EVALUATION METHOD OF DULABILITY IN CONCRETE STRUCTURES

Transcription

1 EVALUATION METHOD OF DULABILITY IN CONCRETE STRUCTURES Koji TAKEWAKA 1 an Koji SAKAI 2 SUMMARY In general, concrete is quite a urable construction material so that a countless number of public structures an builings have been constructe by using it worlwiely. However, when improper selections of use materials an mix proportion of concrete are carrie out, or when the structure is expose to very severe environment, some of eterioration may occur on the concrete structures of which functions ecrease in their levels less than the require ones. Therefore, an evaluation of urability in concrete structure throughout its service life is very important at the esign stage not only for selecting appropriately the materials an the mix proportion of concrete an eciing structural etails of the structure, but also for ensuring the reliability of performance of the esigne structure. This paper escribes the present state of evaluation methos for urability of concrete structures as well as urability esign concepts which are base on the esign methos of the JSCE Stanar Specifications for Concrete Structures. Keywors: Evaluation of urability; urability esign; eterioration of concrete; carbonation; reinforcement corrosion ue to the ingress of chlorie ions; freezing an thawing action; chemical attack; alkali aggregate reaction; JSCE Stanar Specification. INTRODUCTION Concrete is one of the most important materials employe in public works an builing construction projects an a countless number of concrete structures have been constructe worlwie. The conventional esign framework for concrete structures is primarily base on safety but currently focuse on the aspects of urability. Uner the conitions where problems on improper selections of use materials an mix proportion of concrete have arisen an some of the concrete structures are affecte by har eterioration actions in very severe environment, the urability of concrete structures surely be threatene to reuce its performance. Furthermore, it is obvious that environmental aspects shoul be also 1 Associate Professor, Department of Ocean Civil Engineering, Kagoshima University, JAPAN , takewaka@oce.kagoshima-u.ac.jp 2 Professor, Department of Safety Systems Construction Engineering, Kagawa University, JAPAN , sakai@eng.kagawa-u.ac.jp 99

2 incorporate into the esign of concrete structures. From an environmental viewpoint, it can generally be thought that life-extension of a structure is irectly relate to the reuction of environmental impact. Consequently, the establishment of reasonable urability esigns for concrete structures is very important, an now a ay, many efforts have been mae to evaluate the urability of concrete structures. In 2002, the JSCE Stanar Specification for Concrete Structures (JSCE 2002) provie tools for verifying urability of concrete structures numerically, which was the first real coe for the urability esign in the worl. This paper outlines the present state of urability esign, then escribes the urability esign methos of the JSCE PRESENT STATE OF DURABILITY DESIGN We have been facing various an sometimes serious urability problems in concrete structures. In general, urability is efine as the ability of concrete to resist weathering action, chemical attack, abrasion, an other conitions of service. The severity of environmental, chemical an physical attacks on concrete epens on the properties of concrete an its exposure conitions. The actual eterioration phenomena of concrete structures inclue corrosion of reinforcement in concrete ue to the ingress of chlorie ions, carbonation of concrete, freezing an thawing, alkali-aggregate reaction, etc. As the corrosion of reinforcement ue to the ingress of chlorie ions which are supplie from marine environment, use of eicing salt scattere on the roa in the winter season, etc., is a typical an one of the most severe eterioration on concrete structures, it has been unerstoo that concrete cover an its quality are the key an many efforts have been mae to evaluate it quantitatively. The carbonation of concrete ue to carbon ioxie has a isavantage in reinforce concrete because the lower ph in carbonate concrete can not keep the passive conition on steel an introuces corrosion on the reinforcement. Freezing an thawing resistance can be secure by introucing an appropriate air-voi system in concrete. The alkali-aggregate reaction can be prevente by several countermeasures such as the setting of threshol content of alkali in cement, the utilization of blast furnace slag an fly ash as mineral amixtures, etc. Nevertheless, most of the existing urability esign coes o not provie a tool for evaluating the ingress of chlorie ions in concrete, the carbonation of concrete, an other eteriorations. Durability esign may be categorize into three levels as follows: (1) prescriptive esign (2) performance-type esign (3) performance-base esign In the prescriptive esign for the urability of concrete, for example, the maximum water cement ratio an minimum cement content are provie epening on the exposure conitions. Table 1 inicate an example of the recommene limiting values for composition an properties of concrete provie in the EN206 (BS EN ). All requirements are stipulate in a prescriptive manner. It may be sai that prescriptive esign is base on the simplification of safety sie from real performance. However, the backgroun of most provisions is unclear. For example, the water-cement ratio for corrosion protection of 100

