Assignment 1
Assignment 1
1. Derivation
The equations of motion for MDOF systems of n degrees of freedom under ground acceleration and no damping:
[M]{ x␒␒ } + [K]{ x } = - { m }x␒␒g (1.1)
Where:
k11 k12 ⋯ k1n
k k
[K] = 21 22
⫶ ⋱
(1.2)
kn1 knn
m1
m2
[M] = (1.3)
⋱
mn
m1
m2
{m} = (1.4)
⋮
mn
Natural modes of vibration is in absence of external excitation:
[M]{ x␒␒ } + [K]{ x } = 0 (1.5)
Assuming, that vibration is some constant mode shape { 𝜙 } that only varies sinusoidally with time:
{ x } = { 𝜙 }sin(ωt) (1.6)
↓
(1.8) is an eigenvalue problem, and solutions to { x } are only non-trivial when the determinant is zero:
The charactersitic polynomial of (1.9) has n roots (or eigenvalues) ω12 ω22 ω32 ... ωn2 where n is the number of degrees of
freedom. Each root ωi2 corresponds to a particular solution (or eigenvector) { 𝜙 }i to the vector { x } in (1.8), that is, by using (1.8):
Page 1/15
Name: Hazem Emam Kassab CES611 Wind and Earthquake Engineering
ID: 2201011 Assignment 1
The solution to equation (1.1) is then generally a linear combination (superposition) of all modes (1.12) each scaled by a scaler
yn (t) that varies with time:
{ x } = [𝜙]{ y } (1.14)
Where [𝜙] is (1.12) and:
y1
y
{y} = 2
⋮
yn
Similarly:
{ x␒␒ } = [𝜙]{ y␒␒ } (1.15)
Page 2/15
Name: Hazem Emam Kassab CES611 Wind and Earthquake Engineering
ID: 2201011 Assignment 1
Equation (1.1) can be seen as collection of SDOF systems (1.18) when transformed to the normal coordinate system { y }. That is
expansion of (1.18) yields:
Ln
y␒␒ n + ωn2 yn = x␒␒ g (1.22)
mn*
If x␒␒g can be represented as a function, then closed form solutions can be obtained. However, ground acceleration x␒␒g is often a
time history recorded over discrete time intervals that can not be represented as a mathematical formula. In such case,
numerical methods are resorted to. One way to solve for response, is by discretizing x␒␒g to short impulses and then superpose the
responses generated from all impulses. To do this, we first recall the response of free undamped vibration:
x␒ (0)
x(t) = x(0)cos(ωn t) +
ωn
sin(ωn t) (1.23)
From Netwon's law, the force is the change in momentum. For a constant mass:
d
F=m v → Fdt = mdv (1.24)
dt
For an impulse at t = 𝜏 , (1.27) gives the initial velocity that the impulse causes in the aftercoming free vibration state t > 𝜏 .
Thus, setting x␒ (0) = m1 in (1.23) while x(0) = 0 and denoting the response to unit impulse h(t - 𝜏) :
1
h(t - 𝜏) =
mωn
sin(ωn t) (1.28)
Page 3/15
Name: Hazem Emam Kassab CES611 Wind and Earthquake Engineering
ID: 2201011 Assignment 1
Response to all such impulses is then obtained by integration which leads to Duhamel's integral:
t 1 t F(𝜏)
x(t) = ∫ F(𝜏)h(t - 𝜏)d𝜏 = ∫ sin(ωn (t - 𝜏))d𝜏 (1.30)
0 ωn 0 m
Noting that for earthquakes F(𝜏) = mx␒␒g (𝜏) . Therefore (1.30) becomes:
t
1
x(t) = ∫ x␒␒g (𝜏)sin(ωn (t - 𝜏))d𝜏 (1.31)
ωn 0
For the maximum (spectral) values for response (dropping negative signs):
t
1
x(t)max = Sd = ∫0 x␒␒g (𝜏)sin(ωn (t - 𝜏))d𝜏 (1.35)
ωn max
t
