LEC1
LEC1
LEC1
CVL 773
IIT DELHI
CIVIL ENGINEERING MATERIALS.
( Lecture 1)
Introduction & Statistical concepts
B. Bhattacharjee
CIVIL ENGINEERING DEPARTMENT
IIT DELHI
B. Bhattacharjee
DEPARTMENT OF CIVIL ENGINEERING, IIT DELHI
General Outline
Data Reduction √
Normal distribution & examples √
Distribution of sampling mean √
“t” and “F” statistic, example of hypothesis
testing
Binomial distribution and Poisson’s
distribution
Joint and Conditional probability
–1
–18-08-2021
B. Bhattacharjee
3
DEPARTMENT OF CIVIL ENGINEERING, IIT DELHI
15
fi
10
5 Relative frequency
0 f1 f2
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
=fi/Σfi
ΣRelative frequency
Sockets
=Σfi/Σfi=1
B. Bhattacharjee
4
DEPARTMENT OF CIVIL ENGINEERING, IIT DELHI
–2
–18-08-2021
fi
f1 f2
B. Bhattacharjee
5
DEPARTMENT OF CIVIL ENGINEERING, IIT DELHI
–3
–18-08-2021
NDT results
NO. OF SAMPLE: 321 Up to 6th
floor
MEAN: 4.1 Mode 4.17
MEDIAN: 4.1
STANDARD S P
DEV: 0.375
MAXIMUM: 4.9 Mean-3SD
MINIMUM: 1.7 2.94
RANGE: 3.2
NO. OF
CI(3.3LOGN+1) 9
CLASS INT. 0.357
B. Bhattacharjee
DEPARTMENT OF CIVIL ENGINEERING, IIT DELHI
UPVresults Up to 6th
Interval Relative
no. lower upper Frequency frequency
1 1.7 2.06 3 0.01
2 2.06 2.41 0 0
3 2.41 2.77 2 0.01
160
4 2.77 3.13 1 0
140 5 3.13 3.48 5 0.02
6 3.48 3.84 54 0.17
120
7 3.84 4.2 149 0.46
100 8 4.2 4.55 91 0.28
80
9 4.55 4.91 16 0.05
60 Sum 321 1
40
20
0
1 2 3 4 5 6 7 8 9
B. Bhattacharjee
DEPARTMENT OF CIVIL ENGINEERING, IIT DELHI
–4
–18-08-2021
B. Bhattacharjee
9
DEPARTMENT OF CIVIL ENGINEERING, IIT DELHI
NATURE OF VARIATIONS
0.18
(uniform distribution)
0.16
0.1
Distributed population
0.08
fi 1
0.06 relativefrequncy = n
≈
N = ∑ fi
0.04
6
0.02 1
1
Pr( x) = f ( x) =
0
1 2 3 4 5 6
Outcomes
6
–5
–18-08-2021
NATURE OF VARIATIONS
0.18
(triangular distribution)
¾ Two dies
0.16
0.14
0.12
The variations
Relative frequency
0.02
1
0
1 2 3 4 5 6 7 8 9 10 11
1 2 3 4 Outcomes
5 6 7 8 9 10 11
(x) 2 3 4 5 6 7 8 9 10 11 12
NATURE OF VARIATIONS
0.06
(nonlinear distribution)
0.05
¾ Three dies
0.04
The variations
0.03
Leads to a nonlinear
0.02 Distributed population
fi
0.01 relativefrequncy = n
0
N = ∑ fi
1 2 3 4 1
–6
–18-08-2021
NATURE OF VARIATIONS
¾ For large number of
Factors, i.e. x , e.g.,
activities in a project
Frequency
¾The variations
Leads to a normally
Distributed population
Project duration
All critical activities finishes at
All critical activities finishes at
pessimistic duration
optimistic duration
–7
–18-08-2021
NATURE OF VARIATIONS
The error function (also called the Gauss error
function), often denoted by erf, is a function of a
variable z is defined as:
Area under the
curve erf (z) is 1
x−μ for z=∞,
Z= ;
σ ⎛ 1 ⎛ x − μ ⎞2 ⎞
1
⎛ x− μ ⎞
−⎜
2
x−μ
f ( x) = exp ⎜⎜ − ⎜ ⎟ ⎟⎟
σ π ⎝ 2⎝ σ ⎠ ⎠
⎟
⎝ σ ⎠
e = 0 for = −∞ or + ∞ 2
σ
∞
1 ⎛ 1 ⎛ x − μ ⎞2 ⎞ F ( 0, μ ,σ ) = 0
F ( ∞, μ , σ ) = ∫ exp ⎜ − ⎜ ⎟ ⎟⎟dx = 1
⎜
0 σ 2π ⎝ 2⎝ σ ⎠ ⎠
0 ≤ F ( x, μ ,σ ) ≤ 1
NATURE OF VARIATIONS
Mean= Σx/n
s2=Σ(xi-x¯)2/(n-1)
Measure of
σ=s
DEPARTMENT OF CIVIL ENGINEERING, IIT DELHI
16
–8
–18-08-2021
NATURE OF VARIATIONS
Mean= Σx/n
s2=Σ(xi-x¯)2/(n-1)
PERT duration
SD of activity duration.
