Available online at www.sciencedirect.

com

ScienceDirect
Procedia Engineering 193 (2017) 144 – 151

International Conference on Analytical Models and New Concepts in Concrete and Masonry
Structures AMCM’2017

The application of reliability analysis in engineering practice –

reinforced concrete foundation
Izabela Skrzypczaka*, Marta S owikb, Lidia Buda-O óga,
a
Poland Rzeszow University of Technology, 2 Pozna ska, Rzeszów 35-595, Poland
b
Lublin University of Technology, 40 Nadbystrzycka,Lublin 20-618, Poland

Abstract

In recent years, the importance of assessment of structural reliability has increased significantly. This is confirmed by the
recommendations of the standard PN-EN 1990 in which the rules and requirements to ensure safety, serviceability and durability
of the structure are specified. It also sets out the basis for calculation and verification of constructions and provides guidance to
ensure their reliability. Reliability design focuses on the ability to meet specific design requirements, taking into account the
planned period of use. The concept of the planned period of use should be considered as the adopted in the project interval in
which the structure or a part of the structure is to be used for its intended purpose without the need for general repairs. Typically,
reliability is expressed in probabilistic metrics - using an index of reliability or probability of failure. Reliability of building
structures depends on a number of correlated factors, mainly on the quality of materials, building precision and level of control,
protection against environmental influences and maintenance level during exploitation, specific period of use, adopted solutions
for the construction materials, design details and technologies, adopted loads (both their values and combinations), standard
requirements regarding capacity, exploitation and durability, quality of computational models used in the design process and
methods for assessing reliability of the structure. The performed reliability analysis concerns a reinforced concrete foundation,
for which reliability index and probability of failure has been specified using the following methods: analytical method FORM,
simulation methods FORM and SORM as well as Monte Carlo simulation.
©2017
© 2017TheTheAuthors.
Authors. Published
Published by Elsevier
by Elsevier Ltd.is an open access article under the CC BY-NC-ND license
Ltd. This
(http://creativecommons.org/licenses/by-nc-nd/4.0/).
Peer-review under responsibility of the scientific committee of the International Conference on Analytical Models and New
Peer-review
Concepts inunder responsibility
Concrete of the scientific
and Masonry committee of the International Conference on Analytical Models and New Concepts in
Structures.
Concrete and Masonry Structures
Keywords: reinforced concrete foundation; reliability index; FORM; SORM; Monte Carlo simulation;

* Corresponding author. Tel.: +48-17- 8651306

E-mail address: izas@prz.edu.pl

1877-7058 © 2017 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license
(http://creativecommons.org/licenses/by-nc-nd/4.0/).
Peer-review under responsibility of the scientific committee of the International Conference on Analytical Models and New Concepts in Concrete and
Masonry Structures
doi:10.1016/j.proeng.2017.06.197
 Izabela Skrzypczak et al. / Procedia Engineering 193 (2017) 144 – 151 145

