 70 °C shows that in this space, the saturated concrete, instead of contracting with cooling, dilates because the much smaller pores, which until 20 °C contained only super fused water, are filled little by little with ice. Towards 70 °C the finest pores become frozen, and starting from this temperature the variation takes again its normal course. Fig. 6 shows the variation of coefficient of thermal expansion with temperature. The phase of dilation does not appear on the dried concrete.
Temperature ºC
Fig. 3. Variation of compressive strength with temperature [20].
3. Force–deformation relation Fig. 7 shows an evolution very marked towards a purely brittle elastic behavior for the saturated concretes with 170 °C, and hardly less sensitive for the concretes preserved under the normal conditions (to 20 °C and 65% HR) [29,30]. 4. Reinforced steel
Fig. 4. Influence of temperature on modulus of elasticity [19].
Current reinforced concrete steels have a notable increase in their ultimate resistance Fr and their elastic limit F0.2% at the low temperature (Fig. 8). The modulus of elasticity increases by 10% while the thermal dilation coefficient
Fig. 5. Evolution of conductivity for various percentage of moisture [21].
L. Dahmani et al. / Cryogenics 47 (2007) 517–525 Coefficient of thermal expansion, µm/mºC
520
10 9 8
Dry concrete
7 6 5 4 3
Wet concrete
2 1 0 -180
-160
-140
-120
-100
-80
-60
-40
-20
0
20
Temperature ºC
Fig. 6. Variation of coefficient of thermal expansion with temperature [20].
Fig. 9. Prestressed concrete steel [12].
is maintained constant between 70 °C and 165 °C around 10 5/°C [8,9]. One however observes on these steels a marked loss of ductility at low temperature. The improvement of the properties of the steels at cryogenic temperatures is obtained starting from addition, in small quantity of titanium, or of niobium or aluminium (KRYBAR). That makes it possible to preserve ductility even at very low temperature. 5. Prestressing steels
Fig. 7. Stress–strain diagram of Portland concrete [26].
As for the case of current reinforced concrete steels, the steels used in prestressing have a notable increase in their ultimate resistance Fr and their elastic limit F0.2% at low temperature (Fig. 9). The studies showed that [8,10,12,28] for a decrease in temperature until 195 °C, the modulus of elasticity increases by 3–10% compared to the ambient temperature. The losses of tension by relaxation of the steels stretched after 10 h of test were negligible towards 100 °C. 6. Reinforced and prestressed concrete
Fig. 8. Reinforced concrete steel [9].
An important aspect when the structural behavior of reinforced and prestressed concrete is that of thermal expansion, since internal stresses may be built up due to different thermal expansion of concrete and steel [5,9,12]. In fact, the thermal expansion of concrete must be determined down to 165 °C for every concrete used in cryogenic conditions. Experiments show that the thermal expansion of concrete at low temperatures can differ considerably from that of steel, especially for saturated concrete. This is, again, due to the formation of ice that
L. Dahmani et al. / Cryogenics 47 (2007) 517–525
521
the various water states, and their effect on concrete properties during cooling through the freezing transition temperature range. 9. Nuclear magnetic resonance NMR studies
Fig. 10. Variation of ultimate bond strength with temperature for different curing conditions [37].
makes concrete to expand upon cooling – instead of contracting – over certain temperature intervals.
Since its discovery in 1946 nuclear magnetic resonance (NMR) [4,36] has evolved from a scientific curiosity into one of the most powerful spectroscopic techniques. Today it is routinely used as an analytical tool in chemical and bio-chemical research but also in physics, material science and even in geochemistry. It provides information about the physical state of matter. By studying hydrogen nuclear magnetic resonance (1H NMR) signals of water in porous materials, it is possible to determine their effect on concrete properties during cooling through the freezing transition temperature range [29–32]. NMR studies can provide detailed information about the porous microstructure (porosity, pore-size distribution) of these materials [16,27]. The proton nuclear magnetic resonance absorption signal measured at various temperatures for small samples of concrete sealed in glass tube, have identified two type of water in concrete [31,32], alpha water which is tightly bound by chemical hydration, and beta water which is weakly bound by physical adsorption and capillary action. These two types are clearly revealed by NMR spectrum of two peaks, the alpha and beta peak respectively.
