CE 809 - Structural Dynamics

Lecture 4: Response of SDF Systems to Periodic Loading

Semester – Fall 2021

Dr. Fawad A. Najam Prof. Dr. Pennung Warnitchai

Department of Structural Engineering Head, Department of Civil and Infrastructure Engineering
NUST Institute of Civil Engineering (NICE) School of Engineering and Technology (SET)
National University of Sciences and Technology (NUST) Asian Institute of Technology (AIT)
H-12 Islamabad, Pakistan Bangkok, Thailand
Cell: 92-334-5192533, Email: [email protected]
 Periodic Loading
A SDF system is subjected to a “periodic force” 𝑝 𝑡

𝑝 𝑡 = 𝑝 𝑡+𝑇 𝑇 is the period of the periodic force.

𝑝𝑡

𝑇 𝑇 𝑇

• A periodic function is one in which the portion defined over a time 𝑇 repeats itself indefinitely
as shown in the figure.

• Many forces are periodic or nearly periodic. For example, under certain conditions, propeller
forces on a ship, wave loading on an offshore platform, and wind forces induced by vortex
shedding on tall, slender structures are nearly periodic.
2
 Fourier Series Representation of a Periodic Function

Any arbitrary periodic functions can be represented in terms of a summation of simple

sine and cosine functions.

∞ ∞

𝑝 𝑡 = 𝑎𝑜 + 𝑎𝑛 cos 𝑛 𝜔 𝑡 + 𝑏𝑛 sin 𝑛 𝜔 𝑡 ………. (1)

𝑛=1 𝑛=1

Where 𝜔 = 2 𝜋 𝑇 and 𝑎𝑜 , 𝑎1 , 𝑎2 , 𝑎3 , …, 𝑏1 , 𝑏2 , 𝑏3 , … are called Fourier coefficients.

The right hand side of the above expression is called “Fourier series”, i.e. a periodic function
can be separated (decomposed) into its harmonic components in the Fourier series.

3
 Fourier Decomposition
• This concept called Fourier decomposition was first
proposed by Jean-Baptiste Joseph Fourier, a French
physicist and mathematician (1768 - 1830).

• The beginnings on Fourier series can also be found in

works by Leonhard Euler and by Daniel Bernoulli, but it
was Fourier who employed them in a systematic and
general manner in his main work, “Théorie analytique de
la chaleur (Analytic Theory of Heat, Paris, 1822)”.

• It is a very powerful mathematical concept.

Refer to “Advanced Engineering Mathematics” by Erwin Kreszig, 10th Edition).

Joseph Fourier (1768 - 1830)

4
 Fourier Series

If 𝑝(𝑡) is given, the coefficients 𝑎𝑛 and 𝑏𝑛 can be determined by simple integrations as

follows.

𝑡=𝑇 𝑡=𝑇 ∞ ∞

𝑝 (𝑡) 𝑑𝑡 = 𝑎𝑜 + 𝑎𝑛 cos 𝑛 𝜔 𝑡 + 𝑏𝑛 sin 𝑛 𝜔 𝑡 𝑑𝑡 = 𝑎𝑜 𝑇

𝑡=0 𝑡=0 𝑛=1 𝑛=1

𝑡=𝑇
1
𝑎𝑜 = 𝑝 (𝑡) 𝑑𝑡 ………. (2)
𝑇 𝑡=0

5
 Fourier Series
𝑡=𝑇
𝑝(𝑡) cos 𝑚 𝜔 𝑡 𝑑𝑡
𝑡=0

𝑡=𝑇 ∞ ∞
𝑎𝑚 𝑇
= 𝑎𝑜 + 𝑎𝑛 cos 𝑛 𝜔 𝑡 + 𝑏𝑛 sin 𝑛 𝜔 𝑡 cos 𝑚 𝜔 𝑡 𝑑𝑡 =
𝑡=0 2
n=1 n=1

𝑡=𝑇
Therefore, 2 ………. (3)
𝑎𝑚 = 𝑝 (𝑡) cos 𝑚 𝜔 𝑡 𝑑𝑡
𝑇 𝑡=0

Similarly, it can be shown that,

𝑡=𝑇
2
𝑏𝑚 = 𝑝 (𝑡) sin 𝑚 𝜔 𝑡 𝑑𝑡 ………. (4)
𝑇 𝑡=0 6
 Example
Consider a periodic square function as shown below.