3 reinforcement in concrete cannot be easily etermine because it is also irectly relate to the concrete cover. Therefore, it is more reasonable an accurate to consier performance with time. In principle, the require performance shoul be verifie through a irect analysis of time-epenent behavior of a concrete structure uner the assume environmental actions. At present, however, it is ifficult to preict the urability performance of a structure irectly throughout the lifespan because of the inaequacy of the moels necessary for calculations. Further evelopment of research on numerical approaches will pave the way to the realization of performance-base esign. Uner such the current situations, what we can o to verify the urability in our coes is to introuce a performance-type esign in which principal urability performance is consiere with time. The JSCE 2002 provie such a esign concept for urability esign. Framework JSCE STANDARD SPECIFICATION FOR CONCRETE STRUCTURES The JSCE Stanar Specification stipulates its funamental concepts at the esign stage of structures as follows: At the esign stage, structural etails such as the shape, size, reinforcement arrangement, require properties of concrete an reinforcing material, metho of construction (in situ, pre-cast, etc.) an maintenance plan shoul be ecie also taking into account the economic consieration. It shoul also be ensure that the require performances in terms of serviceability, safety, urability an compatibility with the environment, etc. are satisfie over the service life of the structure. Base on these concepts, the following four parts were prepare in the Specifications: (1) Structural performance verification (2) Seismic performance verification (3) Materials an construction (4) Maintenance The proceure for verifying mechanical performance of concrete structures is given in the parts of the structural performance verification an the seismic performance verification. The materials an construction part provies the general framework an sets stanars for general concrete construction with issues relate to use of special concrete or the construction of special structures. Aitionally, this part eals with the performance verification for egraation of a structure on account of the eterioration of concrete an reinforcement. The performance of concrete structures varies over time ue to environmental conitions an other factors. The examination on whether such change is in acceptable range is escribe here. Once the construction is complete, it is ifficult to repair, strengthen or renovate concrete structure, so thorough investigation at the beginning stage of esign, accurate preiction for possible problem in service life an future maintenance are of great importance. The maintenance part provies basic knowlege for the maintenance of concrete structures. Figure 1 inicates the framework of the contents to be covere in each specification. 101

4 Durability Verification of Concrete Structures General concept The performance of concrete structures has to remain the require one throughout its esign service life. Conversely meaning, there will be no problems when utilize uner a certain conition in which the require performance of the concrete structure is satisfie even if concrete or reinforcements partially eteriorate. This is a funamental concept of the urability esign that requires the performance verification of urability for concrete structures, not for concrete. In the urability verification of concrete structures in the JSCE Stanar Specifications, this concept is also provie while the concept in the existing prescriptive urability esign methos is completely ifferent from it. In the Specification, the performance verification methos for the eterioration of structure ue to carbonation, the ingress of chlorie ions, cyclic freezing an thawing action, chemical attack, an alkali aggregate reaction are stipulate as well as the verification metho for water tightness an fire resistance of the structure are ealt with. In the following sections, these verification methos are escribe in etail. Verification for carbonation As the carbon ioxie from the atmosphere penetrates into concrete, the ph in the cover concrete reuces. If such a zone of reuce ph reaches the location of the reinforcing bars in the concrete, they are renere susceptible to corrosion. Once the corrosion is initiate, the formation an eposition of expansive corrosion proucts on the bars may cause the formation of longituinal cracks along the reinforcing bars, which in turn coul accelerate further corrosion an spalling of the cover concrete, an finally a significant reuction in the cross-section of the reinforcement. Therefore, it must be ensure in the urability esign that the performance of the structure is not allowe to fall below the require level by the corrosion of the reinforcing bars. Base on this concept, to check that reinforcement corrosion ue to carbonation in concrete oes not occur, is comparatively easy an essentially on the conservative sie. It is sufficient to verify that the epth of carbonation is less than the critical epth to initiate steel corrosion. Consequently, the verification of a structure for carbonation is conucte by ensuring that y i y lim 1.0 (1) i : Factor representing the importance of the structure. In general, it may be taken as 1.0, but may be increase to 1.1 for important structures. y lim : Critical carbonation epth of steel corrosion initiation. y : Design value of carbonation epth. It is unerstoo in general that corrosion of steel reinforcement may begin before the carbonation epth actually becomes equal to the cover thickness (i.e. the carbonation front reaches the location of the reinforcement). It means that the critical carbonation epth of steel corrosion initiation y lim may be obtaine from Equation 2. ylim = c c k (2) 102

