Effect of Interacted Deterioration Parameters on Service ... · PDF fileRESEARCH REPORT...

RESEARCH REPORT VTT-R-09217-08

Effect of Interacted Deterioration Parameters on Service Life of Concrete Structures in Cold Environments - State of the Art

Authors: Esko Sistonen (TKK) & Erkki Vesikari (VTT)

Confidentiality: Public


1 (47)

Report’s title Effect of Interacted Deterioration Parameters on Service Life of Concrete Structures in Cold Environments - State of the Art Customer, contact person, address Order reference TEKES-project Project name Project number/Short name Effect of Interacted Deterioration Parameters on Service Life of Concrete Structures in Cold Environments

DuraInt

Author(s) Pages Esko Sistonen (TKK) & Erkki Vesikari (VTT) 47 p. Keywords Report identification code Concrete, durability, interacted, interaction, synergistic, deterioration, degradation, service life, modelling, simulation, cold environment, frost, frost-salt, salt, carbonation, corrosion, abrasion, field test, State-of-the-art

VTT-R-09217-08

Summary In Nordic environments, concrete structures like dams, piers, bridges and lighthouses are often affected simultaneously by a wide range of degradation phenomena. However, synergistic effects are overlooked. There is therefore a need to develop a better understanding of the real interaction between these different actions in order to assess the service life of concrete elements exposed to natural conditions.

The objective of this paper is to review different forms of deterioration as they work individually and together and then form models for service life calculation.

When possible, interaction between two or more forms of damage is likely to accelerate degradation. Degree of degradation accounting for interaction when several types of deterioration are acting simultaneously are suggested based on review of physical mechanisms, superposition of degree of degradation with interaction coefficients and coupled degree of degradation equations. A review of some physical mechanisms of deterioration shows that interaction is mainly reducing the service life compared to uniform deterioration. A different approach is proposed by coupling individual service life equations using phenomenological coupling coefficients. Material data and observations of rate of degradation under several types of deterioration in field and/or in accelerated laboratory tests can be used to determine the coupling coefficients.

A general scheme for modelling combined effects of various degradation mechanisms in concrete structures is presented. This method is based on tests both in field and laboratory. As the degradation mechanisms are usually based on some known physical or chemical processes such as diffusion, chemical reactions, abrasion, phase transitions etc. simple degradation models can be derived by theoretical reasoning. The laboratory tests are used in parameterization and quantification of the theoretically created degradation models. Computer simulation is used by combining the effects of constantly changing climatic conditions and other degradation mechanisms on the treated degradation mechanism. Theoretically the computer simulation is a powerful way to treat such complex phenomena. However, the computer simulation must be calibrated with field and laboratory tests. By the use of calibrated computer simulation, service life models for engineering purposes can be developed. In the case of the factor method according to the Finnish national codes, most of the factors of the design formula can be determined using computer simulation. The various degradation mechanisms in structures can be simultaneously simulated throughout the service life. By making sequential computer runs with series of material, structural and environmental parameters the values for material, structural and environmental parameters of the design


3 (47)

Contents

Contents............................................................................................................... 3

1 Introduction ..................................................................................................... 4

2 Uniform deterioration ...................................................................................... 5 2.1 Abrasion of concrete by ice .................................................................... 5 2.2 Frost action ............................................................................................. 7 2.3 De-icer salt scaling ................................................................................. 8 2.4 Chloride induced reinforcement corrosion............................................. 11

3 Interaction ..................................................................................................... 12 3.1 Mechanisms of interaction .................................................................... 12

3.1.1 Internal cracks due to frost and surface scaling due to frost and de-icer salt ................................................................................ 13

3.1.2 Ice abrasion, frost damage and de-icer salt scaling .................. 13 3.1.3 Studded tire abrasion and frost-induced damage ...................... 13 3.1.4 Alkali aggregate reactions and frost .......................................... 14 3.1.5 Carbonation, chloride penetration and reinforcement corrosion 15

3.2 The combined degree of degradation with interaction by superposition 15 3.3 The combined degree of degradation by coupling with phenomenological

coefficients ........................................................................................... 16 3.4 Data to quantify the coupling between the deterministic service life

equations .............................................................................................. 17 3.4.1 Interaction damage table .......................................................... 17 3.4.2 Laboratory data ........................................................................ 18 3.4.3 Field data .................................................................................. 18 3.4.4 Varying values of individual deterioration parameters ............... 18 3.4.5 Other methods .......................................................................... 19

4 Analytic modeling of combined degradation mechanisms of concrete structures ...................................................................................................... 22 4.1 General scheme for modelling combined degradation .......................... 22

5 Phases of development of integrated degradation models ............................ 24 5.1 Laboratory tests .................................................................................... 24 5.2 Field tests ............................................................................................. 24 5.3 Simple analytic models ......................................................................... 25

5.3.1 Carbonation .............................................................................. 25 5.3.2 Chloride penetration ................................................................. 26 5.3.3 Corrosion of Reinforcement ...................................................... 27 5.3.4 Frost Attack .............................................................................. 29 5.3.5 Frost-salt scaling....................................................................... 31

5.4 Modelling of interaction between degradation modes with computer simulation ............................................................................................. 32 5.4.1 Effect of frost attack on carbonation or chloride penetration ...... 32 5.4.2 Effect of frost-salt attack on carbonation or chloride penetration33 5.4.3 Effect of frost-salt scaling and internal frost damage to the active

corrosion time ........................................................................... 34 5.5 Developing service life models using computer simulation ................... 35

5.5.1 Service life models in the Finnish national codes ...................... 35 5.5.2 Computer simulation in specification of factors of service life .... 36

6 Conclusions .................................................................................................. 38

References ......................................................................................................... 40


4 (47)

1 Introduction

In Nordic environments, concrete structures like dams, piers, bridges and lighthouses are often affected simultaneously by a wide range of degradation phenomena. However, in most laboratory procedures, these mechanisms are often isolated from one another and samples are being subjected to a single mechanism. Accordingly, synergistic effects are systematically overlooked. There is therefore a need to develop a better understanding of the real interaction between these different actions in order to assess the service life of concrete elements exposed to natural conditions. A review of the main mechanisms likely to affect concrete structures exposed to sub-zero temperatures is presented. These include frost-induced cracking, de-icer salt scaling, ice abrasion and impact by studded tires. Each of these phenomena is first discussed separately. Then, new approaches that allow taking into account the combined influence of multiple detrimental actions are described. Service life prediction for a concrete structure exposed to severe environment in arctic sea water should include the major degrading forces: ice abrasion, internal frost damage, de-icer salt scaling and reinforcement corrosion. The objective of this paper is to review different forms of deterioration as they work individually and together and then form a model for service life calculation. It is necessary to start as simply as possible in order to limit the amount of material properties and exposure parameters to be used as input for deterministic calculations and model verification before the statistical variability of environment and concrete material parameters are introduced (Sistonen et al., 2006). This does not rule out increasing the complexity of the models as we go along. However, it is ideal to start with a model where it is easy to obtain reliable data that can be used as input for verification of both uniform exposure (i.e. single acting deterioration type) and interaction with several forms of deterioration simultaneously. The outcome of the calculations should be the probability of obtaining a given service life compared to a defined limit state of a concrete structure exposed to the mentioned deteriorating forces (Sistonen et al., 2006).


5 (47)

2 Uniform deterioration

2.1 Abrasion of concrete by ice

The abrasion of concrete by ice has been studied analytically, numerically, experimentally and in the field (Huovinen, 1990a, 1990b, 1993). The surface degradation due to ice sheets crushing against the concrete surface can be divided in three steps (see Figure 1): abrasion of cement paste, abrasion of cement paste and loosening of protruding aggregate stones and abrasion of cement paste when the bond between larger aggregate particles and the cement paste is so weak that the stones loosen upon the first ice impact.

Figure 1. Various steps involved in the ice abrasion of concrete.

The abrasion rate of the cement paste depends on its strength. The relationship between these two factors and the movement of the ice can be described by the following equation (Huovinen, 1990a):

sf

ABRc

a ⋅=3 , (1)

where ABRa is abrasion depth [mm], fc is the compressive strength of the concrete [MPa] and s is the movement of the ice sheet [km].

When the surface degradation has reached the point according to which aggregate particles protrude, the depth of abrasion can be calculated using the following equation:

( ) baRnnaABR ii

sn

iib ⋅−+⋅⋅= ∑∑

=

1loglog

11, [mm], (2)

where ABRb is abrasion depth [mm], ai is the proportional volume of aggregate particles with radius Ri

[-], ∑ai is the total proportional volume of aggregate particles in

concrete [-], ns is the number of ice impacts during ice sheet movements [-],


6 (47)

n1 is the number of ice impacts when Lcr/R = 1 [-], b is the abrasion rate of cement paste [mm] and Lcr is the crack length [mm].

The abrasion depth when the bond strength between larger aggregate particles and the cement paste is so weak that the stones loosen during the first ice impact is given by:

( ) sfa

ABRci

c3

11

⋅−

=∑

, [mm], (3)

where ABRc is abrasion depth [mm], s is the movement of the ice sheet [km], ∑ai is the total proportional volume of aggregate particles in

concrete [-] and fc is the compressive strength of the concrete [MPa].

Typically the ice abrasion of concrete with compressive strength in the range 50 - 100 MPa is in the range 0.0075 - 0.015 mm/km ice passage. The final limit state for service life can be assumed to be an ice-abrasion depth equal to the concrete cover. The above equations, finite element calculations and experimental results have led to the following simplified equation (mean value) for the service life (Huovinen, 1990a):

( )360+⋅

=s

fct cLµ , (4)

where µ(tL) is service life [a], s is the movement of the ice sheet [km], fc is the compressive strength of the concrete [MPa] and c is the cover depth of the concrete [mm].

Fiorio investigated the small-scale effects of the friction-induced wear of concrete (Fiorio, 2005; Fiorio et al., 2002). The results of experimental investigations performed at the ice-concrete contact confirm the three-step mechanism previously described. Measurements of the wear rate of concrete showed that a stable wear rate is typically obtained after erosion of the superficial layer of cement paste. It depends on the sliding speed and the pressure in contact but is independent of the average roughness of the plate.

A review of abrasion of concrete surfaces due to studded tire action as compared to ice abrasion showed that the effect of strength is similar for the two types of abrasion (Jacobsen et al., 2006). Concrete with compressive strength in the range 50 - 100 MPa has studded tire abrasion 0.5 - 3 mm/104 passages. Furthermore, it seems that salt has a negative effect in both cases. This suggests that the deteriorating effect(s) of salt in scaling not only depend on frost. However, there are also dissimilarities between the two types of abrasion, such as the lack of internal damage in studded tire abrasion on high performance concrete (Tveter, 1993). Internal damage in the form of loss of bond between paste and aggregate is an important basis for the modelling of ice abrasion (Huovinen, 1990a, 1990b, 1993; Fiorio, 2005; Fiorio, et al., 2002; Hoff, 1988; Preus, 1962; Keyser, 1970).


7 (47)

The high local loads due to the plastic crushing effect of ice (Huovinen, 1990a; Sodhi, 2006) and possible simultaneous internal frost damage during ice abrasion are probable causes of this difference. Internal frost damage is well known to result in cracks in bond zones between aggregate and cement paste (Pigeon, 1984; Kukko, 1992; Jacobsen et al., 1995a; Nieman, et al., 1996). Studded tire abrasion and ice abrasion studies held together show that concrete strength and good bond with abrasion resistant aggregate are important factors for the concrete to have high abrasion resistance.

2.2 Frost action

Concrete is a quasi-brittle porous material (Bazant, et al., 1998). Under usual exposure conditions, its pore structure is partially (or even fully) saturated with an aqueous solution. If the ambient temperature is reduced below the freezing point of the aqueous solution, ice will progressively form in the saturated pores. The ice formation process is directly affected by the pore size distribution of the material (Helmuth, 1960; Zuber and Marchand, 2000). This partial solidification of the pore solution will often force the remaining unfrozen water to move. Depending on the temperature history, unfrozen water can either be expelled from the freezing sites or attracted to the ice crystals formed within the material’s pore structure (Powers, 1949; Powers, et al., 1953).

These water movements locally generate internal pressures, which may, eventually, exceed the tensile strength of concrete (Helmuth, 1961). Investigations on the subject show how degradation is linked to the degree of saturation of the material (Bager, et al., 1986; Fagerlund, 1971; Marchand, et al., 1995). As previously emphasized, the macroscopic behaviour of concrete upon freezing is controlled by a series of complex phenomena involving phase changes, mass and thermal transfers, non-linear mechanics; activated at several scales that range from a few nanometres to a few centimetres.

Frost damage typically occurs in the form of cracking and/or surface scaling. The latter typically occurs in concrete elements exposed to de-icing salts. In massive structures, it can also lead to spalling of large areas of concrete (Léger, et al., 1995). Elements suffering from frost degradation often require significant repair after only a few years of service. In addition to being costly, repair strategies are often not adapted to the problem and do not contribute to extending the service-life of the structure.

