Vous êtes sur la page 1sur 74

Technical University of Civil Engineering of Bucharest

Reinforced Concrete Department

STRUCTURAL RELIABILITY AND RISK ANALYSIS


Lecture notes

STRUCTURAL RELIABILITY AND RISK ANALYSIS 4th Year FILS

Foreword

These lecture notes provides an insight into the concepts, methods and procedures of
structural reliability and risk analysis considering the presence of random uncertainties. The
course is addressed to undergraduate students from Faculty of Engineering in Foreign
Languages instructed in English language as well as to postgraduate students in structural
engineering. The objectives of the courses are:
-

to provide a review of mathematical tools for quantifying uncertainties using theories


of probability, random variables and random processes;

to develop the theory of methods of structural reliability based on concept of reliability


indices;

to explain the basics of code calibration;

to evaluate actions on buildings and structures due to natural hazards;

to provide the basics of risk analysis;

to provide the necessary background to carry out reliability based design and riskbased decision making and to apply the concepts and methods of performance-based
engineering;

to prepare the ground for students to undertake research in this field.

The content of the lectures in Structural Reliability and Risk Analysis is:
-

Introduction to probability and random variables; distributions of probability

Formulation of reliability concepts for structural components; exact solutions, firstorder reliability methods; reliability indices; basis for probabilistic design codes

Seismic hazard analysis

Evaluation of snow, wind and earthquake loads for design of structures.

UTCB, Technical University of Civil Engineering, Bucharest

STRUCTURAL RELIABILITY AND RISK ANALYSIS 4th Year FILS

Table of Contents
1. INTRODUCTION TO RANDOM VARIABLES THEORY................................................ 6
1.1 Data samples .................................................................................................................... 7
1.2 Indicators of the sample (esantion) .................................................................................. 7
1.3 Probability ........................................................................................................................ 9
1.4 Random variables ........................................................................................................... 10
1.5 Indicators of the probability distributions ...................................................................... 12
2. DISTRIBUTIONS OF PROBABILITY .............................................................................. 15
2.1. Normal distribution ....................................................................................................... 15
2.2. Log-normal distribution ................................................................................................ 19
2.3. Extreme value distributions........................................................................................... 21
2.3.1. Gumbel distribution for maxima in 1 year ............................................................. 22
2.3.2. Gumbel distribution for maxima in N years........................................................... 25
2.4. Mean recurrence interval............................................................................................... 27
3. FUNCTION OF RANDOM VARIABLES ......................................................................... 29
3.1 Second order moment models ........................................................................................ 29
4. STRUCTURAL RELIABILITY ANALYSIS ..................................................................... 33
4.1. The basic reliability problem......................................................................................... 33
4.2. Special case: normal random variables ......................................................................... 35
4.3. Special case: log-normal random variables................................................................... 36
4.4. Calibration of partial safety coefficients ....................................................................... 37
5. SEISMIC HAZARD ANALYSIS........................................................................................ 41
5.1. Deterministic seismic hazard analysis (DSHA) ............................................................ 41
5.2. Probabilistic seismic hazard analysis (PSHA) .............................................................. 42
5.3. Earthquake source characterization............................................................................... 43
5.4. Predictive relationships (attenuation relations) ............................................................. 46
5.5. Temporal uncertainty .................................................................................................... 46
5.6. Probability computations............................................................................................... 47
5.7. Probabilistic seismic hazard assessment for Bucharest from Vrancea seismic source . 47
6. ACTIONS ON STRUCTURES. SNOW LOADS ............................................................... 53
6.1 Introduction .................................................................................................................... 53
6.2 Snow load on the ground................................................................................................ 53
6.3 Snow load on the roof .................................................................................................... 56
6.4 Roof shape coefficients .................................................................................................. 58
6.5 Local effects ................................................................................................................... 60
6.5.1 Local verification .................................................................................................... 60
6.5.2 Exceptional snow drift on the roof .......................................................................... 61
7. ACTIONS ON STRUCTURES. WIND ACTION .............................................................. 62
7.1. General .......................................................................................................................... 62
7.2 Reference wind velocity and reference velocity pressure .............................................. 63
7.3 Terrain roughness and Variation of the mean wind with height .................................... 64
7.4 Wind turbulence ............................................................................................................. 67
7.5 Peak values ..................................................................................................................... 67
7.6 Wind actions................................................................................................................... 68
7.6.1 Wind pressure on surfaces....................................................................................... 69
7.6.2 Wind forces ............................................................................................................ 70
8. ACTIONS ON STRUCTURES. SEISMIC ACTION ......................................................... 71
8.1 Introduction .................................................................................................................... 71
8.2 Representation of the seismic action.............................................................................. 71

UTCB, Technical University of Civil Engineering, Bucharest

STRUCTURAL RELIABILITY AND RISK ANALYSIS 4th Year FILS

8.3 Design spectrum for elastic analysis .............................................................................. 74

References:
1. Structural reliability and risk analysis, Lecture notes (prof.dr.ing. Radu Vacareanu)
http://www.utcb-ccba.ro/attachments/116_Structural%20Reliability%20Lecture%20Notes.pdf
2. Siguranta constructiilor (Dan Lungu & Dan Ghiocel)
3. EN 1991 Eurocode 1: Actions on structures. Part 1-3 General actions Snow Loads
4. EN 1991-1-4 - Eurocode 1: Actions on structures - Part 1-4: General actions - Wind
actions, CEN
5. CR 1-1-4/2012Cod de proiectare. Evaluarea aciunii vntului asupra construciilor.
6. CR 1-1-3/2012 Cod de proiectare. Evaluarea aciunii zpezii asupra construciilor.
7. P100-1/2013 Cod de proiectare seismica - Partea I Prevederi de proiectare pentru
cldiri.

UTCB, Technical University of Civil Engineering, Bucharest

STRUCTURAL RELIABILITY AND RISK ANALYSIS 4th Year FILS

Examination subjects:
1. Random variables, indicators of the sample, histograms
2. Probability, probability distributions function, cumulative density function, indicators
3. Normal distribution
4. Log-normal distribution
5. Extreme value distributions, Gumbel distribution for maxima in 1 year and in N years,
Mean recurrence interval
6. Function of random variables and second order moment models
7. Structural reliability analysis. Special cases
8. Calibration of partial safety coefficients
9. Seismic hazard analysis
10. Snow loads
11. Wind action
12. Seismic action

UTCB, Technical University of Civil Engineering, Bucharest

STRUCTURAL RELIABILITY AND RISK ANALYSIS 4th Year FILS

1. INTRODUCTION TO RANDOM VARIABLES THEORY


Reliability the ability of a system or component to perform its required functions under
stated conditions under a specified period of time.
Structural reliability objective to develop criteria and methods aiming at ensuring that
buildings (structures) built according to specifications will not fail to preserve the functions
for a specified period of time.
Uncertainty
- Random (aleatory) uncertainties that are related to the built in reliability
phenomena and informations that cannot be reduced through better knowledge.
- Epistemic uncertainties that are related to our inability to predict the future
behavior of systems. It can be reduced through better understanding and modeling.
Example: Analysis of a frame structure
Live load + dead loads
Lateral load
(wind,
earthquake)

As
N

Bending
moment

As

Steel area

Compressive
force

Steps: Model (1) Imposed Loads (2) Structural Analysis (3) Sectional Analysis (4)
Design for some limits states (5).
Model (1) - Epistemic uncertainty
Loads (2) dead, wind, live loads are evaluated using the codes.
Random uncertainty (in the codes) - one way to take care, is the use of the safety factors (5%
less than or 5% greater than see the class of concrete)
Structural Analysis (3) - Epistemic uncertainty (linear elastic materials strains proportional
to stresses)
Sectional Analysis (4) - Random + Epistemic uncertainty
Design for some limits states (5) ultimate limit states (safety), serviceability limit states.
For power plants LOCA local lose of cooling agent
Uncertainties can be reduced to physical models:
- random variable (RV) dead load, material strength
- random process wind velocity, ground acceleration.
The random variable (RV) is a quantity whose value cannot be predicted with sufficient
accuracy before performing the experiment.
In engineering statistics one is concerned with methods for designing and evaluating
experiments to obtain information about practical problems, for example, the inspection of
quality of materials and products. The reason for the differences in the quality of products is

UTCB, Technical University of Civil Engineering, Bucharest

STRUCTURAL RELIABILITY AND RISK ANALYSIS 4th Year FILS

the variation due to numerous factors (in the material, workmanship) whose influence cannot
be predicted, so that the variation must be regarded as a random variation.
1.1 Data samples
In most cases the inspection of each item of the production is prohibitively expensive and
time-consuming. Hence instead of inspecting all the items just a few of them (a sample) are
inspected and from this inspection conclusions can be drawn about the totality (the
population).
If X = random variable (concrete compressive strength)
x = value for random variables

If one performs a statistical experiment one usually obtains a sequence of observations. A


typical example is shown in Table 1.1. These data were obtained by making standard tests for
concrete compressive strength. We thus have a sample consisting of 30 sample values, so that
the size of the sample is n=30.
1.2 Indicators of the sample (esantion)
One may compute measures for certain properties of the sample, such as the average size of
the sample values, the spread of the sample values, etc.
_

The mean value of a sample x1, x2, , xn or, briefly, sample mean, is denoted by x (or mx) and
is defined by the formula:
_
1 n
x xj
(1.1)
n j 1
It is the sum of all the sample values divided by the size n of the sample. Obviously, it
_

measures the average size of the sample values, and sometimes the term average is used for x .
The variance (dispersion) of a sample x1, x2, , xn or, briefly, sample variance, is denoted by
sx2 and is defined by the formula:
_
1 n
s x2
(
x

x
)2
(1.2)

j
n 1 j 1
The sample variance is the sum of the squares of the deviations of the sample values from the
_

mean x , divide by n-1. It measures the spread or dispersion of the sample values and is
always positive.
The square root of the sample variance s2 is called the standard deviation of the sample and is
denoted by sx. s x s x2 . The mean, mx and the standard deviation, sx has the same units.
The coefficient of variation of a sample x1, x2, ,xn is denoted by COV and is defined as the
ratio of the standard deviation of the sample to the sample mean
s
(dimensionless)
(1.3)
COV
x

UTCB, Technical University of Civil Engineering, Bucharest

STRUCTURAL RELIABILITY AND RISK ANALYSIS 4th Year FILS

Table 1.1. Sample of 30 values of the compressive strength of concrete, daN/cm2


320
350
370

380
340
390

340
350
370

360
360
370

330
370
400

360
350
360

380
350
400

360
420
350

320
360
360

350
340
390

The statistical relevance of the information contained in Table 1.1 can be revealed if one shall
order the data in ascending order in Table 1.2, first column (320, 330 and so on).
Table 1.2 Frequencies of values of random variable listed in Table 1.1
Compressive
strength
320
330
340
350
360
370
380
390
400
410
420

Absolute
frequency
2
1
3
6
7
4
2
2
2
0
1

Relative
frequency
0.067
0.033
0.100
0.200
0.233
0.133
0.067
0.067
0.067
0.000
0.033

Cumulative
frequency
2
3
6
12
19
23
25
27
29
29
30

Cumulative relative
frequency
0.067
0.100
0.200
0.400
0.633
0.767
0.833
0.900
0.967
0.967
1.000

The number of occurring figures from Table 1.1 is listed in the second column of Table 1.2. It
indicates how often the corresponding value x occurs in the sample and is called absolute
frequency of that value x in the sample.
Dividing it by the size n of the sample one obtains the relative frequency listed in the third
column of Table 1.2.
If for a certain value x one sums all the absolute frequencies corresponding to the sample
values which are smaller than or equal to that x, one obtains the cumulative frequency
corresponding to that x. This yields the values listed in column 4 of Table 1.2. Division by the
size n of the sample yields the cumulative relative frequency in column 5 of Table 1.2.
The graphical representation of the sample values is given by histograms of relative
frequencies and/or of cumulative relative frequencies (Figure 1.1 and Figure 1.2).

Figure 1.1 Histogram of relative


frequencies

Figure 1.2 Histogram of cumulative


relative frequencies

UTCB, Technical University of Civil Engineering, Bucharest

STRUCTURAL RELIABILITY AND RISK ANALYSIS 4th Year FILS

If a certain numerical value does not occur in the sample, its frequency is 0. If all the n values
of the sample are numerically equal, then this number has the frequency n and the relative
frequency is 1. Since these are the two extreme possible cases, one has:
- The relative frequency is at least equal to 0 and at most equal to 1;
- The sum of all relative frequencies in a sample equals 1.
If a sample consists of too many numerically different sample values, the process of grouping
may simplify the tabular and graphical representations, as follows (Kreyszig, 1979).
n1 n2
nj
nm
xmin

x x

xmax

A sample being given, one chooses an interval I that contains all the sample values. One
subdivides I into subintervals, which are called class intervals. The midpoints of these
subintervals are called class midpoints. The sample values in each such subinterval are said to
form a class.
The number of sample values in each such subinterval is called the corresponding class
frequency (absolute frequency - nj).
Division by the sample size n gives the relative class frequency (relative frequency m
n
f
f j j and f j = 1). The normalized relative frequency is f jN j . The normalization is
n
x
j 1
with respect to the x.
The relative frequency is called the frequency function of the grouped sample, and the
corresponding cumulative relative class frequency is called the distribution function of the
j

grouped sample (Fj = f j ).


k 1

If one chooses few classes, the distribution of the grouped sample values becomes simpler but
a lot of information is lost, because the original sample values no longer appear explicitly.
When grouping the sample values the following rules should be obeyed (Kreyszig, 1979):
All the class intervals should have the same length;
The class intervals should be chosen so that the class midpoints correspond to simple number;
If a sample value xj coincides with the common point of two class intervals, one takes it
into the class interval that extends from xj to the right.
1.3 Probability
Probability is an numerical measure of the chance or likelihood of occurrence of an event
relative to other events.
Letting n to move to infinite (n ), frequencies moves to probabilities and consequently
fj probability. If x 0 then histogram of normalized relative frequencies become the
probability density function (PDF) and the histogram of cumulative relative frequencies
become the cumulative distribution function (CDF).
A random experiment or random observation is a process that has the following properties,
(Kreyszig, 1979):
it is performed according to a set of rules that determines the performance completely;

UTCB, Technical University of Civil Engineering, Bucharest

STRUCTURAL RELIABILITY AND RISK ANALYSIS 4th Year FILS

it can be repeated arbitrarily often;


the result of each performance depends on chance (that is, on influences which we cannot
control) and therefore can not be uniquely predicted.
The result of a single performance of the experiment is called the outcome of that trial. The
set of all possible outcomes of an experiment is called the sample space of the experiment and
will be denoted by S. Each outcome is called an element or point of S.
Experience shows that most random experiments exhibit statistical regularity or stability of
relative frequencies; that is, in several long sequences of such an experiment the
corresponding relative frequencies of an event are almost equal to probabilities. Since most
random experiments exhibit statistical regularity, one may assert that for any event E in such
an experiment there is a number P(E) such that the relative frequency of E in a great number
of performances of the experiment is approximately equal to P(E).
For this reason one postulates the existence of a number P(E) which is called probability of an
event E in that random experiment. Note that this number is not an absolute property of E but
refers to a certain sample space S, that is, to a certain random experiment.
1.4 Random variables
Roughly speaking, a random variable X (also called variate) is a function whose values are
real numbers and depend on chance (Kreyszig, 1979).
If one performs a random experiment and the event corresponding to a number a occurs, then
we say that in this trial the random variable X corresponding to that experiment has assumed
the value a. The corresponding probability is denoted by P(X=a).
Similarly, the probability of the event X assumes any value in the interval a<X<b is denoted
by P(a<X<b).
The probability of the event X c (X assumes any value smaller than c or equal to c) is
denoted by P(X c), and the probability of the event X>c (X assumes any value greater than c)
is denoted by P(X>c).
The last two events are mutually exclusive:
P(Xc) + P(X>c) = P(- < X < ) =1

(1.4)

The random variables are either discrete or continuous.