1. Structural assessment and reliability analysis

The assessment of existing structures aims at producing evidence that they will function safely over a specified
residual service life. It is mainly based on estimating the material properties and strength capacity of structural
members taking into account the present state of the structure, and evaluating its ability to withstand anticipated
hazards and future loads. Nowadays, this problem is particularly important in the case of structures and primarily
infrastructure. In fact, the rate and extent of the deterioration of existing structures have lately significantly
increased. Indeed, the current low funding in the building sector of several European countries has forced agencies to
postpone necessary investments in a new load and consequently stretch the service life of their existing old stock.
The prioritization of the distribution of funds among maintenance, repair and rehabilitation activities is a major
problem everywhere [1,2,3].
In the assessment and designing of structures, the probabilistic methods can be used widely. The aim of structural
design is to realize structures that meet the expected performance, which can be often represented by a target
reliability level [4,5,6,7,8]. There are different approaches for reliability verification: deterministic, semi-
probabilistic, and probabilistic.
The most common deterministic safety measure is the global factor of safety, defined as the ratio of the resistance
over the load effect. The concept of the allowable stresses is a traditional deterministic method, where failure of the
structure is assumed to occur when any stressed part of it reaches the permissible stress. Deterministic verification
methods which are based on a single global safety factor do not properly account for the uncertainties associated
with strength and load evaluation. The semi-probabilistic approach is based on the limit state principle and makes
use of partial safety factors for checking the structural safety. The partial factors have been calibrated so that
a structure which satisfies the safety check using a set of design parameters will also satisfy the target reliability
level. The semi-probabilistic verification method is still a simplified method but it can much better account for the
uncertainties of some design parameters. Probabilistic verification procedures are also based on the principle of limit
states, by checking that predefined target structural reliability levels are not exceeded. This approach takes into
account explicitly the uncertainties.
Regardless of the uncertainties in different parameters accounting for the analysis and design of a structure, it is
very difficult to measure its absolute safety using deterministic analysis. Therefore, one of the most important ways
to specify a rational criterion for ensuring the safety of a structure is evaluating its reliability or probability of failure.
The reliability of a structure is defined as its ability to fulfil the design purpose for some specified design lifetime [9,
10]. Reliability is often understood as the probability of not losing the intended function of the structure. The term
failure of a structure does not necessarily mean catastrophic failure but it is used to indicate that the structure does
not perform as desired. In structural reliability calculations, the probability of failure is taken as quantitative measure
of a structural safety.
Most of the modern codes for constructions have recognized the need of using advanced reliability based design
methods that allow taking into account various sources of uncertainty. To verify whether or not a structural design is
acceptable, the uncertainties are modelled by using statistical tools and the failure probability is estimated with
respect to all relevant limit states. The three main documents, that have been drawn on reliability based design, are
the standard ISO 2394 [11], the probabilistic model code developed by the Joint Committee on Structural Safety
(JCSS) [12] and the structural Eurocode [13].
The code ISO 2394 [11] is an important international standard that specifies general principles for the verification
of the reliability of structures subjected to different types of actions. Reliability is considered in relation to the
performance of the structure throughout its design working life. This international standard is applicable in all the
stages of the construction process as well as during the use of the structure, including maintenance and repair,
applicable to the structural appraisal of existing constructions or assessing changes of use.
The probabilistic model code developed by the Joint Committee on Structural Safety [12] represents an important
step in the direction of the necessary standardization of the reliability based method. In 1971, the Liaison
Committee, which coordinates the activities of six international associations of Civil Engineering (FIB, CIB, ECCS,
IABSE, IASS, and RILEM), created the Joint Committee on Structural Safety with the aim of improving the general
knowledge in structural safety. In 1992, the JCSS set as a long term goal the development of a probabilistic model
code for new and existing structures. The JCSS code gives guidance on the modelling of the random variables in
146 Izabela Skrzypczak et al. / Procedia Engineering 193 (2017) 144 – 151

structural engineering and it is intended as the operational part of codes like the ISO 2394 [11], the Eurocodes and
other national codes that allow for probabilistic design but do not give any detailed guidance. The code consists of
three main parts and it deals with general requirements, modelling of loads and modelling of structural properties.
The code gives no information, however, on mechanical models like buckling, shear capacity, foundation failure and
so on. A little or no information is given on other modelling aspects, like for example the wind pressure coefficients
[1].
According to JCSS, reliability method can be classified into three levels. The level 1 summarizes all semi-
probabilistic method like the semi-probabilistic safety concept. The concept lays in the methodological basis for
standard and codes in the field of structural engineering. It uses partial safety factors and characteristic values of
materials and actions to guarantee the required level of reliability. No distribution functions are used and linear limit
state functions are supposed in the level 1. Methods of the level 2 are defined by reliability theories of first and
second order (FORM/SORM). The methods are used for code calibration in the level 1. The level 3 includes
probabilistic methods like numerical integration, stochastic simulation, Monte Carlo and others. The approaches are
used to calibrate the models of level 1 and level 2. Methods of level 2 and level 3 can be used for the direct analyses
of structures [13] but they are rarely applied because of their complexity. The application of the methods makes
sense if the analyses cannot be provided by conventional methods (especially for existing structures), or if the
existing safety margins should be determined more accurately.
Probabilistic structural analysis can be defined as the art of a mathematical model formulating, within which one
can ask and get an answer to the question what is the probability, that the structure behaves in a specified way, given
that one or more of its material properties are random or incompletely known nature and/or that the action on the
structures in some respect have random or incomplete know properties.
Modern engineering design involves two steps: the theory of structures (in order to determinate the way in which
a structure actually carries its loads) and the strength of materials (in order to assess whether the structural response
can safely withstood by material). In practice, the two steps cannot usually be separated and design has to be
iterative.
The JCSS provides a sound basis for decision support for risk and safety management at strategic, normative and
operational levels and can be considered as the best practice for a risk based decision in engineering. The structural
design is an integral part of it.
The basic requirements in structural design rely on such a way of structures designing, constructing and
maintaining that the structures are suited for their use during the design working life and in an economic way.
In particular, they should fulfil the following requirements with appropriate levels of reliability:
• They should remain fit for the use for which they are required (serviceability limit state);
• They should withstand extreme and/or frequently repeated actions occurring during their construction and
anticipated use (ultimate limit state requirement);
• They should not be damaged by accidental events like fire, explosion, impact or consequences of human
errors, to an extent disproportionate to the triggering event.
The expression ‘with appropriate levels of reliability’ used above means that the degree of reliability should be
adopted to suit the use of the structure, the type of structure or structural element and the situation considered in the
design. The choice of the various levels of reliability should take into account the possible consequences of failure in
terms of risk to life or injury, the potential economic losses and the degree of social inconvenience.
In civil engineering, reliability in the generic term for items like safety, vulnerability, robustness as well as
bearing capacity and serviceability of the construction depends on a design process. So, the structural design is an
integral part of risk assessment/analysis of the structure. It affects the structural design as the boundaries of
application are difficult to compare and to integrate [14,15].
As far as existing buildings are concerned, changes in usage and a related stress increase during their lifetime can
result in a reassessment of stability and structure requirements. In most cases, the assessment of the load-bearing
capacity of the existing building is provided according to current standards which have been developed for designing
new buildings. Therefore the codes can be applied to existing buildings with restrictions and modifications. The
static analysis of existing buildings in many cases is not feasible strictly applying the current regulations, since the
specialties of existing structures are not considered in the codes, for example the partial factors where calibrated
 Izabela Skrzypczak et al. / Procedia Engineering 193 (2017) 144 – 151 147