Frost damage can be characterized qualitatively by visual rating, quantified directly as cracking (amount or density of cracks, length, width, concrete length change, loss of mechanical properties like strength, and resilience, etc.) or mass of scaled-off material or depth of scaling. Microscopy is suited to characterize internal frost damage, i.e. degradation resulting from the exposure to subzero temperatures in the absence of de-icing salts.

Various microscopy techniques have been employed to characterize the manifestation of internal frost damage in laboratory and field exposed concrete (Pigeon, 1984; Kukko, 1992; Jacobsen, et al., 1995a, 1995b; Niemann, et al., 1996). The various forms of nanoscale and microscale effects such as cracking and chemical alterations following frost exposure and associated moisture


8 (47)

movements, as well as macroscopic alterations in form of strain, changes in degree of saturation, loss of interfacial bond zones, strength, etc., are reviewed and summarized in (Jacobsen, et al., 2006).

The basic facts of frost durability remain that air-voids work as protection, that the damage is linked to the material degree of saturation and access to water during freeze/thaw due to the freeze/thaw pumping effect and that concrete with sufficiently low water/binder ratio and high strength can be made frost durable even without air entrainment. Following the early work by Powers and Helmuth and the service life model of Fagerlund, varying numerical models to describe the phenomenology have been further developed (Zuber and Marchand, 2000; Hörsch and Wittmann, 2002; Kruschwitz, 2005; Scherer, 1999; Setzer, 2002; Zuber and Marchand, 2004). However, none of these can yet readily be applied to calculate a rate of degradation or deterministic service life for a concrete structure exposed to frost.

2.3 De-icer salt scaling

Scaling is a progressive deterioration occurring when concrete is subjected to freezing and thawing in the presence of a salt solution. This degradation usually manifests itself by the removal of small pieces of mortar from the concrete surface. Scaling can lead to the exposure of coarse aggregates. Numerous mechanisms have been proposed to explain the degradation of concrete by salt scaling but none account for the entire body of results published on the subject although a few general facts seem to remain (Marchand, et al., 1994; Pigeon, et al., 1996).

Scaling resistance is typically assessed by standard tests which attempt to reproduce the most severe conditions found in service. The procedures usually consist in subjecting a concrete sample covered with a salt solution on the exposed horizontal surface to daily freezing and thawing cycles. Scaling is evaluated by rating visually the surface or by regularly weighing the scaled-off particles (kg/m2). This procedure is strongly sensitive to various parameters. The very few systematic comparisons between laboratory and field scaling indicate a relatively poor correlation between the two (Hooton et al., 1997; Marchand, et al., 2005; Bouzoubaâ, 2005). In most cases, laboratory methods are found to overestimate scaling.

It appears clearly that both the quality of the air void system and concrete strength has an influence on scaling (Marchand, et al., 1994; Pigeon, et al., 1996a). Investigations on the topic indicate that if proper air entrainment does not offer full protection against scaling it contributes to strongly reduce the intensity of degradation. Typically, a spacing factor of 230 µm or less is recommended to improve the scaling resistance of concrete. High water/binder ratio concretes have coarser and higher porosity, lower strength and are usually more susceptible to scaling.

The low water/binder ratios of high performance concretes tend to reduce the volume of freezable water and enhance the de-icer salt scaling resistance of the material. The behaviour of concrete also tends to be affected by parameters such as the type of binder and the nature of mineral and chemical additives and


9 (47)

admixtures. For instance, degradation usually increases with the content of fly ash or slag (Marchand, et al., 1994; Pigeon, et al., 1996a). However, the role of these parameters is not clear. Their influence could well be on the capillary pore system (effect on freezable water, ice formation in pores and water movements) or on the microstructure of the surface layer (resistance as a function of type and quantity of hydrates and of air voids). The composition of the concrete mixture can also have a combined effect by modifying the sensitivity of concrete to other parameters.

Scaling is a surface problem sensitive to everything that can modify the superficial layer. Numerous studies have shown the existence of a weak and porous layer on the surface of concrete that would scale more easily (Marchand, et al., 1994; Marchand, et al., 2005; Pigeon, et al., 1996b). The formation of this layer would be due to the bleeding, strongly influenced by casting conditions (mould or granular base, wind, relative humidity, temperature, vibration, etc.) The workability of the concrete combined with the finishing and curing operations can also have an important influence on both the mechanical properties and the microstructure of the surface (Marchand, et al., 1994; Hooton, et al., 1997). However, the surface is so sensitive to these parameters that it is hard to isolate one factor from all these others. These parameters could be partly responsible for the variability and the discrepancy observed between field and laboratory specimens. The so-called Scandinavian slab test avoids surface effects on scaling test results by sawing the test surfaces followed by a standardized drying/rewetting procedure of the test surface prior to salt/frost exposure. This method then gives a pure “bulk material effect” on scaling without special effects in the superficial layer as mentioned above.

Surface exposure has a significant influence on scaling. Studies have shown that drying and wetting cycles, as well as carbonation of the surface, can modify the performance of concrete (Marchand, et al., 1994; Copuroglu, et al., 2004). The characteristics of the freezing and thawing cycles have also a direct bearing on the resistance of concrete to scaling. Studies have shown that no significant scaling occurs if the temperature is maintained above -10 °C (Lindmark, 1998; Marchand, et al., 1996; Sellevold and Farstad, 1991). Once this critical value is reached, damage tends to increase as temperature is reduced. The duration of the freezing period is reported to be more critical than the cooling rate (Marchand, 1993; Stark, 1989). Damage is also sensitive to the concentration of the salt solution in contact with the concrete surface. Deterioration gradually increases as the salt concentration is increased up to a pessimum concentration of roughly 3% by mass, regardless of the type of de-icer used (Marchand, et al., 1994; Pigeon et al., 1996a).

Several mechanisms have been proposed to explain salt scaling degradation. Most of them rely on pressures developed related to the ice formation in the pore system (Marchand, et al., 1994; Pigeon, et al., 1996a). According to these theories, damage does result from pressures exerted by the ice on pore walls or by hydraulic and osmotic pressures developed by the expulsion of water during the growth of ice in the pore system. Others suggest that pressures could be due to the crystallization of salts or expansive products in the pore system. Valenza and Scherer proposed a new mechanism that accounts for several experimental observations (Valenza and Scherer, 2005). The “glue-spall” mechanism explains the “pessimum” salt concentration of around 3% and why the scaled material consists of thin flakes. According to this new approach, scaling would be due to


10 (47)

fractures in the ice layer that would penetrate the concrete surface and bifurcate in such a way as to remove small flakes of mortar. During cooling, salt in the outer solution would create brine pockets in the ice that would act as flaws. The “pessimum” concentration of salt in the outer solution is explained by the fact that pure water ice is too strong to crack and at higher concentrations, the ice/brine mixture is not rigid enough to develop stresses.

The proposed mechanism correlates well with many observations made on the behaviour of samples tested under laboratory conditions. More work is however needed to explain the few cases where scaling occurs in the presence of pure water, the influence of the air void system, the influence of supplementary cementing materials, why scaling is more severe at slow cooling rates (Jacobsen, et al., 1997), why salt has no or much less negative effect on non-cementitious porous building materials such as ceramics (brick/tile/LWA), various natural rocks and asphalt (Jacobsen, 1999), and the observed negative effect on scaling of simultaneous internal cracking and surface scaling (Jacobsen and Sellevold, 1996; Jacobsen, et al., 1997).

A simple formula was developed for predicting service life based on frost-salt scaling (Vesikari, 1991). It separates the material factors from environmental factors. All material factors are incorporated in P-value:

( ) Pkt eL ⋅=µ , (5)

where µ(tL) is the service life [a], ke is the circumstantial factor [-] and P is P-value [-].

According to the Finnish concrete code service life can be determined using the following values of circumstantial factor (BY50, 2004):

25.1:42:2

==

e

e

kXFkXF

P-value is determined from the following equation (Matala, 1991; BY50, 2004):

( ) 11046

2.1

−⋅

⋅⋅=

aWAB

ccP bcur , (6)

( )curcur tLOGc 1017.085.0 ⋅+= , (7)

( )FABSSIQ

Qcb

ohb ⋅+⋅+⋅⋅

−= 01.002.005.01

5.1

2 , (8)

fabssicemb QQQQQ ⋅+⋅+⋅+= 4,08,00,2 , (9)

( )b

oh

QaQ

WAB2102 −⋅+

= , (10)

where P is P-value [-],


11 (47)

ccur is the curing factor [-], cb is the binding factor [-], WAB is the reduced water-air-binder ratio [-], a is the air content [%], tcur is the curing time [d], Qh2o is the effective water content [kg/m3], Qb is the total amount of effective binding material [kg/m3], SI is the silica fume ratio [%], BS is the blast-furnace slag ratio [%], FA is the fly ash ratio [%], Qcem is the cement content [kg/m3], Qsi is the silica fume content [kg/m3], Qbs is the blast-furnace slag content [kg/m3] and Qfa is the fly ash content [kg/m3].

The model is mainly based on empirical results and rests on the above mentioned concrete technological parameters known to affect frost durability. It is also used by Finnish Road Authorities to predict service life (Finnish Road Administration, 2008).

2.4 Chloride induced reinforcement corrosion

Chloride induced reinforcement corrosion is perhaps even more complex than frost action. The damage process and service life is normally modelled as an initiation and a propagation period (Tuutti, 1982) according to the two main processes. First there is a transport period until critical chloride content or sufficiently carbonated material is reached at the steel surface. The subsequent probability of deterioration is determined mainly by corrosion parameters which in turn rely on transport parameters (moisture, oxygen, chlorides, etc.) and fracture parameters. Equation (11) yields for chloride induced reinforcement corrosion of hot dipped galvanised reinforcing steel in concrete (Sistonen, et al., 2006). For this particular case the Tuutti’s general service life model is extended into three phases; initiation, degradation of zinc coating and finally corrosion of the steel bar:

+

+

=

Ørc

rd

kct

sClL

80)(1

2

µ , (11)

where µ(tL) is the mean service life for uniform chloride induced corrosion of hot-dip galvanised reinforcement bar in un-cracked concrete [a],

c is the thickness of concrete cover [mm], kCl is the coefficient of critical chloride content [ amm / ], d is the zinc coating thickness [µm], r1 is the corrosion rate in un-cracked chloride contaminated

concrete (hot-dip galvanised reinforcement) [µm/a], rs is the corrosion rate in un-cracked chloride contaminated

concrete (ordinary steel reinforcement) [µm/a] and Ø is the reinforcement bar diameter [mm].


12 (47)

3 Interaction

Interaction means that two or more types of damage acting simultaneously affect each others rates of deterioration. For example, one can imagine that internal frost damage or de-icer salt scaling speeds up reinforcement corrosion. It is also possible to imagine that significant internal frost damage before ice sheet movement starts increases subsequent ice abrasion compared to a concrete structure with little or no internal frost damage. Some interaction cases were discussed qualitatively to deduce some of the interaction mechanisms and thereby be able to improve estimates of how service life is affected by interaction (Jacobsen, et al., 2006). There is however a need for quantification in order to make service life calculations. For this purpose more general models for interaction are needed and therefore it is necessary look for ways of coupling the degradation mechanisms as in (Gerard, 1996; Johannesson, 1996).

There are several interactions between different types of damage. If most interactions have a negative effect on service life, some may have a neutral or positive influence on the durability of concrete. Furthermore, the question of true damage mechanisms could arise since in practice several types of degradation can be active on a concrete structure. One possible parallel is the testing of “pure compressive strength” which in practice is a combination of compression, tension and shear. Another possible parallel is the flow of heat in partly saturated concrete, which is affected by the moisture content of the material. The moisture content, however, is in turn affected by temperature gradients so that true heat conductivity could be difficult to measure and difficult to predict on the long term. There is interaction or coupling between flow of heat and moisture. Also coupling between other mass transfer properties, chemical degradation, mechanical load and temperature induced strain of concrete have been studied (Gerard, 1996; Johannesson, 1996). Furthermore, one should distinguish between “cause” and “effect” for a type of damage since the word mechanism, within concrete research, often is covering both meanings.

The degree of degradation, D should be expressed (with respect to all degradation modes) with a scale [0....1] so that 0 represents no degradation and 1 degree of degradation at the limit state (corresponding to service life). Furthermore, the combined degree of degradation would be given with the same scale, 0 - 1, and the service life with combined degradation mechanisms would be at its end when the combined degree of degradation attains 1.

3.1 Mechanisms of interaction

In the following sections different cases of interaction are briefly discussed. Already at this level the damage mechanisms seem fairly complex. Therefore, understanding all possible combinations of the four deteriorating mechanisms (ice abrasion, internal frost, de-icer salt scaling and chloride induced reinforcement corrosion) is difficult.