x␒ (t)max = Sv = ∫0 x␒␒g (𝜏)cos(ωn (t - 𝜏))d𝜏 (1.36)
max
t
x␒␒(t)max = Sa = ωn ∫ x␒␒g (𝜏)sin(ωn (t - 𝜏))d𝜏 (1.37)
0 max
Integral in (1.35) and (1.36) are exactly the same except for the trignometric fucntion. It has been shown that bracketed
integral in (1.35) and the spectral velocity integral (1.36) differ very little numerically (see Dynamics of Structures - Ray
Clough, Page 577). Therefore we define the integral in (1.35) be:
t
Spv = ∫0 x␒␒g (𝜏)sin(ωn (t - 𝜏))d𝜏 (1.38)
max
Spv is denoted psueodo-velocity and is very close to the spectral velocity (1.36) except for very long Tn . Therefore, (1.35) is:
1
Sd =
ωn
Spv (1.39)
Similarly, (1.37) is denoted psuedo-acceleration, which - in this special case - is equal to spectral-acceleration (since undamped
motion is assumed):
Spa = Sp = ωn Spv (1.40)
Page 4/15
Name: Hazem Emam Kassab CES611 Wind and Earthquake Engineering
ID: 2201011 Assignment 1
L1
(Spv )1
ω1 m1*
L2
(Spv )2
{ x1 }max { x2 }max ⋯ { xn }max = { 𝜙1 } { 𝜙2 } ⋯ { 𝜙n } ω2 m2* (1.44)
⋱
Ln
(Spv )n
ωn mn*
Or more compactly:
Ln
[x]max = [𝜙] (Spv )n (1.45)
ωn mn*
Ln
[x]max = [𝜙] (S d ) n (1.46)
mn*
The maximum vector of static forces developed can be obtained for a mode n, and recalling that { xn }max = { 𝜙n }(yn )max :
{ fs max }n = [K]{ xn }max = [K]{ 𝜙n }(yn )max (1.47)
Recalling (1.10), then (1.47) becomes (after dropping the minus sign):
Ln
{ fs max }n = [M]{ 𝜙n } ωn (Spv )n (1.49)
mn*
Recalling (1.40), (1.49) becomes:
Ln
{ fs max }n = [M]{ 𝜙n } (Spa )n (1.50)
mn*
Page 5/15
Name: Hazem Emam Kassab CES611 Wind and Earthquake Engineering
ID: 2201011 Assignment 1
L1
(Spa )1
m1*
L2
(Spa )2
{ f1 }max { f2 }max ⋯ { fn }max = [M] { 𝜙1 } { 𝜙2 } ⋯ { 𝜙n } m2* (1.51)
⋱
Ln
(Spa )n
mn*
Noting that [M] = [W] / g ,(1.51) can then also be written as:
L1 (Spa )1
m1* g
L2 (Spa )2
{ f1 }max { f2 }max ⋯ { fn }max = [W] { 𝜙1 } { 𝜙2 } ⋯ { 𝜙n } m2* g (1.52)
⋱
Ln (Spa )n
mn* g
2. Assignment Problem
Given:
200 20.39
200 20.39
200 20.39
{W} =
200
tonf → {m} =
20.39
ton (2.2)
200 20.39
200 20.39
20.39
20.39
20.39
[M] =
20.39
ton (2.3)
20.39
20.39
Page 6/15
Name: Hazem Emam Kassab CES611 Wind and Earthquake Engineering
ID: 2201011 Assignment 1
1 0.522602 144.55
2 0.177642 1251
3 0.11089 3210.5
4 0.084157 5574.1
5 0.071142 7800.3
6 0.064878 9379.2
15 15 15
Height (m)
Height (m)
Height (m)
10 10 10
5 5 5
0 0 0
−0.1 0 0.1
0
0.05
0.1
−0.1
0
0.1
x (m) x (m) x (m)
15 15 15
Height (m)
Height (m)
Height (m)
10 10 10
5 5 5
0 0 0
−0.1
0
0.1
−0.1
0
0.1
−0.1
0
0.1
Page 7/15
Name: Hazem Emam Kassab CES611 Wind and Earthquake Engineering
ID: 2201011 Assignment 1
Response Spectrum
Tn (sec) Spa / g
0 0.25
0.04 0.625
0.23 0.625
0.4 0.355114
0.7 0.248016
0.8 0.190549
Response Spectrum
0.6
0.5
Spa(m/s2)
0.4
0.3
0.2
0 0.2 0.4 0.6 0.8
Tn(sec)
Psuedo-accelaration
Mode Tn (sec) Spa / g Spa m / sec 2
Page 8/15
Name: Hazem Emam Kassab CES611 Wind and Earthquake Engineering
ID: 2201011 Assignment 1
3.05430
6.13125
6.13125
{ Spa } =
6.13125 (2.6)
6.13125
6.13125
{ L } = [𝜙] T { m } (2.8)