B. Bhattacharjee
18
DEPARTMENT OF CIVIL ENGINEERING, IIT DELHI
–9
–18-08-2021
Example of PERT
Uncertainty in activity duration arises out of
variation of durations of component work
elements of the activity
General pattern of variation of activity duration can
be visualized by considering a situation where
same activity is repeatedly performed several times
i.e., sufficiently large number of times.
For a given activity which is repeated large
number of say the minimum duration is Dmin and
Maximum duration is Dmax, one can divide the
range Dmax, - Dmin in to several class interval
{NCI=3.3log(n)+1}, & plot a histogram.
B. Bhattacharjee
19
DEPARTMENT OF CIVIL ENGINEERING, IIT DELHI
Series1
100
50
B. Bhattacharjee
20
DEPARTMENT OF CIVIL ENGINEERING, IIT DELHI
–10
–18-08-2021
Beta distribution
two parameter α=2 β=5
B. Bhattacharjee
21
DEPARTMENT OF CIVIL ENGINEERING, IIT DELHI
Beta distribution
two parameter α=2 β=4
β function
B. Bhattacharjee
22
DEPARTMENT OF CIVIL ENGINEERING, IIT DELHI
–11
–18-08-2021
0.25×(17-5)+5=8=mode
1/3×(17-5)+5=9=mean Beta distribution
two parameter α=2 β=4
β function
B. Bhattacharjee
23
DEPARTMENT OF CIVIL ENGINEERING, IIT DELHI
PERT duration
Thus expected (mean) activity duration is
calculated from 3 time information in PERT,
pessimistic duration tp, most likely duration tm,
and optimistic duration to.
B. Bhattacharjee
24
DEPARTMENT OF CIVIL ENGINEERING, IIT DELHI
–12
–18-08-2021
a= minimum duration
b=maximum duration
–13
–18-08-2021
PERT duration
SD of activity duration.
B. Bhattacharjee
27
DEPARTMENT OF CIVIL ENGINEERING, IIT DELHI
NATURE OF VARIATIONS
¾ For large number of
Factors, i.e. x , e.g.,
activities in a project
Frequency
¾The variations
Leads to a normally
Distributed population
Project duration
All critical activities finishes at
All critical activities finishes at
pessimistic duration
optimistic duration
–14
–18-08-2021
PERT duration
SD of activity duration.
B. Bhattacharjee
29
DEPARTMENT OF CIVIL ENGINEERING, IIT DELHI
PERT duration
SD of activity duration.
B. Bhattacharjee
30
DEPARTMENT OF CIVIL ENGINEERING, IIT DELHI
–15
–18-08-2021
PERT duration
x−μ
Z= ;
σ
99.9% of time project
would be completed
by Mean+3×SD of
project duration, i.e.
Z=3
(95% of project will be
completed
Mean+1.65×SD of
project duration, i.e.
Z=1.65
In general p(D) for
D=μproj +Z×σproj from
graph or table
B. Bhattacharjee
31
DEPARTMENT OF CIVIL ENGINEERING, IIT DELHI
m xj m ∑x ji
Population mean μ = ∑ =∑ 1
j =1 m j =1 mn
DEPARTMENT OF CIVIL ENGINEERING, IIT DELHI
–16
–18-08-2021
–17
–18-08-2021
Mean= Σx/n
sn 2=s2/n
–Sampling mean
distribution
3sn + 3sn
Number of test
• Higher the number of test (repetitions)
less is the error in measurements
• Number of test required to obtain the
mean (true) results within an error E with
a given confidence level (say 99%) is
given by:
For known population standard deviation
2
• n=(Z×σP/E)
• For unknown population standard
deviation
2
• n=(t×s/E)
DEPARTMENT OF CIVIL ENGINEERING, IIT DELHI
36
–18
–18-08-2021
THANK YOU
FOR HEARING
–19