following the uncertainties for new structures. A lot of information is available for existing structures, which can be
useful in the structural analysis. If as-built stocktaking is carried out before starting off with the planning building
measures, reliable information on the type of operation and structural attributes can usually be obtained, decreasing
a lot of uncertainties which are covered by the partial safety factors given by the codes. Thus a reduction in the
partial safety factors for the recalculation of existing structures can be allowed. As mentioned before, the current
codes are based on the semi probabilistic safety concept using partial safety factors and characteristic values of
materials and actions to guarantee the required safety level according to EN 1990 [13]. Therefore the partial safety
factors were determined for building constructions with common sizes and subjected to conventional actions. For the
variation of these random variables general assumptions were made. Special conditions like very thick foundation
with a thickness of a few maters were not considered in the calibration. Thus the partial safety factor for permanent
load of G = 1.35 which considers the variation of the weight and especially the geometry leads to a conservative
design in these cases. Hence, an adaption of safety factors on those special boundary conditions is sometimes useful.
Such a task was described in [14,15] where reduction of the safety factor for permanent load was suggested. In the
recording of an existing structure data like the material strength, the dimensions and spans as well as the size and
location of reinforcement can be measured. The uncertainties can be decreased in comparison to the phase the
building was designed. This benefit can be used in a probabilistic structural analysis. But keeping in mind that
example, uncertainties can also increase due to lost information or uncertainties regarding the assessment techniques.
Especially for concrete an enormous benefit can be taken by as-built stocktaking for concrete compressive or tensile
strength. When cores are taken at different places of the existing structure, the concrete strength can be determined
using statistical methods [13]. However, it is worth to mention the importance that these cores are taken at different
parts of the structure like for example at different points of foundation or over the height of columns. If one concrete
strength class is used in the whole building, cores of different components should be taken, too. To achieve useful
results of the concrete strength, a sufficient number of cores is necessary – different recommendations point out that
minimum 6 cores have to be provided [16, 17]. Thus a possibly hardening of concrete over the years for the recent
structure can be determined and used to allow higher utilization of the structural elements in static analyses. To
avoid damage to the structure, samples of reinforcement should be taken only in special cases due to small variations
of the steel properties. If the steel grade is known, the variation of the tensile strength of the reinforcement in
combination including the variation of the cross sectional area is less than 5%.