13 (47)

3.1.1 Internal cracks due to frost and surface scaling due to frost and de-icer salt

Internal cracks due to frost and surface mass loss due to de-icer salt scaling are both often tested in the laboratory on specimens with a continuously wet surface. It has been observed that the amount of scaled material increases as internal cracking progresses (Jacobsen and Sellevold, 1996; Jacobsen, et al., 1997). It appears that the increase in surface damage is caused by interaction with internal cracking (Powers, 1945). Considering volumetric frost damage near the surface, rather dense micro-cracks (< 0.1 mm) parallel to the surface, as those observed on concrete bridge decks (Nieman, 1996), can probably lead to scaling. In addition, it has been established (Pigeon, 1984; Fagerlund, 1972; Jacobsen and Sellevold, 1996; Jacobsen, et al., 1997) that frost-induced internal cracking can occur without scaling, and that scaling can occur without internal cracking. The observations indicate that it is difficult to distinguish between surface damage in the form of mass loss and volumetric damage in the form of cracks unless both types of damage are monitored simultaneously during exposure. It is possible to then think of frost damage in itself as interaction between surface scaling and volumetric damage.

3.1.2 Ice abrasion, frost damage and de-icer salt scaling

The investigation of the condition of Swedish lighthouses (Janson, 1988) suggested no interaction between abrasion and frost damage since the major part of abrasion damage was found to take place below sea level where there is no freezing and thawing. However, ice abrasion has also been observed above sea level (Huovinen, 1990a). In the latter investigation it was concluded that three different external stresses were imposed on the surface; compression normal to the surface, friction parallel to the surface and tension caused by bond between concrete and ice frozen to the surface. Accordingly, any reduction of mechanical properties of concrete is likely to reduce the abrasion resistance of concrete (Huovinen, 1990b).

As for surface scaling with salt solution (Verbeck and Klieger, 1956; Marchand, et al., 1994) salt water ice tends to increase the degradation by ice abrasion compared to fresh water ice (Huovinen, 1990a). This phenomenon was partly explained by the more plastic nature of salt water ice causing higher loading than tap water ice. The new model for scaling (Valenza and Scherer, 2005) discussed in chapter 2.3, with external ice causing scaling due to the differential thermal expansion between concrete and ice and the fracture as the ice contains some salt solution has been further developed using Finite Element Modelling (Copuroglu, et al., 2006). The latter could increase the understanding of the ice-concrete interaction since the bond between frozen concrete and ice seems very low (roughly 0.3 MPa (Huovinen, 1990a) compared to the concrete tensile strength that must be exceeded to cause scaling according to the model (Valenza and Scherer, 2005).

3.1.3 Studded tire abrasion and frost-induced damage

It seems that the interaction between studded tire abrasion and volumetric damage (cracks) depends on the mechanical strength of concrete. In high strength


14 (47)

concrete, neither cracks extending from the surface to the interior of the concrete nor loss of aggregate bond were found to a significant extent (Tveter, 1993). Consequently, it can be assumed that the abrasion of high performance concrete is mainly a surface phenomenon. In ordinary strength concrete, cracking can be induced by the impact of the studded tires (Hoff, 1988). In this case, interaction between surface damage and internal cracking will probably be similar to that seen between frost degradation and ice abrasion (Huovinen, 1990a). However, the increased abrasion on wet surfaces compared to dry surfaces (Tveter, 1993; Bærland, 1989; Gjørv, 1987) suggests that an additional form of interaction may take place. This could be related to the reduced strength of wet concrete. However, since the same amplification of scaling by a wet surface is seen in frost exposure there could be other effects as well. In both type of damage, the interaction with water would be related to cracks. Thus, by using high performance concrete, one might perhaps reduce several forms of negative interaction. Finally it should be mentioned that there is interaction between studded tire abrasion and presence of salt increasing the abrasion depth compared to uniform abrasion (Preus, 1972).

3.1.4 Alkali aggregate reactions and frost

Interaction between alkali aggregate reactions (AAR) and frost deterioration has been studied experimentally for example in (Trägård, 1996a; Wei, 2005). In this case it almost “goes without saying” that cracks initiated by AAR can be accelerated by frost as was also demonstrated (Trägård, 1996a; Wei, 2005). However, it could also be the other way; that AAR and/or access to moisture is facilitated by internal damage due to frost since existing cracks can trigger alkali aggregate reaction (Bérubé, 1996; Trägård, 1996b; Marchand, 1998; Wei, 2005). Recent studies of concrete cores from dams with AAR showed that the extent of AAR damage quantified as a cracking index increased with the degree of capillary saturation measured on concrete specimens with in-situ moisture condition (Lindgård, 2006). The cracking index was measured by optical microscopy on fluorescent impregnated polished sections. The observed monotonic increase of AAR with increased saturation is in contrast to the “classical view” on uniform AAR from laboratory studies; that maximum damage in terms of expansion in accelerated laboratory tests occurs at a somewhat lower than capillary saturated condition. Most field conditions of larger structures, on the other hand, probably involve interaction with other types of damage (frost, delayed ettringite formation, leaching of calcium etc.) depending on the actual exposure condition and concrete quality. All dams studied in (Lindgård, 2006) had been exposed to frost. The exact mechanism by which AAR and frost interacts cannot be deduced directly from (Lindgård, 2006). If AAR initiated the damage there could clearly be an interaction with frost if degree of saturation is increased. This interaction could either be due to water transport in cracks and/or by the hygroscopic gel adsorbing more water. This could explain the observed monotonic increase of damage with increasing saturation. However, also in standard expansion tests in the laboratory of (supposedly) uniform alkali reactivity of various rocks, interaction with water is suggested. A linear correlation between the absorption of the reactive aggregate and mortar bar expansion was observed in (Wigum, 1995). However, it is not necessarily an interaction but merely an indication of increased reactivity as the porosity of the reactive rock increases. This could also be the cause for the


15 (47)

increased damage by AAR with increasing saturation in (Lindgård, 2006) and as such not necessarily taken as evidence for an interaction between frost and AAR.

3.1.5 Carbonation, chloride penetration and reinforcement corrosion

An example of interaction between three types of damage is the carbonation of concrete, chloride transport and corrosion of reinforcing steel (Broomfield, 1997). Carbonation and chloride transport are in this case mainly controlling the initiation period and the chloride threshold value for corrosion whereas the damage itself is caused by reinforcement corrosion. However, for calculation of service life they have all been classified as types of damage. The topic of chloride binding is complex as described in a review (Justnes, 1998) and will depend on both type of chloride salt, e.g. NaCl or CaCl2, whether it is mixed into the fresh concrete or penetrating the concrete structure, type of binder etc. In brief the reaction products tricalcium aluminate (C3A) reacts with chlorides. The resulting chloro-aluminates, or whatever compound that is produced between chlorides and binder (Justnes, 1998), might not be capable of inflicting corrosion of reinforcing steel. Carbonation might decrease the pH of the concrete and release chlorides from chloro-aluminate and also decrease the amount of OH- - ions of the concrete pore solution. The increasing ratio of free chloride ions to OH- - ions can exceed the tolerance limit of chloride based corrosion of steel. The interaction of chlorides and carbonation can in this way accelerate the corrosion substantially. However, carbonation also strengthens and compacts OPC concrete improving transport properties such as sorption (Østmoen, 1993) whereas the opposite might be the case with slag concrete so that at least concrete mix design should be known in addition to the chloride profile and depth of carbonation to assess interaction.

3.2 The combined degree of degradation with interaction by superposition

In (Jacobsen et al., 2006) a new method is proposed to calculate the combined degree of degradation with interaction between different damage mechanisms using superposition of individual degrading phenomena or by amplifying or reducing the individual degrading mechanisms by using interaction coefficients kint ≠ 1 as presented in Equation (12):

∑=

=n

iiDkD

1int , (12)

where D is the combined degree of degradation which includes an interaction effect [-],

Di is the degree of degradation of individual type of degradation [-] and

kint is the interaction coefficient [-].


16 (47)

3.3 The combined degree of degradation by coupling with phenomenological coefficients

Calculation of the combined degree of degradation when several degrading forces are active should be done in such a way that it can be verified or calibrated with data from field and/or laboratory exposure. A possible approach is to use the phenomenological coupling principles of irreversible thermodynamic processes (Onsager, 1931, 1932). In this case it means to couple together individual degradation models with phenomenological coupling coefficients (Lij) (Jacobsen, et al., 2007). In the following is suggested how to couple the degradation models for ice abrasion, internal frost damage, de-icer salt scaling and chloride induced reinforcement corrosion. For these four types of degradation a system of coupled model results:

=−iceabreractDint Clcorrsaltfrostfrosticeabr DLDLDLDL ⋅+⋅+⋅+⋅ 14131211 (13)

=− frosteractDint Clcorrsaltfrostfrosticeabr DLDLDLDL ⋅+⋅+⋅+⋅ 24232221

=−saltfrosteractDint Clcorrsaltfrostfrosticeabr DLDLDLDL ⋅+⋅+⋅+⋅ 34333231

=−ClcorreractDint Clcorrsaltfrostfrosticeabr DLDLDLDL ⋅+⋅+⋅+⋅ 44434241

In Equation (13) a possible coupling between the four types of degradation is quantified by a set of phenomenological L-coefficients for a specific concrete material quality. To obtain this, the degree of degradation must be measured (laboratory, field) following simultaneous action of the degrading forces. There is no coupling if all Lij = 1 for i = j (i, j ≠ 0) and Lij = 0 for i ≠ j, (i, j ≠ 0). Then the individual types of degradation remain. If the deterioration processes interact with each other it may be assumed that they are coupled according to the reciprocal relations (Onsager, 1931, 1932):

L14 = L41, L13 = L31, L12 = L21, L24 = L42, L23 = L32 and L34 = L43. (14)

There are 15 possible combinations of exposure in the degree of degradation matrix Equation (13): four uniform, six combinations of two simultaneous, four combinations of three simultaneous and one with four simultaneous. The number of phenomenological coefficients is thus reduced to ten when Equation (13) is coupled by Equation (14). To solve Lij it is necessary to obtained values for the degree of degradation from laboratory and/or field data, and also the uniform types of degradation. It also is also possible to collected data for the four types of degradation for any of the 15 possible combinations of exposure that can result in degradation measured either as damaged by abrasion, frost, scaling or corrosion. Some of these combinations will give zero degradation (no ice abrasion in a case with only chloride transport and corrosion at temperatures above 0 °C). Hopefully, enough types of degradation can be obtained from single material qualities with multiple attacks. For field data it is necessary to look at different locations on the structure with uniform and multiple degradation (ice abrasion, frost, de-icer salt scaling and reinforcement corrosion).


17 (47)

3.4 Data to quantify the coupling between the deterministic service life equations

3.4.1 Interaction damage table

When designing experiments to quantify interaction between two degrading phenomena, it is suggested to use an interaction table as shown in Table 1 (Jacobsen, et al., 2007).

Table 1. Interaction of two uniform types of damage.

Rei

nfor

cem

ent c

orro

sion

Car

bona

tion

of c

oncr

ete

Chl

orid

e di

ffus

ion

Exte

rnal

fros

t dam

age

in c

oncr

ete

- su

rfac

e sc

alin

g

Inte

rnal

fros

t dam

age

in c

oncr

ete

Abr

asio

n of

con

cret

e by

ice

Abr

asio

n of

con

cret

e by

stud

ded

tires

Ettr

ingi

te r

eact

ion

Alk

ali a

ggre

gate

rea

ctio

ns

Def

orm

atio

n an

d cr

acki

ng o

f co

ncre

te

Che

mic

al lo

ads

Leac

hing

of C

oncr

ete

Mic

ro o

rgan

isms

Reinforcement corrosion I I I I I I I I I I I I

Carbonation of concrete I I I I I I I I I I I D/I

Chloride diffusion I D/I I I I I I I I I I I

External frost damage in concrete - surface scaling I D/I D/I I I I I I I I I I

Internal frost damage in concrete I D/I I I I I I I I I I I

Abrasion of concrete by ice I D/I I I I - I I I I I -

Abrasion of concrete by studded tires I D/I I I I - I I I I I -

Ettringite reaction I D/I I I I I I I I I I I

Alkali aggregate reactions I/- D/I/- I/- I/- I/- I/- I/- I/- I/- I/- I/- I/-

Deformation and cracking of concrete I D/I I I I I I I I I I I

Chemical loads I/- D/I/- I/- I/- I/- I/- I/- I/- I/- I/- I/- I/-

Leaching of concrete I/- I/- I/- I/- I/- I/- I/- I/- I/- I/- I I/-

Micro organisms I/- I D D/I/- I/- D D - - I I I

Interacted influencing damage typeConcrete structures in generalI = Deterioration increases;D = Deterioration decreases;- = Indicates no interaction

Uniform influenced damage type

Table 1 gives an overview of possible deteriorating phenomena and what is known or expected will be the outcome of interaction between two simultaneously occurring types of damage: increases (I) or decreases (D) deterioration or indicates no interaction (-) compared to the uniform damage type. Logic of the table is to compare simultaneous damage mechanism with the uniform damage mechanism. The following examples guide to understand how to interpret the table. Uniform damage type at the third row in the table is chloride diffusion. It can be noticed that only carbonation might have a decreasing effect on chloride diffusion (second column, third row). Reason for that is that by carbonation the OPC concrete may compact and strengthen. However, carbonation might have an increasing effect on chloride diffusion by releasing bound chlorides deeper into the concrete structure. In addition to carbonation, leaching and sulfate corrosion releases the binding chlorides (Puatatsananon, 2002; Fagerlund, 2003; Gaal, 2004) due to the reduced alkalinity of concrete, and in the case of sulfate corrosion due to the preferential formation of calcium sulphoaluminate (Martin-Pérez, 1999)


18 (47)

(eleventh and twelfth column, third row). Interacted damage type leaching of concrete may lead to increased porosity of concrete and therefore increasing intake of water. That in turn may lead up to alkali aggregate reactions in concrete with reactive aggregate and binder with high alkali content (twelfth column, ninth row). However, leaching of concrete by itself does not increase the alkali reactivity of aggregate or the alkali content of the binder to levels which are the prerequisites for alkali aggregate reactions together with high moisture level of concrete. In that case there is no indication of interaction (twelfth column, ninth row). Thus, probability of damage increases when there are damage types affecting at the same time. The physical mechanisms causing the interaction have been discussed more thoroughly in (Jacobsen, et al., 2006).

