Vous êtes sur la page 1sur 6

10DBMC International Confrence On Durability of Building Materials and Components

LYON [France] 17-20 April 2005

Durability vs. Reliability of RC Structures


B. Tepl, Z. Kerner, P. Rovnank, M. Chrom
BUT Brno, Faculty of Civil Engineering
17 ikova, CZ-60200 Brno, Czech Republic
teply.b@fce.vutbr.cz
TT4-42

ABSTRACT
The failure criteria of Serviceability Limit States are linked to design service life. Moreover, different
levels of reliability should be adopted for different types of Limit States. The choice of level of
reliability for a particular structure and material should take account of the relevant factors, including
possible consequences of failure and potential costs of safety measures. This principle is not fully
accommodated in current codes.
In the context of durability of reinforced concrete structures, the European codes EN 206-1 and
EN1992-1-1 together with EN1990 provide recommendations for different exposition and structural
classes as to the minimum content of cement, maximum water/cement ratio, and, optionally,
minimum concrete strength class. Limiting concrete cover is required simultaneously. The relevance
of these requirements and associated values of the index of reliability should be ensured. By utilizing
analytical models for carbonation progress in concrete and taking account of the uncertainties of
concrete mixture, carbon dioxide concentration and other input data, a statistical description of
carbonation progress is gained. Random or deterministic input parameters are involved. Statistical
analysis is performed allowing for calculation of the reliability index relevant to different service lifes.
The time passing before the initiation of rebar corrosion is considered as the limiting condition.
The results of the investigation of the reliability index profiles and its trends concerning different
conditions/classes are provided utilizing an interactive web-site covering service life, concrete cover
and reliability index assessment. The non-uniformity of reliability level is shown.
Some conclusions (and generated questions) are discussed.
KEYWORDS
Concrete structures, durability, carbonation, reliability index, Eurocodes

10DBMC International Confrence on Durability of Building Materials and Components


LYON [France] 17-20 April 2005
1 INTRODUCTION
Design for durability is coming considerably into the focus of researchers and recently, designers too.
This has been clearly demonstrated at several international and/or local conferences, and by numerous
papers (let us mention e.g. [Siemes 2002]). ISO activity (TC98) is currently following this trend by
working to produce a new document, ISO/NP 13823 General principles in the design of structures
for durability. Also, Integrated Design and its subsiduary, Performance-Based Design (PBD), are
leading trends in structural engineering design; these approaches deal with durability and reliability
issues, which rank amongst the most decisive structural performance characteristics.
Unfortunately, the prescriptive approach of current standards (Eurocodes) does not allow simply for
design focused on specific service life and/or specific level of reliability. Such tasks necessarily
require the utilization of stochastic approaches, analytical models and also simulation techniques. The
theoretical apparatus of this approach has been developed already but the practising engineer is usually
not equipped with the appropriate knowledge, routines and instruments or software. In order to
partially solve this problem the authors have recently introduced [Tepl et al. 2004] a simple tool for
the designing process for concrete structures taking account of the considerations of durability and
reliability thus attempting to furnish the designer with a user friendly instrument for dealing with
such problems. It is an interactive web page called RC_LifeTime, freely accessible on http://rclifetime.stm.fce.vutbr.cz/.
The goal of the present paper is to assess the feasibility of modern codes of practice to design
reinforced concrete structures for durability and also to demonstrate the features of the RC_LifeTime
web page.
2 DESIGN FOR DURABILITY OF RC STRUCTURES
2.1 Eurocodes
The service life of a building or structure is determined by its design, construction, ageing and
maintenance during use. The combined effect of structural performance and ageing should be
considered.
The modern codes (Eurocodes) generally do not allow for a design subjected to a specific (target)
service life. Some exceptions concern RC structures: combining tables 2.1 with tables 4.3N and 4.4N
[EN1992-1-1] the value of nominal concrete cover can be determined with relevance to structural class
(S1-S6) and exposition class (e.g. XC1XC4 in cases of concrete carbonation). The indicative strength
class is given too. Moreover, considering tables 1 and F.1 [EN206-1] the requirements of maximum
water-cement ratio and minimum content of cement (together with minimum concrete strength class
as an additional specification) are recommended in accordance with exposition classess referring to the
use of CEM I and an intended working life of 50 years only! Immediately some questions appear: the
direct dependance of strength class on water and cement content is not a unique one, and, the relation
of concrete durability to strength is not generally to be taken as granted. The increasing proportion of
cement may lead to concretes being more crack-prone. Also the type and the grain size of used
cement, supplementary cementing materials and aggregatte may have a strong influence see e.g.
[Mehta 1997]. It should be noted also that the above-mentioned structural recommendations of
EN1992-1-1 and EN206-1 do not take into account the indicative design working life categories see
the basic structural Eurocode EN1990, table 2.1 and the reliability requirement!
Structural design based on modern building codes deals with limit states (both ultimate and
serviceability ULS and SLS). The level of reliability in the context of durability is not stated there.
When considering the degradation of reinforced concrete structures, the corrosion of reinforcement is
the dominating effect. In the context of corrosion the following limit states can be recognized: (i)
TT4-042, Durability vs. Reliability of RC Structures, B. Tepl, Z. Kerner, P. Rovnank, M. Chrom

