0 Голоса «за»0 Голоса «против»

Просмотров: 018 стр.Sep 11, 2018

© © All Rights Reserved

PDF, TXT или читайте онлайн в Scribd

© All Rights Reserved

Просмотров: 0

© All Rights Reserved

- Neuromancer
- The E-Myth Revisited: Why Most Small Businesses Don't Work and
- How Not to Be Wrong: The Power of Mathematical Thinking
- Drive: The Surprising Truth About What Motivates Us
- Chaos: Making a New Science
- The Joy of x: A Guided Tour of Math, from One to Infinity
- How to Read a Person Like a Book
- Moonwalking with Einstein: The Art and Science of Remembering Everything
- The Wright Brothers
- The Other Einstein: A Novel
- The 6th Extinction
- The Housekeeper and the Professor: A Novel
- The Power of Discipline: 7 Ways it Can Change Your Life
- The 10X Rule: The Only Difference Between Success and Failure
- A Short History of Nearly Everything
- The Kiss Quotient: A Novel
- The End of Average: How We Succeed in a World That Values Sameness
- Made to Stick: Why Some Ideas Survive and Others Die
- Algorithms to Live By: The Computer Science of Human Decisions
- The Universe in a Nutshell

Вы находитесь на странице: 1из 18

Systems Subjected to Periodic Loading

Asrat Worku

Associate Professor, Department of Civil Engineering,

Addis Ababa University, Ethiopia

Currently Operations Manager for Geotechnical Engineering, Gibb

International, Nairobi, Kenya

asratie@gmail.com

ABSTRACT

The dynamic analysis of multi-degree-of-freedom systems interacting with the foundation soil

and acted by a system of periodic loads is considered. Such systems belong to the category of

non-classically damped systems. The method of modal superposition is employed for

coordinate transformation. The resulting coupled algebraic equations are easily solved

analytically without resorting to iterations, which would otherwise be necessary in time-

domain analyses for irregular excitations like earthquake ground motions. A practical method

of compiling the non-proportional damping matrix is adapted for the intended purpose. The

proposed method of solution is convenient to use on spreadsheets, with MATLAB or to

integrate it with existing software for routine structural analyses. The application is illustrated

in detail on a simplified model of a symmetrically-loaded frame founded on a rigid mat. The

method is equally applicable to systems the superstructure of which is discretized using finite-

element method and the foundation system consisting of a set of isolated rigid footings or

piles under isolated rigid pile caps.

KEYWORDS: Non-proportional damping, geometric damping, modal superposition,

periodic loading, soil-structure interaction.

INTRODUCTION

The method of modal superposition is widely recognized as an efficient technique of

transforming the coupled differential equations of motion of classically damped linear multi-

degree-of-freedom (MDF) structural systems in geometric coordinates into a set of uncoupled

differential equations in modal coordinates. A major reason for its efficiency is that one can

represent the damping matrix in such systems proportional to the mass and the stiffness matrices

so that the uncoupling succeeds. Such damping often employed in practice is the familiar

Rayleigh damping. One is justified to assume a proportional damping as far as the energy

dissipating mechanism is fairly uniform in the entire vibrating system (Clough and Penzien 1993;

Chopra 1995; Wilson 2000).

- 3835 -

Vol. 17 [2012], Bund. Z 3836

Cases of vibrating systems, where the use of a proportional damping is not valid or justified,

include structures, the foundations of which interact with the soil, structures constructed of

different materials in distinctly different zones, and structures with additional artificial damping

elements (Clough and Penzien 1993; Chopra 1995). Due to the inherent difference in the

mechanism of energy dissipation in different parts of the system, the damping matrix of such

systems cannot be set proportional to the mass and/or stiffness matrices. Such systems may also

not generally have orthogonal natural modes and are commonly referred to as non-proportionally

or non-classically damped systems.

The dynamic analysis of non-proportionally damped systems is generally more complex than

that of classically damped systems. In structures with foundations interacting with the underlying

soil, for example, the damping in the soil is of entirely different nature from that in the

superstructure. Energy is dissipated in soils in form of geometric and material damping.