3.4.2 Laboratory data

In accelerated laboratory tests the rate of deterioration should be easily measurable and with clear limit states in order to obtain the combined degree of degradation for D. In the case of de-icer salt scaling and abrasion, the maximum rate of material loss from the surface [mm/year] can be used, in the case of internal frost attack, maximum strain or permanent length increase [mm/mm] can be used, whereas initiation time or maximum rate of corrosion could be used in the last case.

3.4.3 Field data

It is a great advantage with the use of the degree of degradation Di that field inspection of real concrete structures can be used to collect data. By careful inspection of structures with known age and concrete quality, valuable data can be obtained and the calculation method can be verified by combining the field inspections with sampling to determine the material parameters.

In order to determine the interaction coefficients both uniform (individual) and simultaneous deterioration must be measured. Furthermore, common sense should be applied when selecting interacting mechanisms to be tested and evaluated. If it seems probable that one mechanism will be dominant in terms of rate of deterioration, such as ice abrasion due to very large ice sheet movement compared to slow corrosion rate under the same conditions, then a fair service life estimate may perhaps be made with the dominant mechanisms only.

After determining coupling coefficients Lij in Equation (14) service life is calculated using the stochastic process (Sistonen, et al., 2006) to determine the probability that a certain service life is obtained.

3.4.4 Varying values of individual deterioration parameters

One method is to change values of the deterioration parameters in the individual service life calculation formula so that interaction is taken into account. An example is presented in Figure 2. In the figure interaction of the carbonation of concrete and chlorides to the decreasing corrosion rate of reinforcing steel is presented with arrows. The formulae used for service life (Sistonen, 2001) are also shown. It can be assumed, based on the interaction damage table, that


19 (47)

carbonation of concrete releases bound chlorides deeper into the concrete structure (marking I in Table 1; third row and second column). This increase is assumed to be larger than the reduction in external supply due to reduced chloride diffusion of OPC concrete. Thus, service life due to reinforcement corrosion is reduced. (Note that it is also deduced the opposite effect of carbonation on chloride diffusion). Based on these two interactions, it can be estimated that corrosion rate, ks shall increase for instance from a mean value 20 am /µ to a higher value 30 /m aµ . Also, it can be estimated that coefficient of critical chloride content, kcl might increase from mean value 3 /mm a to higher value 4

amm / where a is time in years. Those interactions mean in this example that service life would decrease approximately 47 per cent (From 57 years to 30 years calculated with stochastic process method with five percent of probability of damage) (Sistonen, 2001).

Service life based on steel corrosion: Chloride diffusion of uncracked concreteLog-normal distribution function (probability of damage 5%)

1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

0 10 20 30 40 50 60 70 80 90 100Target service life tLtarg [ a ]

Coe

ffici

ent o

f crit

ical

chl

orid

e co

nten

t kcl

[mm

/a0.

5 ]

ks = 20 µm/(a)½ks = 30 µm/(a)½ks = 40 µm/(a)½ks = 50 µm/(a)½ks = 60 µm/(a)½ks = 70 µm/(a)½ks = 80 µm/(a)½ks = 90 µm/(a)½ks = 100 µm/(a)½ks = 110 µm/(a)½

[ ]ee

)Y(

)Y(

argLtt σ⋅β

µ

=

( )2

s

2

clL øk

c80kct

⋅⋅

+

=µ

c = 30 mm (ν = 0,25)φ = 12 mm (ν = 0,15)ks = variable (ν = 0,4)kcl = variable (ν = 0,15)

Figure 2. Effect of interaction between chloride diffusion and carbonation on service life by changing values of deterioration parameters in individual service life calculation formula.

3.4.5 Other methods

One way of quantifying how interaction affects rate of deterioration and thus service life for various combinations of deterioration types as shown in Table 1 is to use computer aided coupled nonlinear finite-difference program (Puatatsananon, 2005). In that study the latter interaction case (chapter 3.1.5) is studied. External reinforced concrete elements exposed to chloride and/or CO2 will eventually have a lower pH, which in turn will depassivate the reinforcement and initiate corrosion, thus causing spalling. The study seeks to address the complex multi-physics nature of concrete environmental damage using coupled nonlinear partial differential equations. Heat, relative pore humidity, chloride, and carbonation are all implemented in a two-dimensional coupled nonlinear finite-difference code. Coupling between carbonation and chloride diffusion is explored in the context of both homogeneous and heterogeneous concrete models and simulated. This could be a possible further way since it also provides a systematic way of testing sensitivity to different parameters.


20 (47)

One method is, in brief, to use laboratory tests and computer software programs to analyze how interaction affects material properties. Changes in material parameters can be used as new values in computer calculations, for example changing every five or ten years. An example is presented in Figure 3 (Al-Neshawy, et al., 2005). In the laboratory test, interaction between two variables, (UV-radiation and freeze-thaw (UP); UV-radiation and carbonation (UK); carbonation and freeze-thaw (KP)) are compared with interaction between three variables (UV-radiation, carbonation and freeze-thaw (UKP)). Material properties that are reasonable to study in interaction laboratory tests are: strength properties (compressive and tensile strength, modulus of elasticity, adhesion/bond), permeability properties (carbonation depth, diffusivity, gas permeability measurements, capillarity (water uptake, drying), porosity and pore size distribution etc.), durability properties (frost resistance, air pore distribution, degree of degradation, thin sections).

Figure 3. Laboratory test to analyze interaction of deterioration damage processes (left) and interaction combinations (right).

Water absorption is presented as an example of an interaction laboratory test in Figure 4. The absorption of the ten rock based materials is clearly affected very differently by the three isolated exposure conditions compared to the absorption following interaction between all three exposure forms (UV, freeze/thaw and carbonation).

Figure 4. Change in four hour absorption of 10 rock based materials after uniform exposure, in % of absorption following interaction between the 3 exposure forms.


21 (47)

Note that experimental results in the approach can probably also be used for input in the previous approaches. That is, to determine interaction coefficients and/or to what extent superposition can be used. Once again it is important to emphasize that such experiments should elucidate how the interaction works at least by analyzing the data, reviewing and discussing the mechanisms of the uniform damage types and comparing them in terms of interaction (Hengjing Ba, 2004).

The interaction effect can also be taken into account in the factorial method (ISO 15686-1, 2000) by adding an interaction coefficient as variable in the calculation formula. Service life planning based on the factorial method (Sarja, 1996; ISO 15686-1, 2000) is included in the Finnish concrete code (By 50, 2004).


22 (47)

4 Analytic modeling of combined degradation mechanisms of concrete structures

4.1 General scheme for modelling combined degradation

Structures exposed to outdoor climate are usually subject to several types of degradation mechanisms simultaneously. If, for example, a concrete structure is exposed to normal Nordic climate, it is inevitably attacked by both carbonation and frost. If the structure is also exposed to seawater or de-icing chlorides the penetration of chlorides into concrete may effect on both carbonation and frost attack of concrete and the penetration process itself may be affected by the other degradation mechanisms. Later corrosion of reinforcement may be activated as a result of carbonation, chlorides and frost attack. The cracking of concrete cover as a result of corrosion on steel reinforcement, which is usually considered to determine the service life of a concrete structure, is a complicated result of various degradation mechanisms both in concrete and steel.

The single degradation mechanisms, such as carbonation, chloride penetration and frost attack, are relatively well understood. There are also fairly good models for such degradation when treated separately. However, the interaction of these mechanisms is not well understood and there are hardly any models for such combined effects. However, to be able to evaluate the service life of concrete structures understanding on combined effects of degradation mechanisms is necessary.

To create improved degradation and service life models, field tests, laboratory tests and theoretical/analytic reasoning on the degradation mechanisms is needed. Field tests are absolutely necessary for obtaining reliable degradation and service life data for reference. Laboratory tests are useful for studying single and coupled degradation mechanisms to learn about the mechanisms themselves. Theoretical/analytic reasoning provides understanding on the basic physical phenomena behind the degradation such as diffusion, convection, erosion, abrasion, phase transitions, electrochemical corrosion etc. Parameters to the theoretical framework are usually found by experimental research both in laboratory and in field.

As a first step in the theoretical reasoning process, simple time related model functions on degradation are developed. As a second step computer simulation can be used for profound understanding of the combined effects of degradation. In computer simulation the ‘first step’ model functions are used but as differential approximations. That means that the changes in temperature and moisture conditions and the effects of other degradation mechanisms can be re-evaluated in every time step (typically 1 hour) according to the real situation in climatic conditions (Finnish Meteorological Institute, 1991) and the real progress of other degradation mechanisms. Computer simulation in this case refers to the following: 1. theoretical emulation of ambient climatic conditions, 2. determination of the temperature and moisture variations in a cross-section of

a concrete structure and 3. application of temperature and moisture sensitive degradation models so that

the degradation over time and the service life can be predicted.


23 (47)

Computer simulation must be calibrated with tests results from both laboratory and field. Calibration means that the simulation results of degradation are made fit with the experimental test results by changing the parameters of simulation models.

Finally when full agreement between test results and computer simulation is reached the computer simulation can be used for making simple (practical) service life models to be used by engineers. By computer simulation material, structural and environmental ‘factors’ for simple models of service life can be determined including ‘interaction factors’ between degradation mechanisms.

Figure 5 shows a general scheme for the development of integrated degradation models. The process starts from tests in laboratory and in field. Laboratory tests are conducted both as ‘single’ and as ‘coupled’. Based on these laboratory tests and theoretical/analytic reasoning degradation models for ‘single’ and ‘coupled’ degradation mechanisms are developed.

Simple laboratory testswithout interaction ofdegradation mechanisms

Laboratory tests withcoupled degradation mechanisms

Field tests

Simple analytic modelswithout interaction

Computer simulation withinteractive degradation models and natural climatic conditions

Yes

No

Practical service life models using computer simulation

Are the results of computer simulation consistent with field test results?

Figure 5. Development of integrated degradation and service life models.

Computer simulation is used for studying degradation when the temperature and moisture conditions are constantly changing and the structure is subject to several degradation modes. Results of the ‘coupled’ tests are used to calibrate the simulation. In the next step the results of the simulation in natural weather conditions are compared with the results of field tests. If the results are not consistent changes in the degradation models and other parameters of computer simulation are made to improve the fit. The models can be considered satisfactory only when the results of theoretical calculations and the test results are essentially consistent.


24 (47)

5 Phases of development of integrated degradation models

5.1 Laboratory tests

Laboratory tests are conducted using test methods that support the theoretical understanding on the degradation processes. When the degradation is based on diffusion in concrete (carbonation, chloride penetration etc.), the diffusion coefficient or other characteristic parameter describing the rate of the process is to be determined from the test results. When the degradation is manifested as loss of material from the concrete surface (erosion, abrasion, frost-salt scaling etc.), the test result should be expressed as a measure of lost material from the surface. The result of the test should be comparable with the corresponding loss of material in the field tests.

If the type of degradation is frost attack resulting in internal cracking of concrete the degree of damage is evaluated by ultrasonic wave velocity. Also the capillary suction rate and the critical degree of saturation should be determined during the test or using separate tests.