10DBMC International Confrence on Durability of Building Materials and Components


LYON [France] 17-20 April 2005
depassivation of reinforcement due to carbonation or chloride ion penetration; (ii) cracking; (iii)
spalling of concrete cover; (iv) decrease in the effective reinforcement area (leading to excessive
deformation and finally to collapse). Types (i)-(ii) are in the SLS category of limit states, (iii) might
fall in both the ULS and SLS categories depending on the location and grade of degradation, and, (iv)
is in the ULS (load bearing capacity) or SLS (deformation capacity) categories. As stated in the basic
design code [EN1990], the recommended value of the reliability index for SLS (irreversible state) is
= 1.5, which is again relevant to a 50-year service life.
2.2 RC_LifeTime
Rather often, case (i), i.e. the depassivation of reinforcing steel due to carbonation (pH 9.5), is
considered conservatively as a limiting condition. This condition is considered in RC_LifeTime as:
the time to depassivation = initiation period irreversible serviceability limit state. In other words,
at this stage the reinforcement is no longer secure against corrosion. Under the condition of the
presence of a certain level of moisture in concrete (necessary for cathodic reaction), hydroxide ions
form corrosion products.
RC_LifeTime performs statistical analyses and offers the following options:
(i)
Service Life Assessment provides an evaluation of service life based on the equality of
carbonation depth and concrete cover
(1)
xc = c
The assessment of xc is based optionally on one of the two models shown in [Papadakis et al.
1992 (2a,b)]; both models are enhanced within RC_LifeTime by the humidity function
extracted from [Matouek 1977] see also [Kerner et al. 2004]. The input data are the values
of concrete cover c together with 12 model variables (or 5 according to the chosen model;
optionally deterministic or random variables). The statistical characteristics of relevant service
life (mean and standard deviation) are the output data. The target value of the reliability index
may be the additional input value describing the probability of reaching condition (1)
and then the corresponding service life is the output value.
(ii)
Concrete Cover Assessment provides an evaluation of concrete cover appropriate to equality
(1). The input data are the value of the target service life (as a deterministic value) together
with model variables (again, deterministic or random). Statistical characteristics of relevant
concrete cover (mean and standard deviation) are output data. The value of required concrete
cover c may be inputted too and the relevant reliability index is then an output value
(describing the reliability of reinforcement depassivation).
It has to be noted: (a) both models are capable of describing the carbonation of concrete made from
CEM I only, while other models are the focus of ongoing work; (b) the values of should depend also
on the consequences of failure and on the relative cost of safe design; (c) the condition (1) is assessed
for the structure/member in general, not considering the issue of a critical cross-section and the
degree of statical redundancy; (d) within RC_LifeTime the reliability index is of the Cornells type.
In the view of comments (b) and (c) the reliability level given by = 1.5 seems to be too severe as
it is relevant for limit states described by e.g. excessive deflections or excessive crack width as well.
This is why ISO 2394, table E.2, shows SLS values from = 0 (for reversible) up to = 1.5 (for
irreversible states).
3 RELIABILITY STUDY
Utilizing RC_LifeTime and following the recommendations of the Eurocodes described above,
parametric studies have been performed assessing the reliability index values for exposition clases and
structural classes. The corresponding input data and statistical characteristics are listed in tables 1 and
2. Comments to Tab. 1: the mean values of cover are shown, the two-parametric lognormal probability
density function is assumed (origin in zero point) with COV = 20%. In this case slab geometry COV
TT4-042, Durability vs. Reliability of RC Structures, B. Tepl, Z. Kerner, P. Rovnank, M. Chrom

10DBMC International Confrence on Durability of Building Materials and Components


LYON [France] 17-20 April 2005
= 5% only. Table 2: the mean values of cement content are the minima and water-cement ratios the
maxima recommended by EN 206-1; all probability distributions are normal.
Exposure
class
XC1
XC2/XC3
XC4

Exposure
class
XC1
XC2
XC3
XC4

Reliability index [-]

Structural class
S1
S2
S3
S4
S5
20
20
20
25
30
20
25
30
35
40
25
30
35
40
45
Table 1. Nominal concrete cover in mm recommended by EN1992-1-1
for exposure and structural classes.
Water/
Water
Cement content
3
Cement
[kg/m3]
[kg/m ]
ratio
Mean
Standard
Standard
[]
value
deviation
deviation
260
7
0.65
4
280
8
0.60
4
280
8
0.55
4
300
9
0.50
3
Table 2. Random input variables

XC1

XC2

XC3

S6
35
45
50

Relative Humidity
[%]
Mean
Standard
value
deviation
55
10
85
5
75
10
65
10

XC4

5
4
3
2
1
0
-1

50

100

slab

Figure 1. Reliability index for exposure classes


Fig. 1 depicts the reliability index for exposure classes and three cases with reference to EN1992-1-1:
case 50 is the basic situation, i.e. the structural class S4, with a design working life of 50 years;
100 has a design working life of 100 years; slab is relevant to members with slab geometry (the
position of reinforcement is not affected by the construction process or members where special quality
control of concrete has been ensured see table 4.3N of EN1992-1-1). Fig. 2 shows reliability profiles
together with the reliability level of = 1.5 demanded by EN 1990. All these results clearly document
the remarkable non-uniformity of the reliability level hidden in the Eurocodes.