5 c : Expecte value of cover thickness (mm). In general, it may be taken as the esign cover thickness (mm) c : Remaining non-carbonate cover thickness (mm). k The remaining non-carbonate cover thickness is efine as a thickness of uncarbonate concrete still remaining in the neighborhoo of the steel reinforcement even when the reinforcement corrosion just initiates, as shown in Figure 2. Usually, there are inee very rare cases in which corrosion is able to impair the performance of a structure having so long as the remaining non-carbonate cover was more than 10mm. However, when chlorie ions are present in concrete, corrosion may start even at larger remaining non-carbonate cover thickness (more than the 10mm) because the chlorie ions can easily cause amage to the passivation film of the reinforcement in the conition having the high alkalinity. Therefore, in such cases, higher remaining non-carbonate cover thickness (i.e. 10 to 25mm) is recommene. For estimating the epth of carbonation, the square-root law (i.e. the epth of carbonation varies linearly with relate to the square-root of time) is normally use, as it is consiere to agree with previous stuies an is easy to use. Design value of carbonation epth y, therefore, may be obtaine from Equation 3. y = cb α t (3) α : Design carbonation rate ( mm year ), t : Designe service life of structure (year). cb : Safety factor to account for the variation in the esign value of carbonation epth. Normally it may be taken as In the case of high fluiity concrete, it may be taken as 1.1. In the JSCE Specification, the esign carbonation rate α is also given as Equation 4. α = α k β e c (4) α k : Characteristic value of carbonation rate ( mm year ) β e : Coefficient representing the extent of environmental action. It may be taken as 1.0 for environments in which the surface of structure is ifficult to be rie out, or for the north-facing surface. It may be increase to 1.6 for environments in which structures can be easily rie out or for the south-facing surface c : Factor to account for the material properties of concrete. In general it may be taken as 1.0, but shoul be taken as 1.3 for upper portions of the structure. Figure 3 shows the relation between the effective biner ratio an coefficient of carbonation spee. The ata consist of ifferent types of biner, incluing fly ash an blast furnace slag. Although there is large scattering in the ata, the following equation can be introuce to estimate the characteristic value of carbonation rateα k : k = W/B (5) where, W/B: water biner ratio It is obvious that the carbonation of concrete is epenent on the exposure environment of the 103

6 structure. To consier the effect of exposure environment, environmental factor βe is introuce in Equation 4. As inicate in Figure 4, the values of β e for concrete in a ry environment or when concrete faces south irection an that in a wet environment or when the concrete faces north irection are 1.6 an 1.0, respectively. Figure 5 shows the require minimum cover thickness at ifferent water cement ratios an service years calculate accoring to the verification metho mentione above, in which the corrosion of reinforcing bars ue to carbonation of concrete is prevente. Verification for reinforcement corrosion ue to the ingress of chlorie ions Chlorie ions can easily penetrate into concrete from outer environment an cause the corrosion of steel reinforcements in concrete. In cases of the concrete structures in marine environment, the ones on which the eicing agents are use, an etc., it is very important to verify the reinforcement corrosion ue to the ingress of chlorie ions. As well as the matter mentione in the clause of the verification for carbonation, if the structural performance oes not appear to be impaire, it may still be consiere serviceable even if there are some signs of reinforcement corrosion on the structure. In other wors, the structural integrity of the structure may consiere to be intact so long as corrosion inuce longituinal cracks are not forme along the reinforcing bars, even if there are some other signs of reinforcement corrosion. However, as far as the verification of the performance of a structure with respect to chlorie penetration is concerne, a conition that chlorie inuce corrosion of the reinforcement shoul not occur uring the service life of the structure, is relatively easy to unerstan an on the safe sie. It may be sufficient to carry out to ensure uring verification that the chlorie ion concentration at the location of the reinforcement is below the critical concentration that coul initiate corrosion in the reinforcement. In the Stanar Specification, therefore, the verification of a structure for reinforcement corrosion ue to the ingress of chlorie ions is conucte by ensuring that C i 1.0 (6) Clim i : Factor representing the importance of the structure. In general, it may be taken as 1.0, but shoul be increase to 1.1 for important structures C lim : Critical chlorie concentration for initiation of steel corrosion. C : Design value of chlorie ion concentration at the epth of reinforcement. C lim The critical chlorie concentration in the neighborhoo of the reinforcement, which coul initiate corrosion in the reinforcement, has been reporte to be about kg/m 3 of concrete by the past research works. For example, values of about kg/m 3 have been reporte from accelerate tests carrie out with chlories ae to the fresh concrete, an values of kg/m 3 from exposure tests carrie out in the fiel. These ifferences in the critical chlorie concentrations reporte on the basis of accelerate an exposure tests, etc., can be attribute to factors such as ifferences in the water-cement ratio of the concrete, cover to the reinforcement, etc. The effect of high temperatures sometimes use in accelerate tests can also influence conclusions. Taking all the factors into account, the Specification sets a limit of 1.2 kg/m 3 (of chlorie per cubic meter of concrete) as the critical chlorie 104