P( x X x dx) f x ( x)dx - Definition of probability density function (PDF)


b

P(a X b) f (u )du

(1.5)

Hence this probability equals the area under the curve of the density f(x) between x=a and
x=b, as shown in Figure 1.3.

UTCB, Technical University of Civil Engineering, Bucharest

10

STRUCTURAL RELIABILITY AND RISK ANALYSIS 4th Year FILS

f(x)
P(a<X<b)

-4

-3

-2

0 a

-1

1b

Figure 1.3 Example of probability computation


Properties of PDF:
- fx(x) >0
- As x - fx(x) 0
- As x fx(x) 0

- f ( u )du 1

(the most important property)

(1.6)

If X is any random variable, then for any real number x there exists the probability P(X x)
corresponding to X x (X assumes any value smaller than x or equal to x) that is a function of
x, which is called the cumulative distribution function of X, CDF and is denoted by F(x).
Thus F(x) = P(X x).
Since for any a and b > a one has P(a X <b) = P(X b) - P(X a)
One shall now define and consider continuous random variables. A random variable X and the
corresponding distribution are said to be of continuous type or, briefly, continuous if the
corresponding distribution function F(x) = P(X x) can be represented by an integral in the
form
x

F ( x ) f ( u )du

cumulative distribution function (CDF)

(1.7)

where the integrand is continuous and is nonnegative. The integrand f is called the probability
density or, briefly, the density of the distribution. Differentiating one notices that
F(x) = f(x)
In this sense the density is the derivative of the distribution function.
Fx(x)

Shape of
CDF

Fx(a)
0
a

Properties of CDF:

UTCB, Technical University of Civil Engineering, Bucharest

11

STRUCTURAL RELIABILITY AND RISK ANALYSIS 4th Year FILS

- Fx(x) >0
- As x - Fx(x) 0
- As x Fx(x) 1
- if x2 x1 Fx(x2) Fx(x1) it is an increasing function
- Since for any a and b > a Fx(a) = P(X <a); Fx(b) = P(X <b) one has:
b

P( a X b ) f ( u )du F ( b ) F ( a )
a

1.5 Indicators of the probability distributions


- Refers to the population The mean value or mean of a distribution is denoted by and is defined by
xj f( xj )
(discrete distribution)

(1.8a)

xf ( x )dx

(continuous distribution)

(1.8b)

where f(xj) is the probability function of discrete random variable X and f(x) is the density of
continuos random variable X.
The mean is also known as the mathematical expectation of X and is sometimes denoted by
E(X).
The variance of a distribution is denoted by x2 and is defined by the formula:
(1.9)
The positive square root of the variance is called the standard deviation and is denoted by x.
Roughly speaking, the variance is a measure of the spread or dispersion of the values, which
the corresponding random variable X can assume.
The coefficient of variation of a distribution is denoted by Vx and is defined by the formula
Vx

x
x

(1.10)

The central moment of order i of a distribution is defined as:


(1.11)
2 = x2 the variance is the central moment of second order.
The skewness coefficient is defined with the following relation:

(1.12)
Skewness is a measure of the asymmetry of the probability distribution of a random variable.
A distribution is said to be symmetric with respect to a number x = c if for every real x,
f(c+x) = f(c-x).
(1.13).

UTCB, Technical University of Civil Engineering, Bucharest

12

STRUCTURAL RELIABILITY AND RISK ANALYSIS 4th Year FILS

An asymmetric distribution does not comply with relation (1.13).


Asymmetric distributions are with positive asymmetry (skewness coefficient larger than zero)
and with negative asymmetry (skewness coefficient less than zero).
A negative skewness coefficient indicates that the tail on the left side of the probability
density function is longer than the right side and the bulk of the values (including the median)
lies to the right of the mean.
A positive skew indicates that the tail on the right side of the probability density function is
longer than the left side and the bulk of the values lie to the left of the mean. A zero value
indicates that the values are relatively evenly distributed on both sides of the mean, typically
implying a symmetric distribution.

Figure 1.4. Asymmetric distributions with positive asymmetry (left) and negative asymmetry
(right) (www.mathwave.com)

The mode of the distribution X is the value of the random variable that corresponds to the
peak of the distribution (the most likely value).
Distributions with positive asymmetry have the peak of the distribution shifted to the left
(mode smaller than mean); distributions with negative asymmetry have the peak of the
distribution shifted to the right (mode larger than mean).
~

The median of the distribution, X , xm is the value of the random variable that have 50%
chances of smaller values and, respectively 50% chances of larger values.
P(X > xm)= P(X xm) = 0.5 = Fx(xm)
For a symmetric distribution the mean, the mode and the median are coincident and the
skewness coefficient is equal to zero.
The fractile xp is defined as the value of the random variable X with p non-exceedance
probability (P(X xp) = p).

UTCB, Technical University of Civil Engineering, Bucharest

13

STRUCTURAL RELIABILITY AND RISK ANALYSIS 4th Year FILS

P(X xp) = p = Fx(xp) = Shaded arias

P(X xp) = 1- p = Fx(xp) = Shaded arias


f x (x )

xp

F X (x )

xp

F X (x )

Cumulative distribution
function

Cumulative distribution
function

Probability distribution
function

Probability distribution
function

f x (x )

FX ( x p )

xp

xp

x
Inferior fractile

Superior fractile

The loads are superior fractiles.


The resistances are inferior fractiles.
If a random variable X has mean X and variance x2 , then the corresponding variable
Z = (X - X)/X has the mean 0 and the variance 1. Z is called the standardized variable
corresponding to X.

UTCB, Technical University of Civil Engineering, Bucharest

14

STRUCTURAL RELIABILITY AND RISK ANALYSIS 4th Year FILS

2. DISTRIBUTIONS OF PROBABILITY
2.1. Normal distribution
The continuous distribution having the probability density function, PDF
1 x X
X


1
2
f ( x)

e
2 X

(2.1)

is called the normal distribution or Gauss distribution.


A random variable having this distribution is said to be normal or normally distributed. This
distribution is very important, because many random variables of practical interest are normal
or approximately normal or can be transformed into normal random variables. Furthermore,
the normal distribution is a useful approximation of more complicated distributions.
In Equation 2.1, is the mean and is the standard deviation of the distribution. The curve of
f(x) is called the bell-shaped curve. It is symmetric with respect to and is biparametric (X
and x). Figure 2.1 shows f(x) for same and various values of (and various values of
coefficient of variation V).

0.012

f(x)

Normal distribution
0.01
V=0.10
V=0.20

0.008

V=0.30
0.006

0.004

0.002

0
100

200

300

400

500

600

Figure 2.1. PDFs of the normal distribution for various values of V


The smaller (and V) is, the higher is the peak at x = and the steeper are the descents on
both sides. This agrees with the meaning of variance. (The inflection points of the PDFs are
at left and right of x = x =

UTCB, Technical University of Civil Engineering, Bucharest

15

STRUCTURAL RELIABILITY AND RISK ANALYSIS 4th Year FILS

From (2.1) one notices that the normal distribution has the cumulative distribution function, CDF

F (x)

1
1

2 X

1 v X

2 X

dv

(2.2)

Figure 2.2 shows F(x) for same and various values of (and various values of coefficient of
variation V).
1
F(x)

0.9

Normal distribution

0.8
V=0.10

0.7

V=0.20

0.6

V=0.30

0.5
0.4
0.3
0.2
0.1
0
100

200

300

400

500

600

Figure 2.2. CDFs of the normal distribution for various values of V


From (2.2) one obtains
1 b 1 v
e 2
2 X a
1

P(a X b) F (b) F (a )

dv

(2.3)

In Microsoft Excel: NORMDIST (x, X and x ) normal distribution


Change of variable:
The integral in (2.2) cannot be evaluated by elementary methods.
v x
du 1
In fact, if one sets
or (dv = du x, and one has to integrate the
u , then

x
dv x
x x
integral (2.2) from - to z =
.
x
X x x x
Fx ( x) P ( X x) P (

) P ( Z z ) Fz ( z ) the same type of the distribution


x
x
function. dFx ( x) dFz ( z ) f x ( x)dx f z ( z )dz and P(x X < x+dx)=P(z Z < z+dz)
1

From (2.2) one obtains F ( x)


2 X

( x ) /

u2
2 du

; drops out, and the expression on the

right equals:

UTCB, Technical University of Civil Engineering, Bucharest

16

STRUCTURAL RELIABILITY AND RISK ANALYSIS 4th Year FILS

1
( z )
2

u2
2

du

(2.4)

which is the distribution function of the standardized variable with mean 0 and variance 1 and
x x
has been tabulated and where z =
,but can be represented in terms of the integral.
x
x X
F ( x)
X

(2.5).

The density function and the distribution function of the normal distribution with mean 0 and
variance 1 are presented in Figure 2.3.
0.40

1
2

f(z)
Standard normal
distribution

0.35
0.30
0.25
0.20
0.15
0.10
0.05

z
0.00
-4

1.0

-3

-2

-1

(z)
Standard normal
distribution

0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1

0.0
-4

-3

-2

-1

Figure 2.3. PDF and CDF of the normal distribution with mean 0 and variance 1
1

f(z)

1
z2
e 2 -

normal standard PDF(probability distribution function)

( z ) f ( u )du - normal standard CDF (cumulative distribution function)

In Microsoft Excel: NORMSDIST (z, z=0 and z = 1) normal standard distribution

UTCB, Technical University of Civil Engineering, Bucharest

17

STRUCTURAL RELIABILITY AND RISK ANALYSIS 4th Year FILS

From (2.3) and (2.5) one gets:


b X
P(a X b) F (b) F (a )
X

a X

(2.6).

In particular, when a = XX and b = XX , the right-hand side equals (1) - (-1);


to a = XX and b = XX there corresponds the value (2) - (-2), etc. Using
tabulated values of function one thus finds
(a) P(X-X < X X+X) 68%
(b) P(X-2X < X X+2X) 95.5%

(2.7).

(c) P(X-3X < X X+3X) 99.7%


Hence one may expect that a large number of observed values of a normal random variable X
will be distributed as follows:
(a) About 2/3 of the values will lie between X-X and X+X
(b) About 95% of the values will lie between X-2X and X+2X
(c) About 99 % of the values will lie between X-3X and X+3X.
Practically speaking, this means that all the values will lie between X-3X and X+3X;
these two numbers are called three-sigma limits.
The fractile xp that is defined as the value of the random variable X with p non-exceedance
probability ( P(X xp) = p=Fx (xp)) is computed as follows:
xp = X + kpX

(2.8).

The meaning of kp becomes clear if one refers to the reduced standard variable z = (x - .
Thus, x = +z and kp represents the value of the reduced standard variable for which:
(z) = p.
Ex: x0,05

p=0,05 (non-exceedance probability)

p < 0,5 kp <0


p > 0,5 kp >0
p = 0,5 kp =0 we define the median value
The most common values of kp are given in Table 2.1.
Table 2.1. Values of kp for different non-exceedance probabilities p (from Tables)
p
kp

0.01
-2.326

0.02
-2.054

0.05
-1.645

UTCB, Technical University of Civil Engineering, Bucharest

0.95
1.645

0.98
2.054

0.99
2.326

18

STRUCTURAL RELIABILITY AND RISK ANALYSIS 4th Year FILS

2.2. Log-normal distribution


The lognormal distribution of random variable x is the normal distribution of the random
variable lnx.
The log-normal distribution (Hahn & Shapiro, 1967) is defined by its following property: if
the random variable lnX is normally distributed with mean lnX and standard deviation lnX,
then the random variable X is log-normally distributed. Thus, the cumulative distribution
function CDF of random variable lnX is of normal type:
F (ln x )

2 ln X

ln x

1 ln v ln X

2 ln X

d (ln v )

2 ln X

1 ln v ln X

2 ln X

1
dv
v

(2.9).

Since:
x

F (ln x ) f ( v )dv

(2.10)

the probability density function PDF results from (2.9) and (2.10):
1

f(x)

2 ln X

1 ln x ln X
ln X

1 2
e
x

(2.11).

The lognormal distribution is asymmetric with positive asymmetry, i.e. the peak of the
distribution is shifted to the left. The skewness coefficient for lognormal distribution is:
1 3V X V X3

(2.12)

where VX is the coefficient of variation of random variable X. Higher the variability, higher
the shift of the lognormal distribution.
The mean and the standard deviation of the random variable lnX are related to the mean and
the standard deviation of the random variable X as follows:
ln X ln

X
1 V X2

ln X ln( 1 V X2 )

(2.13)

(2.14).

In the case of the lognormal distribution the following relation between the mean and the
median holds true:
ln X ln x0.5

(2.15).