TT4-042, Durability vs. Reliability of RC Structures, B. Tepl, Z. Kerner, P. Rovnank, M. Chrom

10DBMC International Confrence on Durability of Building Materials and Components


LYON [France] 17-20 April 2005
XC1
5

3
2
1
0

5
4

-1

3
2
1
0
-1

-2

-2
20

40
60
Time [years]

80

100
S1
S2
S3
S4
S5
S6
EN 1990

XC3
5
4
3
2
1
0

100

80

100

2
S1
S2
S3
S4
S5
S6
EN 1990

1
0

-2
80

100

-2
40
60
Time [years]

80

-1

20

40
60
Time [years]
XC4

-1

20

Reliability index [-]

Reliability index [-]

S1
S2
S3
S4
S5
S6
EN 1990

XC2

Reliability index [-]

Reliability index [-]

S1, S2, S3
S4
S5
S6
EN 1990

20

40
60
Time [years]

Figure 2. Reliability index vs. service life


4 CONCLUSIONS QUESTIONS
Carbonation is a particularly important form of deterioration and its impact on existing or newly
designed structures is evident. Its vast variability due to variability in concrete quality and
environmental changes is remarkable and should be taken into account. It could be assessed by
RC_LifeTime which serves as a simple tool for the Performance-Based approach of reinforced
concrete structure design. However, it is based on a most conservative limit state.
The authors believe that apart from the verification shown in [Kerner et al. 2004], some additional
research concerning the current model and especially other models for concretes from blended
cements is necessary.
The results of the approach and the study shown above generate some questions:
?1 . how should relevant durability limit states of RC structures be defined?
?2 . what should the appropriate values of the reliability index associated with different durability
limit states of RC structures be?
?3 . is it necessary to make required concrete cover thickness and required service life compatible
with a certain balanced level of reliability?

TT4-042, Durability vs. Reliability of RC Structures, B. Tepl, Z. Kerner, P. Rovnank, M. Chrom

10DBMC International Confrence on Durability of Building Materials and Components


LYON [France] 17-20 April 2005
Definitely, these questions require further study and discussion. The authors also believe that design
for durability (and for a specific target service life) needs a probabilistic approach and
appropriate/individual reliability assessments.
5 ACKNOWLEDGMENTS
This paper was produced with the financial support of the project 103/03/1350 backed by the Grant
Agency of the Czech Republic.
6 REFERENCES
EN 1990:2002 Basis of Structural Design (European Standard)
EN 206-1:2000 Concrete Part 1: Specification, performance, production and conformity (European
Standard)
EN 1992-1-1:2003 (E) Eurocode 2: Design of concrete structures Part 1.1: General rules and rules
for buildings
Gehlen, Ch. 2000, Probabilistishe Lebensdauerbemessung von Stahlbeton bauwerken, Deutsher
Ausschuss fuer Stahlbeton, Heft 510, Berlin.
Kerner, Z., Rovnankov, P., Tepl, B. & Novk, D. 2004, Design for durability: An interactive tool
for RC structures, Proc. Life cycle assessment, behaviour and properties of concrete structures
LC 2004, Brno, Czech Republic, 9 11 November 2004, (in print).
Maage, M. & Smeplass, S. 2001, Carbonation A probabilistic approach to derive provisions for
EN 206-1 DuraNet workshop, Tromso, Norway, June 2001.
Matouek, M. 1977 Effects of some environmental factors on structures. PhD thesis, Technical
University of Brno, Brno (in Czech).
Mehta, P.K. Durability Critical Issues for the Future, Concrete International, July 1997, 27-33.
Papadakis, V. G., Fardis, M. N. & Vayenas, C. G. 1992, Effect of Composition, Environmental
Factors and Cement-lime Mortar Coating on Concrete Carbonation, Materials and Structures,
1992, 25[149], 293304.
Siemes, T. 2002, Overview Of The Service Life And Maintenance Problem Probabilistic Design,
Proc. 9DBMC (CD ROM), Brisbane, paper No. 260, 10 pages.
Tepl, B., Krlov, H. & Stewart, M. G. 2003, Ambient Carbon Dioxide, Carbonation and
Deterioration of RC Structures, International Journal of Materials & Structural Reliability, 1,
3136.
Tepl, B., Rovnank, P., Kerner, Z. & Rovnankov P. 2004, Support to durability design of RC
structures, Beton TKS, 3/2004, 3840 (in Czech).

TT4-042, Durability vs. Reliability of RC Structures, B. Tepl, Z. Kerner, P. Rovnank, M. Chrom

Vous aimerez peut-être aussi