7 C lim concentration, which can initiate reinforcement corrosion of which level is consiere etrimental to the performance of the structure. Mathematical formulations base on the iffusion theory are most commonly use to moel chlorie penetration in concrete. When using the equations such as Equation 7 base on the Fick s secon law of iffusion to moel, the esign value of chlorie ion concentration at the epth of reinforcement is consiere satisfactory. C C (1 c = cl C0 erf ( )) (7) D t C 0 : Assume chlorie ion concentration at concrete surface (kg/m 3 ). Generally, it may be obtaine from Table 2 c : Expecte value of concrete cover thickness (mm). In general, the esigne cover thickness may be selecte. t : Design service life of the structure (year). : Safety factor, to account for the variation in the esign value of the chlorie ion cl concentration at the epth of reinforcement C. Normally, it may be set at 1.3, but in the case of high fluiity concrete, a value of 1.1 may be selecte. D : Design value of iffusion coefficient of chlorie ions into concrete (cm 2 /year) erf(s) : Error function, efine as 2 s erf s = e 2 η ( ) η 1/ 2 0 This equation is base on the concept that chlorie ions iffuse equally towar steel. When bening crack is generate on concrete cover, chlorie ion concentration also epens on the istance from the crack, even though the variation of concentration of chlorie ion may tren to be small when crack with is very narrow. In the JSCE Specification, therefore, such influence of the crack on the iffusibility of chlorie ions can be evaluate by introucing a concept of the average iffusion coefficient of chlorie ion on cover concrete, escribe as Equation 8, in which the influence of both concrete quality an concrete crack are consiere. w w D = c Dk + D0 l w (8) a c : Factor to account for the material properties of concrete. In general, it may be set at 1.0 but shoul be taken as 1.3 for upper portions of the structure. However, if there is no ifference in the quality of concrete in structure an that of specimens cure in laboratory, this value may be taken as 1.0 even for all portions of the structure. D k : Characteristic value of iffusion coefficient of chlorie ion in concrete (cm 2 /year) D o : Constant to express the influence of crack on the movement of chlorie ions into concrete (cm 2 /year). In general, it may be taken as 200cm 2 /year w: Crack with (mm). w a : Allowable crack with (mm). w/l: Ratio of crack with to crack interval The ratio of crack with to crack interval, w/l is introuce in orer to express the influence of crack averagely. This averaging metho is effective when crack with is controlle 2 105

8 smaller than allowable crack, but the resistance to ingress of chlorie ions may be rapily influence by the crack if crack with is large. (w/w a ) 2 in the equation inicates the influence of crack with. In the calculation, the chlorie iffusion coefficient must be set as the characteristic value D k. In the Specification, the chlorie iffusion coefficient is essentially etermine on the basis of the actual measurement of chlorie ion concentration istributing into concrete. However, from the values reporte on the basis of the surveys from a large number of structures as shown in Figure 6, it is clear that there is a large variation in the iffusion coefficient etermine from the chlorie concentration istributions in these structures. This means that the actual concrete has a iffusion coefficient for which the characteristic value for esign is on the safety sie as escribe in Equation 9. D p 1.0 (9) p Dk D : Preicte value of iffusion coefficient of actual concrete (cm 2 / year). p p : Safety factor to account for the accuracy in etermining D p In the Specification, the following equation for the preiction of D p, were introuce as the preicte values: (a) When orinary Portlan cement is use; 2 log = 3.9( W / C) + 7.2( W / C) 2.5 (10) D p (b) When blast furnace slag cement is use; 2 log = 3.0( W / C) + 5.4( W / C) 2.2 D p Figure 7 shows the require minimum cover thickness at ifferent water cement ratios for concrete structures suffering from the chlorie attack uner 50 years of esign service life. In the calculations, Equation (10) an (11) is use for preicting the iffusion coefficients of the concretes mae of orinary Portlan cement an blast furnace slag cement, respectively. Verification for cyclic freezing an thawing action In col regions, a cyclic freezing an thawing action is essential as the cause of eterioration on concrete structures, an raises pop-outs, scaling, an formation of micro cracks at the concrete surface. The egree of these amages on concrete structures epens not only on the quality of concrete but also several other factors, such as the number of cycles of freezing an thawing actions, lowest temperature an the egree of water saturation of the concrete. However, up to now, virtually no quantitative information is available to relate these amages to any change in the performance of the structure uner the freezing an thawing action. This means that it is ifficult to specify a threshol level of allowable eterioration an to utilize such the permissible level as inices or criteria for the verification of the require performance of the structures subjecte to this action. Only a realistic manner is to examine whether or not eterioration ue to the freezing an thawing action is likely to occur in a structure, or how much egree of eterioration will progress insie concrete. In general, it can be consiere that the structures will keep their require performance when the concrete has sufficient resistance to cyclic freezing an thawing action. Therefore, in the normal concrete structures uner the action, a eterioration level in which some eterioration (11) 106