Combining (2.13) and (2.15) it follows that the median of the lognormal distributions is
linked to the mean value by the following relation:
x 0 .5

X
1 V X2

UTCB, Technical University of Civil Engineering, Bucharest

(2.16)

19

STRUCTURAL RELIABILITY AND RISK ANALYSIS 4th Year FILS

If VX is small enough (VX 0.1), then:

ln X ln X

(2.17)

ln X V X

(2.18)

The PDF and the CDF of the random variable X are presented in Figure 2.4 for different
coefficients of variation.
0.012

f(x)

Log-normal distribution
0.01
V=0.10
V=0.20

0.008

V=0.30
0.006
0.004
0.002
0
100

200

300

400

500

600

F(x)
0.9

Log-normal distribution

0.8
0.7

V=0.10
V=0.20

0.6

V=0.30

0.5
0.4
0.3
0.2
0.1
0
100

200

300

400

500

600

Figure 2.4. Probability density function, f(x) and cumulative distribution function, F(x)
of the log-normal distribution for various values of V
The mode median mean of the distribution.
UTCB, Technical University of Civil Engineering, Bucharest

20

STRUCTURAL RELIABILITY AND RISK ANALYSIS 4th Year FILS

Change of variable:
If one uses the reduced variable

ln v ln v
du
1 1
u , then

or (dv = du v lnv,
ln v
dv ln v v

and one has to integrate from - to z =

1
1
(z)

2 ln v

(ln v ln v ) / ln v

ln v ln v
. From (2.9) one obtains:
ln v

u2
2

1
1
ln v v du
v
2

u2
2

du

(2.19)

Standard normal cumulative distribution function (CDF)


The fractile xp that is defined as the value of the random variable X with p non-exceedance
probability (P(X xp) = p) is computed as follows, given lnX normally distributed:
ln(xp) = lnX + kplnX

(2.20)

From (2.20) one gets:


xp e

ln X k p ln X

(2.21)

where kp represents the value of the reduced standard variable for which (z) = p.
xm x0 ,5 e ln X ; ln x0.5 ln X relation between the median and the mean

b X
X

Normal distribution for X : P(a X b) F (b) F (a)

a X

ln b ln x
ln x

LogNormal distribution for X: P( a X b ) F ( b ) F ( a )

ln a ln x

ln x

2.3. Extreme value distributions


The field of extreme value theory was pioneered by Leonard Tippett (19021985). Emil
Gumbel codified the theory of extreme values in his book Statistics of extremes published
in 1958 at Columbia University Press.
Extreme value distributions are the limiting distributions for the minimum or the maximum of
a very large collection of random observations from the same arbitrary distribution. The
extreme value distributions are of interest especially when one deals with natural hazards like
snow, wind, temperature, floods, etc. In all the previously mentioned cases one is not
interested in the distribution of all values but in the distribution of extreme values, which
might be the minimum or the maximum values. In Figure 2.5 it is represented the distribution
of all values of the random variable X as well as the distribution of minima and maxima of X.
Year 1: y11, y21, y31, ym1 - x1
Year 1: y12, y22, y32, ym2 - x2
..
Year n: y1n, y2n, y3n, ymn - xn
Xi = max (y1i, y2i, y3i, ymi) maximum annual value

UTCB, Technical University of Civil Engineering, Bucharest

21

STRUCTURAL RELIABILITY AND RISK ANALYSIS 4th Year FILS

0.009

f(x)

0.008

minima
distribution

maxima
distribution

0.007

0.006
all values
distribution

0.005

0.004

0.003

0.002

0.001

0
0

100

200

300

400

500

600

700

800

900

x
1000

Figure 2.5. Distribution of all values, of minima and of maxima of random variable X
2.3.1. Gumbel distribution for maxima in 1 year
The Gumbel distribution for maxima is defined by its cumulative distribution function, CDF:
F ( x ) e e

( x u )

(2.22)

where u and are the parameters:


u = x

n 1

; = 0,5772 Euler constant (Euler-Mascheroni); lim lnn

n k 1 k

6 x

The final form is: u = x 0.45x mode of the distribution (Figure 2.8) and
= 1.282 / x dispersion coefficient (shape factor).
The skewness coefficient of Gumbel distribution is positive constant ( 1 1.139 ), i.e. the
distribution is shifted to the left. In Figure 2.6 it is represented the CDF of Gumbel
distribution for maxima for the random variable X with the same mean x and different
coefficients of variation Vx.
The probability distribution function, PDF is obtained straightforward from (2.22):
( x u )
dF ( x )
(2.23)
f(x)
e ( x u ) e e
dx
The PDF of Gumbel distribution for maxima for the random variable X with the same mean
x and different coefficients of variation Vx is represented in Figure 2.7.

UTCB, Technical University of Civil Engineering, Bucharest

22

STRUCTURAL RELIABILITY AND RISK ANALYSIS 4th Year FILS

Gumbel distribution
for maxima in 1 year

F(x)

0.9
V=0.10

0.8

V=0.20
V=0.30

0.7
0.6
0.5
0.4
0.3
0.2
0.1

0
100

200

300

400

500

600

700

Figure 2.6. CDF of Gumbel distribution for maxima for the random variable X
with the same mean x and different coefficients of variation Vx
0.014

f(x)
Gumbel distribution
for maxima in 1 year

0.012

V=0.10
V=0.20

0.01

V=0.30

0.008
0.006
0.004
0.002
x
0
100

200

300

400

500

600

700

Figure 2.7. PDF of Gumbel distribution for maxima for the random variable X
with the same mean x and different coefficients of variation Vx
One can notice in Figure 2.7 that higher the variability of the random variable, higher the shift
to the left of the PDF.

UTCB, Technical University of Civil Engineering, Bucharest

23

STRUCTURAL RELIABILITY AND RISK ANALYSIS 4th Year FILS

0.005

f(x)

0.0045
0.004
0.0035
0.003
0.0025
0.002

0.45

0.0015
0.001

0,164 x

0.0005

0
100

xm x
400

u
300

200

500

600

700

Figure 2.8. Parameter u in Gumbel distribution for maxima


Fractile xp:
Given X follows Gumbel distribution for maxima, the fractile xp that is defined as the value of
the random variable X with p non-exceedance probability (P(X xp) = p) is computed as
follows:

F ( x p ) P( X x p ) p e e

( x p u )

From Equation 2.24 it follows: ln( p ) e


xp u

(2.24).

( x p u )

( x p u ) ln( ln p )

x
1
ln( ln p ) x 0.45 x
ln( ln p ) x k Gp x

1.282

(2.25)

where:

k Gp 0.45 0.78 ln( ln p )

(2.26).

The values of kpG for different non-exceedance probabilities are given in Table 2.2.
Table 2.2. Values of kpG for different non-exceedance probabilities p
p
kpG

0.50
-0.164

0.90
1.305

0.95
1.866

0.98
2.593

xm=x - 0,164 x (median value)

UTCB, Technical University of Civil Engineering, Bucharest

24

STRUCTURAL RELIABILITY AND RISK ANALYSIS 4th Year FILS

2.3.2. Gumbel distribution for maxima in N years


All the previous developments are valid for the distribution of maximum yearly values. If one
considers the probability distribution in N (N>1) years, the following relation holds true (if
one considers that the occurrences of maxima are independent events):
F(x)1 year = P(X x) in 1 year
F(x)N years = P(X x) in N years= P(X x)year1 P(X x)year2 P(X x)year N
F(x)N years = P(X x) in N years = [P(X x) in 1 year]N = [F(x)1 year]N

(2.27)

where:
F(x)N years CDF of random variable X in N years
F(x)1 year CDF of random variable X in 1 year.
The Gumbel distribution for maxima has a very important property the reproducibility of
Gumbel distribution - i.e., if the annual maxima (in 1 year) follow a Gumbel distribution for
maxima then the maxima in N years will also follow a Gumbel distribution for maxima:
1( x u1 )

F ( x )N F ( x )1 N e e

ee

e Ne

1( x u1 )ln N

1( x u1 )

ee

1( x ( u1

ln N
))
1

ee

( x u N )

(2.28)

where:
u1 mode of the distribution in 1 year
dispersion coefficient in 1 year
uN = u1 + lnN / mode of the distribution in N years
dispersion coefficient in N years
The PDF of Gumbel distribution for maxima in N years is translated to the right with the amount
lnN / with respect to the PDF of Gumbel distribution for maxima in 1 year, Figure 2.9.
0.0063

f(x)
Gumbel distribution
for maxima
N yr.
1 yr.

0.0042
lnN /

0.0021

lnN /

0
100

200

u1
300

m1
400

500

uN
mN
600
700
800

900

x
1000

Figure 2.9. PDF of Gumbel distribution for maxima in 1 year and in N years

UTCB, Technical University of Civil Engineering, Bucharest

25

STRUCTURAL RELIABILITY AND RISK ANALYSIS 4th Year FILS

V x ,1

x
; V x , N x ; Vx,N < Vx,1 (COV is decreasing with increasing the years)
x ,1
x,N

Also, the CDF of Gumbel distribution for maxima in N years is translated to the right with the
amount lnN / with respect to the CDF of Gumbel distribution for maxima in 1 year, Figure 2.10.
1

F(x)
Gumbel distribution
for maxima

0.9
0.8
0.7

N yr.
1 yr.

0.6
0.5
0.4
0.3
0.2
0.1
0
100

200

300

400

500

600

700

800

900

x
1000

Figure 2.10. CDF of Gumbel distribution for maxima in 1 year and in N years
Important notice: the superior fractile xp (p >> 0.5) calculated with Gumbel distribution for
maxima in 1 year becomes a frequent value (sometimes even an inferior fractile if N is large,
N 50) if Gumbel distribution for maxima in N years is employed, Figure 2.11.
0.0063

f(x)
Gumbel distribution
for maxima

N yr.
1 yr.

0.0042
probability of
exceedance of x p
in N years

0.0021
probability of
exceedance of x p
in 1 year
0
100

200

300

400

xp
500
600

700

800

900

x
1000

Figure 2.11. Superior fractile xp in 1 year and its significance in N year


UTCB, Technical University of Civil Engineering, Bucharest

26

STRUCTURAL RELIABILITY AND RISK ANALYSIS 4th Year FILS

2.4. Mean recurrence interval


The loads due to natural hazards such as earthquakes, winds, waves, floods were recognized
as having randomness in time as well as in space. The randomness in time was considered in
terms of the return period or recurrence interval. The recurrence interval also known as a
return period is defined as the average (or expected) time between two successive statistically
independent events and it is an estimate of the likelihood of events like an earthquake, flood
or river discharge flow of a certain intensity or size. It is a statistical measurement denoting
the average recurrence interval over an extended period of time. The actual time T between
events is a random variable.
Values of a random
variable (natural hazard)
RP return period

RP it is not the same


(random variable)

years

The mean recurrence interval, MRI of a value larger than x of the random variable X may be
defined as follows:
MRI ( X x )

1
1
1
1

P1 year X x 1 P1 year X x 1 Fx x 1 p

(2.29)

where:
p is the annual probability of the event (Xx) and FX(x) is the cumulative distribution function of X.
Thus the mean recurrence interval of a value x is equal to the reciprocal of the annual
probability of exceedance of the value x. The mean recurrence interval or return period has an
inverse relationship with the probability that the event will be exceeded in any one year.
For example, a 10-year flood has a 0.1 or 10% chance of being exceeded in any one year and
a 50-year flood has a 0.02 (2%) chance of being exceeded in any one year. It is commonly
assumed that a 10-year earthquake will occur, on average, once every 10 years and that a 100year earthquake is so large that we expect it only to occur every 100 years. While this may be
statistically true over thousands of years, it is incorrect to think of the return period in this
way. The term return period is actually misleading. It does not necessarily mean that the
design earthquake of a 10 year return period will return every 10 years. It could, in fact, never
occur, or occur twice. This is why the term return period is gradually replaced by the term
recurrence interval. Researchers proposed to use the term return period in relation with the
effects and to use the term recurrence interval in relation with the causes.
The mean recurrence interval is often related with the exceedance probability in N years. The
relation among MRI, N and the exceedance probability in N years, Pexc,N is:
Pexceedance, N (>x) = 1 Pnon-exceedance, N (x) = 1 - [Pnon-exceedance, 1 (x)]N =1-pN

UTCB, Technical University of Civil Engineering, Bucharest

27

STRUCTURAL RELIABILITY AND RISK ANALYSIS 4th Year FILS


N

1
T

e
Pexceedance, N (>x) = 1 - 1

T ( x )
1
1
MRI ( X x p )

; MRI ( X x p ) = T ( X x p ) ;
1 P1year X x 1 p

example: x0.98 similar xT 50 ; MRI

(2.30)

1
50years
1 0,98

Usually the number of years, N is considered equal to the lifetime of ordinary buildings, i.e.
50 years. Table 2.3 shows the results of relation (2.30) for some particular cases considering
N=50 years.
Table 2.3 Correspondence amongst MRI, Pexc,1 year and Pexc,50 years
Mean recurrence interval,
years MRI, T ( x )
10
30
50
100
225
475
975
2475

Probability of exceedance
in 1 year, Pexc,1 year

Probability of exceedance
in 50 years, Pexc,50 years

0.10
0.03
0.02
0.01
0.004
0.002
0.001
0.0004

0.99
0.81
0.63
0.39
0.20
0.10
0.05
0.02

The modern earthquake resistant design codes consider the definition of the seismic hazard
level based on the probability of exceedance in 50 years. The seismic hazard due to ground
shaking is defined as horizontal peak ground acceleration, elastic acceleration response
spectra or acceleration time-histories. The level of seismic hazard is expressed by the mean
recurrence interval (mean return period) of the design horizontal peak ground acceleration or,
alternatively by the probability of exceedance of the design horizontal peak ground
acceleration in 50 years. Four levels of seismic hazard are considered in FEMA 356
Prestandard and Commentary for the Seismic Rehabilitation of Buildings, as given in Table
2.4. The correspondence between the mean recurrence interval and the probability of
exceedance in 50 years, based on Poisson assumption, is also given in Table 2.4.
Table 2.4. Correspondence between mean recurrence interval and probability of exceedance
in 50 years of design horizontal peak ground acceleration as in FEMA 356
Seismic Hazard
Mean recurrence interval
Probability of
Level
(years)
exceedance
SHL1
72
50% in 50 years
SHL2
225
20 % in 50 years
SHL3
475
10 % in 50 years
SHL4
2475
2 % in 50 years

UTCB, Technical University of Civil Engineering, Bucharest

28

STRUCTURAL RELIABILITY AND RISK ANALYSIS 4th Year FILS

3. FUNCTION OF RANDOM VARIABLES


3.1 Second order moment models
Let us consider a simply supported beam, Figure 3.1:
q

Figure 3.1. Simple supported beam


Considering that:
q, y are described probabilistically (normal, lognormal, Gumbel for maxima distributions)

and l, W are described deterministically, the design condition (ultimate limit state condition)
is:
M max M cap ;

ql 2
y W
8

and considering
S sectional effect of load, and
R sectional resistance
it follows that:

ql 2
; R y W .
8

The following question rises:


If q and y are described probabilistically, how can one describes S and R probabilistically?
If X1, X2, Xi,.. Xn are random variables defined by fx1 ( x ) , fx2 ( x ) fxn (x ) ;
mx1 , mx2 mxn ; x1 , x2 xn

If Y (X1, X2, Xi,.. Xn) is the function of random variables how to find fy, my , y
To answer the question, two cases are considered in the following:
1.

The relation between q and S ( y and R) is linear

2.

The relation is non linear.

Case 1: Linear relation between random variables, X, Y


X Y a X b

(a and b constants, X random variable)

UTCB, Technical University of Civil Engineering, Bucharest

29

STRUCTURAL RELIABILITY AND RISK ANALYSIS 4th Year FILS

For the random variable X one knows: the probability density function, PDF, the cumulative
distribution function, CDF, the mean, m and the standard deviation, . The unknowns are the
PDF, CDF, m and for the random variable Y.
If one applies the equal probability formula, Figure 3.2 (a linear transformation keeps the
distribution):

f X ( x)dx fY ( y )dy P( x X x dx) P( y Y y dy )


x

(3.1)

1
1
1 y b
y b
f X ( x) f X
; fY ( y ) f X ( x )

dy
a
a
a a

(3.2)

dx
y
y=ax+b

fY ( y ) dy

y dy

fY ( y ) ?

f X ( x)dx
x
x dx

f X (x)

Figure 3.2. Linear relation between random variables X and Y


Distribution of Y Distribution of X
Developing further the linear relation it follows that:

mY

yf

( y )dy

( x)dx a

2
Y

( a x b) f

(y m

xf

( x)dx b

( x)dx a mX b

) fY ( y )dy (a x b a mX b) 2 f X dx
2

a 2 ( x m X ) 2 f X ( x )dx a 2 X2

Y a X
mY a mX b
VY

(3.3)

Y
a X

mY a mX b

UTCB, Technical University of Civil Engineering, Bucharest

30

STRUCTURAL RELIABILITY AND RISK ANALYSIS 4th Year FILS

Case 2: Non-linear relation between random variables, X, Y


Let the random variable Y Y ( X 1 , X 2 ,, X i , X n ) , the relation being non-linear, how to
find fy, my , y
Observations:
Let the random variable Y X 1 X 2 X 3 X n . If the random variables Xi

1.

are normally distributed then Y is also normally distributed.