Coupled tests should be performed with all combinations of studied degradation mechanisms. These tests should be made so that they support theoretical reasoning and the needs of computer simulation.

5.2 Field tests

The conditions in the field tests should be representative to the conditions of structures for which the degradation and service life models are developed. The results of the field tests should be unambiguous and comparable to the corresponding test results in the laboratory. Care should be taken that the degradation mechanisms in the field tests and those in the corresponding laboratory tests are the same. This agreement should be considered in the selection the laboratory tests.

When the field tests are carried out for structures exposed to de-icing salts (bridges, etc.) a suitable test site would be by a motorway where the specimens are exposed to splashing water and chlorides from the road (Figure 6). If the field tests are conducted for structures that are not exposed to chlorides the test field can be anywhere afar from sea and roads. In this case the specimens should be tested both as protected from rain (to get the maximum carbonation rate) and exposed to rain (to get the combined effect of carbonation and frost).

Figure 6. Field tests for concrete specimens beside a motorway near Kotka, Finland.


25 (47)

5.3 Simple analytic models

Theoretical/analytic degradation modelling is based on general physical and chemical theories of phenomena such as diffusion, dissolution, convection, abrasion, chemical reactions, electrochemical corrosion, phase transition etc. In the following the theoretical bases of the most usual degradation mechanisms in concrete are studied more closely.

5.3.1 Carbonation

The fact that the depth of carbonation is approximately proportional to the square root of time can be theoretically derived by applying the diffusion theory with a ‘moving boundary’. In this theory the carbon dioxide is diffused through the already carbonized layer of concrete and reacts with the non-carbonated calcium minerals at the ‘moving boundary’, that is at the distance of Xc (depth of carbonation) from the surface of a structure. The carbon dioxide content between the surface and the moving boundary is assumed to be linear. Then the flux of carbon dioxide towards the moving boundary can be evaluated as:

carb

S

XCDJ ∆

= , (15)

where J is flux of carbon dioxide [g/m2s], D is the diffusion coefficient of concrete with respect to CO2

[m2/s], Xcarb is distance of the moving carbonation boundary from the surface

of the structure [m], ∆CS = Cs-Cxc [kgCO2/m3], CS is CO2 content of air at the surface of concrete [kgCO2/m3] and Cxc is CO2 content of air at the moving boundary [kgCO2/m3].

The carbon dioxide flux into concrete must be in balance with the rate of reaction occurring at the boundary. The rate of reaction depends on the amount of non-carbonated calcium in concrete:

dtdXaJ carb

R = , (16)

where JR is reaction rate of carbon dioxide [g/m2s], a CO2-binding capacity of concrete [kgCO2/m3] and t time [s].

By combining the above Equations (15) and (16) (J = JR) and integrating over time (XC = 0 when t = 0), the following solution is obtained (Schieβl, 1979):

atCDX S

Carb⋅∆⋅

=2 . (17)

From Equation (17) it is seen that at constant conditions the depth of carbonation is proportional to the square root of time. If ∆Cs is constant in Equation (17), the equation can be presented in a more simplified form as:


26 (47)

tkX CarbCarb = (18)

where kCarb is coefficient of carbonation [ amm / ].

The values of coefficient of carbonation can be found experimentally.

5.3.2 Chloride penetration

Another phenomenon in the concrete that can initiate corrosion of reinforcement is chloride penetration. Chlorides in sufficiently rich concentrations can break the passive film around the reinforcement, even in very alkaline solutions. The chloride penetration into concrete can be assumed to comply with the Fick's second law of diffusion. Accordingly in a semi-infinite wall the following solution can be derived for chloride content in concrete:

⋅−=

−−

tDxerf

CCCC

appS 21

0

0 (19)

where C0 is the initial chloride content of concrete pore water [mol/m3], CS is chloride content at the surface of concrete [mol/m3], x is distance from surface [m], Dapp is apparent diffusion coefficient [m2/s] and t is time [s].

According to Bazant the error-function solution (Equation (19)) can be closely approximated by the following parabola function (Bazant, 1979):

2

321

⋅⋅−⋅=

tDXCC

app

ClS . (20)

When Equation (20) is solved for XCl and the chloride content C is replaced by the critical chloride content Ccrit the following function is obtained for the depth of critical chloride content:

tDCCX app

S

critCl ⋅⋅

−= 321 (21)

where XCl is depth of critical chloride content at moment t [m].

In this form it is obvious that the depth of the critical chloride content approximately complies with the ‘square-root-of-time’ relationship in the same way as the depth of carbonation:

tkX clCl *= (22)

where


27 (47)

−⋅=

s

criteffcl C

CDk 132* (23)

where kcl* is coefficient of chloride penetration [ amm / ] and t is time [a].

The Bazant-simplification offers a mathematically easy way to treat the problem of chloride ingress. The coefficient of chloride penetration depends on the type of cement and it can be found experimentally. From Equations (18) and (22) follows that in both cases the following differential equation can be applied (Fagerlund, 1994):

tXk

ttkXc

ccC ∆⋅⋅=∆⋅⋅⋅≈∆

−2

21

21

21 (24)

where Xc is depassivative depth (either carbonation depth or the depth of critical chloride content) [mm] and

kc is coefficient of depassivation (either carbonation or the chloridepenetration) [ amm / ] and

t is time [a].

Both the carbonation depth and the depth of critical chloride content can be also modelled with two parameters as follows:

BtAtX ⋅=)( (25)

where X is the depth of carbonation or critical chloride content [mm], and A, B are parameters depending on the climatic conditions, material

properties and possible structural measures [-].

In that case the differential approximation during ∆t would be:

tXABttABX BBBc ∆⋅⋅=∆⋅⋅≈∆

−−111

1 . (26)

5.3.3 Corrosion of Reinforcement

In aerial conditions corrosion can start only as a result of physical or chemical changes in concrete around the reinforcement. Physical changes are cracking and deterioration of concrete. As a result of cracking and deterioration a part of the steel is exposed and left without any physical or chemical protection. Chemical changes in concrete are carbonation and penetration of chlorides. Corrosion starts when either the front of carbonation or the critical chloride content reaches the depth of reinforcement.

The process of corrosion is electrochemical. This means that electrons together with chemical compounds take part in the reactions. Two reactions are necessary to build up an electrochemical reaction: (1) anodic reaction, which generates electrons, and (2) cathodic reaction, which consumes electrons. In concrete the


28 (47)

anode and cathode areas of the reinforcement form an electric circuit with concrete as the electrolyte.

In corrosion reactions oxygen together with iron ions are chemically bound to ferrous or ferric hydroxides. The volume of corrosion products is much greater than the volume of the original metal. As a result of the expanding layer of corrosion products, pressure is produced in concrete at the anodic areas of rusting steel. If corrosion is allowed to continue the pressure will grow so great that it exceeds the tensile strength of concrete. As a result the concrete cover cracks. With continued corrosion and an increasing layer of corrosion products, concrete will spall from the surface of the structure. During this phase concrete cover completely loses its ability to protect the steel. In addition, the capacity of the structure may become questionable as a result of the reduced bond between steel and concrete. Furthermore the aesthetic appearance suffers from cracking and corrosion stains. So for many reasons the phases of corrosion cracking and spalling can be regarded as the limit states for service life of a concrete structure.

Applying Faraday’s law, the corrosion rate can be used from direct electrical measurement:

corrcorr ir ⋅= 6.11 , (27)

where rcorr is corrosion rate [µm/a] and icorr is mean value of the corrosion current [µA/cm²].

The rate of corrosion in aerial conditions slows down with time. The reduction of corrosion rate is a result of corrosion products formed around the steel. Corrosion products seemingly act as extra electric resistance. The depth of corrosion can normally be presented in the form:

Bcorr tAs ⋅= , (28)

where scorr is depth of corrosion [µm], t is time [a] and A, B are coefficients [-].

Equation (28) is of the same form as Equation (25). In computer simulation it can be treated in a differentiated form as Equation (26). The coefficients A and B can be studied experimentally. Normally the exponent of time, B, is smaller than 1. However, if the rate of corrosion is controlled by oxygen diffusion, as is the case for wet concrete, B is close to 1. For the maximum allowable depth of corrosion several limit states may be applied:

• smax = 0 • smax = scrack (limit state of cracking in concrete cover) • smax = scapacity (limit state based on the load bearing capacity of the structure).

In some cases, as in the case of prestressing steel, no corrosion can be allowed (smax = 0). It means that there is no active corrosion time allowed. Then the service life is determined only based on the depassivation time (t0) which depends on the carbonation and/or chloride penetration processes. In general the service life can be determined as follows:


29 (47)

10 tttL += , (29)

where tL is service life [a], t0 is depassivation time of corrosion [a] and t1 is active corrosion time (propagation time) [a].

Normally the active corrosion time is defined using the limit state of cracking in concrete cover. The critical depth of corrosion corresponding to the limit state of cracking can be evaluated by the formula:

100crackcsd

= ⋅ , (30)

where scrack is depth of crack [µm], c is thickness of concrete cover [mm] and d is diameter of steel rod [mm].

5.3.4 Frost Attack

Frost attack is caused by freezing of water inside concrete. The vulnerability of concrete to internal frost attack can be evaluated in a laboratory using the theory of critical degree of saturation (Fagerlund, 1977). The degree of water saturation is defined as the relation of water filled volume of pores to the total volume of pores:

tot

e

VwS = , (31)

where S is degree of water saturation [-], we is volume of evaporating water in concrete [mm3] and Vtot is total volume of pores in concrete (air pores included) [mm3].

According to the theory, the occurrence of frost failure depends on the degree of water saturation at the moment of freezing. The damage is always confined to such parts of the concrete where the degree of water saturation exceeds the critical value during freezing. The damage causes cracking, reduction in the strength and reduction in the E-modulus of concrete. Normally the critical degree of saturation is exceeded only if all the capillary pores and a part of the smallest air pores have been filled with water (Fagerlund, 1977):

tot

crit

tot

crittotcrit V

aV

aVS −=−

= 1 , (32)

where Scrit is critical degree of water saturation [-] and acrit is critical volume of (air filled) air pores [mm3].

There are theoretical methods by which it is possible to evaluate the critical volume of air pores if the air pore distribution of concrete is known. By this approach, the critical degree of saturation can be evaluated from Equation (32).


30 (47)

The critical degree of saturation can also be determined by a direct freezing test (Fagerlund, 1977).

To study the rate of water uptake into concrete a thin piece of concrete is placed on a grating so that it can continually absorb water through one of its surfaces. By weighing the specimen one can follow the increase of the water saturation which is nearly proportional to the square root of time. However there is a clear nick-point in the curve indicating the point where all the capillary pores are filled and only air pores, which have been capsulated by the capillary water are still air filled. The growth of the degree of water saturation can be mathematically modelled as:

Do

D tCStCASortBStBAS

⋅+≈⋅+=

⋅+≈⋅+=

'

loglog 0 , (33)

where A, A’, B, C and D are parameters obtained from results of a capillary water saturation test (Fagerlund, 1977). Using Equation (33) it is possible to evaluate the time, which is needed to reach the critical degree of saturation. In outdoor conditions the periods of moisture absorption and drying keep changing. The internal frost damage of concrete depends on the length of wet periods in relation to the time needed to reach the critical degree of saturation. So, the ‘failure’ depends on both the outdoor conditions and the properties of concrete. Frost damage is manifested as changes in strength, modulus of elasticity, weight and volume of concrete. In practical frost resistance tests frost damage is often evaluated by the transition time of an ultrasonic wave through a concrete specimen as this can be measured without destroying the specimen. The dynamic modulus of elasticity of concrete is proportional to the square of the velocity of the ultrasonic wave:

2

00

=

vv

EE NN , (34)

where EN is dynamic modulus of elasticity after N freeze-thaw cycles [MPa], E0 is dynamic modulus of elasticity in the beginning of a test [MPa], vN is velocity of ultrasonic wave after N freeze thaw cycles [m/s], and v0 is velocity of ultrasonic wave in the beginning of a test [m/s].

If the degree of saturation is smaller than the critical saturation practically no reduction in the modulus of elasticity takes place as a result of freezing. If the degree of saturation is greater than the critical one the diminishing of dynamic modulus is clear. The degree of damage in concrete is proportional to the difference between the actual degree of saturation and the critical degree of saturation. According to Fagerlund the degree of damage in concrete might be defined in the following way (Fagerlund, 1996):

)(10

critNN SSK

EED −⋅=−= , (35)

where D is degree of damage (0 < D < 1) [-], N is number of critical freezing events [-],


31 (47)

S is degree of saturation at the moment of freezing [-], Scrit is critical degree of saturation [-] and KN is coefficient of degradation [-].

The coefficient of degradation can be determined from the formula:

NBNAKN +⋅

= , (36)

where A, B are constants [-].