9 may occur on concrete but any egraation in the functions of the structure oes not arise can be consiere as a stanar threshol level to keep the urability in structures. The relationship between the results of the accelerate tests an the change in the level of performance in an actual structure subjecte to cyclic freezing an thawing action is somewhat better unerstoo on the basis of past research works an fiel ata. It recommens that parameters such as the relative ynamic moulus of elasticity, which are measure in the accelerate tests, can be use as inices for the verification of the resistance of concrete to the action. Consequently, in the JSCE Specification, the verification of a structure for cyclic freezing an thawing action is conucte by ensuring that Emin i 1.0 (12) E i : Factor representing the importance of the structure. In general, it may be taken as 1.0, but may be increase to 1.1 for important structures. E : Design value of relative ynamic moulus of elasticity an is given as Equation 13. Ek E = (13) c where, E k : Characteristic value of the relative ynamic moulus of elasticity. c : Material factor. In general, it may be set at 1.0, but shoul be taken as 1.3 for upper positions of the structure. However, if there is no ifference in the quality of concrete in the structure (in situ) an that of laboratory-cure specimens, this value may be set at 1.0 for all E min positions. : Critical minimum value of relative ynamic moulus of elasticity to ensure require performance of the structure uner cyclic freezing an thawing action. In general, it may be obtaine from Table 3. As for preicting the relative ynamic moulus of elasticity, the accelerate test in accorance with the metho specifie in JIS A 1148 (A metho) The freeze-thaw test metho of concrete (the freeze-thaw test metho of concrete unerwater) is recommene. The relationship between the characteristic value of relative ynamic moulus of elasticity of concrete E k an the preicte one by the accelerate test E p is shown in Equation 14. E p 1.0 (14) p Ek E p : Preicte value of relative ynamic moulus of elasticity of concrete. p : Safety factor to account for the accuracy in etermining E p. When tests are carrie out in accorance with JIS A 1148 (A metho), it may be taken to be 1.0 If actual freezing an thawing conitions are more severe than the conition stipulate in JIS A 1148 (A metho), or if the structure is being esigne for a very long service life, the test conitions in terms of the freezing an thawing temperatures, the time urations for the freezing an thawing cycles, etc. shoul be appropriately change in the acceleration test conition. 107

10 In orer that ams an other important structures satisfy the level of urability require, it may be essential to not allow the relative ynamic moulus of elasticity to fall below 80%. Further, in cases where it is require that the structure retains its original appearance an sounness over the entire esign service life, no change in the value of the relative ynamic moulus of elasticity may be acceptable. In such cases, the relative ynamic moulus of elasticity may be kept at a level higher than 95%. Verification for chemical attack When aggressive chemicals come in contact with or penetrate the concrete, they may react with the hyration proucts of cement causing issolution of the concrete or formation of expansive proucts which causes cracking in concrete an is sometime followe by spalling of the cover concrete etc. However, ue to the limite unerstaning on how the eterioration of concrete uner chemical attack results in the egraation of the structure performance, unfortunately quantitative evaluation has not been realize. Therefore, only conceptual provisions are introuce in the JSCE Specification for the verification of a structure for chemical attack. The egraation in performance an the change in performance with time in the structure uner the chemical attack are basically relate to the change in the performances of the concrete itself as a constituent material. Thus, in orer to ensure the urability of the structure uner the chemical attack, it may be enough to ensure that the eterioration of concrete is kept below a certain pre-etermine level (so as not to cause unacceptable levels of changes in the structural performance). It also be convenient an rational to actually set a certain level of chemical resistance in the concrete in orer to preserve the integrity of the structure. Consequently, in case the concrete meets the criteria for resistance against chemical attack, the structure may be assume that its performance will not be impaire on account of chemical attack In an actual verification for resistance to chemical attack in concrete, accelerate tests, exposure tests, or any other suitable tests shall be performe on concrete specimens at the conitions as close to the actual conitions as possible. Then, the verification shoul be carrie in a manner to ensure that there is no notable eterioration in concrete on account of the chemical attack, or that the eterioration is confine to levels that o not significantly affect the performance of the structure. In general, the require performance level for structures in a relatively mil chemical environment, such as aci rain, marine structures, etc., may be set as a level such that any eterioration ue to chemical attack is not easily seen by the nake eye. However, in cases when the chemical attack is likely to be severe, such as in the case of structures in contact with sewage or in the hot springs, etc., the require performance level may be set at a level, where the eterioration oes not affect the structural performance of the structure. In all cases the importance of the structure shoul be uly accounte for in the process of etermining the require level of performance. When it is require that the eterioration of concrete is just within the level that oes not affect the require performance of the structure, it is allowe in the specification to use the following maximum water-cement ratio, instea of the verification on the resistance to chemical attack of concrete irectly: (a) When concrete are in contact with soil or water, which contain 0.2% or more of a 108