Let the random variable Y X 1 X 2 X 3 X . If the random variables Xi are

2.

log-normally distributed then Y is also log-normally distributed.


Let the random variables Xi with known mean and standard deviations:
mX i
X i ; i 1, n
X i
f ( x ) f (a)

f (a)
f (a)
( x a)
( x a) 2 .... Taylor series
1!
2!

The first order approximation


FORM First order reliability method
If one develops the function Y in Taylor series around the mean m =( mx1 , mx2 mxi .. mxn ),
and keeps only the first order term of the series, then the mean and the standard deviation of
the new random variable Y, mY , Y are approximated by:

Y Y ( X 1 , X 2 ,, X i , X n ) Y( mx1 , mx2 mxi . mxn ) +

Y
( X i m X i ) ... linear
i 1
i m
n

relation
mY Y ( mx1 , mx2 mxi .. mxn ) - First order approximation of mean

y2

(3.4)

Y 2
X i - First order approximation of the variance
i 1
i m
n

X
2

n y
y
y
y
2
X2
X2
Y2
Xn
1 x 2
2
i 1 xi
x1 m
xn m
m

2
Xi
m

(3.5)

Relations 3.4 and 3.5 are the basis for the so-called First Order Second Moment Models,
FOSMM.
Few examples of FOSMM are provided in the following:
Y X1 X 2
mY Y ( mx1 , mx2 ) = m X m X
1

y
y
X2
X2 12 X2 12 X2 X2 X2 ;
Y2
1
2
1
2
1
2
x1 m
x2 m

UTCB, Technical University of Civil Engineering, Bucharest

Y X2 1 X2 2

31

STRUCTURAL RELIABILITY AND RISK ANALYSIS 4th Year FILS

Y a X b
mY a m X b
Y2 a 2 X2 ; Y a X

Y X1 X 2

mY Y ( mx1 , mx2 ) = m X m X
1

y
y
X2
X2 x1 2m X2 x2 2m X2 m X2 X2 m X2 X2

1
2
1
2
2
1
1
2
x1 m
x2 m
2
Y

m X2 2 X2 m X2 1 X2
X2
X2
Y
1
2
1
VY

2 2 VX21 VX22
mY
mX mX
m X2
mX
1

UTCB, Technical University of Civil Engineering, Bucharest

32

STRUCTURAL RELIABILITY AND RISK ANALYSIS 4th Year FILS

4. STRUCTURAL RELIABILITY ANALYSIS


4.1. The basic reliability problem
The basic structural reliability problem considers only one load effect S resisted by one
resistance R. Each is described by a known probability density function, fS( ) and fR( )
respectively. It is important to notice that R and S are expressed in the same units.
For convenience, but without loss of generality, only the safety of a structural element will be
considered here and as usual, that structural element will be considered to have failed if its
resistance R is less than the load effect S acting on it. The probability of failure Pf of the
structural element can be stated in any of the following ways, (Melchers, 1999):
Pf = P(RS)

(4.1a)

=P(R-S0)

(4.1b)

=P(R/S1)

(4.1c)

=P(ln R-ln S0)

(4.1d)

or, in general
=P(G(R ,S)0)

(4.1e)

where G( ) is termed the limit state function and the probability of failure is identical with the
probability of limit state violation.
Re= P(R>S) - probability of reliability
Pf + Re=1
(R>S safe; R=S limit state; R<S failure)
Quite general density functions fR and fS for R and S respectively are shown in Figure 4.1
together with the joint (bivariate) density function fRS(r,s). For any infinitesimal element
(rs) the latter represents the probability that R takes on a value between r and r+r and S a
value between s and s+s as r and s each approach zero.

Figure 4.1. Joint density function fRS(r,s), marginal density functions fR(r) and fS(s)
and failure domain D, (Melchers, 1999)
UTCB, Technical University of Civil Engineering, Bucharest

33

STRUCTURAL RELIABILITY AND RISK ANALYSIS 4th Year FILS

In Figure 4.1, the Equations (4.1) are represented by the hatched failure domain D, so that the
probability of failure becomes:
Pf PR S 0 f RS r , s dr ds
(4.2).
D

When R and S are independent, fRS(r,s)=fR(r)fS(s) and relation (3.2) becomes:


P f P R S 0

sr

f R r f S s dr ds

F x f x dx
R

(4.3)

Relation (4.3) is also known as a convolution integral with meaning easily explained by
reference to Figure 4.2. FR(x) is the probability that Rx or the probability that the actual
resistance R of the member is less than some value x. Let this represent failure. The term fs(x)
represents the probability that the load effect S acting in the member has a value between x
and x+x as x 0. By considering all possible values of x, i.e. by taking the integral over all
x, the total probability of failure is obtained. This is also seen in Figure 4.3 where the density
functions fR(r) and fS(s) have been drawn along the same axis.
1 F R (x), f S (x)
0.9
0.8
0.7
R
S

0.6
0.5
0.4
0.3
0.2

P(R<x)

0.1
0
-4 -3 -2 -1

R=x
2 3

x
9 10 11

Figure 4.2. Basic R-S problem: FR( ) fS( ) representation


0.0063

f R (x), f S (x)
amount of overlap
of f R( ) and fS()
rough indicator of pf

0.0042
R
S
0.0021

0
100

200

S
300

400

S500
k Rk
600

R
700

800

900

x
1000

Figure 4.3. Basic R-S problem: fR( ) fS( ) representation


UTCB, Technical University of Civil Engineering, Bucharest

34

STRUCTURAL RELIABILITY AND RISK ANALYSIS 4th Year FILS

If S is distanced from R we deal with Pf small.


If S is near to R we deal with Pf large.
An alternative to expression (4.3) is:

Pf

1 F

( x) f R ( x)dx

(4.4)

which is simply the sum of the failure probabilities over all cases of resistance for which the
load effect exceeds the resistance.
4.2. Special case: normal random variables
For a few distributions of R and S it is possible to integrate the convolution integral (4.3)
analytically.
One notable example is when both are normal random variables with means R and S and
variances R2 and S2 respectively.
The safety margin Z=R-S (Cornell approach) follow an normal distribution with a mean and
variance given by :
Z = R - S

(4.5a)

Z2 = R2 + S2

Z R2 S2

(4.5b)

Equation (4.1b) then becomes

0 Z
Pf PR S 0 PZ 0 FZ ( z 0 )
Z

(4.6)

where ( ) is the normal standard cumulative distribution function (for the standard normal
variate with zero mean and unit variance).
0.035

f Z (z)

0.03
Z<0
0.025

Z>0

Failure

Safety

Failure domain
(z<0)

0.02

Safe domain
(z>0)

0.015

0.01

0.005

Pf

0 0
-60

z=r-s=0
Limit state
function

-50

-40

-30

-20

-10

10

20
z

Figure 4.4. Distribution of safety margin Z = R S

UTCB, Technical University of Civil Engineering, Bucharest

35

STRUCTURAL RELIABILITY AND RISK ANALYSIS 4th Year FILS

The random variable Z = R - S is shown in Figure 4.4, in which the failure region Z 0 is
shown shaded. Using (4.5) and (4.6) it follows that (Cornell, 1969):

0
R
S
1
Pf
2
2
R S
where =

(4.7)

z
is defined as reliability (safety) index.
z

If either of the standard deviations R and S or both are increased, the term in square brackets
in (4.7) will become smaller and hence Pf will increase. Similarly, if the difference between the
mean of the load effect and the mean of the resistance is reduced, Pf increases. These
observations may be deduced also from Figure 4.3, taking the amount of overlap of fR( ) and fS()
as a rough indicator of Pf.
Pf
10-3
10-4
10-5
10-6

3.09
3.72
4.27
4.75

4.3. Special case: log-normal random variables


The log-normal model for structural reliability analysis was proposed by (Rosenbluth &
Esteva, 1972). Both random variables R and S have lognormal distribution with the following
parameters: means lnR and lnS and variances lnR2 and lnS2 respectively.
R
The safety margin Z= ln then has a mean and a standard deviation given by:
S
R
Z R ln
(4.8a)
ln

S
S
Z

ln

ln R ln

R
S

V R V R2 VS2

(4.8b)

R
1 VR2

ln R ;

ln R ln(1 VR2 ) VR ;
Relation (4.1d) then becomes

ln S ln

S
1 VS2

ln S

ln S ln(1 VS2 ) VS

0 Z
R

Pf P ln 0 PZ 0
S

(4.9)

where ( ) is the standard normal distribution function (zero mean and unit variance). The
R
random variable Z= ln is shown in Figure 4.5, in which the failure region Z 0 is shown
S
shaded.
Using (4.8) and (4.9) it follows that

UTCB, Technical University of Civil Engineering, Bucharest

36

STRUCTURAL RELIABILITY AND RISK ANALYSIS 4th Year FILS


R

0 ln
S
Pf
1
V 2 V 2
S
R

where =

(4.10)

z
is defined as reliability (safety) index,
z

R
S

ln

V R2 VS2

0.035

(4.11)

f Z (z)

0.03
Z<0
0.025

Z>0

Failure

Safety

0.02

0.015

0.01

0.005

Pf

0 0
-60

-50

-40

-30

-20

-10

Figure 4.5. Distribution of safety margin Z= ln

10

20
z

R
S

4.4. Calibration of partial safety coefficients


In USA: LRDF Load and resistance factor design

ln

R
S

V R2 VS2

Lindt proposed the following liniarization:

VS2 VR2 VS VR with 0,70,75 for

1 VR

3 . Given Lindts linearization it follows that:


3 VS

UTCB, Technical University of Civil Engineering, Bucharest

37

STRUCTURAL RELIABILITY AND RISK ANALYSIS 4th Year FILS

ln R
S

VR VS

(4.12).

The calibration of partial safety coefficients used in semi-probabilistic design codes is


accomplished using the log-normal model using FOSM.
From Equation 4.12 one has:

ln R
S

R e VR VS R e VR e VS R e VR S e VS (4.13)
VR VS
S
S
where e VR and e VS are called safety coefficients, SC with respect to the mean.
Rq and Sp are characteristics values (fractiles) of the random variables R respectively S.

Rq e

ln R k q ln R

ln R

Sp e

ln S k p ln S

ln S

k V

R e q R e

S p S e k p VS
Rq

( k )V

k q VR

R e

k p VS

S e

(VR VS )

k q VR

k p VS

k q VR

k p VS

( k )V

p
S
e
( kq )V
R
e

( k )V

q
p
R
S
Rq e
S p e

p
q

q e

( k q )VR

p e

- Partial safety coefficient for sectional resistance

( k p )VS

- Partial safety coefficient for the effects of the load

But one needs the SC with respect to the characteristic values of the loads and resistances, the
so-called partial safety coefficients, PSC.
kq= -1,645 (p=0.05)
kp= 1,645 (p=0.95)
kq= -2,054 (p=0.02)
kq= 2,054 (p=0.98)
To this aim, one defines the limit state function used in the design process:

S0.98 0.98 R0.05 0.05

Rdesign
S design

(4.14)

where 0.98 and 0.05 are called partial safety coefficients, PSC.
Assuming that S and R are log-normally distributed, one has:
ln R 1.645 ln R

ln S 2.054 ln S

R0.05 e
S 0.98 e

ln R

e 1.645VR R e 1.645VR

(4.15)

ln S

e 2.054VS S e 2.054VS

(4.16)

VR

R0.05 R e 1.645VR e

S 0.98 S e 2.054VS

e S e 1.645VR
e 2.054VS

UTCB, Technical University of Civil Engineering, Bucharest

(4.17)

38

STRUCTURAL RELIABILITY AND RISK ANALYSIS 4th Year FILS


( 2.054 )V

( 1.645 )V

R
R0.05 e
S0.98 eS
0.98
0.05

0.05 e

( 1.645 )V R

0.98 e

( 2.054 )V S

(4.18)

- Partial safety coefficient for sectional resistance

(4.19)

- Partial safety coefficient for the effects of the load (4.20)

The partial safety coefficients 0.05 and 0.98 as defined by Equations 4.19 and 4.20 depend on
the reliability index and the coefficients of variation for resistances and loads, respectively.
If the reliability index is increased, the partial safety coefficient for loads 0.98 increases
while the partial safety coefficient for resistance 0.05 decreases. The theory of partial safety
coefficients based on lognormal model is incorporated in the Romanian Code CR0-2013
named Cod de proiectare. Bazele proiectarii structurilor in constructii (Design Code. Basis
of Structural Design).
g
0.98

2.0

0.9

F0.05

1.8

0.8
1.6

1.4

0.7
1.2
VS
1.0
0

0.2

0.4

VR
0.6

0.6

0.05

0.1

0.15

0.2

Figure 4.6. Variation of partial safety coefficients with the coefficient of variation of load
effect (left) and of resistance (right)
Eurocode 1 (EN1-1990) Basis of Structural design
= 4,7 ultimate Limit states (ULS)
= 2,9 Serviceability Limit states (SLS)

Rd Ed (design condition)

Actions (F) : Permanent actions (G), Variable actions (Q), Accidental action (A)
Fk characteristic value (upper fractile)
Fd design value
f - partial safety coefficient applied to the action
Fd = f Fk
Ed = Sd E(Fd)
- design value of section effort of the load
E(Fd) sectional effect of the design action (load)
Sd - safety coefficient considering the errors involved in the model for calculation of the
sectional effort of the load

UTCB, Technical University of Civil Engineering, Bucharest

39

STRUCTURAL RELIABILITY AND RISK ANALYSIS 4th Year FILS

Ed = E( Sd f Fk) = E( F Fk)
Xk characteristic value of the strength of the material (inferior fractile)
Xd design value of the strength of the material
m - partial safety coefficient applied to the strength of the material
Xd =

1
m

Xk

Rd design value of the sectional resistance


Rd =

1
1
1
1
R( X d ) R(

X k ) R(
Xk )
Rd
Rd m
M

Other representative values of variable actions used in the design, Figure 4.7 are:
the combination value, represented as a product of 0Qk used for the verification of
ultimate limit states and irreversible serviceability limit states;

the frequent value, represented as a product 1Qk, used for the verification of ultimate
limit states involving accidental actions and for verifications of reversible
serviceability limit states; this value is closed to the central value of the statistical
distribution;

the quasi-permanent value, represented as a product 2Qk, (2 1) used for the


verification of ultimate limit states involving accidental actions and for the verification
of reversible serviceability limit states. Quasi-permanent values are also used for the
calculation of long-term effects.
25

15

Valoare caracteristica, Qk

Valoare de combinatie, 0Qk

Valoare frecventa, 1Qk

Valoare cvasipermanenta, 2Qk

10

20

20

15

Valoare
instantanee Q

10

25

0
0

0.01

0.02

Timp
0.03

0.04

fQ

10

20

30

40

50

60

70

Figure 4.7 Values of variable actions


n

j 1

i2

Ed = G , j Gk , j p P Q ,1Qk ,1 Q ,i 0 ,i Qk ,i - fundamental combination


n

j 1

i 1

Ed = Gk , j P AEd 2 ,i Qk ,i

UTCB, Technical University of Civil Engineering, Bucharest

- seismic combination

40

STRUCTURAL RELIABILITY AND RISK ANALYSIS 4th Year FILS

5. SEISMIC HAZARD ANALYSIS


The final product of the seismic hazard analysis is the hazard curve for a given site.

