Section Durability And Section Durability And Concrete Cover Cover To Reinforcement

C4 The rules on design for durability in EC2 are substantially different than in the past. Previously the concrete cover was prescribed in dependence of the environmental class, but independent of the concrete quality. In the actual version of EC2 (EN 1992-1-1) the cover required depends not only on the environmental class, but as well on the concrete strength class, the required design working life and the quality control applied. In the following, background information is given with regard to those choices. The values for the cover in EN 1992-1-1 are a result of increased understanding in the processes of deterioration, as revealed by the Duracrete studies [1], and practical experience. In the following theoretical considerations the most important deterioration processes are explained and parameter studies illustrate the mutual dependencies. Further background information is found in FIB Bulletin n. 34 "Model code for service life design" [2].

4.1.1 Introduction

In a European research project [1] a probabilistic based durability design procedure of concrete structures has been developed with the objective to set up a similar concept as in structural design, where the resistance of the structure is compared to the acting load. Related to corrosion protection of reinforcement the resistance of a concrete structure is mainly determined by the thickness and the quality of the concrete cover.

In the European design code for concrete structures EN 1992-1-1 the designer has to determine the nominal concrete cover, which consists of the minimum concrete cover (dependent on the relevant environmental class) plus an allowance in design for tolerance. According to the first draft of prEN 1992-1 the allowance in design for tolerance was also dependent on the environmental class - 10 mm for XC0 and XC1 and 15 mm for all other classes. This rule was changed with respect to the general requirement that values or rules specified in other Eurocodes should not explicitly given again in eN 1992-1-1 but they should only be referred to. Thereafter in the December 99 draft of prEN 1992-1 the allowance in design for tolerance was determined with respect to the execution standard prENV 13670 where the execution tolerance is uniformly defined to 10 mm. This means a reduction of the nominal cover if the values for the minimum cover are not increased accordingly.

In order to find out which value for the nominal concrete cover is adequate a durability design was performed, based in the model in the European project. In particular, the reliability index p was determined and evaluated for different concrete mixes and different nominal concrete covers. The concrete mixes have been chosen with respect to the environmental classes given in EN 206-1.

4.1.2 Description of Deterioration Models and Probabilistic Durability Design

This chapter only gives a short overview on those models, since the deterioration models related to reinforcement corrosion and the safety concept of durability design have already been thoroughly described in the literature.

4.1.2.1 Deterioration Models

The corrosion process can be divided into two time periods: the initiation period describes the time until the reinforcement is depassivated either by carbonation or by penetrating chlorides reaching a critical chloride content.

After depassivation, corrosion will start if sufficient oxygen and moisture are available. As a result of corrosion a reduction of the steel cross section, cracking or spalling of the concrete cover will occur. This time period is described as the propagation period.

In the literature deterioration models [2], by which the processes of the initiation period can be described, are well established. The process of the propagation period is much more complex and so far no unanimously accepted models exist.

In the following the time-dependent description of the carbonation progress and the time dependent diffusion-controlled penetration of chlorides are briefly presented.

4.1.2.1.1 Carbonation-Induced Corrosion

The CEB Task Group V model by which the carbonation process in the initiation period can be predicted is given in equation (4.1):

where:

xc(t) is the carbonation depth at time t

DEff,0 effective diffusion coefficient of dry concrete for carbon dioxide in defined environment (20°C, 65% rel. humidity)

a the amount of CO2 for complete carbonation [kgCo2/m3]

AC the concentration difference of CO2 at the carbonation front and in the air, which usually means the carbon dioxide content of the surrounding air c0

ke parameter for micro climatic conditions, describing the mean moisture content of concrete kc parameter to describe the curing conditions w parameter (exponent) for micro climatic conditions at the concrete surface, describing wetting and drying t0 reference period,

4.1.2.1.2 Chloride-Induced Corrosion

The model for predicting the initiation period in the case of chloride-induced reinforcement corrosion is defined by equation (4.2):

where:

x(t) depth with a corresponding chloride content (here C(crit)) at time (t)

D0 effective chloride diffusion coefficient under defined compaction, curing and environmental conditions, measured at time t0 Drcm,0 chloride migration coefficient under defined compaction, curing and environmental conditions, measured at time to

Ccrit chloride threshold level n factor which takes the influence of age on material property into account kt constant which transforms the measured chloride migration coefficient Drcm,0

into a chloride diffusion coefficient D0

ke constant which considers the influence of environment on D0

erf-i inverse of the error function

CSN surface chloride level t time to reference period (28 days)

4.1.2.2 Probabilistic Durability Design 4.1.2.2.1 Safety Concept

The simplest design problems have only one resistance variable and one action variable. They are generally solved by facing the two variables R and S:

where Z is the reliability of the structure, R the resistance and S the action: both variables R and S have their averages and standard deviations, and in this example they are normally distributed. Z is a variable itself (see Figure 4.i) and is also normally distributed with a mean ^z and a standard deviation az according to equations (4.6) and (4.7).

0 0

Post a comment