Geometric damping could be the most important source of energy dissipation in thick deposits of

soils providing support to machine foundations, in which the deformations are in the order of

fractions of a millimeter. Material damping, which represents energy dissipation due to the cyclic

nature of the loading and inelastic deformation, increases with increasing level of strains. As a

result of this, it could be a major source of energy dissipation in systems subjected to strong

excitations like earthquakes (Gazetas 1991; Pais and Kausel 1988).

Numerical solution methods based on the modal superposition approach are available for

both linear and non-linear systems vibrating under periodic and more general loading but are

limited to classically damped systems (Villaverde 1988; Villaverde and Hanna 1992; Wilson

2000).

This paper deals with the dynamic response of a non-proportionally damped MDF system

subjected to a set of periodic loads. Modal superposition is employed for the coordinate

transformation using the natural modes of the undamped system as the basis. The modes of non-

proportionally damped systems may not generally be orthogonal to each other. This is particularly

true in structures, the foundations of which interact with soil. In such systems, the foundation

impedance functions, and subsequently the stiffness matrix of the system, are frequency

dependent. Simplifying assumptions are necessary to obtain orthogonal real modes.

The orthogonality of the natural modes achieved in this manner results in diagonalization of

the generalized stiffness and mass matrices but not the generalized damping matrix. Due to the

periodic nature of the loading, however, the coupling with respect to damping does not pose a

major problem. As presented in this work in details, it has been possible to arrive at a closed form

solution. For loadings of general nature like earthquake ground motion, an iterative approach

would have been unavoidable as far as analysis in the time domain is of interest (Worku 1996).

The proposed method results in a number of equations in modal coordinates double the

number of modes considered for coordinate transformation, but generally much less than the total

number of degrees of freedom for large structures. The doubling of the number of equations to be

solved is a result of the phase shift between the loading and the response. The modal coordinates

are readily determined by employing conventional methods of solutions of coupled linear

algebraic equations.

The application of the method is illustrated using a simple model of a framed reinforced-

concrete machine foundation subjected to a set of periodic loads. A proportional damping is

employed for the portion of the structure above the foundation and a geometric damping for the

substructure. The latter is based on a wealth of foundation impedance functions compiled from

research works conducted over a period of several decades for different foundation types and

Vol. 17 [2012], Bund. Z 3837

modes of vibration (Das 1993; Beredugo and Novak 1972; Gazetas 1983; Gazetas 1991; Luco

and Westmann 1971; Luco 1974; Novak and Beredugo 1972; Novak and Sachs, 1973; Pais and

Kausel 1988; Rücker, 1982; Veletsos and Wei, 1971; Wong and Luco 1971; Worku 1996).

It is not the intention of this paper to get into the details of how the available impedance

functions are used to determine dynamic spring and dashpot coefficients, as this is

straightforward and readily available in the literature (Das 1993; Gazetas 1983; Worku 1991).

To the knowledge of the author, the problem of non-classically damped systems subjected to

periodic loads has not been solved in the past in such a complete closed form as presented in this

paper.

Anticipated areas of application of the proposed method include framed machine foundations

and other structures subjected to wind, sea wave and acoustic forces of periodic nature resulting

in deformations within the elastic limit. The method can be easily extended to cover non-linear

response following such techniques as devised by Villaverde (1988) and Villaverde and Hanna

(1992). However, it is unlikely that a closed-form solution would be obtained. This aspect is not

within the scope of the paper.

APPROACH TO SOLUTION

Equations of Motion

The vibration of an elastic MDF system subjected to externally applied dynamic loads is

expressed by the well known matrix equation of

where [m] , [c] and [k ] are the mass, damping and stiffness matrices of the system, and {u} and

{ f (t )} are the displacement and external force vectors, respectively.

Modal superposition has proved to be a very efficient tool for solving vibration problems of

both linear and nonlinear dynamic systems (Villaverde 1988; Villaverde and Hanna 1992; Wilson

2000). Widely used commercial software like SAP and ETABS make use of the tool extensively

(Wilson 2000).