Annual exceedance rate, l(PGA )

1.E-01

PSHA - Bucharest
1.E-02

1.E-03

1.E-04
100

150

200

250
300
PGA , cm/s2

350

400

Hazard curve for Bucharest from Vrancea seismic source

1
; y mean annual rate of ground exceedance of mean parameter (PGA-peak
y
ground acceleration, SA-spectral acceleration, peak ground displacement, etc)
The two methods of seismic hazard analysis are deterministic and probabilistic ones.
MRI ( y )

5.1. Deterministic seismic hazard analysis (DSHA)


The deterministic seismic hazard analysis involves the development of a particular seismic
scenario, i.e. the postulated occurrence of an earthquake of a specified size at a specific
location (longitude, latitude, magnitude and depth are deterministic). The DSHA is
developed in four steps (Reiter, 1990):
1. Identification and characterization of all earthquake sources geometry and position
of the sources and the maximum magnitude for all sources;
Epicenter - epicentral
distance
h depth

Site

R hypocentral
distance

M- Source (focus)
2. Selection of source-to-site distance parameters (either epicentral or hypocentral
distance) for each source zone;
3. Selection of controlling earthquake (expected to produce the strongest shaking at the
site); use predictive (attenuation) relations for computing ground motion produced at
the site by earthquakes of magnitudes given in step 1 occurring at each source zone;

UTCB, Technical University of Civil Engineering, Bucharest

41

STRUCTURAL RELIABILITY AND RISK ANALYSIS 4th Year FILS

4. Define the seismic hazard at the site in terms of peak ground acceleration PGA,
spectral acceleration SA, peak ground velocity PGV, etc.
All the steps are summarized in Figure 5.1.
Source 3
M3

Source 1
M1

R1
R3
Site
R2
Source 2
M2
Step 1

Step 2

M
Ground
3
Motion
Parameter, M
2
M
Y

Y1
Y
2
Y Y3


YN

R2

R3
Step 3

R1

Distance
Step 4

Figure 5.1. Steps in DSHA

5.2. Probabilistic seismic hazard analysis (PSHA)


The PSHA (Cornell, 1968, Algermissen et. al., 1982) is developed in four steps (Reiter, 1990):
1. Identification and characterization of earthquake sources. Besides the information
required in step 1 of DSHA, it is necessary to obtain the probability distribution of
potential rupture location within the source and the probability distribution of source
to-site distance (longitude, latitude, magnitude and depth are random variable);
2. Definition of seismicity, i.e. the temporal distribution of earthquake recurrence
(average rate at which an earthquake of some size will be exceeded);
3. Use predictive (attenuation) relations for computing ground motion produced at the
site by earthquakes of any possible size occurring at any possible point in each source
zone; uncertainty in attenuation relations is considered in PSHA;

UTCB, Technical University of Civil Engineering, Bucharest

42

STRUCTURAL RELIABILITY AND RISK ANALYSIS 4th Year FILS

4. Uncertainties in earthquake location, size and ground motion prediction are combined
and the outcome is the probability that ground motion parameter will be exceeded
during a particular time period.
All the steps are summarized in Figure 5.2.
Source 3
M3

Source 1
M1

Average
rate,
log

3
R

Site

f
R
Source 2
M2

Magnitude

Step
1

Step
2
P(Y>y*
)

Ground
Motion
Parameter,
Y

Distance, R

Step
3

Parameter value, y*

Step
4
Figure 5.2. Steps in PSHA

5.3. Earthquake source characterization


The seismic sources can be modeled as:
point sources if the source is a short shallow fault
area sources if the source is a long and/or deep fault
volumetric sources.
The spatial uncertainty of earthquake location is taken into account in PSHA. The earthquakes
are usually assumed to be uniformly distributed within a particular source zone. The uniform
distribution in the source zone does not often translate into a uniform distribution of sourceto-site distance.
Another important source of uncertainties is given by the size of the earthquake and by the
temporal occurrence of earthquakes. The recurrence law gives the distribution of earthquake
sizes in a given period of time.

UTCB, Technical University of Civil Engineering, Bucharest

43

STRUCTURAL RELIABILITY AND RISK ANALYSIS 4th Year FILS

Gutenberg & Richter (1944) organized the seismic data in California according to the number
of earthquakes that exceeded different magnitudes during a time period. The key parameter in
Gutenberg & Richters work was the mean annual rate of exceedance, M of an earthquake of
magnitude M which is equal to the number of exceedances of magnitude M divided by the
length of the period of time. The Gutenberg & Richter law is (Figure 5.3):
lg M = a - b M

(5.1)

where M - mean annual rate of exceedance of an earthquake of magnitude M,


M - magnitude,
a and b numerical coefficients depending on the data set.
If NM is the number of earthquakes with magnitude higher than M in T years M =

NM
T

The physical meaning of a and b coefficients can be explained as follows:


0 =10a mean annual number of earthquakes of magnitude greater than or equal to 0 (a
gives the global seismicity)
b describes the relative likelihood of large to small earthquakes. If b increases the
number of larger magnitude earthquakes decreases compared to those of smaller earthquakes
(b is the slope of the recurrence plot).
lg M

10a
1

b small
b

b large

Figure 5.3. The Gutenberg-Richter law


The a and b coefficients are obtained through regression on a database of seismicity form the
source zone of interest. Record of seismicity contains dependent events (foreshocks,
aftershocks) that must be removed form the seismicity database because PSHA is intended to
evaluate the hazard form discrete, independent releases of seismic energy.
ln M
lg M =
=a-bM
; M = 10ab M
(5.2)
ln 10
ln M = a ln10 - b ln10 M = - M

(5.3)

M = e M

(5.4)

where = a ln10 = 2.303 a and = b ln10 = 2.303 b.

UTCB, Technical University of Civil Engineering, Bucharest

44

STRUCTURAL RELIABILITY AND RISK ANALYSIS 4th Year FILS

The original Gutenberg & Richter law (5.1) is unbounded in magnitude terms. This leads to
unreliable results especially at the higher end of the magnitude scale. In order to avoid this
inconsistency, the bounded recurrence law is used.
The bounded law is obtained and defined hereinafter.
The form (5.4) of Gutenberg & Richter law shows that the magnitudes follow an exponential
distribution. If the earthquakes smaller than a lower threshold magnitude Mmin are eliminated,
one gets (McGuire and Arabasz, 1990):
FM(M) = P[Mag. M M Mmin] = 1 - P[Mag. > M M Mmin] =
= 1

fM(M) =

M min

=1-

e M
M min

= 1 - e ( M M min )

(5.5)

dFM ( M )
= e ( M M min) .
dM

(5.6)

Mmin is the mean annual rate of earthquakes of magnitude M larger or equal than Mmin.
If both a lower threshold magnitude Mmin and a higher threshold magnitude Mmax are taken
into account, the probabilistic distribution of magnitudes can be obtained as follows (McGuire
and Arabasz, 1990); (Bounded Guttenberg-Richter law).
The cumulative distribution function must have the unity value for M = Mmax. This yields:
FM ( M )
1 e ( M M min )
=
(5.7)
FM (M ) = P[Mag. M Mmin M Mmax] =
FM ( M max ) 1 e ( M max M min )

dFM ( M )
e ( M M min )
=
.
f M (M ) =
dM
1 e ( M max M min )

(5.8)

The mean annual rate of exceedance of an earthquake of magnitude M is:

M M min [1 FM ( M )] = M min

e ( M M min ) e ( M max M min )


1 e ( M max M min )

(5.9)

where M min = e M min is the mean annual rate of earthquakes of magnitude M larger or
equal than Mmin.
Finally one gets (McGuire and Arabasz, 1990):
M min

M = e

e ( M M min ) e ( M max M min )


1 e ( M max M min )

e ( M M min ) [ 1 e ( M max M ) ] M 1 e ( M max M )


= e M min
=e
1 e ( M max M min )
1 e ( M max M min )
M

M = e

(5.10)

1 e ( M max M )
1 e ( M max M min )

UTCB, Technical University of Civil Engineering, Bucharest

45

STRUCTURAL RELIABILITY AND RISK ANALYSIS 4th Year FILS

5.4. Predictive relationships (attenuation relations)


The predictive relationships usually take the form Y = f(M, R, Pi), where Y is a ground motion
parameter, M is the magnitude of the earthquake, R is the source-to-site distance and Pi are
other parameters taking into account the earthquake source, wave propagation path and site
conditions. The predictive relationships are used to determine the value of a ground motion
parameter for a site given the occurrence of an earthquake of certain magnitude at a given
distance. The coefficients of the predictive relationships are obtained through least-square
regression on a particular set of strong motion parameter data for a given seismic region. This
is the reason for not extrapolating the results of the regression analysis to another seismic
region. The uncertainty in evaluation of the ground motion parameters is incorporated in
predictive relationships through the standard deviation of the logarithm of the predicted
parameter. Finally, one can compute the probability that ground motion parameter Y exceeds a
certain value, y* for an earthquake of magnitude, m at a given distance r (Figure 5.4):

PY y* | m, r 1 PY y* | m, r 1 FY y *
where F is the CDF of ground motion parameter, usually assumed lognormal.

(5.11)

Y
P(Y>y*|m,r)
y*

fY(y|m,r)
r

Figure 5.4. Incorporation of uncertainties in the predictive relationships


5.5. Temporal uncertainty
The distribution of earthquake occurrence with respect to time is considered to have a random
character. The temporal occurrence of earthquakes is considered to follow, in most cases, a
Poisson model, the values of the random variable of interest describing the number of
occurrences of a particular event during a given time interval.
The properties of Poisson process are:
1. The number of occurrences in one time interval is independent of the number of
occurrences in any other time interval.
2. The probability of occurrence during a very short time interval is proportional to the
length of the time interval.
3. The probability of more than one occurrence during a very short time interval is
negligible.
If N is the number of occurrences of a particular event during a given time interval, the
probability of having n occurrences in that time interval is:
n e
PN n
(5.12)
n!
where is the average number of occurrences of the event in that time interval.
UTCB, Technical University of Civil Engineering, Bucharest

46

STRUCTURAL RELIABILITY AND RISK ANALYSIS 4th Year FILS

5.6. Probability computations


The results of the PSHA are given as seismic hazard curves quantifying the annual probability
of exceedance of different values of selected ground motion parameter.
The probability of exceedance a particular value, y* of a ground motion parameter (GMP) Y is
calculated for one possible earthquake at one possible source location and then multiplied by
the probability that that particular magnitude earthquake would occur at that particular
location. The process is then repeated for all possible magnitudes and locations with the
probability of each summed:
(5.13)
PY y * PY y* | X P X PY y* | X f X x dx
where X is a vector of random variables that influence Y (usually magnitude, M and source-tosite distance, R). Assuming M and R independent, for a given earthquake recurrence, the
probability of exceeding a particular value, y*, is calculated using the total probability
theorem (Cornell, 1968, Kramer, 1996):
(5.14)
P (Y y*) P (Y y* | m, r ) f M (m) f R (r )dmdr
where:
- P(Y>y*|m,r) probability of exceedance of y* given the occurrence of an earthquake of
magnitude m at source to site distance r.
- fM(m) probability density function for magnitude;
- fR(r) probability density function for source to site distance.
Site
Epicenter - epicentral
distance
h depth

R hypocentral
distance

M- Source (focus)
The attenuation relation for subcrustal earthquakes (Mollas & Yamazaki, 1995):
ln PGA = c0 + c1 Mw + c2 lnR +c3R +c4 h +
where: PGA is peak ground acceleration at the site,
Mw- moment magnitude,
R - hypocentral distance to the site,
h - focal depth,
c0, c1, c2, c3, c4 - data dependent coefficients
- random variable with zero mean and standard deviation = ln PGA .

P( PGA PGA*) P( PGA PGA* | M w , R ) f M w ( M w ) f R ( R )dM w dR

5.7. Probabilistic seismic hazard assessment for Bucharest from Vrancea seismic source
The Vrancea region, located when the Carpathians Mountains Arch bents, is the source of
subcrustal (60-170km) seismic activity, which affects more than 2/3 of the territory of
Romania and an important part of the territories of Republic of Moldova, Bulgaria and
Ukraine. According to the 20th century seismicity, the epicentral Vrancea area is confined to a
rectangle of 40x80km2 having the long axis oriented N45E and being centered at about 45.6o
Lat.N 26.6o and Long. E.

UTCB, Technical University of Civil Engineering, Bucharest

47

STRUCTURAL RELIABILITY AND RISK ANALYSIS 4th Year FILS

Two catalogues of earthquakes occurred on the territory of Romania were compiled, more or
less independently, by Radu (1974, 1980, 1995) and by Constantinescu and Marza (1980,
1995) Table 5.1. The Radus catalogues are more complete, even the majority of significant
events are also included in the Constantinescu and Marza catalogue. The magnitude in Radu
catalogue is the Gutenberg-Richter (1936) magnitude, MGR. The magnitude in Constantinescu
& Marza catalogue was the surface magnitude, MS. Tacitly, that magnitude was later
assimilated as Gutenberg-Richter magnitude (Marza, 1995).
Table 5.1. Catalogue of subcrustal Vrancea earthquakes (Mw 6.3 ) occurred during the 20th century
Date

1903 13 Sept
1904

6 Feb

1908 6 Oct
1912 25 May
1934 29 March
1939

5 Sept

Time
(GMT)
h:m:s

Lat. N

Long.
E

26.6

RADU Catalogue,
1994
h, km
>60

I0 MGR Mw
7 6.3 -

MARZA www.infp.ro
Catalogue, Catalogue,
1998
1980
I0
Ms
Mw
6.5
5.7
6.3

08:02:7

45.7

02:49:00

45.7

26.6

75

5.7

6.3

6.6

21:39:8
18:01:7
20:06:51

45.7
45.7
45.8

26.5
27.2
26.5

150
80
90

8
7
7

6.8
6.0
6.3

8
7
8

6.8
6.4
6.3

7.1
6.7
6.6

06:02:00

45.9

26.7

120

5.3

6.1

6.2

1940 22 Oct
06:37:00 45.8
26.4
122 7 / 8 6.5 - 7
6.2
6.5
1940 10 Nov
01:39:07 45.8
26.7
1501)
9 7.4 - 9
7.4
7.7
1945 7 Sept
15:48:26 45.9
26.5
75 7 / 8 6.5 - 7.5
6.5
6.8
1945 9 Dec
06:08:45 45.7
26.8
80
7 6.0 - 7
6.2
6.5
1948 29 May
04:48:55 45.8
26.5
130 6 / 7 5.8 - 6.5
6.0
6.3
1977 4 March 2)
19:22:15 45.34 26.30
109 8 / 9 7.2 7.5 9
7.2
7.4
1986 30 Aug
21:28:37 45.53 26.47
133
8 7.0 7.2 7.1
1990 30 May
10:40:06 45.82 26.90
91
8 6.7 7.0 6.9
1990 31 May
00:17:49 45.83 26.89
79
7 6.1 6.4 6.4
1)
Demetrescus original (1941) estimation: 150Km; Radus initial estimation (1974) was 133 km
2)
Main shock
Nov.10 , 1940 destruction in the epicentral area and in Moldavia, around 1000 deaths, the
largest RC building in Bucharest (Carlton) collapsed;
March 4, 1977 more than 1500 deaths, more than 11000 injured, more than 2 billion
$ losses, 31 buildings with more than 4 stories collapsed.
As a systematization requirement for seismic hazard assessment, usually it is recommended
the use of the moment magnitude, Mw. For Vrancea subcrustal events the relation between
two magnitudes can be simply obtained from recent events data given in Table 5.1:
Mw MGR+ 0.3
6.0 < MGR < 7.7
(5.15)
Even the available catalogues of Vrancea events were prepared using the Gutenberg-Richter
magnitude MGR, the recurrence-magnitude relationship was herein newly determined using the
moment magnitude Mw. The relationship is determined from Radus 20th century catalogue of
subcrustal magnitudes with threshold lower magnitude Mw=6.3.
The average number per year of Vrancea subcrustal earthquakes with magnitude equal to and
greater than Mw, as resulting also from Figure 5.5, is:
log n(Mw) = 3.76 - 0.73 Mw