In the derivation of Equations (35) and (36) it has been assumed that the degree of water saturation stays constant at successive critical freezing events. In natural weather conditions, however, the moisture content of concrete varies all the time. So the degree of water saturation is different in successive freezing events and Equation (35) cannot be applied as such. That is why a modified Equation (37) was created to be applied in normal outdoor conditions (Vesikari, 1999; Vesikari, 1998):

∑=

−⋅∆

=N

NcritN

sat

NN ww

wKD

1)( , (37)

where DN is degree of damage after N critical freezing events [-], wN is moisture content of concrete at the moment of Nth critical

freezing [kg/m3], wcrit is critical moisture content (corresponding to critical degree of

saturation) [kg/m3] and wsat is moisture content of completely saturated

concrete(corresponding to total volume of pores) [kg/m3].

The quantity ∆KN depends on the number of critical freezing events and is determined as:

)1()1(

−+−⋅

−+⋅

=∆NBNA

NBNAKN . (38)

The critical moisture content wcrit and the constants A and B can be determined experimentally by direct freezing tests and by an analysis of test results. The service life is ended when the degree of damage reaches the maximum allowable value (normally ∼ 30%).

5.3.5 Frost-salt scaling

Frost-salt scaling is caused by freezing on a concrete surface when it is in contact with a chloride solution. First the initial scaling occurs in cement paste while the aggregate grains are intact. Due to the gradually deeper and deeper scaling, coarser aggregate grains can also be lost. The depth of scaling is related to the number of freeze-thaw cycles in the surface of concrete. The relationship between the freeze-thaw cycles in a frost-salt test and the depth of scaling can be evaluated by Equation (39) (Fagerlund, 1996):


32 (47)

kscsc Nkx ⋅= , (39)

where xsc is scaling depth of concrete [mm], N is number of freeze-thaw cycles [-], ksc is coefficient of freeze-thaw scaling [mm] and k is acceleration factor [-].

The rate of scaling can be

• retarding (k < 1) • even (k = 0) • accelerating (k > 1).

An accelerating rate of scaling is usually a sign of poor frost resistance of concrete. A retarding scaling rate is a sign of inferior quality of concrete on the surface compared to the quality inside. In homogeneous concrete an even scaling rate is expected. The rate of scaling in natural weather conditions cannot be evaluated exactly. However, rough evaluations can be made on the bases of laboratory test results. Assuming homogeneous concrete and a linear scaling rate, Equation (40) for the depth of scaling may be applied in natural conditions (Fagerlund, 1996):

test

eqsctestsc n

nxx ∑⋅= , (40)

where xsc is depth of frost scaling in natural conditions [mm], xsctest is depth of frost scaling in a laboratory test [mm], ntest is number of freeze-thaw cycles in a laboratory test [-] and Σneq is equivalence number of freeze thaw cycles in natural

conditions (corresponding to the cycles in a laboratory test) [-].

5.4 Modelling of interaction between degradation modes with computer simulation

Computer simulation offers a possibility to better manage the effects of constantly changing climatic conditions (Finnish Meteorological Institute, 1991) and simultaneous degradation processes. Computer simulation utilizes theoretical/analytic degradation models as exemplified above. However they are applied as differential approximations in order to take into account the momentary condition and exposure stresses and the momentary progress of various degradation mechanisms in the structure. In the following the theoretical methods of coupling degradation mechanisms are discussed (Vesikari, 1999; Vesikari, 1998).

5.4.1 Effect of frost attack on carbonation or chloride penetration

Carbonation depth or the depth of critical chloride content with time is expressed by the following model formula (ref. Eq. (25)):


33 (47)

BtAtX ⋅=)( , (41)

where X is depth of carbonation or critical chloride content [mm], A is coefficient [mm/aB], B is exponent of time [-] and t is time [a].

The coefficient A depends on the quality of concrete, type of cement, temperature and moisture content of concrete etc. The exponent B is dependent on the moisture content of concrete. Then the differential approximation of the degradation formula is of the following form (ref. Eq. (26)):

iRHB

iRHB

i tXRHTConcARHBX ∆⋅⋅⋅=∆−

−)(

11

1)(

1

),,()( , (42)

where i refers to the time interval in sequence [-].

By this means the contribution of the time interval ∆t is determined using the actual values of concrete temperature and relative humidity in parameters A and B during the time interval i. By a computer simulation the thermal and moisture movements in structures are imitated closely as they occur in natural environmental conditions. The simulation program calculates hour by hour the temperature and moisture content in various points of a cross-section having the real weather conditions or in-use conditions as initial data on the surfaces of the structure. The weather is modelled based on long-term statistics of weather parameters such as temperature, relative humidity, wind, rain and solar radiation. Frost action accelerates the processes of carbonation and chloride penetration in concrete by increasing the permeability of concrete. So it is obvious that it affects the coefficient A in Equation (42). Accordingly, coefficient A is a function of degree of frost damage, Dfrost, (ref. Eq. (37)). The effect of frost action on carbonation or chloride penetration can so be determined using Equation (43):

iRHB

iRHB

ifrosti tXDRHTConcARHBX ∆⋅⋅⋅=∆−

−)(

11

1)(

1

; ),,,()( , (43)

where Dfrost;i is the degree of frost damage during time interval i.

If carbonation and chloride penetration have any effects on each other their contribution to the coefficient A is added in the same way as that of frost. In that case the coefficient A is not only dependent on Dfrost but also on the depth of carbonation or on the depth of critical chloride content. The function A with all its dependencies is studied by experimental tests. The total depth of carbonation or depth of critical chloride content is obtained by totalling the differential depths during all intervals of ∆ti.

5.4.2 Effect of frost-salt attack on carbonation or chloride penetration

Frost-salt attack accelerates the penetration of carbonation or critical chloride content by removing material from the surface of concrete. Having the depth of scaled concrete from the surface of concrete Xscal the contribution of surface


34 (47)

scaling on the penetration process during time interval i can be theoretically expressed as presented in Equation (44):

iRHB

iscaliRHB

ifrosti tXXDRHTConcARHBX ∆⋅−⋅⋅=∆−

−−)(

11

1:1)(

1

; )(),,,()( , (44)

where Xscal;i is the depth of scaled concrete as a result of frost-salt attack in time interval i [mm].

If the carbonation or chloride penetration is modelled using the ‘square root of time’ relationship (ref. Eq. (18) and (22)), Equation (44) can be replaced by Equation (45) (Vesikari, 1998):

iiscali

ifrosti t

tXtXDRHTConck

X ∆⋅−

⋅=∆−− )()(

),,,(21

1;1

2: . (45)

The effect of any other degradation mode that causes loss of concrete on the surface of the structure (erosion, abrasion etc.) could be modelled in the same way.

5.4.3 Effect of frost-salt scaling and internal frost damage to the active corrosion time

Frost damage affects the physical properties of concrete. As corrosion cracking is dependent on both the thickness and the quality of concrete cover, frost damage must have an influence on the amount of corrosion necessary to cause cracking. Both surface scaling and internal frost damage are taken into account in the reduction of the critical depth of corrosion (ref. Eq. (30)). The reduction in scrack is considered to be proportional both to the reduction in concrete cover and to the reduction in dynamic E modulus of concrete (1-Dfrost):

)1(100 ;;

ifrostiscal

crack DdXc

ms −⋅−

⋅= µ , (46)

where scrack is average depth of corrosion necessary to cause a crack in concrete cover [µm],

c is thickness of concrete cover [mm], and Xscal;i is depth of scaling of concrete surface at the beginning of ith

time increment [mm], and d is diameter of steel rod [mm]. Dfrost;i degree of damage at the beginning of ith time increment [-].

As scrack is reduced with time the active corrosion time is also reduced.


35 (47)

5.5 Developing service life models using computer simulation

5.5.1 Service life models in the Finnish national codes

The Finnish national codes presents simple models for service life design of concrete structures based on carbonation and frost attack (EN 206-1, 2004). The design life span is calculated by Equation (47):

tL = tLr · A · B · C · D · E · F · G, (47)

where tL is design life span [a], tLr is comparison life span of 50 years [a] and A to G are coefficients taking into consideration all of the variables

affecting the design life span [-].

The design parameters affecting the life span with respect to frost attack are presented in Table 2.

Table 2. Coefficients affecting the freeze-thaw life span of a concrete structure. (EN 206-1, 2004) Coefficient Affecting feature Design parameters A Material properties, porosity Water-cement ratio, air content B Design, structural features Structural member, possible surface treatment C Workmanship Curing measures D Internal climate -

E External climate Freeze-thaw class, geographical direction and location

F Design loads - G Maintenance Inspection and service sequence

Every coefficient, A to G, can be determined from a respective table. As an example the first coefficient A can be selected from Table 3 as a function of the maximum aggregate size and concrete air content measured from fresh concrete. As can be seen from the table, the maximum value of the coefficient is 4 which enables a 200 year design life span because tL = 50 years. In a full service life analysis, the other coefficients (B to G) would be assessed in the same manner.


36 (47)

Table 3. Values of coefficient A in environmental class XF1 (façade elements). The values in the shaded area can be used only in design failure situations. (EN 206-1, 2004)

Air content, % Maximum aggregate

Coefficient A

diameter, mm Effective water-cement ratio 8 12 >16 0.3 0.35 0.4 0.45 0.5 0.55 0.6 0.65 0.7

2.5 2.0 1.5 1.56 1.03 0.78 0.64 0.55 0.48 0.43 0.39 0.36 3.0 2.5 2.0 4.00 1.95 1.26 0.95 0.77 0.66 0.57 0.51 0.46 3.5 3.0 2.5 4.00 3.76 1.89 1.29 1.00 0.82 0.70 0.62 0.55 4.0 3.5 3.0 4.00 4.00 2.86 1.71 1.24 0.99 0.82 0.71 0.63 4.5 4.0 3.5 4.00 4.00 4.00 2.25 1.52 1.16 0.95 0.81 0.71 5.0 4.5 4.0 4.00 4.00 4.00 3.01 1.85 1.35 1.08 0.90 0.78 5.5 5.0 4.5 4.00 4.00 4.00 4.00 2.25 1.57 1.22 1.00 0.86 6.0 5.5 5.0 4.00 4.00 4.00 4.00 2.76 1.81 1.36 1.10 0.93 6.5 6.0 5.5 4.00 4.00 4.00 4.00 3.43 2.09 1.52 1.21 1.01 7.0 6.5 6.0 4.00 4.00 4.00 4.00 4.00 2.42 1.70 1.32 1.09 7.5 7.0 6.5 4.00 4.00 4.00 4.00 4.00 2.82 1.90 1.44 1.18 8.0 7.5 7.0 4.00 4.00 4.00 4.00 4.00 3.31 2.12 1.57 1.26 8.5 8.0 7.5 4.00 4.00 4.00 4.00 4.00 3.93 2.36 1.71 1.35 9.0 8.5 8.0 4.00 4.00 4.00 4.00 4.00 4.00 2.65 1.86 1.45 9.5 9.0 8.5 4.00 4.00 4.00 4.00 4.00 4.00 2.98 2.03 1.55

10.0 9.5 9.0 4.00 4.00 4.00 4.00 4.00 4.00 3.37 2.21 1.66

In the case of life span assessment for carbonation, a similar formula as Equation (47) is used and the coefficients A to G are obtained in a similar manner. The problem with the service life design according to the Finnish national codes is that it does not present any model for structures in chloride environment and that the interaction between degradation modes is not considered satisfactorily.

5.5.2 Computer simulation in specification of factors of service life

After calibration with field and laboratory tests, computer simulation can be used for quantification of factors in the service life models with respect to various degradation mechanisms. Considering the models in the Finnish national codes (Eq. (47)) the simulation method is especially suitable for quantification of factors A, B, D and E.

The simulation program of VTT allows a possibility for easy quantification of the environmental factor E. The climatic data have been gathered from different observation sites so that the factors for coastal Finland, Middle Finland and Northern Finland can be determined using actual weather parameters in these regions. The weather models are based on the data gathered by the Finnish Meteorological Institute and they consist of data on temperature, relative humidity, velocity and direction of wind, amount of rain and intensity of solar radiation.


37 (47)

The VTT simulation program contains also design parameters of hardened concrete and concrete mix. The effects of nominal compressive strength, water-cement-ratio, air content, quality of binder etc. on the degradation rate and service life can thus be studied by the program. Not only the effects of air content of concrete but also the quality of air i.e. the air pore distribution can be studied by the program. The effect of concrete cover on service life of structures can of course also be studied.

Different model structures can be used in computer simulation. A concrete sandwich element has been used for studying degradation of concrete in a non-chloride environment. For a chloride environment, simple concrete slabs and walls representing bridge slabs and walls have been used as model structures. The outdoor weather conditions form the boundary conditions of a thermal and moisture mechanical problem to solve the temperature and moisture distributions in a structural cross-section. Although the calculation covers the whole cross-section of the structure the main interest is focused on the edge parts where the degradation is greatest.