2. Case study of the reliability of reinforced concrete foundation

Due to the presence of uncertainties in various parameters accounting for the analysis and design, it is very
difficult to measure safety for a structure from deterministic analysis. The main objective of this research is to
evaluate the safety in terms of reliability of reinforcement foundation design using EN 1990 [13]. A foundation as a
structural member can fail due to punching shear, flexural shear or flexural moment (structural failure), or exciding
bearing capacity and excessive settlement of foundation (geotechnical failure). If one of the types of failure occurs,
the foundation fails. A damage of foundation leads to the collapse of the structure. The failure probability or the
reliability of foundation depends on a failure mode. Therefore, it is essential to determine the reliability or margin of
safety and corresponding failure probability of foundation taking all the probabilities of the foundation failure into
considerations.
Probability of failure for different levels was assessed for the designed foundation. The flexural capacity has been
analyzed. Due to the high portion of dead load at the total load, a probabilistic approach seems to be promising. The
foundation is 1.8 m high, 9 m wide and 18 m long [14], see Fig. 1. Variations due to the soil properties in this
example do not influence to results significantly and they have not been considered at the model uncertainty of the
stress results.
As the result the probability of failure of the most utilized section has been derived and it was compared with the
limit value Pf,lim = 10-6 according to EN 1990 [13]. A nonlinear and plastic calculations were not conducted, thus the
result provided an upper limit value of the probability of failure. A time dependent analysis was conducted due to
the lack of information of the required stochastic parameters of the loads. Apart from a live load on the foundation
and a dead load also a wind load had to be regarded.
148 Izabela Skrzypczak et al. / Procedia Engineering 193 (2017) 144 – 151

Fig. 1. Scheme of the reinforced concrete foundation

The analyzed limit state was defined as the failure of the reinforcement at the bottom of the foundation caused by
bending. Failure of the concrete was not relevant in this example. The limit state function Z is defined as (1):

Z=R-E (1)

In this case, the limit state function has the form (2):

(2)

Taking into account the uncertainty of resistance model and effects model, the limit state function has the form (3):

(3)

The input basic random variables are shown in the Table 1. The complete results of calculations by different
methods are shown in Figures 2 to 4 and in the Table 2.

Table 1. Basic random variables, according to [14]

Variable Distribution Mean value Standard deviation
R [-] N 1.0 0.10
2
fy [MN/m ] LN 500 35
2
As [m ] det 0.0061 –
h [m] N 1.80 0.02
d1 [m] N 0.09 0.01
fc [MN/m2] LN 38 4.86
E [-] LN 1.0 0.05
MG [MNm] N 1.947 0.077
MQ [MNm] Gum 0.474 0.190
MW [MNm] Gum 0.075 0.075
 Izabela Skrzypczak et al. / Procedia Engineering 193 (2017) 144 – 151 149

compressive strength steel

high of foundation effective high of foundation

dead load
live load

wind

Fig. 2. Histogram and distribution for basic variables

Fig. 3. Histogram for the limit state function

150 Izabela Skrzypczak et al. / Procedia Engineering 193 (2017) 144 – 151

Fig. 4. Histogram and normal distribution for the limit state function

Table 2. The comparison of results calculated by different methods

Method/Sensitivity
Variable FORM FORM SORM Monte Carlo
Analytical method Simulation method Simulation method Simulation
R – – 0.480 0 380
fc – - 0.013 0.018 0.047
fy - 0.7298 - 00307 0.389 0.970
As 0 0 0 0
h 0 - 0.064 0.089 0.181
d1 0 - 0.032 - 0.068 -0.108
E – – -0.273 -0.159
MG 0.6837 0.078 - 0.054 0.008
MQ – 0.941 - 0.567 0.008
MW – 0.083 - 0.078 - 0.033
– Index of reliability 5.4363 5.966 6.637 8.145
Pf – Probability of failure 2.72 x10-8 1.215 x10-9 1.611 x10-11 1.890 x10-16
Pf lim – Limit value of failure probability 10-6

3. Conclusions

In recent years, the importance of assessment of structural reliability has increased significantly. This fact is
confirmed by the recommendations of the standard EN 1990 [13]. The standard specifies rules and requirements to
ensure safety, serviceability and durability of the structure. According to the standards recommendations, reliability
of the reinforced concrete foundation was calculated. The received value of the failure probability calculated using
reliability methods was smaller than the limit value of Pf,lim = 10-6 according to EN 1990. Using the probabilistic
approach, it was possible to verify a sufficient safety of the foundation although the live load for the foundation was
raised up. Without the probabilistic design, a limitation of the live load would have been required.
The reliability indices for most geotechnical components and systems lie between 1.0 and 5.0, corresponding to
probabilities of failure ranging from about 0.16 to 3x10-7, as shown in the Table 3 [18,19].