UTCB, Technical University of Civil Engineering, Bucharest

(5.16)

48

STRUCTURAL RELIABILITY AND RISK ANALYSIS 4th Year FILS

Cumulative number, n(>M) per yr

20 th century Radu's catalogue

0.1

log n (>M w ) = 3.76 - 0.73M w

0.01
1 e 1 .687(8 .1 M w )
n M w e8 .654 1 .687M w
1 e1 .687( 8. 1 6 .3 )

M w, max = 7.8

8.1

0.001
6.3
6.0

6.7
6.4

7.1
6.8

7.5
7.2

7.9
7.6

8.3
8.0

Moment magnitude, M w

Figure 5.5. Magnitude recurrence relation for the subcrustal Vrancea source (M w6.3)
The values of surface rupture area (SRA) and surface rupture length (SRL) from Wells and
Coppersmith (1994) equations for "strike slip" rupture were used to estimate maximum
credible Vrancea magnitude. According to Romanian geologists Sandulescu & Dinu, in
Vrancea subduction zone: SRL 150200 km, SRA<8000 km2. Based on this estimation, from
Table 5.2 one gets:
Mw,max= 8.1.

(5.17)

Table 5.2. Application of Wells and Coppersmith equations to the Vrancea source (mean values)
M Mw

Event

Experienced
SRA, km2

Wells & Coppersmith equations


logSRA=-3.42+0.90Mw logSRL=-3.55+0.74Mw
SRA, km2
SRL, km
759
43
1148
60
2138
100
7413
278

6.7 7.0 May 30,1990


11001)
7.0 7.2 Aug. 30,1986
14001)
7.2 7.5 March 4, 1977 63 x 37=23312)
8.1 Max. credible
1)

As cited in Tavera (1991)

2)

Enescu et al. (1982)

If the source magnitude is limited by an upper bound magnitude Mw,max, the recurrence
relationship can be modified in order to satisfy the property of a probability distribution,
Equation 5.18:
n M w e

M w

1 e
1 e

M w ,max M w

M w ,max M w0

(5.18)

and, in the case of Vrancea source (Elnashai and Lungu 1995):


n M w e 8.654 1.687 M w

1 e 1.687 (8.1 M w )
1 e 1.687 (8.1 6.3)

UTCB, Technical University of Civil Engineering, Bucharest

(5.19)

49

STRUCTURAL RELIABILITY AND RISK ANALYSIS 4th Year FILS

In Eq.(5.18), the threshold lower magnitude is Mw0=6.3, the maximum credible magnitude of
the source is Mw,max=8.1, and = 3.76 ln10 = 8.654, = 0.73 ln10 =1.687.
The maximum credible magnitude of the source governs the prediction of source
magnitudes in the range of large recurrence intervals, where classical relationship (5.16) does
not apply, Table 5.3.
Table 5.3. Mean recurrence interval (MRI) of Vrancea magnitudes, (Mw)=1/n(Mw)
Date

GutenbergRichter
magnitude, MGR

10 Nov. 1940

7.4

4 March 1977

7.2

30 Aug. 1986
30 May 1990

7.0
6.7

Moment
MRI from Eq.
magnitude,
(5.18),
Mw
years
8.1
8.0
778
475
7.9
356
7.8
217
7.7
148
100
7.6
108
7.5
82
7.4
63
50
7.3
50
7.2
40
7.0
26

MRI from Eq.


(5.16),
years
150
127
107
91
76
65
55
46
37
33
23

The depth of the Vrancea foci has a great influence on the experienced seismic intensity. The
damage intensity of the Vrancea strong earthquakes is the combined result of both magnitude
and location of the focus inside the earth.
The relationship between the magnitude of a destructive Vrancea earthquake (Mw6.3) and the
corresponding focal depth shows that higher the magnitude, deeper the focus:
ln h = - 0.866 + 2.846 lnMw - 0.18 P

(5.20)

where P is a binary variable: P=0 for the mean relationship and P=1.0 for mean minus one
standard deviation relationship.
The following model was selected for the analysis of attenuation (Mollas & Yamazaki, 1995):
ln PGA = c0 + c1 Mw + c2 lnR +c3R +c4 h +

(5.21)

where: PGA is peak ground acceleration at the site, Mw- moment magnitude, R - hypocentral
distance to the site, h - focal depth, c0, c1, c2, c3, c4 - data dependent coefficients and - random
variable with zero mean and standard deviation = ln PGA, Table 5.4. Details are given
elsewhere (Lungu et.al., 2000, Lungu et. al. 2001). To obtain the values of table 5.4 we used the
records from 1977, 1986 and 1990s earthquakes.
Table 5.4. Regression coefficients inferred for horizontal components of peak ground
acceleration during Vrancea subcrustal earthquakes, Equation (4.21)
c0
3.098

c1
1.053

c2
-1.000

c3
-0.0005

c4
-0.006

lnPGA
0.502

The application of the attenuation relation 5.21 for the Vrancea subcrustal earthquakes of
March 4, 1977, August 30, 1986 and May 30, 1990 is represented in Figures 5.6, 5.7 and 5.8.

UTCB, Technical University of Civil Engineering, Bucharest

50

STRUCTURAL RELIABILITY AND RISK ANALYSIS 4th Year FILS

500
Attenuation relation - March 4, 1977; Mw=7.5, h=109km

450
400

median

PGA, cm/s^2 ..

350

median+stdev

300
250
200
150
100
50
0
0

50

100

150

200

250

D, km

Figure 5.6. Attenuation relation applied for March 4, 1977 Vrancea subcrustal source
250
Attenuation relation - August 30 1986; Mw=7.2, h=133km

PGA, cm/s^2..

200

150

100
median
median+stdev

50

0
0

50

100

150

200

250

D, km

Figure 5.7. Attenuation relation applied for August 30, 1986 Vrancea subcrustal source
400
Attenuation relation - May 30, 1990; Mw=7.0, h=91km
350

PGA, cm/s^2..

300
250
200
150
median
median+stdev

100
50
0
0

50

100

150

200

250

D, km

Figure 5.8. Attenuation relation applied for May 30, 1990 Vrancea subcrustal source

UTCB, Technical University of Civil Engineering, Bucharest

51

STRUCTURAL RELIABILITY AND RISK ANALYSIS 4th Year FILS

For a given earthquake recurrence, the mean annual rate of exceedance of a particular value of
peak ground acceleration, PGA*, is calculated using the total probability theorem (Cornell,
1968, Kramer, 1996):
( PGA PGA*) M min P ( PGA PGA* | m, r ) f M (m) f R (r )dmdr

(5.22)

where:
(PGA>PGA*) mean annual rate of exceedance of PGA* ;
- Mmin is the mean annual rate of earthquakes of magnitude M larger or equal than Mmin;
- P(PGA>PGA*|m,r) probability of exceedance of PGA* given the occurrence of an
earthquake of magnitude m at source to site distance r. This probability is obtained from
attenuation relationship (4.21) assuming log-normal distribution for PGA;
- fM(m) probability density function for magnitude;
- fR(r) probability density function for source to site distance.
The probability density function for magnitude is obtained from Eq. (4.8) (Kramer, 1996).
The probability density function for source to site distance is considered, for the sake of
simplicity, uniform over the rectangle of 40x80km2 having the long axis oriented N45E and
being centered at about 45.6o Lat.N and 26.6o Long. E.
The mean annual rate of exceedance of PGA the hazard curve - for Bucharest site and
Vrancea seismic source is represented in Figure 5.9.
The hazard curve can be approximated by the form H ko a g k , where ag is peak ground
acceleration, and ko and k are constants depending on the site (in this case ko=1.176E-05,
k=3.0865).

Annual exceedance rate, l(PGA )

1.E-01

PSHA - Bucharest
1.E-02

1.E-03

1.E-04
100

150

200

250
300
PGA , cm/s2

350

400

Figure 5.9. Hazard curve for Bucharest from Vrancea seismic source

UTCB, Technical University of Civil Engineering, Bucharest

52

STRUCTURAL RELIABILITY AND RISK ANALYSIS 4th Year FILS

6. ACTIONS ON STRUCTURES. SNOW LOADS


6.1 Introduction
In Europe, the norm EN 1991-1-3 gives guidance to determine the values of loads due to
snow to be used for the structural design of buildings and civil engineering works. The norm
does not apply for sites at altitudes above 1 500 m. In Romania the norm is implemented as
the law CR 1-1-3/2012 Cod de proiectare. Evaluarea aciunii zpezii asupra
construciilor approved by the Romanian Ministry of Public Works and published by the
Official Gazette of Romania.
Some aspects are not covered by the norm, for example:
- impact snow loads resulting from snow sliding off or falling from a higher roof;
- the additional wind loads which could result from changes in shape or size of the
construction works due to the presence of snow or the accretion of ice;
- loads in areas where snow is present all year round;
- ice loading;
- lateral loading due to snow (e.g. lateral loads exerted by drifts);
- snow loads on bridges.
Snow loads shall be classified as variable, fixed, and static actions.
Depending on geographical locations for the particular condition, exceptional snow loads and
the loads due to exceptional snow drifts may be treated as accidental actions.
The relevant snow loads shall be determined for each design situation identified, in
accordance with EN 1990:2002 in Europe and CR0-2013 in Romania.
The design situations have to be considered for the two types of local site conditions:
a) Normal conditions (without considering exceptional snow drifts on the roof)
The transient/persistent design situation should be used for both the undrifted and the drifted
snow load arrangements.
b) Exceptional conditions (locations considering exceptional snow drifts on the
roof)
i) the transient/persistent design situation should be used for both the undrifted and the drifted
snow load;
ii) the accidental design situation (the snow load is accidental action) should be used when is
considered accidental drifted snow load.
For local effects (ex. drifting at projections and obstructions, the edge of the roof, snow
fences) the persistent/transient design situation should be used.
Due to the temperate climate, for the design of buildings in Romania, the exceptional
conditions relating to areas with exceptional snow falls on the ground (characterized by a very
low probability of occurrence) are not taken into account.
6.2 Snow load on the ground
According to international practice the snow load on the ground can be studied by adopting a
reference period of one year to collect the statistical data, because it believes that the annual

UTCB, Technical University of Civil Engineering, Bucharest

53

STRUCTURAL RELIABILITY AND RISK ANALYSIS 4th Year FILS

meteorological data are statistically independent. The statistical analysis use annual maximum
values, which in this case is the maximum associated with winter snow. Although in some
geographic regions over a long period of time can identify certain trends in climate evolution,
the actual practice setting snow load on the ground they do not take into account.
Romanian law CR 1-1-3/2012, is following the SR EN 1991-1-3, and establish that
characteristic value of snow load on the ground is defined as the snow load on the ground
based on an annual probability of exceedence of 0,02, i.e. IMR=50 years, which is an upper
fractile value of a random variable whose values are measured as annual maximum, using the
Gumbel probability distribution for maxima.
The Gumbel probability distribution for maxima was the EN 1991-1-3 recommendation
because after analyzing the data available at European level when drawing up standard (2003),
this probability distribution proved to be most suitable modeling snow in most European
countries (Switerland, Italy, Grece, Norway, Sueden, Finland, Island, Germany, France, Great
britain) weather stations.
The characteristic values of snow load on the ground in Romania, sk, are indicated in the
zoning map in Figure 6.1. The values shown are valid for designing the action for altitudes of
the sites A 1000 m. The values in Figure 6.1 are minimum values used in the design action
of snow load.

Figure 6.1 Characteristic value of snow load on the ground zoning map, sk, kN/m2, for
altitudide of the sites A 1000 m
The altitude of the site is height above mean sea level, of the site where the structure is to be
located, or is already located for an existing structure.
For the altitude of the site 1000m < A 1500m, the characteristic value of snow load on the
ground is:
sk(1000m < A 1500m) = 2,0 + 0,00691 (A-1000)

UTCB, Technical University of Civil Engineering, Bucharest

for sk(A1000m)=2,0 kN/m2

(6.1)

54

STRUCTURAL RELIABILITY AND RISK ANALYSIS 4th Year FILS

for sk(A1000m)=1,5 kN/m2

sk(1000m < A 1500m) = 1,5 + 0,00752 (A-1000)

(6.2)

Ground level snow loads for any mean recurrence interval different to that for the
characteristic snow load, sk, (which by definition is based on annual probability of exceedence
of 0,02) may be adjusted to correspond to characteristic values by application of relation 6.3.

ln( ln p )

1 0,45
V1
1,282

sp
sk
1 2,593 V1

(6.3)
where:
sk is the characteristic snow load on the ground (kN/m2), with a return period of 50 yearsannual probability of exceedence of 0,02 (annual probability of nonexceedence p=0.98) - in
accordance with EN 1990:2002
sp is the snow load on the ground having the p annual probability of nonexceedence
V1 is the coefficient of variation of annual maximum snow load (in Romania V1 is 0,351,0).
Figure 6.2 are exemplified ratios of the snow load on the ground having MRI = 75years,
respectively MRI = 100years and with characteristic snow load on the ground (MRI=50
years) for different values of the coefficient of variation V1.

Raportul sIMR/sIMR=50ani

1.2

sIMR=100ani/sIMR=50ani

1.15

1.1

sIMR=75ani/sIMR=50ani
1.05
0.3

0.4

0.5

0.6

0.7

0.8

0.9

Coeficientul de variatie V1

Figure 6.2 Ratios of snow load on the ground having MRI = 75years and MRI = 100 years
and with characteristic snow load on the ground (MRI=50 years)
The exceptional snow load on the ground is the load of the snow layer on the ground
resulting from a snow fall which has an exceptionally infrequent likelihood of occurring due
to the temperate climate is not applicable for Romania. If is the case, the EN 1991-1-3
recommended to double value of the sk.
The bulk weight density of snow varies. In general it increases with the duration of the snow
cover and depends on the site location, climate and altitude. Indicative values for the mean
bulk weight density of snow on the ground given in Table 6.1 may be used.