Substituting (2.1) (transposed) and (2.2) in (2.8):
10.312
-3.3015
1.8140
{L} =
-1.1093
(2.9)
-0.6572
0.3086
Obtaining modal mass matrix M* by substituting (2.3) and (2.1) in (1.19) and :
m1*
1
m2*
1
m3* 1
M * T
= [𝜙] [M][𝜙] = =
1
= [I] (2.10)
m4*
1
m5*
1
m6*
Page 9/15
Name: Hazem Emam Kassab CES611 Wind and Earthquake Engineering
ID: 2201011 Assignment 1
Ln
Calculating the matrix of diagonal elements (S d ) n using (2.7), (2.9) and (2.10):
mn*
L1
(S d ) 1
m1*
L2
(S d ) 2
m2*
0.217962
L3
(S d ) 3 -0.01619
m3* 0.003465
L4
=
-0.00122
(2.11)
(S d ) 4 -0.00052
m4*
0.000202
L5
(S d ) 5
m5*
L6
(S d ) 6
m6*
Finally, max displacements in material cooridnates x can be obtained by (1.46). Multiplying (2.1) by (2.11):
Floor/Mode 1 2 3 4 5 6 RSS
6 2.66 × 10 -2 -1.86 × 10 -3 3.5 × 10 -4 -9.94 × 10 -5 2.95 × 10 -5 -5.93 × 10 -6 2.66 × 10 -2
Page 10/15
Name: Hazem Emam Kassab CES611 Wind and Earthquake Engineering
ID: 2201011 Assignment 1
Each column in (2.14) represents maximum forces for each mode shape. Reversing the order of (2.14) so that first row is
last floor and so on.
Column Shear can be cacluated by cumulating (from top to bottomm) maximum story forces.
Floor/Mode 1 2 3 4 5 6 RSS
6 78.32 -47.41 22.93 -11.30 4.69 -1.13 95.18
5 152.09 -70.98 16.26 2.72 -5.33 2.01 168.74
4 217.02 -58.84 -11.40 10.64 1.36 -2.42 225.42
3 269.34 -17.11 -24.34 -5.29 3.78 2.28 271.07
2 306.01 33.23 -5.87 -9.37 -5.66 -1.62 308.06
1 324.89 66.85 20.18 7.55 2.65 0.58 332.41
Table 2: Maximum column shear (tonf)
Page 11/15
Name: Hazem Emam Kassab CES611 Wind and Earthquake Engineering
ID: 2201011 Assignment 1
18 18 18
15 15 15
Height (m)
Height (m)
Height (m)
12 12 12
9 9 9
6 6 6
3 3 3
0 0 0
−50 0 50 −20 0 20
0
100
200
300
18 18 18
15 15 15
Height (m)
Height (m)
Height (m)
12 12 12
9 9 9
6 6 6
3 3 3
0 0 0
−10 0 10 −5 0 5 −2 0 2
Page 12/15
Name: Hazem Emam Kassab CES611 Wind and Earthquake Engineering
ID: 2201011 Assignment 1
18 18 18
15 15 15
Height (m)
Height (m)
Height (m)
12 12 12
9 9 9
6 6 6
3 3 3
0 0 0
−500 0 500 −100 0 100
−40
−20
0
20
40
18 18 18
15 15 15
Height (m)
Height (m)
Height (m)
12 12 12
9 9 9
6 6 6
3 3 3
0 0 0
−10
−5
0
5
10
−4
−2
0
2
4
−20
−10
0
10
20
Page 13/15
Name: Hazem Emam Kassab CES611 Wind and Earthquake Engineering
ID: 2201011 Assignment 1
Floor/Mode 1 2 3 4 5 6
6 1409.79 -853.42 412.72 -203.36 84.43 -20.42
5 1106.56 -353.47 -100.01 210.32 -150.29 47.15
4 779.18 145.62 -331.89 95.01 80.32 -53.18
3 470.87 375.58 -116.53 -143.37 21.74 42.35
2 219.99 302.02 110.85 -24.46 -56.62 -23.40
1 56.64 100.87 78.15 50.74 24.92 6.61
∑ (M o ) i 4043.03 -282.79 53.29 -15.12 4.49 -0.90
i=1
Page 14/15
Name: Hazem Emam Kassab CES611 Wind and Earthquake Engineering
ID: 2201011 Assignment 1
T1 = 0.5226 sec > 2Tc = 0.46 sec and number of floors is greater than two. Therefore, 𝜆 = 1
Base shear is:
F b = S d (T 1 ) · 𝜆 W / g (2.17)
Where Sd (T1 ) = SPa 1 = 2.8971 m / sec 2 and W / g = 20.39 × 6 = 122.34 ton . Therefore, (2.17) is:
Fb = 2.8971 × 1 × 122.34 = 354.43 tonf (2.18)
Page 15/15