𝑘 𝑓𝑜𝑟 0 < 𝑡 < 𝜋

𝑝𝑡 =
−𝑘 𝑓𝑜𝑟 𝜋 < 𝑡 < 2𝜋

𝑝𝑡
𝑘 𝑘 𝑘

−𝑘 −𝑘 −𝑘
2𝜋
7
 Example
Conducting the integrations as shown be the equations (2), (3) and (4), we obtain,

𝑝𝑡
𝑘 𝑘 𝑘 𝑎𝑜 = 0
𝑎𝑛 = 0
𝑡
2𝑘
𝑏𝑛 = 1 − cos 𝑛 𝜋 𝑛 = 1, 2, 3, … ∞
−𝑘 −𝑘 −𝑘 𝑛𝜋
2𝜋

That is,
4𝑘 4𝑘 4𝑘
𝑏1 = , 𝑏2 = 0, 𝑏3 = , 𝑏4 = 0, 𝑏5 = ,…
𝜋 3𝜋 5𝜋

8
𝑆1 = 𝑏1 sin(𝜔 𝑡)

𝑆3 = 𝑏1 sin(𝜔 𝑡) + 𝑏3 sin(3 𝜔 𝑡)

𝑆5 = 𝑏1 sin(𝜔 𝑡) + 𝑏3 sin 3 𝜔 𝑡 + 𝑏5 sin(5 𝜔 𝑡)

The first three partial sums of the corresponding

Fourier series of the given square periodic function9
10
10

5
𝑆5 5
𝑆13

2 4 6 8 10 12 2 4 6 8 10 12

5
5

10
10

10 10

𝑆9 𝑆25
5 5

2 4 6 8 10 12
2 4 6 8 10 12

5
5

10
10
10
 Example
The series coverage quickly to the square function.

Theoretically, an infinite number of terms are required for the Fourier series to converge to 𝑝(𝑡) .

In practice, however, a few terms are sufficient for good convergence.

Therefore, in many practical applications, it is not necessary to evaluate ∞ series. Only a finite
series is good enough.
𝑁

𝑝 (𝑡) ≅ 𝑏𝑛 sin 𝑛 𝜔 𝑡 Where N is finite, not ∞

𝑛=1

11
 Response to a Periodic Loading

Response to a Response to the Fourier

periodic loading
= series of the loading

Superposition

the sum of the responses to each

= sine and cosine loadings in the series

12
 Response to a Periodic Loading
Superposition

Let 𝑢1(𝑡) be response to 𝑝1(𝑡) loading i.e.

𝑚 𝑢1(𝑡) + 𝑐 𝑢1(𝑡) + 𝑘 𝑢1 (𝑡) = 𝑝1 (𝑡)

And 𝑢2 (𝑡) be the response to 𝑝2(𝑡) i.e.

𝑚 𝑢2(𝑡) + 𝑐 𝑢2(𝑡) + 𝑘 𝑢2(𝑡) = 𝑝2 (𝑡)

Then 𝑢1(𝑡) + 𝑢2 (𝑡) is the response to 𝑝1(𝑡) + 𝑝2 (𝑡).

𝑚 𝑢1(𝑡) + 𝑢2(𝑡) + 𝑐 𝑢1(𝑡) + 𝑢2(𝑡) + 𝑘 𝑢1(𝑡) + 𝑢2(𝑡) = 𝑝1(𝑡) + 𝑝2 (𝑡)

13
 Steady-state Response to a Periodic Loading
∞ ∞

𝑝 (𝑡) = 𝑎𝑜 + 𝑎𝑛 cos 𝑛 𝜔 𝑡 + 𝑏𝑛 sin 𝑛 𝜔 𝑡

𝑛=1 𝑛=1

𝑎𝑜
𝑢𝑜𝑎 =
𝑘
Define 𝛽𝑛 = 𝑛 𝜔/𝜔 and use the result obtained from the previous section.