The effects of coatings and other protective measures can also be studied by simulation (Vesikari, 2002). The effect of coatings is complex as two kinds of effects can be distinguished: (1) direct way by retarding the penetration of harmful agents such as chlorides and CO2 into concrete and (2) indirect way by retarding the moisture penetration into concrete. The indirect effect may decrease degradation (chloride penetration, frost attack) or increase it (carbonation). In many cases the total effect is hardly evaluable unless computer simulation is used.

The effects of other degradation modes on the service life can be studied using the interaction parameters in the degradation models as presented in Chapter 5.4. As the service life can be determined with and without parallel degradation modes ‘the interaction factors can be determined.

The incremental time in the computer simulation is normally one hour. The total calculation may cover some months, some years or even some hundreds of years from the lifetime of the structure although the time of calculation by itself takes only some hours.


38 (47)

6 Conclusions

Ice abrasion, internal frost, de-icer salt scaling and chloride induced reinforcement corrosion are major forms of damage for concrete exposed to seawater in arctic and Nordic environments. An increase of the compressive strength in addition to good bond between cement paste and abrasive resistant aggregate appears to be the best way to obtain ice abrasion resistant concrete. All mechanisms governing the deterioration in de-icer salt scaling and partly also internal frost action are still not clear even though it is known that proper mixture design including low water/binder ratio and high strength as well as adequate air entrainment give the most durable materials. In addition the access to water must be limited so as to limit the materials degree of saturation. This partly also holds for chloride induced reinforcement corrosion. There service life is governed by both material transport parameters and environment during the initiation phase and in addition by corrosion parameters in the propagation phase. These are taken care of in standard procedures for concrete composition for varying exposure conditions and some deterministic service life models based on the basic parameters are presented.

When possible, interaction between two or more forms of damage is likely to accelerate degradation. Degree of degradation accounting for interaction when several types of deterioration are acting simultaneously are suggested based on review of physical mechanisms, superposition of degree of degradation with interaction coefficients and coupled degree of degradation equations. A review of some physical mechanisms of deterioration shows that interaction is mainly reducing the service life compared to uniform deterioration. Finally a different approach is proposed by coupling individual service life equations using phenomenological coupling coefficients (Lij). Material data and observations of rate of degradation under several types of deterioration in field and/or in accelerated laboratory tests can be used to determine the coupling coefficients, Lij.

A general scheme for modelling combined effects of various degradation mechanisms in concrete structures is presented. The method is based on tests both in field and laboratory. The laboratory tests are conducted as ‘single’ tests consisting of only one degradation mechanism and as ‘coupled’ consisting of two degradation mechanisms at the same time. As the degradation mechanisms are usually based on some known physical or chemical processes such as diffusion, chemical reactions, abrasion, phase transitions etc. simple degradation models can be derived by theoretical reasoning. The laboratory tests are used in parameterization and quantification of the theoretically created degradation models. Computer simulation is used by combining the effects of constantly changing climatic conditions and other degradation mechanisms on the treated degradation mechanism. Theoretically the computer simulation is a powerful way to treat such complex phenomena. However, the computer simulation must be calibrated with field and laboratory tests. To improve the consistency between the observed and calculated degradation, parameters of the simulation models are changed until full agreement is reached. By the use of calibrated computer simulation, service life models for engineering purposes can be developed. In the case of the factor method according to the Finnish national codes, most of the factors of the design formula can be determined using computer simulation. The various degradation mechanisms in structures can be simultaneously simulated throughout the service life. By making sequential computer runs with series of


39 (47)

material, structural and environmental parameters the values for material, structural and environmental parameters of the design formula can be determined including ‘interaction parameters‘ of various degradation mechanisms.


40 (47)

References

Al-Neshawy, F., Sistonen, E., Piironen, J., Huovinen, S. (2005), Effect of Maintenance Measures on Service Life of Concrete Facades, Nordic Concrete Research - Research projects 2005, Proceedings Nordic Concrete Research Meeting, XIX Symposium on Nordic Concrete Research & Development - A meeting place for research and practice; Norway, Sandefjord: 12. - 15.6.2005, Publication No. 33, pp. 217 - 219.

Bager, D.H., Sellevold, E. J. (1986), Ice formation in hardened cement paste - Part I: Room temperature cured paste with variable moisture contents, Cement and Concrete Research, Vol. 16, No. 5, pp. 709 - 720.

Bazant, Z. P. (1979), Physical model for steel corrosion in concrete sea structures - application. Journal of the Structural Division (ASCE). 1979: June, pp. 1155 - 1165.

Bazant, Z. P., Chern, J. C., Rosenberg, A.M., Gaidis, J. M. (1988), Mathematical model for freeze-thaw durability of concrete, Journal of the American Ceramic Society, Vol. 71, No. 9, p. 776 - 783.

Bazant, Z. P., Planas, J. (1998), Fracture and size effects in concrete and other quasibrittle materials, CRC Press, Boca Raton, USA.

Bérubé, M. A., et al. (1996), Proc. 10th Int. Conf AAR in Concrete, Melbourne, Australia, pp. 1056 - 1063.

Bouzoubaâ N., Bilodeau A., Fournier B. (2005), De-Icing Salt Scaling Resistance of Concrete Incorporating Supplementary Cementing Materials, Conmat’05 Third international conference, Vancouver, Canada, August 22-24 2005.

Broomfield, J. P. (1997), Corrosion of Steel in Concrete. London, United Kingdom: E & FN Spon, ISBN 0-419-19630-7. 238 p.

Bærland T. (1989), High performance concrete for roads, Norcem, Norway 68 p. (in Norwegian)

By 50 (2004), Concrete Code 2004, Concrete Association of Finland, 2004. (in Finnish).

Çopuroglu O., Fraaij A. L. A., Bijen J. M. J. M. (2004), Effects of curing conditions on freeze-thaw de-icing salt resistance of blast furnace slag cement motars, High Perf Structures and Mat. II, C.A. Brebbia & W.P. De Wilde (Editors), Netherlands, p. 233 - 241.

Çopuroglu O., Schlangen E., van Breugel K. (2006), Frost salt scaling modeling of cement paste, Concrete Repair, Rehabilitation and Retrofitting - Ed.: M. Alexander, Taylor and Francis Group, London, ISBN 0 415 39654 9, pp. 127 - 133.

EN 206-1. (2004), Annex A: Concrete - Part 1: Specification, performance, production and conformity, European Standard, draft Finnish National Annex, Helsinki, to be accepted 2004.


41 (47)

Fagerlund G. (1971) Critical degree of saturation - A tool for the assessment of the frost resistance of construction materials, Materials and Structures/Matériaux et Constructions, pp. 271-285, (in French).

Fagerlund G. (1972), Critical degrees of saturation at freezing of porous and brittle materials, Lund Inst. of Techn., Div of build. Mtrls report 32, 411 p. (in Swedish).

Fagerlund, G. (1977), Critical degree of saturation method of assessing the freeze/thaw resistance of concrete. Tentative RILEM recommendation. Prepared on the behalf of RILEM Committee 4 CDC. Materiaux et Constructions, 1977, No. 58, pp. 217 - 229.

Fagerlund, G. (1996), Service-life calculations for concrete structures. Lund, Lunds Tekniska Högskola, Byggnadsmaterial. TVBM-3070, 124 p. (in Swedish).

Fagerlund, G., Somerville, G., Tuutti, K. (1994), The residual service life of concrete exposed to the combined effect of frost attack and reinforcement corrosion. Int. Conf. Concrete Across Borders. Odense 1994. Danish Concrete Association. Proceedings, pp. 351 - 365.

Fagerlund, G. (2003), Synergetic Effects of Combined Destructive Action on Concrete, Proceedings of the International Symposium: Role of Concrete In Sustainable Development, 3-4 Sept. 2003, Dundee, Scotland.

Finnish Association of Civil Engineers (1995). RIL 183-4.9-1995. Service life of construction materials and structures. methods of appraisal. 4.9: Service life design of concrete structures. Vantaa, Finland: Finnish Association of Civil Engineers, RIL, ISBN 951-758-344-3, 120 p. (In Finnish).

Finnish Meteorological Institute. (1991), Climatological Statistics in Finland 1961 - 1990. Supplement to the Meteorological Yearbook of Finland. Volume 90 Part 1 - 1990. Helsinki 1991. 125 p.

Finnish Road Administration. (2008), The P-rate method for concrete bridges. Helsinki, Finland. Finnish Road Administration, 26 p., ISBN 978-952-221-024-1, TIEH 2200054-v-08 (in Finnish).

Fiorio, B. (2005), Wear characterisation and degradation mechanisms of a concrete surface under ice friction, Construction and Building Materials, Vol. 19, No. 5, p. 366 - 375.

Fiorio, B., Meyssonnier, M., Boulon, M. (2002), Experimental study of the friction of ice over concrete under simplified ice-structure interaction conditions, Can. J. Civil Eng., (9) p. 347 - 359.

Gaal, G. C. M. (2004), Prediction of Deterioration of Concrete Bridges Corrosion of Reinforcement due to Chloride Ingress and Carbonation, Ph.D. Thesis, Delft University of Technology, Delft, Netherlands.

Gerard B. (1996), Modelling the long term durability of concrete barriers for radioactive waste disposals: couplings between mechanics, chemistry and transport properties, ENS Cachan/U. Laval-CRIB 291 p + 79 app. (in French).

http://www.sciencedirect.com/science?_ob=JournalURL&_cdi=5702&_auth=y&_acct=C000049220&_version=1&_urlVersion=0&_userid=952938&md5=6ef13fe55a498465e6f43d91fc6cb1fd

http://www.sciencedirect.com/science?_ob=IssueURL&_tockey=%23TOC%235702%232005%23999809994%23566358%23FLA%23&_auth=y&view=c&_acct=C000049220&_version=1&_urlVersion=0&_userid=952938&md5=47811dbcf14fab89207654c380667000

http://www.sciencedirect.com/science?_ob=IssueURL&_tockey=%23TOC%235702%232005%23999809994%23566358%23FLA%23&_auth=y&view=c&_acct=C000049220&_version=1&_urlVersion=0&_userid=952938&md5=47811dbcf14fab89207654c380667000


42 (47)

Gjørv O.E., Bærland T., Rønning H., (1987), High strength concrete for highway pavements and bridge decks, Proc. Utilization of High Strength Concrete, (Eds.: Holand I., Helland S., Lenschow R., Jakobsen B.), Tapir, Norway p. 111 - 122.

Helmuth, R. A. (1960), Capillary size restrictions on ice formation in hardened cement paste, Proceedings of the Fourth International Congress on the Chemistry of Cement, Washington, D.C., USA, Monograph 43, Vol. II, Session VI, Paper VI-S2, p. 855 - 869.

Helmuth, R. A. (1961), Dimensional changes of hardened Portland cement pastes caused by temperature changes, Highway Research Board Proceedings, Vol. 40, p. 315 - 336.

Hengjing Ba, Zhao Xiaolong, Yang Yingzi (2004), Research on durability of concrete under the actions of multidamage factors, Proc. International conference on durability of high performance concrete and final workshop of CONlife, M.J.Setzer Ed., AEDIFICATIO Publishers, Freiburg (2004), pp. 277 - 285.

Hoff, G. C. (1988), Resistance of Concrete to Ice Abrasion - A Review, ACI SP-109, p. 427 - 455.

Hooton, R. D., Boyd, A. (1997), Effects of finishing, forming and curing on de-icer salt scaling resistance of concretes, Frost Resistance of Concrete, Ed. by M. J. Setzer and R. Auberg, E & FN SPON, London, U.K., p. 174 - 183.

Hörsch, T., Wittmann, F. H., (2002), Simulation of damage and crack formation in the composite structure of concrete under freeze/thaw cycles, RILEM PRO 24 (Eds.: M. J. Setzer, et al.) p. 225 - 234.

Huovinen, S. (1990a), Abrasion of concrete by ice in arctic sea structures, VTT Publications 62, (Doctoral thesis), Espoo, 110 p, app. 31 p.

Huovinen, S. (1990b), Ice Abrasion of Concrete in Arctic Sea Structures, ACI-Materials Journal, Vol. 87, No. 3, May-June, p. 266 - 270.

Huovinen, S. (1993), Abrasion of Concrete Structures by Ice, Cement and Concrete Research, Vol. 23, No. 1, pp. 69 - 82.

ISO 15686-1. (2000), Buildings and constructed Assets-Service Life Planning-Part 1: General Principles. ISO/TC59/SC14. 2000. 41 p.

Jacobsen, S., Sellevold, E. J. Sæther, D. H. (1997), Frost testing high strength concrete: frost/salt scaling at different cooling rates, RILEM/Materiaux et Constructions, Vol. 30, No. 195 p. 33 - 42

Jacobsen, S., Gran H., Sellevold, E. J., Bakke J. A., (1995a), High strength concrete-freeze/thaw testing and cracking, Cement and Concrete Research, Vol. 25, No. 8, pp. 1775 - 1780.