11 sulfate such as SO 4, maximum water cement ratio is 50%. (b) When eicing salts are use, maximum water cement ratio is 45%. When the chemical attack on the concrete is very severe, it may be ifficult to secure the performance of the structure against chemical attack simply by increasing the concrete cover thickness an increasing the resistance of concrete. Certain sewer facilities an structures near hot springs coul be examples of such concrete structures. In such cases, it may be more realistic an rational to recommen the use of corrosion resistant reinforcement materials an/or proviing surface coating for the concrete. Verification for alkali-aggregate reaction Depening on the kin of reactive mineral present in the aggregate that reacts with the alkali in the concrete, alkali aggregate reaction can be ivie basically into two types, which are alkali silica reaction an alkali-carbonate reaction. However, from the occurrence of alkali aggregate reaction in ifferent regions an countries in the worl, it is clear that most of the alkali aggregate reaction is in fact the alkali-silica reaction, so that, it has been assume that it may be sufficient to carry out appropriate verification for the alkali-silica reaction. Deterioration ue to alkali aggregate reaction in concrete is characterize as the eterioration to evelop cracks in concrete by the formation of reaction proucts which are generate from the reaction of alkali in the concrete with the reactive minerals in the aggregate an then absorb water accompanying an expansion of the prouct. However a clear relationship between the egree of the cracking in concrete an the egraation rate in performance or its changes with time of the concrete structure has not been establishe yet. There are also reports that the structural loa-carrying capacity of reinforce concrete structures is not impaire by the occurrence of alkali aggregate reaction, proviing that they have at least a certain amount of reinforcement. However, it has been consiere appropriately to set the minimum require performance level of a structure susceptible to alkali aggregate reaction as a level which oes not cause onset of the reaction inuce cracks in concrete uring the service life of the structure. The following implications of the crack formation have been kept in min in fixing this limit (i) onset of these cracks accompanies an appearance of a white gel-like substance which is aesthetically unappealing, the cracks provie accelerate access to other eleterious material, such as gases, water an ions, an (iii) it is very ifficult to stop the alkali aggregate reaction through repair methos once cracks have been forme. Furthermore, in some structure, breakage of the reinforcing steels has been foun out ue to the large expansion of concrete. As same as in the case of egraation in the performance of a structure uner chemical attack escribe in the previous clause, the egraation in performance an the changes in performance with time in a structure ue to the alkali aggregate reaction are also basically relate to the changes in the performance of the concrete itself as a constituent material. Thus, in orer to ensure the urability of a structure susceptible to alkali aggregate reaction, it may be enough to ensure that the eterioration in the properties of concrete is kept below a certain pre-etermine level so as not to cause unacceptable levels of changes in the structural performance. In the JSCE Specification, therefore, it is stipulate that in case the concrete meets criteria for resistance against alkali aggregate reaction, the structure may be assume that its performance will not be impaire on account of alkali aggregate reaction. 109