In this paper, the solution of Eq. (1) is sought for non-proportionally damped systems

subjected to a system of periodic loads in which a coordinate transformation is made using the

conventional modal substitution of

where [φ ] is the modal matrix of the undamped system and {z} is the vector of modal

coordinates.

The eigenvalue problem of the undamped system is given by the following equation:

Vol. 17 [2012], Bund. Z 3838

([k ] − ω n

2

[m]){φ } = {0} (3)

The natural frequencies, ω n , and the corresponding natural modes, {φn } , for conventional

systems are determined using standard methods. However, the solution procedure for non-

conventional systems is not always straightforward. For example, in cases of structures, the

foundations of which interact with the soil, the foundation impedance functions, and subsequently

the stiffness matrix, are frequency dependent. The solution thus generally demands an iterative

procedure.

One approach for the problem at hand is to establish the frequency-dependant stiffness matrix

based on the frequency, ω0 , of the fundamental harmonic of the exciting force. Another

alternative is to establish the stiffness matrix based on the fundamental frequency of the system,

in which case iteration is unavoidable. The choice is at the discretion of the user. It is known that

the natural modes so obtained satisfy the orthogonality conditions with respect to both the

stiffness and mass matrices (Worku 2005).

Modal Superposition

The coordinate transformation is accomplished by substituting Eq. (2) in Eq. (1) and pre-

multiplying the resulting equation by [φ]T to obtain

Because of the orthogonality conditions, the generalized mass and stiffness matrices in Eq.

(4) are diagonalized, and their elements are M m = {φm } [m]{φm } and K m = {φ m } [k ]{φ m } ,

T T

respectively. The modal force vector has Fm = {φm } { f } as its elements, where {φm } is the mth

T

natural mode.

The modal damping matrix, [C ] = [φ ] [c ][φ ] , remains non-diagonalized because of the non-

T

proportional damping of the system under consideration. Thus, the differential equations of Eq.

(4) remain coupled through the off-diagonal damping terms (Chopra 1995; Clough and Penzien

1993; Worku 1996).

A set of periodic loads with the fundamental excitation period, T0, is considered given by

{ f (t )} = { f (t + jT0 )} (5)

The Fourier-series expansion of Eq. (5) is given by

∞ ∞

{ f (t )} = {a0 } + a j cos( jω0t ) + b j sin( jω0t )

j =1 j =1 (6)

where ω0 = 2π T0 is the frequency of the fundamental harmonic (Chopra 1995; Clough and

Penzien 1993; Kreyszig 2006; Paz 1997).

Vol. 17 [2012], Bund. Z 3839

The individual Fourier coefficients in the vectors of the right-hand side of Eq. (6) are

expressed in terms of the given force function f(t) as

1

f (t )dt

T0

a0 =

T0 0

1

f (t )cos( jω t )dt

T0

aj = 0 (7)

T0 0

1

f (t )sin ( jω t )dt

T0

bj = 0

T0 0

The coefficient a0 is the average value of the function f(t), whereas aj and bj are the

amplitudes of the jth cosine and sine harmonics of frequency jω0, respectively (Chopra 1995;

Clough and Penzien 1993; Kreyszig 2006; Paz 1997).

METHOD OF SOLUTION

Response to periodic loading

When Eq. (6) is substituted back in Eq. (4), one obtains for the mth mode

∞ ∞

M m zm + n Cmn zn + K m zm = a0 m + a jm cos( jω0t ) + b jm sin ( jω0t ) (8)

j =1 j =1

where the coefficients in the modal force of the right-hand side of Eq. (8) are given by

T T T

(9)

and Cmn are the elements of the mth row of the non-diagonal generalized damping matrix. For

linearly elastic systems, the solution of Eq. (8) can be obtained as the superposition of the

solutions to the individual forces on the right-hand side. For a non-linear response, a technique

devised by Villaverde (1988) and Villaverde and Hanna (1992) can be adapted in which the non-

linear terms are taken to the right-hand side of the equation to modify the loading step by step

with due account for the initial conditions at each stage.