Table 3. The range of geotechnical reliability index [18,19].

Reliability index, Probability of failure, Pf Expected performance reliability level
1.0 0.16 Hazardous
1.5 0.07 Unsatisfactory
2.0 0.023 Poor
2.5 0.006 Below average
3.0 0.001 Above average
4.0 0.00003 Good
5.0 0.0000003 High
 Izabela Skrzypczak et al. / Procedia Engineering 193 (2017) 144 – 151 151

The received values of probability of failure and reliability index from the carried out analyzes have shown, that
the expected performance reliability level for the analyzed foundation is high (Fig. 5).

Fig. 5. Relationship between reliability index and probability of failure Pf , and received values of probability of failure and reliability index

References

[1] S. Arangio, Reliability based approach for structural design and assessment: performance criteria and indicators in current European codes
and guidelines, Int. J. Lifecycle Performance Engineering. 1 (1) (2001) 64-91.
[2] J.R. Casas, Bridge management: actual and future trends, in Bridge Management, Life Cycle Performance and Cost, Taylor and Frances,
London, 2006, pp.21–30.
[3] D. Frangopol, P. Das, Management of bridge stocks based on future reliability and maintenance cost, Current and Future Trends in
Bridge Design, Construction and Management, pp.45–58, Thomas Thelford, 1999.
[4] M. Gwó d , A. Machowski, Wybrane badania i obliczenia konstrukcji metodami probabilistycznymi, Wydawnictwo Politechniki
Krakowskiej, Kraków 2011.
[5] J. Murzewski, Niezawodno konstrukcji in ynierskich, Arkady, Warszawa, 1989.
[6] J. Pawlikowski, Podstawy projektowania probabilistycznego konstrukcji z betonu, Prace naukowe ITB, WITB, Warszawa 2004.
[7] Z. Sakka, I. Assakkaf, T. Al-Yaqoub, J. Parol, Structural Reliability of Existing Structures: A Case Study, International Journal of Civil,
Environmental, Structural, Construction and Architectural Engineering. 8(11) (2014) 1173-1179.
[8] J. Schneider, Introduction to safety and reliability of structures’, IABSE – AIPC – IVBH, 1997.
[9] A.S. Nowak, K.R. Collins, Reliability of Structures, McGraw-Hill Higher Education, 2000.
[10] Sz. Woli ski, K. Wróbel, Niezawodno konstrukcji budowlanych, Oficyna Wydawnicza Politechniki Rzeszowskiej, Rzeszów 2000.
[11] ISO 2394:2015 General principles on reliability for structures. 2015 ISO.
[12] JCSS, Probabilistic Model Code, Part I - Basis of design, http://www.jcss.byg.dtu.dk/Publications/Probabilistic_Model_Code.aspx.
[13] EN 1990:2002 Eurocode – Basis of structural design. 2002 CEN.
[14] S. Glowienka, A. Fischer, M. Krauss, Implementation of probabilistic method in structural design, Proceedings of 9-th IPW, eds. H.
Budelmann, A. Holst, D. Proske, 17-18 November 2011, pp. 274-286.
[15] S. Glowienka, H. Hansen, Teilsicherheitsfaktoren für die Berechnung von Großkraftwerken, Beton und Stahlbetonbau. 7(105) (2010) 455-
462.
[16] L. Brunarski , Postawy matematyczne kszta towania kryteriów zgodno ci wytrzyma o ci materia ów, Prace ITB, WITB, Warszawa 2009.
[17] I. Skrzypczak, L. Buda-O óg, Wska nik jako ci produkowanego betonu a niezawodno w odniesieniu do próby o ma ej liczebno ci,
Budownictwo i Architektura. 13(2) (2014) 135-141.
[18] Sukanta Kumer Shill, Reliability analysis of reinforced concrete shallow footings designed using BNBC 2006, Master thesis, Department of
Civil Engineering, Dhaka University of Engineering and Technology, Gazipur, February, 2015.
[19] US ACE, Engineering and design: Introduction to probability and reliability methods for use in geotechnical engineering, U.S. Army Corps
of Engineers, Engineer Technical Letter 1110-2-547, Dept. of the Army, Washington, D.C., 1997.