UTCB, Technical University of Civil Engineering, Bucharest

55

STRUCTURAL RELIABILITY AND RISK ANALYSIS 4th Year FILS

Table 6.1: Mean bulk weight density of snow


Type of snow
Bulk weight density [kN/m3]
Fresh
1,0
Settled (several hours or days after its fall)
2,0
Old (several weeks or months after its fall)
2,5 - 3,5
Wet
4,0
6.3 Snow load on the roof
There is still relatively little data from measurements on the snow load on the roof and
measurement procedures are not standardized. In addition, there are practical difficulties of
making multiple measurements. Therefore uncertainties associated with the snow load on the
roof are greater than the uncertainties associated with the snow load on the ground. It has
highlighted the existence of a large number of types, different roofs. The rules and codes tries
to collect and standardize the types of roofs, but obviously it is impossible to be considered all
possible configurations for them. The design shall recognise that snow can be deposited on a
roof in many different patterns. Properties of a roof or other factors causing different patterns
can include:
a) the shape of the roof;
b) its thermal properties;
c) the roughness of its surface;
d) the amount of heat generated under the roof;
e) the proximity of nearby buildings;
f) the surrounding terrain;
g) the local meteorological climate, in particular its windiness, temperature, variations,
and likelihood of precipitation (either as rain or as snow).
In regions with possible rainfalls on the snow and consecutive melting and freezing, snow
loads on roofs should be increased, especially in cases where snow and ice can block the
drainage system of the roof.
The load should be assumed to act vertically and refer to a horizontal projection of the roof
area.
When artificial removal or redistribution of snow on a roof is anticipated the roof should be
designed for suitable load arrangements.
The following two primary load arrangements shall be taken into account:
1. undrifted snow load on the roof - load arrangement which describes the uniformly
distributed snow load on the roof, affected only by the shape of the roof, before any
redistribution of snow due to other climatic actions).
2. drifted snow load on the roof - load arrangement which describes the snow load
distribution resulting from snow having been moved from one location to another
location on a roof, e.g. by the action of the wind).
The characteristic value of snow load on the roof is the product of the characteristic snow
load on the ground and appropriate coefficients.
The snow loads on roofs for the persistent/transient design situations shall be determined as
follows:
s = Is i Ce Ct sk

UTCB, Technical University of Civil Engineering, Bucharest

(6.2)

56

STRUCTURAL RELIABILITY AND RISK ANALYSIS 4th Year FILS

where:
Isis the importance/exposure coefficient for the snow load;
i is the snow load shape coefficient;
sk is the characteristic value of snow load on the ground
Ce is the exposure coefficient
C t is the thermal coefficient.
The snow loads on roofs for the accidental design situations where exceptional snow drift is
the accidental action shall be determined as follows:
s = Is i sk

(6.3)

where:
Isis the importance/exposure coefficient for the snow load;
i is the snow load shape coefficient;
sk is the characteristic value of snow load on the ground
The exposure coefficient Ce should be used for determining the snow load on the roof. The
choice for Ce should consider the future development around the site. Ce should be taken as
1,0 unless otherwise specified for different topographies. He characterizes the overall effect of
the wind on snow deposit on the building by the surrounding topography and natural
environment and / or built in the vicinity of the building.
Table 6.2 Values of Ce for different topographies
Topographies
Windswept

Ce
0,8

flat unobstructed areas


exposed on all sides
without, or little shelter
afforded by terrain, higher
construction works or trees

Normal

1,0

areas where there is no


significant removal of
snow by wind on
construction work,
because of terrain, other
construction works or
trees.

Sheltered

1,2

areas in which the


construction work being
considered is considerably
lower than the surrounding
terrain or surrounded by
high trees and/or
surrounded by higher
construction works

The thermal coefficient C t, is the coefficient defining the reduction of snow load on roofs as
a function of the heat flux through the roof, causing snow melting. The thermal coefficient C t

UTCB, Technical University of Civil Engineering, Bucharest

57

STRUCTURAL RELIABILITY AND RISK ANALYSIS 4th Year FILS

should be used to account for the reduction of snow loads on roofs with high thermal
transmittance (> 1 W/m2K) for some glass covered roofs, because of melting caused by heat
loss. Based on the thermal insulating properties of the material and the shape of the
construction work the thermal coefficient value is determined by special studies, see ISO
4355 - 1998. Ct should be taken as 1,0.
The roughness of the roof surface influences the snow sliding and is difficult to be evaluated.
For example, in some areas of the roof, the small objects might prevent natural snow slide
(other than parapets). However, since it is considered that the snow slides off the roof entirely
(when there are no obstacles or parapets) if the roof angle above 60, those shape coefficients
are zero for these parts of roofs.
6.4 Roof shape coefficients
The roof snow load shape coefficient is ratio of the snow load on the roof to the undrifted
snow load on the ground, without the influence of exposure and thermal effects.
European values of coefficients of EN 1991-1-3 were calibrated based on the analysis results
of experimental studies both on site (in-situ) and in the wind tunnel and the comparative
analysis of coefficients prescriptions from different countries, see Formichi, P., 2008. EN
1991 Eurocode 1: Actions on structures. Part 1-3 General actions Snow Loads,
presentation at Workshop Eurocodes. Background and Applications, 18-20 Feb., Brussels,
60p.
The results of in-situ measurements in the US, Canada, Norway and England were
completed with the results of measurements conducted special campaigns in Europe for
studying deposits of snow on roofs (winter 1998/1999). The measurements were very
detailed, both in terms of meteorological parameters (wind speed, wind direction, air
temperature, air humidity, solar radiation, rainfall regime, etc.) and in terms of roof types
(shape dimensions, inclinations, the surface heat transfer from inside the building, roof
insulation, etc.), altitude, exposure (wind, sun), drifts of snow on the roof at various points,
etc. In England measurements were performed on 25 types of roofs in 18 different sites at
altitudes from 5m to 656 m. In the Italian Alps measurements were performed on 13 rooftops
in 7 different sites at altitudes from 88m to 1340m and in the Dolomites on sites. In Germany
measurements were performed on three roofs in 2 different sites at altitudes of 141m and
880m, and in Switzerland on 35 rooftops at 8 different sites at altitudes from 570m to 1628m.
In total, 81 measurements were carried out on the roof, see Sanpaolesi L., 1999. Scientific
Support Activity In The Field Of Structural Stability Of Civil Engineering Works - Snow
Loads, Final Report, Commission Of The European Communities, DGIII - D3, Contract n
500990/1997, 172p. These in-situ measurement information were completed with results of
laboratory tests at "Jules Verne climatic tunnel" Centre Scientifique et Technique du's
Btiment (CSTB), Nantes.
The snow load shape coefficients that should be used for roofs are given in Table 6.3 and in
Figure 6.3. Where snow fences or other obstructions exist or where the lower edge of the roof
is terminated with a parapet, then the snow load shape coefficient should not be reduced
below 0,8.

UTCB, Technical University of Civil Engineering, Bucharest

58

STRUCTURAL RELIABILITY AND RISK ANALYSIS 4th Year FILS

Table 6.3 Snow load shape coefficients used for monopitch roofs, pitched roofs and for multispan roofs
Angle of pitch of roof, 0

00 300

300 < < 600

600

0,8

0,8 (60 - )/30

0,0

0,8 + 0,8 /30

1,6

Figure 6.3 Snow load shape coefficients used for monopitch roofs, pitched roofs and for
multi-span roofs
Case (i)

1(2)

1(1)

1(2)

Case (ii) 0.51(1)


Case (iii)

1(1)

0.51(2)

Figure 6.4 Snow load shape coefficients and the load arrangements used for monopitch roofs,
pitched roofs and for multi-span roofs
In the case of monopitch roofs, the load arrangement of Figure 6.4 should be used for both the
undrifted and drifted load arrangements.
In the case of pitched roofs the load arrangement of Figure 6.4 are as follows:
- The undrifted load arrangement which should be used is case (i);
- The drifted load arrangements which should be used are cases (ii) and (iii), unless
specified for local conditions.

undrifted snow on pitched roofs


UTCB, Technical University of Civil Engineering, Bucharest

drifted snow on pitched roofs


59

STRUCTURAL RELIABILITY AND RISK ANALYSIS 4th Year FILS

For multi-span roofs the load arrangement of Figure 6.4 are as follows:
- The undrifted load arrangement which should be used is shown in case (i);
- The drifted load arrangement which should be used is shown in case (ii), unless
specified for local conditions.

undrifted snow on multi-span roofs

drifted snow on multi-span roofs

The snow load shape coefficients that should be used for cylindrical roofs, in absence of snow
fences, are given in the Figure 6.5.
0,8

Cazul (i)

Cazul (ii)

0,53
ls/4

ls/4

ls/4

ls/4

ls

Figure 6.5 Snow load shape coefficients for cylindrical roof


For cilindrical roofs the load arrangement of Figure 6.5 are as follows:
- The undrifted load arrangement which should be used is shown in case (i);
- The drifted load arrangement which should be used is shown in case (ii), unless
specified for local conditions.
6.5 Local effects
6.5.1 Local verification
For the local verifications the applied forces used in the design situations have to be
considered are persistent/transient.
In windy conditions drifting of snow can occur on any roof which has obstructions as these
cause areas of aerodynamic shade in which snow accumulates.
UTCB, Technical University of Civil Engineering, Bucharest

60

STRUCTURAL RELIABILITY AND RISK ANALYSIS 4th Year FILS

The snow load shape coefficients and drift lengths for quasi-horizontal roofs should be taken
as follows (see Figure 6.1), unless specified for local conditions:
1 = 0,8
2 = h / sk with restrictions 0,8 2 2,0
- weight density of snow (2 kN/m3)
ls = 2 h with restrictions 5 m ls 15 m.

The design of those parts of a roof cantilevered out beyond the walls should take account of
snow overhanging the edge of the roof, in addition to the load on that part of the roof. The
loads due to the overhang may be assumed to act at the edge of the roof and may be calculated
as follows:

se

se = k s2 /
se - snow load per metre length due to the overhang
s - the most onerous undrifted load case
- weight density of snow (3 kN/m3)
k - a coefficient to take account of the irregular shape of the
snow
k = 3/d with restrictions k d ,

6.5.2 Exceptional snow drift on the roof


The load due to exceptional snow drift is the load arrangement which describes the load of the
snow layer on the roof resulting from a snow deposition pattern which has an exceptionally
infrequent likelihood of occurring.
It should be assumed that they are exceptional snow drift loads and that there is no snow
elsewhere on the roof.
The accidental design situation (the snow load is accidental action) should be used when the
load is considered as accidental drifted snow load.
The snow load shape coefficients used for evaluation of exceptional snow drift on the roof are
given in the Annex B of the EN 1991-1-3:2003 or in the Chapter 7 of the CR 1-1-3/2012.

UTCB, Technical University of Civil Engineering, Bucharest

61

STRUCTURAL RELIABILITY AND RISK ANALYSIS 4th Year FILS

7. ACTIONS ON STRUCTURES. WIND ACTION


The wind action on buildings and structures is presented according to the EUROCODE 1:
Actions on structures Part 1-4: General actions Wind actions and in the Romanian
design code CR 1-1-4/2012.
7.1. General
Wind effects on buildings and structures depend on the exposure of buildings, structures and
their elements to the natural wind, the dynamic properties, the shape and dimensions of the
building (structure).
Wind actions fluctuate with time and act directly as pressures on the external surfaces of
enclosed structures and, because of porosity of the external surface, also act indirectly on the
internal surfaces. They may also act directly on the internal surface of open structures.
Pressures act on areas of the surface resulting in forces normal to the surface of the structure
or of individual cladding components. Additionally, when large areas of structures are swept
by the wind, friction forces acting tangentially to the surface may be significant.
The wind action is represented by a simplified set of pressures or forces whose effects are
equivalent to the extreme effects of the turbulent wind.
The wind actions calculated using EN 1991-1-4 are characteristic values. They are determined
from the basic values of wind velocity or the velocity pressure having annual probabilities of
exceedence of 0,02, which is equivalent to a mean return period of 50 years.
The wind velocity and the velocity pressure are composed of a mean and a fluctuating
component. The random field of natural wind velocity is decomposed into a mean wind in the
direction of air flow (x-direction) averaged over a specified time interval and a fluctuating
and turbulent part with zero mean and components in the longitudinal (x-) direction, the
transversal (y-) direction and the vertical (z-) direction.
V(z,t)

v(z,t)

Gusts, velocity fluctuations from the mean

v(z,t)

vm(z)

Mean wind velocity

V(z,t) = vm(z) + v(z,t)


0

Averaging time interval


of mean velocity (10 min)

UTCB, Technical University of Civil Engineering, Bucharest

vm(z)

62

STRUCTURAL RELIABILITY AND RISK ANALYSIS 4th Year FILS

7.2 Reference wind velocity and reference velocity pressure


The fundamental value of the basic wind velocity, vb , is the characteristic 10 minutes mean
wind velocity, irrespective of wind direction and time of year, at 10 m above ground level in
open and horizontal terrain exposure (category II) with an annual risk of being exceeded of
0, 02 (the mean recurrence interval MRI=50years).
For other than 10 min averaging intervals, in open terrain exposure, the following
relationships may be used:
10min
(7.1)
1,05 v 1h
0,84 v 1min
0,67 v b3s
b vb
b
The coefficient of variation of maximum annual wind velocity, V1 = 1 / m1 depends on the
climate and is normally between 0.10 and 0.40; the mean of the maximum annual wind
velocities is usually between 10 m/s to 50 m/s. The sequence of maximum annual mean wind
velocities can be assumed to be a Gumbel distributed sequence with possibly direction
dependent parameters.
The reference wind velocity pressure can be determined from the reference wind velocity
(standard air density =1.25kg/m3) as follows:
qb Pa

1
vb2 0.625 vb2 m / s
2

(7.2)

The conversion of velocity pressure averaged on 10 min into velocity pressure averaged on
other time interval can be computed from relation (9.1):
10min
1.1q1h
0.7q1min
0.44qb3 s .
b qb
b

(7.3)

The reference wind velocity pressure qb (in kPa) with mean recurrence interval MRI=50 yr.
averaged on 10 min. at 10 m above ground in open terrain is presented in Figure 7.1.