The steady-state response to the constant modal force, a0m , for a damped system is given by

(Chopra 1995)

z0 m = a0 m K m (10)

The steady-state response to the single harmonic cosine force of a jm cos( jω 0 t ) is also

harmonic with a phase shift that can be expressed as

z cjm = Acjm sin ( jω 0 t ) + B cjm cos( jω 0 t ) (11)

c c

In Eq. (11), A jm and B jm are coefficients yet to be determined. Inserting Eq. (11) and its

derivatives in Eq. (8), in which only a single cosine harmonic force is considered on the right-

Vol. 17 [2012], Bund. Z 3840

hand side, comparing coefficients of the sine and cosine functions, and solving for the vector of

c c

the unknown coefficients A jm and B jm , one obtains for the mth mode:

{ }

A cjm 1 {a }

c = [E ]−j1 jm

{ } (12)

B jm jω 0 {0}

where, the coefficient matrix [E]j is given by

[C ]

1

[D]

jω 0

[E ] j = (13)

1 [D ] − [C ]

jω 0

The sub-matrix [D] in Eq. (13) is a diagonal matrix with its elements given by

( )

Dmm = M m ω m2 − j 2ω 0 , and the sub-matrix [C] is the non-diagonalized generalized damping

2

c c

With the coefficients A jm and B jm so determined, the modal response of Eq. (11) to the

single harmonic cosine force a jm cos( jω 0 t ) is known.

The steady-state response to the single harmonic sine force of b jm sin ( jω 0 t ) is also

s s

harmonic given by Eq. (11) with another set of coefficients A jm and B jm . Thus,

{ }

A sjm 1 {0}

s = [E ]−j1

{ }

{b jm }

(15)

B jm jω 0

Finally, the steady-state modal response to the given system of periodic excitation is

determined as the combination of the responses to the individual terms given by Eq. (10), Eq.

(11) and Eq. (14). Therefore,

a0m ∞ c ∞

zm = ( ) (

+ A jm + A sjm sin ( jω 0 t ) + B cjm + B sjm cos( jω 0 t )

K m j =1

) (16)

j =1

Once the modal responses are determined as such, the displacement vector is found by

superposing the contributions of all significant modes as

N

{u} = {φ n }z n (17)

n =1

where N is the number of modes significantly contributing to the total response. Note that the

number of algebraic equations in Eq. (12) and Eq. (15) is double the number of modes included in

Vol. 17 [2012], Bund. Z 3841

the solution. Since the number of modes significantly contributing to the overall response is

mostly small, the number of linear algebraic equations to be solved is generally smaller than the

total number of degrees of freedom of the MDF model. This is especially true when the number

of degrees of freedom in the model is large which is the case when the superstructure is modeled

using finite-element discretization or the structure is complex and big in size.

If a given MDF system is instead subjected to a set of harmonic loads, say f i (t ) = ai sin ωt

acting in the ith degree of freedom, with the frequency ω, then Eq. (4) reduces to

M m zm + n C mn z n + K m z m = a m sin ωt (18)

where a m = {φ m } {a} and {a} is the vector of the force amplitudes. The solution of Eq. (18) for

T

the steady-state condition is similar to Eq. (11) and Eq. (14) and given by

z m = Am sin ωt + Bm cos ωt (19)

Inserting Eq. (19) and its derivatives in Eq. (18), and solving for the vector of the open

coefficients, one obtains:

{Am } −1 {a m }

= ω [E ]

−1

(20)

{Bm } {0}

in which

[E ] = ω [D ] −−1[C ]

−1

(21)

[C ] ω [D ]

The sub-matrix [D] in Eq. (21) is a diagonal matrix with its elements given by

( )

Dmm = M m ω m2 − ω 2 , and the sub-matrix [C] is the generalized damping matrix.

Analogously, the displacement vector is found through superposition according to Eq. (17).

This problem was solved earlier separately (Worku 2005) and can thus be considered as a simpler

variant of the present problem of a system subjected to periodic loads.

The non-diagonal generalized damping matrix, which is needed in the coefficient matrix [E]j

of Eq. (12) and Eq. (15) and in [E] of Eq. (20), can be assembled following a procedure outlined

in Clough and Penzien (1993). Accordingly, for structures interacting with the foundation soil,

the damping in the superstructure can be assembled as a proportional damping, whereas the

damping in the substructure is established in accordance with principles of foundation dynamics.