7. Bond strength A good bond between reinforcement and concrete is required so that the two materials can act together in the best way [37]. Fig. 10 shows the ultimate bond strength (su) plotted against temperature for different curing conditions. It appears that the test temperature and the moisture content of the concrete have a clear influence on su. 8. Behavior of water in the concrete during the freezing process Many researches [11,15,17,18,22,26] believed that the degree of saturation is a key parameter for the frost resistance of concrete. Due to its porous nature, it may hold considerable amounts of water that may cause frost damage. Cycles of freezing and thawing might result in progressive disruption of the cement paste due to the expansion of the absorbed water on freezing. The water in concrete can be in three forms [13,22– 25,31]:
10. NMR study in a saturated moist cured concrete sample with w/c ratio of 0.37 [31–33] 10.1. Effect of reducing temperature By reducing the temperature from 12 °C to 73 °C, the observed spectra for a saturated moist cured concrete sample with a water cement ratio of 0.37 is shown in Fig. 11, a change in the relative size of the alpha and beta peak is apparent. The beta peak diminishes in magnitude and totally disappears at temperature of 53 °C as the beta water freeze to ice.
(a) chemically bound water; (b) physically adsorbed water and (c) rewetting water. By using some experimental parameters [31–34] like nuclear magnetic resonance variation (NMR), enthalpy– temperature (or heat capacity) variation, and thermal strain–temperature measurement, we can resolve between
Fig. 11. Proton NMR in a saturated moist-cured concrete sample W/ C = 0.37. Effect of reducing temperature [31].
522
L. Dahmani et al. / Cryogenics 47 (2007) 517–525
Fig. 12. NMR in dried and rewetted concrete sample. Effect of cooling to 53 °C [31,32].
11. NMR study in a dried and rewetted concrete sample 11.1. Effect of cooling The saturated moist cured concrete sample now is dried to a constant weight and then rewetted in the glass tube and sealed. On cooling to 53 °C, the beta peak in Fig. 12 diminish with decreasing temperature, again as the beta water freeze to ice. 11.2. Effect of heating On heating to 120 °C, the beta peak first grows in relation to the alpha peak, and then diminishes with time. Fig. 13, indicates the relative magnitude of the alpha and beta line after 1/2 h, 1 h, and 10 h, respectively. 12. Concluding remarks It is concluded that beta-water, corresponding to the beta peak is at least part of the evaporable water which is more rapidly removed by heating to 120 °C. It should also be noted that on re-wetting the dried sample, the beta peak reappears hence the beta water. These examples indicate how NMR spectra provide a qualitative means for observing the behavior of different type of water in the concrete.
Fig. 13. NMR in dried and rewetted concrete sample. Effect of heating to 120 °C [31,32].
13. Enthalpy–temperature measurements The samples used to produce enthalpy–temperature plot down to 80 °C were circular cylinders 96 mm diameter 200 mm high [22,31,32]. The heat capacity measurements were carried out in a large calorimeter. All samples were moist cured. The enthalpy versus temperature (H–T) curve for dry (sample D), saturated (samples A, B, and C) and re-wetted (sample E) concrete samples are shown in Figs. 14 and 15, the respective differential enthalpy (heat capacity) plots are shown in Figs. 16 and 17. We note that the H–T plot for the dry sample (D) is a straight line and that the heat capacity is a constant. It is concluded that there is no change of phase tacks place in the water remaining in a dried sample. The normalized H–T curves for dry, saturated and rewetted concrete samples and water ice are shown in Fig. 18. It can be seen that the normalized enthalpy plots are closely similar for samples A, B, and C with different water/ cement ratio.