Jacobsen, S., Marchand, J., Hornain H., (1995b), Sem observations of the microstructure of frost deteriorated and self-healed concretes, Cement and Concrete Research, Vol. 25 No. 8, pp. 1781 - 1790.


43 (47)

Jacobsen, S., Sistonen, E., Huovinen, S., Marchand, J., Tremblay, M.-H. (2006), Abrasion and scaling of concrete by ice, studded tires and frost-salt: interaction and service-life modeling, RILEM proceedings PRO 51 (Eds.: Marchand, J., et al.) paper 26, 23 p.

Jacobsen, S. (1999), Recycled and porous aggregate in wet frost testing, 3rd Nordic Res. seminar on the Frost Resistance of Building Materials, LTH, Sweden Aug.31-Sept.1 1999, report TVBM 3087, pp. 69 - 74, LTH, Sweden.

Jacobsen, S., Sellevold, E.J. (1996), Frost testing high strength concrete: scaling and cracking, 4th Int. Symposium on Utilization of High Strength/-High Performance Concrete, (Eds. F.d.Larrard et al), Presses d'ecole de ponts et Chausses, Paris, pp. 597 - 605.

Jacobsen, S., Sistonen E., Huovinen S., Marchand J., Tremblay, M.-H. (2007), Ice Abrasion, Frost, De-Icer Salt Scaling and Reinforcement Corrosion on Concrete Structures: Interaction and Service-life, CONSEC’07, Concrete under severe conditions: Environment and loading, Tours, France, (Eds.: Toutlemonde et al.) June 4-6, 2007. pp. 1137 - 1152.

Janson, J. E. (1988), Long term resistance of concrete offshore structures in ice environment, 7th Int. Conf. OMAE, Houston, pp. 225 - 231.

Johannesson, B. (1996), Combined effects on concrete exposed to mechanical load and environmentally induced strains, Durability of building materials and components; Proceedings of the 7th International Conf. 7DBMC, Stockholm 19-23 May 1996.

Justnes, H. (1998), A review of chloride binding in cementitious systems, Nordic Concrete Research Publication. 16 p.

Keyser, J. H. (1970), Effect of studded tires on the durability of road surfacing, Highway Research Record No.331 Transportation Research Board p. 41 - 52.

Kruschwitz, J., Bluhm, J. (2005), Modeling of ice formation in porous solids with regard to the description of frost damage. Computational Materials science, Vol. 32, No. 2, pp. 407 - 417.

Kukko, H. (1992), Frost effects on the microstructure of high strength concrete, and methods for their analyses VTT publications 126. 133 p. + app.

Léger, P., Côté, M., Tinawi, R. (1995), Thermal protection of concrete dams subjected to freeze-thaw cycles, Canadian Journal of Civil Engineering, Vol. 22, No. 3, pp. 588 - 602.

Lindgård, J., Rodum, E., Pedersen, B. (2006), Alkali Silica Reactions in Concrete - Relationship between water Content and Damage on structures, paper to be presented at CANMET/ACI Durability Conference, Montreal, Canada. 20 p.

Lindmark, S. (1998), Mechanisms of salt frost scaling of Portland cement-bound materials: studies and hypothesis, Ph.D. thesis (Report TVBN 1017), Lund institute of technology, Sweden, 266 p.


44 (47)

Marchand, J., Pigeon, M., Boisvert, L. (1996), Influence of the minimum temperature of freezing and thawing cycle on the concrete scaling deterioration due to de-icing salts, Revue Canadienne de Génie Civil, Vol. 23, No. 3, pp. 595 - 601. (in French).

Marchand, J. (1993), Contribution to the study of the de-icer salt scaling deterioration of concrete, Ph.D. Thesis, École Nationale des Ponts et Chaussées, France.

Marchand, J., Sellevold, E. J., Pigeon, M. (1994), The de-icer salt scaling deterioration of concrete - An overview, ACI-SP-145, pp. 1 - 46.

Marchand, J., Pleau, R., Gagné, R. (1995), Deterioration of concrete due to freezing and thawing, Materials Science of Concrete - IV, American Ceramic Society, pp. 283 - 354.

Marchand, J., Jolin, M., Machabée, Y. (2005), Influence of the supplementary cementing materials on the de-icer salt-scaling resistance: Ten years of field performance, Cement Combination for Durable Concrete, Thomas Telford, pp. 579 - 590.

Marchand, J, Jacobsen S., Gerard, B. (1998), Concrete cracks I: Durability and self healing - a review, Consec ’98 E&FN Spon, London, ISBN 0 419 23860 3, pp. 217 - 231.

Martín-Pérez, B. (1999), Service Life Modelling of R.C. Highway Structures Exposed to Chlorides, Ph.D. dissertation, University of Toronto, 1999.

Matala, S. (1991), Calculation methods of service life. In: Finnish Association of Civil Engineers. RIL K145-1991. Durability, damage and repair of concrete structures. Helsinki, Finland. pp. 67 - 79, ISBN 951-758-271-4 (In Finnish).

Niemann, C., Lehtonen, V., (1996), How does frost action break down concrete in the laboratory test and in field circumstances, Lund Inst of techn. rep TVBM-3072 (Ed.: Sture Lindmark), pp. 1 - 10.

Onsager, L. (1931), Reciprocal relations in irreversible processes I. Phys.Rev. 37, pp. 405 - 426.

Onsager, L. (1932), Reciprocal relations in irreversible processes II. Phys.Rev. 38, pp. 2265 - 2279.

Pigeon, M. (1984), Microstructure and frost resistance of cements and concretes, Univ. Pierre et Marie Curie, 311 p. (in French).

Pigeon, M., Marchand, J., Pleau, R. (1996a), Frost resistant concrete, Construction and Building Materials, Vol. 10, No. 5, pp. 339 - 348.

Pigeon, M., Talbot C., Marchand J., Hornain H. (1996b), Surface microstructure and scaling resistance of concrete, Cement and Concrete Research, Vol. 26, No 10, pp. 1555 - 1566.


45 (47)

Powers, T.C., with discussion by R. Terzagi et al, (1945), J. ACI, Vol. 16, No. 4, pp. 245 - 272.

Powers, T.C. (1949), The Air Requirement of Frost Resistant Concrete, Proceedings of the Highway Research Board, Vol. 29, p. 184 - 211.

Powers, T.C., Helmuth R. A. (1953), Theory of Volume Changes in Hardened Cement Paste During Freezing, Proceedings of the Highway Research Board, Vol. 32, p. 285 - 297.

Preus, C. K. (1972), Studded tire effects on pavements and traffic safety in Minnesota, Highway Research Record No. 418 Transportation Research Board, pp. 44 - 54.

Puatatsananon, W. (2002), Deterioration of Reinforced and Massive Concrete: A Multi-Physics Approach, Doctural Thesis, University of Colorado, Boulder.

Puatatsananon, W., Saouma, V. E. (2005), Nonlinear Coupling of Carbonation and Chloride Diffusion in Concrete. Journal of Materials in Civil Engineering (ASCE)/May/June 2005. pp. 264 - 275.

Sarja, A., Vesikari, E. (Editors) (1996), Durability design of concrete structures. RILEM Report of TC 130-CSL. RILEM Report Series 14. E&FN Spon, Chapman & Hall, 165 p.

Scherer, G. W. (1993), Freezing gels, Journal of Non-Crystalline Solids, Vol. 155, pp. 1 - 25.

Scherer, G. W. (1999), Crystallization in pores, Cement and Concrete Research, Vol. 29, No. 8, pp. 1347 - 1358.

Schieβl, P. (1979), Regarding the question of maximum crack width and the required concrete cover in reinforced concrete structures paying particular attention to the carbonation of the concrete, Heft 255. 175 p.

Sellevold, E. J., Farstad, T. (1991), Frost/salt testing of concrete: Effect of test parameters and concrete moisture history, Nordic Concrete Research, Vol. 10, pp. 121 - 138.

Sellevold E. J., Jacobsen, S., Bakke, J. A. (1993), High Strength Concrete without air entrainment: effect of rapid temperature cycling above and below 0 °C, Int.worksh. resistance of concrete to scaling due to freezing with deicing salts (Eds.: J.Marchand et al) CRIB, Québec, pp. 153 - 164.

Setzer, M. J. (2002), Development of micro-ice-lens model, RILEM proceedings on Frost resistance of concrete, Edited by Setzer Auberg, pp. 193 - 200.

Sistonen, E. (2001), The Main Corrosion Parameters and their Influence on the Durability of Outdoor Concrete Structures, Licentiate Th., Espoo, TKK, 84 p. + app. 20 p. (in Finnish).

Sistonen E., Kari O., Tukiainen P., Huovinen S. (2006), Stochastic method and Monte-Carlo simulation for predicting service life of hot-dipped galvanized


46 (47)

reinforcement, RILEM proceedings PRO 51 (Eds.: Marchand et al.) paper 45, 23 p.

Sodhi D. S. (2006), Ice forces on structures: crushing failure of ice, Power point presentation, CRREL, New Hampshire, September 20, 59 p.

Stark, D. (1989), Effect of length of freezing period on durability of concrete, Portland Cement Association, Research and Development Bulletin RD096, 9 p.

Tuutti, K. (1982), Corrosion of steel in concrete, Swe. Cem&Conc. Res. Institute Fo4, 469 p.

Trägård, J., Lagerblad B. (1996), Influence of ASR expansion on the frost resistance of concrete, Lund Inst of technology report TVBM-3072 (Ed.: Sture Lindmark), pp. 111 - 120.

Trägård, J., Lagerblad B. (1996), Proc. 10th Int. Conf AAR in Concrete, Melbourne, Australia pp. 853 - 860.

Tveter T. (1993), Studded tire abrasion resistance of High Strength Concrete PhD thesis 62, Norwegian Inst of Technology, Trondheim, 194 p + app (in Norwegian).

Valenza, J. J., Sherer, G. W. (2005), Mechanisms of salt scaling, Materials&Structures. Vol. 38, pp. 479 - 488.

Verbeck, G., Klieger P. (1956), Studies of Salt Scaling of Concrete, HRB Bull. No.150, 13 p.

Vesikari, E. (1991), Frost resistance and service life of concrete. Technical Research Centre of Finland VTT, Research Reports 749. Espoo, Finland, 55 p. (in Finnish).

Vesikari, E. (1999), Computer simulation technique for prediction of service life in concrete structures. Proc. Int. Conf. Life Prediction and Ageing Management of Concrete Structures, July 1999. RILEM Expertcentrum. Bratislava, pp. 17 - 23.

Vesikari, E. (1999), Prediction of service life of concrete structures with regard to frost attack by computer simulation. Proc. Nordic Res. Sem. on Frost Resistance of Building Materials, Aug. 31 - Sept. 1 1999, Lund. Lund Institute of Technology, Division of Building Materials. 13 p.

Vesikari, E. (1998), Prediction of service life of concrete structures by computer simulation. Helsinki University of Technology. Faculty of Civil and Environmental Engineering. Licentiate's thesis. 131 p. (in Finnish)

Vesikari, E. (2002), The Effect of Coatings on the Service Life of Concrete Facades. Proc. 9th International Conference on Durability of Building Materials and Components. Brisbane Convention & Exhibition Centre, Australia, 17–21 March 2002.


47 (47)

Wei Sun, (2005), Binggen Zhan, Jianfang Sha, Experimental Investigation of Damage in Concrete under Combined Effect of Freeze-Thaw and ASR, ConMat’05 UBC, Canada, 10 p.

Wigum, B. J. (1995), Alkali Aggregate Reactions in concrete, properties, classification and testing of Norwegian cataclastic rocks, dr.ing dissertation Norw. Inst. of Techn., 227 p.

Zuber, B., Marchand, J., Delagrave, A., Bournazel, J.P. (2000), Ice formation mechanisms in normal and high-performance concrete mixture, ASCE Materials Journal, Vol. 12, No. 1, pp. 16 - 23.

Zuber, B., Marchand, J. (2000), Modeling the deterioration of hydrated cement systems exposed to frost action, Cement and Concrete Research, Vol. 30, p. 1929 - 1941.

Zuber, B., Marchand, J. (2004), Predicting the volume instability of hydrated cement systems upon freezing using poro-mechanics and local phase equilibria, Materials and Structures/Concrete Science & Engineering, Vol. 37, No. 268, p. 257 - 270.

Østmoen, et al., (1993), Chloride durability of Norwegian coastal bridges, Norw. Publ. Roads Administration 154 p. (In Norwegian).

Effect of Interacted Deterioration Parameters on Service ... · PDF fileRESEARCH REPORT...

Documents

Transcript of Effect of Interacted Deterioration Parameters on Service ... · PDF fileRESEARCH REPORT...