Figure 7.1. The reference wind velocity pressure, [kPa] with MRI=50 yr.
wind velocity averaged on 10 min. at 10 m above ground in open terrain. The mapped values
are valid for sites situated bellow 1000 m

UTCB, Technical University of Civil Engineering, Bucharest

63

STRUCTURAL RELIABILITY AND RISK ANALYSIS 4th Year FILS

The mean wind velocity vm(z) at a height z above the terrain depends on the terrain roughness
and orography and on the basic wind velocity, vb and should be determined using Expression
(7.4).
vm(z) = co(z) cr (z ) vb

(7.4)

where:
co(z) - is the orography factor, taken as 1,0
cr (z ) -is the roughness factor
7.3 Terrain roughness and Variation of the mean wind with height
The roughness of the ground surface is aerodynamically described by the roughness length, zo,
(in meters), which is a measure of the size of the eddies close to the ground level. Various
terrain categories are classified in Table 7.1 according to their approximate roughness lengths.
Table 7.1. Roughness length zo, in meters, for various terrain categories 1)
Terrain
Terrain description
category
0
Sea or coastal area exposed to the open sea
I
Lakes or flat and horizontal area with negligible vegetation and
without obstacles
II
Area with low vegetation such as grass and isolated obstacles
(trees, buildings) with separations of at least 20 obstacle heights
III
Area with regular cover of vegetation or buildings or with
isolated obstacles with separations of maximum 20 obstacle
heights (such as villages, suburban terrain, permanent forest)
IV
Area in which at least 15 % of the surface is covered with
buildings and their average height exceeds 15 m
1)

2)

zo [m] zmin [m]


0.003
0.01

1
1

0.05

0.3

1.0

10

Smaller values of zo produce higher mean velocities of the wind

2)

For the full development of the roughness category, the terrains of types 0 to III must prevail in the
up wind direction for a distance of at least of 500m to 1000m, respectively. For category IV this
distance is more than 1 km.

z
k r ln

z0
cr z

cr z min

for z min z z max 200 m


(7.5)

for z z min

Where,
kr is terrain factor depending on the roughness length calculated with:

z
k r 0,189 0
0,05

0 , 07

(7.6).

z0 and zmin values are given in the Table 7.1. The values kr are indicated in Table 7.2.
UTCB, Technical University of Civil Engineering, Bucharest

64

STRUCTURAL RELIABILITY AND RISK ANALYSIS 4th Year FILS

Table 7.2. The kr i kr2 factors for various terrain categories


Terrain
category
kr
kr2

II

III

IV

0,155
0,024

0,169
0,028

0,189
0,036

0,214
0,046

0,233
0,054

(a)

(a)

(b)

(b)

Figure 7.2a Category 0 - Sea or coastal area exposed


to the open sea (z0 = 0,003 m)

Figure 7.2b Category II - Area with low vegetation


such as grass and isolated obstacles (trees,
buildings) with separations of at least 20 obstacle
heights (z0 = 0,05 m)

(a)

(a)

(b)

(b)

Figure 7.2c Category III - Area with regular cover


of vegetation or buildings or with isolated obstacles
with separations of maximum 20 obstacle heights
(such as villages, suburban terrain, permanent
forest) (z0=0,3 m)

Figure 7.2d Categoria IV - Area in which at least


15 % of the surface is covered with buildings and
their average height exceeds 15 m (z0 = 1,0 m)

UTCB, Technical University of Civil Engineering, Bucharest

65

STRUCTURAL RELIABILITY AND RISK ANALYSIS 4th Year FILS

The logarithmic profile of the mean wind velocity is valid for moderate and strong winds
(mean velocity > 10 m/s) in neutral atmosphere (where the vertical thermal convection of the
air may be neglected).
Wind velocity will be the same
(almost constant)

Atmospheric boundary layer


The mean wind velocity is increasing
- gradient height (2km)
zref=10m

(almost constant)

The roughness factor is represented in Figure 7.2.

Height above ground z, m

200
180

Terrain category IV

160

Terrain category III

140

Terrain category II

120

Terrain category I

100
Terrain category 0

80
60
40
20
0
0

0.5

1.5

2.5

3.5

Roughness factor, cr(z)

Figure 7.2. Roughness factor, cr ( z ) k r ( z 0 ) ln

z 0

Consequently, the mean wind velocity pressure at height z is defined by:


qm z co2 cr2 z qb

(7.7)

z
2
k r z 0 ln for z min z z max 200 m

2
z 0
cr z

c 2 z z
for z z min
min
r

(7.8)

Where:

UTCB, Technical University of Civil Engineering, Bucharest

66

STRUCTURAL RELIABILITY AND RISK ANALYSIS 4th Year FILS

7.4 Wind turbulence


The turbulence intensity Iv is measure of the fluctuations of wind velocity around the mean
value. The turbulence intensity Iv (z) at height z is defined as the standard deviation of the
turbulence (of the wind velocity v(z,t)) divided by the mean wind velocity at height z, (vm(z)).
I v z

v
v m z

(7.9)

The recommended rules for the determination of Iv (z) are given in Expression (7.10):

for z min z z max 200 m

2,5 ln z
z
I v z
0

for z z min
I v z z min

(7.10)

The values of (Table 7.3) depends of terrain roughness factor (z0, m):

4,5 4,5 0,856 ln z 0 7,5

(7.11)

Table 7.3. Values of


Terrain category

0
2,74

I
2,74

II
2,66

III
2,35

IV
2,12

7.5 Peak values


The peak value of the wind velocity is the maximum expected value to occur within
10minutes of averaging period. Considering a normal distribution of the wind velocity, the
peak velocity is:
vp z vm z g v vm z 1 g I v z

(7.12)

v p z c pv z v m z ;

(7.13)

c pv z

v p z

v m z

1 g I v z 1 3,5 I v z

(7.14)

Where:
g is the peak factor, recommended value g=3,5
cpv(z) is gust factor for mean wind velocity.
The peak velocity pressure qp(z) at height z, which includes mean and short-term velocity
fluctuations, should be determined by:
q p z

1
2
2
vm z 1 g I v z qm z 1 7 I v z
2

UTCB, Technical University of Civil Engineering, Bucharest

67

STRUCTURAL RELIABILITY AND RISK ANALYSIS 4th Year FILS

The gust factor for velocity pressure is the ratio of the peak velocity pressure to the mean
wind velocity pressure.

c pq z

q p z

qm z

1 2 g I v z 1 7 I v z

(7.15)

The peak velocity pressure at the height z above ground is the product of the gust factor, the
roughness factor and the reference velocity pressure.

qp z cpq z qm z cpq z cr2 z qb

(7.16)

q p z ce z q b

(7.17)

The exposure factor is defined as the product of the gust and roughness factors:

ce z cpq z cr2 z

(7.18)

If we cannot neglect the orography effect, the exposure factor ce(z) is:

ce z c02 cr2 z cpq z

(7.19)

200

Inaltimea deasupra terenului

z, m

180
160

Teren categoria IV

140

Teren categoria III

120
100

Teren categoria II

80

Teren categoria I

60

Teren categoria 0

40
20
0
1.0

1.5

2.0

2.5

3.0

3.5

4.0

4.5

5.0

Factorul de expunere, ce(z)

Figure 7.3 Illustrations of the exposure factor ce(z)


7.6 Wind actions
Wind action is the product of wind pressures on the surfaces of buildings and structures, or
the forces produced by wind on buildings and structures. The wind actions are variable
actions in time and act both directly as pressure / suction on exterior surfaces of buildings and
enclosed structures and indirectly on the inner surfaces of buildings and other facilities closed
due to porosity outer surfaces. The pressure / suction and may act directly on the inner
surfaces of buildings and other open facilities. The pressure / suction forces are acting normal
on the surface of construction. In addition, when large areas of the buildings are exposed to

UTCB, Technical University of Civil Engineering, Bucharest

68

STRUCTURAL RELIABILITY AND RISK ANALYSIS 4th Year FILS

the wind, the horizontal frictional forces acting tangentially to the surface can have a
significant effect.
The wind action is classified as a fixed variable action; actions are assessed as wind pressure /
suction or forces and are represented by their characteristic values.
Wind actions on structures and structural elements shall be determined taking account of both
external and internal wind pressures.

7.6.1 Wind pressure on surfaces


The wind pressure acting on the external surfaces,
we Iw c pe q p z e

(7.20)

where:
qp(ze)
ze
cpe
Iw

is the peak velocity pressure;


is the reference height for the external;
is the pressure coefficient for the external;
importance exposure factor for wind actions.

The wind pressure acting on the internal surfaces of a structure, wi, should be obtained from
wi Iw c pi q p z i
(7.21)
where:
qp(zi) is the peak velocity pressure ;
zi
is the reference height for the internal pressure;
cpi
is the pressure coefficient for the internal pressure.

Figure 7.4 Pressure on surfaces


The net pressure on a wall, roof or element is the difference between the pressures on the
opposite surfaces taking due account of their signs. Pressure, directed towards the surface is
taken as positive, and suction, directed away from the surface as negative. Examples are given
in Figure 7.4.

UTCB, Technical University of Civil Engineering, Bucharest

69

STRUCTURAL RELIABILITY AND RISK ANALYSIS 4th Year FILS

7.6.2 Wind forces


The wind forces for the whole structure or a structural component should be determined:
- by calculating forces using force coefficients (i)or
- by calculating forces from surface pressures (ii).
(i)
The wind force Fw acting on a structure or a structural component may be determined
directly by using Expression 7.22:
Fw Iw c d c f q p z e Aref

(7.22)

or by vectorial summation over the individual structural elements by using Expression (7.23):

Fw Iw cd

qp ze Aref

(7.23)

elements

Where:
qp(ze)
cd
cf
Aref

is the peak velocity pressure reference height ze;


is the structural factor;
is the force coefficient for the structure or structural element;
is the reference area of the structure or structural element.

(ii)
The wind force, Fw acting on a structure or a structural element may be determined by
vectorial summation of the forces Fw,e and Fw,i, calculated from the external and internal
pressures using Expressions (7.24) and (7.25) and the frictional forces resulting from the
friction of the wind parallel to the external surfaces, calculated using Expression (7.26).
external forces:
Fw,e cd

w z A
e

ref

(7.24)

surfaces

internal forces:
Fw,i

w z A
i

ref

(7.25)

surfaces

friction forces:
F fr Iw c fr q p z e A fr

(7.26)

where:
cd
we(ze)
wi(zi)
Aref
cfr
Afr

is the structural factor;


is the external pressure on the individual surface at height ze;
is the internal pressure on the individual surface at height zi;
is the reference area of the individual surface;
is the friction coefficient;
is the area of external surface parallel to the wind.

UTCB, Technical University of Civil Engineering, Bucharest

70

STRUCTURAL RELIABILITY AND RISK ANALYSIS 4th Year FILS

8. ACTIONS ON STRUCTURES. SEISMIC ACTION


8.1 Introduction
For implementing the EN 1998, national territories are subdivided into seismic zones,
depending on the local hazard. By definition, the hazard within each zone is assumed to be
constant. For most of the applications of EN 1998, the hazard is described in terms of a single
parameter, i.e. the value of the reference peak ground acceleration.
Seismic hazard level indicated by the National Authorities is a minimum for design.
In Romania, the reference peak ground acceleration, chosen by the National Authorities for
each seismic zone, corresponds to the reference return period T=225years, requirement (or
equivalently the reference probability of exceedance in 50 years, P=20%).

Figure 8.1 Zonation of design peak ground acceleration, ag


The AEd, design value of seismic action is equal with AEk, characteristic value of the seismic
action for the reference return period multiplied with the importance/exposure factor of the
construction:

AE d = I,e AE k

(8.1)

8.2 Representation of the seismic action


Within the EN 1998, the earthquake motion at a given point on the surface is represented by
an elastic ground acceleration response spectrum, henceforth called an elastic response
spectrum. The horizontal seismic action is described by two orthogonal components assumed
as being independent and represented by the same response spectrum.
The elastic response spectrum Se (T) is defined by the following expressions:

UTCB, Technical University of Civil Engineering, Bucharest

71

STRUCTURAL RELIABILITY AND RISK ANALYSIS 4th Year FILS

S e ( T ) a g T

(8.2)

where
Se (T) is the elastic response spectrum;
T
is the vibration period of a linear single-degree-of-freedom system;
ag
is the design ground acceleration (m/s2), from zonation map;
T is the normalized elastic response spectrum of absolute accelerations.
In the case of 0,05% viscous damping, the normalized elastic response spectrum of absolute
accelerations is defined by the following relations, Figure 8.3:

0 1 T

0 T TB

(T) 1

TB < T TC

TC < T TD

T
(T) 0 C
T
T T
(T) 0 C D
T2

TD < T 5s
where:
0
T
TB
TC
TD

TB

(8.3)
(8.4)
(8.5)
(8.6)

maximum dynamic amplification factor of the ag by the structure, 0 =2,5;


is the vibration period of a linear single-degree-of-freedom system;
is the lower limit of the period of the constant spectral acceleration branch;
is the upper limit of the period of the constant spectral acceleration branch;
is the value defining the beginning of the constant displacement response range
of the spectrum.

The values of the periods TB, TC , TD describing the shape of the elastic response spectrum
depend upon the ground type. In Romania, based on seismic records, the zonation of control
period TC (s) is given in Figure 8.2.

Figure 8.2 Zonation of control period TC (s)

UTCB, Technical University of Civil Engineering, Bucharest

72

STRUCTURAL RELIABILITY AND RISK ANALYSIS 4th Year FILS

Table 8.1 Values of the periods TB, TC , TD describing the horizontal elastic response spectra
TC (s)
TB (s)
TD (s)

0,7
0,14
3,0

1,0
0,2
3,0

1,6
0,32
2,0

3
= 2.5
=0,05

2.5

(T)

2
1.75/T

1.5

5.25/T 2

1
T C =0.7s

0.5
0

T B =0.14
0

0.5

T D =3
1

1.5
2
2.5
Perioada T , s

3.5

3
=2.5

=0,05

2.5

(T)

2
2.5/T

1.5

7.5/T 2

1
0.5
T B =0.2 T C =1.0s

0
0

0.5

T D =3

1.5
2
2.5
Perioada T , s

3.5

3
=2.5

=0,05

2.5
4/T

(T)

2
1.5

8/T 2

1
0.5
T B =0.32

0
0

0.5

T C =1.6s
1

T D =2

1.5
2
2.5
Perioada T , s

3.5

Figure 8.3 Shapes of normalized elastic response spectrum for TC = 0,7s, 1,0s and 1,6s (5%
damping)

UTCB, Technical University of Civil Engineering, Bucharest

73

STRUCTURAL RELIABILITY AND RISK ANALYSIS 4th Year FILS

The elastic displacement response spectrum, SDe(T) (in meters), shall be obtained by direct
transformation of the elastic acceleration response spectrum, Se(T), using the following
expression:

T
S De ( T ) S e ( T )
2

(8.7)

8.3 Design spectrum for elastic analysis


The capacity of structural systems to resist seismic actions in the non-linear range generally
permits their design for resistance to seismic forces smaller than those corresponding to a
linear elastic response. To avoid explicit inelastic structural analysis in design, the capacity of
the structure to dissipate energy, through mainly ductile behaviour of its elements and/or other
mechanisms, is taken into account by performing an elastic analysis based on a response
spectrum reduced with respect to the elastic one, henceforth called a ''design spectrum''. This
reduction is accomplished by introducing the behaviour factor q.
Behavior factor q is a factor used for design purposes to reduce the forces obtained from a
linear analysis, in order to account for the non-linear response of a structure, associated with
the material, the structural system and the design procedures. The values of the behaviour
factor q, which also account for the influence of the viscous damping being different from
5%, are given for various materials and structural systems according to the relevant ductility
classes in the various Parts of EN 1998.
For the horizontal components of the seismic action the design spectrum, Sd(T) (in m/s2) shall
be defined by the following expressions:
0

q
S d ( T ) a g 1
T
0 < T TB
(8.9)
TB

T > TB

Sd (T) a g

(T )
0 ,2 a g
q

UTCB, Technical University of Civil Engineering, Bucharest

(8.10)

74

Vous aimerez peut-être aussi