𝑢𝑏𝑛 (𝑡) = steady-state response to 𝑏𝑛 sin 𝑛𝜔𝑡

𝑏𝑛 1
𝑢𝑏𝑛 (𝑡) = 1 − 𝛽𝑛 2 sin 𝑛 𝜔 𝑡 − 2 𝜉 𝛽𝑛 cos 𝑛 𝜔 𝑡
𝑘 1−𝛽 2 2+ 2𝜉𝛽 2
𝑛 𝑛

14
 Steady-state Response to a Periodic Loading

𝑏𝑛 1
𝑢𝑏𝑛 (𝑡) = 1 − 𝛽𝑛 2 sin 𝑛 𝜔 𝑡 − 2 𝜉 𝛽𝑛 cos 𝑛 𝜔 𝑡
𝑘 1−𝛽 2 2+ 2𝜉𝛽 2
𝑛 𝑛

𝑎𝑛 1
𝑢𝑎𝑛 (𝑡) = 2 𝜉 𝛽𝑛 sin 𝑛 𝜔 𝑡 + 1 − 𝛽𝑛 2 cos 𝑛 𝜔 𝑡
𝑘 1−𝛽 2 2+ 2𝜉𝛽 2
𝑛 𝑛

15
 Steady-state Response to a Periodic Loading

The combined response would be,

1 0
𝑢 (𝑡)
𝑡 = 𝑎
𝑘
∞
1
+ 𝑎𝑛 2 𝜉 𝛽𝑛 + 𝑏𝑛 1 − 𝛽𝑛 2 sin 𝑛 𝜔 𝑡
2 2 2
𝑛=1 1 − 𝛽𝑛 + 2 𝜉 𝛽𝑛

16
 Example

Response of an SDF structure with 𝜔 = 5

rad/sec when subjected to a periodic loading of

triangular waveform (𝜔 = 1 rad/sec)

Inputs: 𝜔 = 1, 𝜔 = 5 rad/sec
Fourier Series:

𝜔 𝜔 𝜔
𝛽1 = = 0.2, 𝛽3 = 3 = 0.6, 𝛽5 = 5 = 1, …
𝜔 𝜔 𝜔

For 𝛽5 term, the response will be dominated by the

resonance response at frequency 5 𝜔.
An example steady state response of an
input triangular force
17
Appendix

Fourier Series of some Common Periodic Functions

18
Function:

𝑎 𝑓𝑜𝑟 0 < 𝑥 < 𝜋

𝑓𝑓 𝑥𝑥 =
−𝑎 𝑓𝑜𝑟 − 𝜋 < 𝑥 < 0

Fourier series:

4𝑎 sin 𝑥 sin 3𝑥 sin 5𝑥

𝑓𝑥 = + + +⋯
𝜋 1 3 5

19
Function:

𝑎 𝑓𝑜𝑟 𝑑 < 𝑥 < 𝜋 − 𝑑

𝑓𝑓 𝑥𝑥 =
−𝑎 𝑓𝑜𝑟 𝜋 + 𝑑 < 𝑥 < 2𝜋 − 𝑑

Fourier series:

4𝑎 1 1
𝑓𝑓 𝑥𝑥 = cos 𝑑 sin 𝑥 + cos 3𝑑 sin 3𝑥 + cos 5𝑑 sin 5𝑥 + ⋯
𝜋 3 5

20
Function:

𝑎 𝑓𝑜𝑟 𝑑 < 𝑥 < 2𝜋 − 𝑑

𝑓 𝑥𝑥 =
0 𝑓𝑜𝑟 0 < 𝑥 < 𝑑 , 2𝜋 − 𝑑 < 𝑥 < 2𝜋

Fourier series:

2𝑎 𝜋 − 𝑑 sin 𝜋 − 𝑑 sin 2 𝜋 − 𝑑 sin 3 𝜋 − 𝑑

𝑓𝑓 𝑥𝑥 = − cos 𝑥 + cos 2𝑥 − +⋯
𝜋 2 1 2 3

21
Function:

2𝑎𝑥 𝜋 𝜋
𝑓𝑜𝑟 − ≤ 𝑥 ≤
𝑓𝑓 𝑥𝑥 = 𝜋 2 2
2𝑎(𝜋 − 𝑥) 𝜋 3𝜋
𝑓𝑜𝑟 ≤ 𝑥 ≤
𝜋 2 2

Fourier series:

8𝑎 sin 𝑥 sin 3𝑥 sin 5𝑥

𝑓𝑓 𝑥 = 2 − 2
+ 2
−⋯
𝜋 1 3 5

22
Function:

𝑎𝑥
𝑓𝑜𝑟 0 ≤ 𝑥 ≤ 𝜋
𝜋
𝑓𝑓 𝑥 = 𝑎(2𝜋 − 𝑥)
𝑓𝑜𝑟 𝜋 ≤ 𝑥 ≤ 2𝜋
𝜋

Fourier series:

𝑎 4𝑎 cos 𝑥 cos 3𝑥 cos 5𝑥

𝑓𝑓 𝑥 = − 2 + 2
+ 2
+⋯
2 𝜋 1 3 5

23
Function:

𝑎𝑥
𝑓𝑓 𝑥𝑥 = 𝑓𝑜𝑟 − 𝜋 < 𝑥 < 𝜋
𝜋

Fourier series:

2𝑎 sin 𝑥 sin 2𝑥 sin 3𝑥

𝑓𝑓 𝑥𝑥 = − + −⋯
𝜋 1 2 3

24
Function:

𝑎𝑥
𝑓𝑜𝑟 0 ≤ 𝑥 ≤ 𝜋
𝑓𝑓 𝑥𝑥 = 𝜋
0 𝑓𝑜𝑟 𝜋 ≤ 𝑥 ≤ 2𝜋

Fourier series:

𝑎 2𝑎 cos 𝑥 cos 3𝑥 cos 5𝑥 𝑎 sin 𝑥 sin 2𝑥 sin 3𝑥

𝑓𝑓 𝑥 = − 2 + 2
+ 2
+⋯ + − + −⋯
4 𝜋 1 3 5 𝜋 1 2 3

25
Function:

𝑎𝑥
𝑓𝑓 𝑥𝑥 = 𝑓𝑜𝑟 0 < 𝑥 < 2𝜋
2𝜋

Fourier series:

𝑎 𝑎 sin 𝑥 sin 2𝑥 sin 3𝑥

𝑓𝑓 𝑥𝑥 = − + + +⋯
2 𝜋 1 2 3

26
Function:

𝑓𝑓 𝑥𝑥
𝑎𝑥
−𝑏 ≤ 𝑥 ≤ 𝑏
𝑑
𝑎 𝑏 ≤𝑥 ≤𝜋−𝑏
= 𝑎 𝜋−𝑥
𝑓𝑜𝑟 𝜋 − 𝑏 < 𝑥 ≤ 𝜋 + 𝑏
𝑑
−𝑎 𝑓𝑜𝑟 𝜋 + 𝑏 < 𝑥 ≤ 2𝜋 − 𝑏

Fourier series:

𝑓𝑓 𝑥𝑥 =?
Find yourself 27
Function:

𝑎 sin 𝑥 𝑓𝑜𝑟 0 ≤ 𝑥 ≤ 𝜋
𝑓𝑓 𝑥𝑥 =
0 𝑓𝑜𝑟 𝜋 ≤ 𝑥 ≤ 2𝜋

Fourier series:

2𝑎 1 𝜋 sin 𝑥 cos 2𝑥 cos 4𝑥 cos 6𝑥

𝑓𝑓 𝑥 = + − − − −⋯
𝜋 2 4 1×3 3×5 5×7

28
Function:

𝑎 cos 𝑥 𝑓𝑜𝑟 0 < 𝑥 < 𝜋

𝑓𝑓 𝑥𝑥 =
−𝑎 cos 𝑥 𝑓𝑜𝑟 − 𝜋 < 𝑥 < 0

Fourier series:

8𝑎 sin 2𝑥 2 sin 4𝑥 3 sin 6𝑥

𝑓𝑓 𝑥𝑥 = + + +⋯
𝜋 1×3 3×5 5×7

29
Function:

𝑓𝑓 𝑥𝑥 = 𝑥 2 𝑓𝑜𝑟 − 𝜋 ≤ 𝑥 ≤ 𝜋

Fourier series:

𝜋2 cos 𝑥 cos 2𝑥 cos 3𝑥

𝑓𝑓 𝑥 = −4 − 2
+ 2
−⋯
3 1 2 3

30
Function:

𝑓𝑓 𝑥𝑥 = 𝑎 sin 𝑥 𝑓𝑜𝑟 − 𝜋 < 𝑥 < 𝜋

Fourier series:

2𝑎 4𝑎 cos 2𝑥 cos 4𝑥 cos 6𝑥

𝑓𝑓 𝑥 = − + + +⋯
𝜋 𝜋 1×3 3×5 5×7

31
Thank you