0
6
-20
5
-40
4
∂H/ ∂ T kj / kg. ºC
Enthalpy kj/kg
L. Dahmani et al. / Cryogenics 47 (2007) 517–525
-60
-80
Sample A Sample B Sample C Sample D
W/C = 0.30 W/C = 0.37 W/C = 0.40 Dried
Sample D Sample E
3
2
-50
-40
-30
-20
-10
0
0 -60
+10
-50
-40
0
-10
0
10
100
Normalised Enthalpy of water kj/kg.
-20
-40
Enthalpy kj/kg
-20
Fig. 17. Specific heat of rewetted concrete sample [32].
Fig. 14. Enthalpy of concrete samples A, B, C, and D [29].
-60
-80 Sample D Sample E
Dried Rewetted
-100
-50
-40
-30
-20
Temperature
-10
0
+10
0
Sample A Sample B Sample C Sample E
W/C = 0.30 W/C = 0.37 W/C = 0.40 Dried and rewetted
-100
-200
-300
-400
-500
-60
-50
-40
ºC
-30
-20
-10
0
10
Temperature ºC
Fig. 15. Enthalpy of concrete samples D and E [31].
Fig. 18. Normalised enthalpy of water in concrete samples [32].
The dried and re-wetted sample E shows a totally different behavior. The excess heat capacity is confined to a narrower temperature range from 0 °C to 10 °C (90% of the freezing is confined to a narrower temperature range from 0 °C to 10 °C). The remaining 10% of the freezing appears to take place between 35 °C and 45 °C.
2.0
∂ H/ ∂ T kj / kg .ºC
-30
Temperature ºC
Temperature ˚C
-120 -60
Dried Rewetted
1
-100
-120 -60
523
1.0 Sample A Sample B Sample C Sample D
-60
-50
-40
-30
14. Concluding remarks
W/C = 0.30 W/C = 0.37 W/C = 0.40 Dried
-20
-10
Temperature ºC Fig. 16. Specific heat of concrete samples [32].
0
-10
Some of the evaporable water is strongly physically bound and does not exhibit a latent heat of freezing. For saturate moist cured concrete samples A, B, and C, the freezing of the non-chemically bound water takes place in a similar manner over the whole temperature range between 0 and 60 °C.
524
L. Dahmani et al. / Cryogenics 47 (2007) 517–525
Sample C resoacked
-400
Sample C
-200
Saturated sample After 3 weeks in dry chamber
-400 warming
-200
0
Thermal strain μ strain
Thermal strain μ strain
Cooling
-200
-400
-600
0
-200
-400
-600
Cooling
-800
-1000 -80
-800
-60
-40
-20
0
+20
Temperature ºC
Fig. 19. Thermal strain of moist-cured concrete samples W/C = 0.40 [32].
For a dried and re-wetted concrete sample E, 90% of the freezing is confined to a narrower temperature range from 0 °C to 10 °C. The remaining 10% of the freezing appears to take place between 35 °C and 45 °C. 15. Strain-temperature measurements Fig. 19 shows the thermal strain behavior of a saturated moist cured concrete (sample C) with a water cement ratio of 0.4. There is a dilation on freezing for a saturated sample, however after drying for three weeks in a drying chamber, there is no dilation on freezing. Fig. 20 shows the thermal strain of the re-wetted concrete sample. The dilation and hysteretic behavior has returned. 16. General conclusion The behavior of concrete at low temperature is governed by its porosity. The changes at low temperatures are influenced primarily by the moisture content of the material. For moist concrete, the strength increases as the temperature is reduced. It is due to the formation of ice in the pores of the material. However, concrete performance, although better at low temperature, become clearly poorer after the concrete returns to room temperature. Damage appears that manifests itself as a degradation of strength and stiffness, the more severe the larger the water content of the concrete.