This results in a non-proportional damping matrix of the entire system under consideration.

A proportional damping sub-matrix of the superstructure, [c]s, is commonly established as a

Rayleigh damping:

[c]s = α [k ]s + β [m]s (22)

Vol. 17 [2012], Bund. Z 3842

By specifying two modal damping ratios ξ r and ξ s for two selected modes – the rth and sth

modes - the two constants α and β are determined from

2(ξ s ω s − ξ r ω r ) 2ω s ω r (ξ r ω s − ξ s ω r )

α= ; β= (23)

ωs 2 − ωr 2 ωs 2 − ωr 2

in which ω r and ω s are the corresponding frequencies. If the two modal damping ratios are

specified to be one and the same, say ξ , then Eq. (23) simplify to

2ξ 2ω r ω s ξ

α= ; β= = ω r ω sα (24)

ωr + ωs ωr + ωs

The value of ξ in the above procedure can be easily selected from recommended values in

the literature depending on the nature of the material and structural system. Use of the

fundamental frequency as one of the controlling frequencies is common. The other controlling

frequency is normally recommended as one among the higher frequencies (Clough and Penzien

1993).

The damping sub-matrix corresponding to the substructure is assembled similar to the way

the stiffness sub-matrix is assembled. The elements of this frequency-dependent sub-matrix can

be directly evaluated from available impedance functions depending on the foundation shape,

embedment depth, soil stiffness, stratification, and soil homogeneity (Beredugo and Novak 1972;

Das 1993; Gazetas 1983; Gazetas 1991; Luco and Westmann 1971; Luco 1974; Novak and

Beredugo 1972; Novak and Sachs 1973; Pais and Kausel 1988; Rücker 1982; Veletsos and Wei

1971; Wong and Luco 1971; Worku 1996).

NUMERICAL EXAMPLE

The solution procedure presented above is illustrated now using an example problem of an

idealized framed machine foundation.

The concrete platform shown in Figure 1(a), which has a thickness of 30cm and a rectangular

shape of 3.5m by 10.5m, is subjected to a periodic horizontal force of amplitude 120 kN (Figure

1(a)) and period T0=0.02 seconds as shown in Figure 1(b).

Vol. 17 [2012], Bund. Z 3843

(a)

p(t)

p0

T0/2 T0 2T0

t

(b)

periodic excitation of amplitude p0 and period T0

The platform is supported by six reinforced concrete columns of size 30cm by 30cm and

height 2.5m arranged in three rows parallel to the plane of the paper. A 40cm-thick and 4.3m by

11.3m rigid mat embedded 1.0m into the underlying deep uniform clay formation support the

columns.

The vibration amplitudes of the upper slab are to be determined using a simple planar (2D)

two-mass oscillator model with due consideration of the interaction of the rigid foundation with

the surrounding soil (Figure 2).

Vol. 17 [2012], Bund. Z 3844

u2

u1

mr

mf

Kh

u4

Cr u3

Kr Ch

Consideration of the interaction of the rigid foundation with the soil results in a four-degree-

of-freedom system as shown in Figure 2. The soil may be assumed to have a uniform behavior

with depth and exhibiting a unit weight of 16.8 kN/m3, a Poisson ratio of 0.42, and an initial-

tangent shear modulus of 60 MPa. For the concrete in the superstructure and the foundation, a

unit weight of 24 kN/m3 and an elastic modulus of 25 GPa may be used.

The stiffness and mass matrices are easily established as:

12 EI 6 EI − 12 EI 6 EI

h3

mr 0 0 0 h2 h3 h2

0

6 EI 4 EI − 6 EI 2 EI

Iθr 0 0

[m] = ; [k ] = h

2

h h2 h (E1)

0 0 mf 0 − 12 EI − 6 EI 12 EI − 6 EI

h3 + Kh

0 0 0 Iθf h2 h3 h2

6 EI 2 EI − 6 EI 4 EI

+ Kr

h2 h h2 h

The dynamic stiffness coefficients Kh and Kr corresponding to the horizontal and rocking

movements, respectively, of the foundation are determined from the literature, which provide

these quantities in form of impedance functions (Beredugo and Novak 1972; Gazetas 1983;

Gazetas 1991; Luco 1974; Pais and Kausel 1988; Rücker 1982; Worku 1996). The impedance

functions are generally frequency dependent and are evaluated at the fundamental excitation

frequency in this case.