12 The most reliable metho to verify the resistance of a structure to alkali aggregate reaction is to cast concrete specimens in the same conitions as the real structure, then expose them in similar environmental conitions an confirm the possibility of crack formation. However, consiering various involving factors, such as the testing time as well as the require expense, the nee to test concrete for various kins of materials an proportions, etc., this real exposure test is not always feasible. Therefore, at present the verification for the resistance to alkali aggregate reaction is usually carrie out on the basis of the accelerate tests using concrete specimens. The JCI AAR-3 Test metho for evaluation of aggregate reactivity in concrete is one of such methos that may be use as a reference in carrying out accelerate tests. It has been confirme from laboratory experiments an fiel investigations that an expansion level of less than 0.1% at the age of 6 months in the test carrie out base on this metho may not cause appreciable egraation in the performance of concrete structures. This level can thus, be use as a criterion to ecie whether or not the expansion ue to alkali aggregate reaction coul be etrimental. Namely, the verification for resistance to alkali aggregate shall be conucte out by insuring that: L p p 1.0 (15) Lmax L p : Preicte expansion of concrete ue to alkali aggregate reaction (%). Generally, this is set equal to the expansion at 6-month of test perios in accorance with JCI AAR-3. L max : Maximum permissible expansion rate at which concrete still satisfies the require resistance against alkali aggregate reaction. Generally, it may be set at 0.10%. p : Safety factor to account for the accuracy in etermining. When tests are carrie out in accorance with JCI AAR-3, it may be taken to be 1.0 When using aggregates which have been conforme to non-reactive as per the test metho for evaluation of aggregate reactivity in concrete, an when restraining alkali aggregate reaction in concrete using appropriate cement, the verification for the resistance to alkali aggregate reaction can be omitte. On the other han, it shoul be borne in min that egraation in performance of a structure suffering from alkali aggregate reaction coul be accelerate in the presence of chlorie ions penetrating into concrete from outsie. The fact shoul be aequately accounte for a higher level of resistance to alkali aggregate reaction require of such a concrete. CONCLUDING REMARKS In 2002, the JSCE Stanar Specification was the first to provie a concept for urability esign of concrete structures as a part of the performance-base esign concept, while most of the existing esign coes in the worl have not yet provie such tools. This paper introuce the outline of the verification methos of the urability escribe in the Specification. However, the metho mentione here is not a final goal but the avance methos for preicting irectly the urability performance of the structure throughout the lifespan is require for the realization of performance-base esign. Concrete structures an their components shoul be planne, esigne, constructe, an operate, inspecte, maintaine an repaire in such a way that they maintain their require 110

13 performance uring their esign lives with sufficient reliability for the safety an the intene use of the structure, etc., uner expecte environmental conitions. As a result, the preicte service lives, t S, of the structure an its components shoul meet or excee their esign lives, t D, though t S may be possible to be less than t D for the components that are inspectable an replaceable. Figure 7 shows one of the mathematical moels for preicting the service life. In the moel, the verification of the urability of the structure in the performance-base esign concept is conucte by ensuring that P { f } = P{ R( td ) S( td ) < 0} < Pt arg et t D or { f } t = P{ ts td} Pt et P D arg R( t D ): Resistance capacity of the structural component at the esign life, t D S( t D ): Cumulative egraation of the component at the esign life, t D P f : Probability of failure in the esigne structure at the esign life, t D { } t D Pt arg et : Specifie target probability (16) (17) { } D The probability of failure in the esigne structure P f t in Equation 16 is inicate in Figure 8 by the shae overlap area of the probability ensity curves for R(t) an S(t) on a vertical axis, while the P{ f } t D in Equation 17 inicate by the shae area of the probability ensity function shown on the horizontal axis. For completion of this moel, however, it is important that R(t) an S(t), which are the moels evaluating the time-epenent behaviors in the performances of the structure an the progress of the egraation ue to eterioration actions respectively, have to be evelope as quantitative analysis moels on the basis of the probability theory. Furthermore, the concept of life cycle cost is also important as the factors eciing the specifie target probability Pt arg et, so that a quantitative evaluation metho of the life cycle cost shoul be establishe as soon as possible as well. There is no nee to mention that concrete structures have contribute to the social an economic activities of human beings. On the other han, civil engineering an builing structures consume enormous resources an emit huge amount of greenhouse gases. Durability problems of concrete structures are irectly linke to environmental impacts because the shortening of the lifespan will result in the wasteful utilization of limite natural resources. Thus, in the viewpoints not only of the safety of the structures in their service lives but also of the preservation of environmental conitions, it will become more important to improve urability esign methos. REFERENCES British Stanar EN206-1, Concrete Part 1: Specification, performance, prouction an conformity, Japan Society of Civil Engineers, Environmental impact evaluation of concrete ( ), Concrete Engineering Series 62, ISO (WD), General Principles on the Design of Structures for Durability, April

14 Table 1 An example of recommene limiting values for composition an properties of concrete EN206 Maximum w/c Minimum strength class Minimum cement content (kg/m 3 ) No risk of corrosi on or attack Carbonation-inuce corrosion Exposure classes Chlorie-inuce corrosion Sea water Chlorie other than sea water Freeze/thaw attack Aggressive chemical environments X0 XC1 XC2 XC3 XC4 XS1 XS2 XS3 XD1 XD2 XD3 XF1 XF2 XF3 XF4 XA1 XA2 XA C12/15 C20/25 C25/30 C30/37 C30/37 C30/37 C35/45 C35/45 C30/37 C30/37 C35/45 C30/37 C25/30 C30/37 C30/37 C30/37 C30/37 C35/ Minimum air content ( ) Other requirements 4.0 1) 4.0 1) 4.0 1) Freeze/thraw resisting aggregates in accorance with the recommenations in pren 12620:1996 Surfateresisting cement 2) 1) Where the concrete is not air entraine, the performance of concrete shoul be teste accoring to ISO FFF-1 in comparison with a concrete for which freeze/thaw resistance for the relevant exposure class is proven. 2) When SO 4 1eas to exposure classes XA2 an XA3, it is essential to use sulfate-resisting cement. Where cement is classifie with respect to sulfate resistance, moerate or high sulfate-resisting cement shoul be use in exposure class XA2 (an in exposure class XA1 when applicable) an high sulfate-resisting cement shoul be use exposure class XA3. 112