-80
warming
st
Saturated resoacked sample 1 cycle Saturated resoacked sample 2nd cycle Further resoack in water 1st cycle
-60
-40
-20
0
+20
Temperature ºC
Fig. 20. Thermal strain of rewetted concrete samples W/C = 0.40 after repeated cycles [32].
By using proton (hydrogen) nuclear magnetic resonance (NMR), heat capacity and thermal strain measurement on concrete samples, the following remarks can be made: – Only some of the physically bound water (beta water) change phase from liquid to solid with a latent heat of freezing. – The chemically bound water (alpha water) does not change state. – There is a sharp change of state for a rewetting water (beta water) just below °C. The beta water is present in a saturated moist-cured concrete or in a dried and rewetted concrete as a mixture of free water and capillary water which constitute a major portion of evaporable water. These studies have shown that beta water is almost exclusively responsible for the thermal strain hysterisis behavior and for cracking, etc. When beta water is absent,1 the cryogenic concrete has repeatable properties of strength, permeability, resistance to cracking, etc.
1 A 10 cm cube sample [32] survive without damage under extreme thermal cycling by repeated dunking in liquid nitrogen.
L. Dahmani et al. / Cryogenics 47 (2007) 517–525
The solution to manufacturing reliable cryogenic concrete is therefore to remove beta water (by appropriate drying) after curing, and to stop beta water re-entering subsequently. The air entrainment is not necessary. References [1] Bamforth PB, Murray WT, Browne RD. The application of concrete at cryogenic temperature to LNG tank design. In: 2nd international confe´rence on cryogenic concrete, Amsterdam; 1983. [2] Berner D, Gerwick BC, Polivka M. Static and cyclic behavior of structural lightweight concrete at cryogenic temperature. In: Temperature effects on concrete. ASTM Sp Tech Publ No 858, Philadelphia (PA); 1983. p. 21–37. [3] Browne RD, Bameorth PB. In: Proceedings of the 1st international conference on cryogenic concrete, Newcastle, (March); 1981. [4] Callaghan PT. Principles of nuclear magnetic resonance microscopy. New York: Oxford Science Publications; 1991. [5] Canadian Prestressed Concrete Institute. Concrete sea structures in Artic regions. In: Proc calgary symposium. 1984. p. 109. [6] Cryogenic Concrete. In: Proceedings of the 1st international conference of the Concrete Society, March. Newcastle upon Tyne: Construction Press; 1981. p. 336. [7] Cryogenic Concrete. In: Proceedings of the 2nd international conference of the concrete societies of the Netherlands and UK, October. Amsterdam/London: Devon House; 1983. [8] Elices M, Rostasy FS, Faas WM, Wiedemann G. Cryogenic behaviour of materials for prestressed concrete. FIP State-of-theArt Report. Slough: Wexham Spring; 1982. p. 84. [9] Elices M, Corres H, Planas J. Behaviour at cryogenic temperatures of steel for concrete reinforcement. J ACI. Technical paper 83-40; 1986. p. 405–11. [10] Elices M. Cryogenic prestressed concrete: fracture aspects. Theoretical and applied fracture mechanics, vol. 7. North-Holland; 1987. p. 51–63. [11] Etzler FM, Conners JJ. Temperature dependence of the heat capacity of water in small pores. Langmuir 1990;6(7):1250–3. [12] Federation International De La Precontrainte. Cryogenic behaviour of materials for prestressed concrete, State of art report, FIP; 1982. [13] Geiker M, Thaulow N. Ingress of moisture due to freeze/thaw exposure. In: Lindmark S, editor. Frost resistance of building materials. Report TVBM-3072, Lund Institute of Technology, Div of Building Materials; 1996. [14] GotoY, Miura T. Mechanical properties of concrete at very low temperature. In: Proceedings of the 21st Japan congress on materials research; 1978. p. 157–9. [15] Hanaor A. Microcracking and permeability of concrete to liquide nitrogen. J Am Concr Inst 1985. [16] Hansen EW, Stocker M, Schmidt R. Low-temperature phase transition of water confined in mesopores probed by NMR. Influence on pore size distribution. J Phys Chem 1996;100:2195–200. [17] Hedenblad G. Moisture permeability of mature concrete, Cement mortar and cement paste. Lund Institute of Technology, Div of Building Materials, Report TVBM-1014; 1993.