15 Table 2 Chlorie ion concentration at concrete surface C 0 (kg/m 3 ) Distance from the coastline (km) Tial an splash zone Near coastline Consiering the effect of elevation above the water surface, an increase of 1m in elevation may be consiere to be equivalent to a horizontal istance of 25m. The equivalent C 0 may then be calculate from the above table. Table 3 Minimum critical level, E min (%), of relative ynamic moulus of elasticity to ensure a satisfactory performance of the structure uner cyclic freezing an thawing action Climate Section Severe weather conitions or frequent cyclic freezing an thawing action Not so severe weather conitions, atmospheric temperature rarely rop to below 0 C Exposure Thin 2) General Thin 2) General of structure (1) Immerse in water or often saturate with water 1) (2) Not covere in item (1) above an subjecte to normal exposure conitions ) Structures close to the water surface or in contact with water such as waterways, water-tanks, abutments of brige, brige piers, retaining walls, tunnel linings, etc. Besies, structures such as slabs, beams etc not close to the water surface but may be expose to snow, water flow, spray, etc., also belong to this category. 2) Members with thickness less than 20cm may be consiere thin. 3) Generally, the verification in section (2) may be omitte in cases when the characteristic value of relative ynamic moulus of elasticity, E k, is higher than

16 Require performance of structure Construction conition Design of structure Design Earthquake resistance esign volume Maintenance management manual Setup of structure Setup of concrete Construction Construction Plan Setup of maintenance management Selection of reinforcing material Setup of construction Verification of structural performance Structural performance Durability performance Figure 1 Works at esign stage Only CaCO 3 CaCO 3 + Ca(OH) 2 Only Ca(OH) 2 ph value Concrete cover Critical epth Corrosion occurs on rebar ph profile in concrete at esign service life Remaining non-carbonation cover thickness (about 10mm) Distance from concrete surface Concrete color changes by phenolphtalein solution Figure 2 Schematic figure on ph istribution in carbonate concrete 114

17 1.5 1 Coefficient of carbonation spee (cm/ year) Fly ash committee OPC Rapi PC Low heat PC Fly ash C Blast furnace C W/B Coefficient of carbonation spee (cm/ year) East West South North Southeast Northeast = ( W/B ) Effective water biner ratio (%) Figure 3 Relation between water biner ratio an carbonation spee Effective water biner ratio (%) Figure 4 Relation between environment an carbonation spee Orinary portlan cement 7 6 Blast furnace cement (Slag content 40%) Years 100 Concrete cover (cm) Wet environment, e=1.0 Dry environment, e=1.6 Years Concrete cover (cm) Wet environment, e=1.0 Dry environment, e= W/C Figure 5 Minimum cover at ifferent water cement ratio an service years for concrete structures uner carbonation W/C 115

18 Diffusion coefficient, Dp (cm 2 /year) E+02 1.E+01 1.E+00 1.E-01 Water cement ratio (W/C) Orinary Portlan cement concrete log Dp=-3.9(W/C) (W/C) Diffusion coefficient, Dp (cm 2 /year) E+01 1.E+00 1.E-01 Water cbiner ratio (W/C) Concrete using blast furnace slag an/or silica fume log Dp=-3.0(W/C) (W/C) E-02 1.E-02 Figure 6 Diffusion coefficient Concrete cover (cm) Orinary Portlan cement Design service life: 50 years Splash zone Coast line 0.25km 0.5km 1km Concrete cover (cm) W/C W/C Figure 7 Minimum cover at ifferent water cement ratio require for concrete structure uner chlorie attack Blast furnace slag 50% Design service life: 50 years Splash zone Coast line 0.25km 0.5km 1km

19 R, S Probability ensity function of R(tD) Mean R(t) P[f]tD Mean S(t) Probability ensity function of S (td) Mean service life Time Design Service Life td Target failure probability P[f]tD Probability ensity function of service life Figure 8 An example of mathematical moel for preicting service life of structures 117