525
[18] Marchand J, Pleau R, Gagne´ R. Deterioration of concrete due to freezing and thawing. In: Skalny Jan, Mindess Sidney, editors. Materials science of concrete IV; 1995. [19] Marechal JC. Variation in the modulus of elasticity and poisson’s ratio with temperature. In: International seminar on concrete for nuclear reactors. ACI SP-34, Detroit (MI); 1972. p. 495–503. [20] Marshal AL. Cryogenic concrete. Cryogenics. Butterworth & Co (Publishers) Ltd.; 1982. p. 555–65. [21] Menzes NC. Etude du comportement des structures en be´ton soumises a` des tempe´ratures cryoge´niques. The`se de IIIeme` cycle. Paris VI, Paris: Universite´ Pierre et Marie Curie; 1986. [22] Mohd Yusof KB. The effects of moisture on the thermal properties of concrete between 80 °C and 0 °C. PhD thesis, University of Southampton, UK; 1984. [23] Pel L, Kopinga K, Hazrati K. Water distribution and pore structure in concrete as determined by NMR. In: Proc 9, Feuchtetag, Weimar, September 17–18, 1997. p. 294–300. [24] Okada T, Iguro M. Bending behaviour of prestressed concrete beam under low tempe´rature. J Jpn Prestressed Concr Eng Assoc 1978. [25] Pel L. Moisture transport in porous building materials. Ph D thesis, Eindhoven University of Technology; 1995. [26] Pigeon M, Pleau R. Durability of concrete in cold climates. London: Chapman & Hall; 1995. [27] Prado PJ, Balcom BJ, Beyea SD, Bremner TW, Armstrong RL, Grattan-Bellew PE. Concrete freeze/thaw as studied by magnetic resonance imaging. Cement Concr Res 1998;28(2):261–70. [28] Rostasy FS. Prestressing steels and systems for storage tanks for refrigerated liquefied gases. In: FIP symposium, New Delhi; 1986. [29] Rostazy FS, Wiedemann G. Strength, deformation and thermal strains of concrete at cryogenic conditions. In: First international conference cryogenic concrete, Newcastle upon Tyne; 1981. [30] Rostazy FS, Wiedemann G. Stress–strain behaviour of concrete at extremely low temperature. Int J Cem Concr Res 1980;10:565–72. [31] Scurlock RG, Yusof Kamarudin M. Thermal behaviour of concretes between 80 °C and +20 °C. In: Proceedings of the 2nd international conference on cryogenic concrete, Amsterdam; 1983. [32] Scurlock RG, Mohd Yusof KB. Cryogenic and frostproof concrete made from conventional aggregates and Portland cement. Adv Cryo Eng 2000;45:1779. [33] Scurlock RG. Dewars low loss. Cryogenic and frostproof concrete, low loss dewars and tanks. Cryogenic Society of America. 2005; p. 35–7. [34] Scurlock RG. Cryogenics at the University of Southampton: A review. Institute of Cryogenics, University of Southampton, Southampton S09 5N H, UK, 19 February, 1988. p. 439–66. [35] Van Der Veen C. In: Delft TU, editors. Properties of concrete at very low temperatures. A survey of literature; 1987. [36] Wolter B, Dobmann G. Nuclear magnetic resonance as a tool for the characterisation of concrete in different stages of its development. In: Proceedings of the international symposium on non-destructive testing in civil engineering (NDT-CE), Berlin, Germany, vol. 1; 1995. p. 181. [37] Vandewalle L. Bond between a reinforcement bar and concrete at normal and cryogenic temperatures. J Mater Sci Lett 1989;8: 147–9.