After inserting the appropriate structural and soil properties in EQ. (E1) and solving the

eigenvalue problem, the spectral and modal matrices are obtained as:

Vol. 17 [2012], Bund. Z 3845

0.0025 0 0 0

0 0.0402 0 0

[λ ] = × 105 sec − 2 (E2a)

0 0 0.3979 0

0 0 0 1.0155

- 0.1049 - 0.1589 - 0.0071 0.0018

[Φ] = (E2b)

0.0006 - 0.0069 0.1448 - 0.0020

- 0.0005 0.0016 0.0017 0.1168

where the elements of [λ] are the eigenvalues λ = ω 2 . The fact that none of the eigenvalues is

identically zero indicates that the natural modes of the non-classically damped system considered

here are orthogonal with respect to the mass and stiffness matrices as indicated earlier.

The natural modes in Eq. (E2b) are sketched in Figure 3, where the associated natural

frequencies are also indicated.

The non-proportional damping matrix for the entire system is found by dividing it into two

sub-matrices: the upper left two-by-two sub-matrix for the superstructure and the lower right two-

by-two sub-matrix for the foundation system. The superstructure damping sub-matrix is

assembled as a proportional damping matrix following a standard procedure established for

classically damped systems, in which a 5% critical damping ratio is employed for this purpose

(Clough and Penzien 1993). The damping sub-matrix corresponding to the foundation system is

established by directly evaluating the frequency-dependent impedance functions, ch (ω ) and

cr (ω ) , available in the literature at the fundamental frequency of the system for the horizontal

translation and rocking degrees of freedom of the foundation (Beredugo and Novak 1972; Das

1993; Gazetas 1983; Gazetas 1991; Luco 1974; Pais and Kausel 1988; Rücker 1982; Worku

1996). In the latter case, the small coupling existing between the two degrees of freedom is often

neglected. The resulting matrix is a non-proportional damping matrix obtained as

Vol. 17 [2012], Bund. Z 3846

0.0576 0.0226 0 0

0.0226 0.078 0 0

[c] = × 10 3 (E3)

0 0 223.48 0

0 0 0 683.52

As the treatment of the impedance functions is beyond the scope of this paper, the reader is

referred to the above list of literature for a detailed treatment, the choice being dependent on the

type of foundation, embedment depth and the geological stratification. The generalized mass,

stiffness and damping matrices are then calculated as:

1 0 0 0 0.0025 0 0 0

0 1 0 0 0 0.0402 0 0

[M ] = = [I ] ; [K ] = × 105

0 0 1 0 0 0 0.3979 0

0 0 0 1 0 0 0 1.0155

- 0.0014 0.0158 - 0.2223 0.1287

[C ] = × 103 (E4)

0.0189 - 0.2223 4.6901 0.0699

- 0.0383 0.1287 0.0699 9.3259

Since the generalized mass and stiffness matrices in Eq. (E4) are identical to the unit and

spectral matrices, respectively, the modal vectors in [Φ] are orthonormal modes. In contrast, it is

important to note that the generalized damping matrix is not diagonalized due to its non-

proportional nature as stated above.

The given loading function of Figure 1(b) can be expressed as

2 p0

T t ; for 0 ≤ t ≤ T0 2

p(t ) = 0 (E5)

2 p0 (1 − t T0 ); for T0 2 ≤ t ≤ T

The coefficients in the Fourier series expansion of this loading function are evaluated using

the integrals in Eq. (7). Accordingly,

Vol.

V 17 [2012], Bund.. Z 384

47

1 2 p0 p0

(1 − t T0 )ddt =

T0 2 T0

a 01 =

T0 T0

0

tdt + 2 p 0

T0 2

2

= 660 kN

2 2 p0

t cos( jω 0 t )dt + 2 p 0 (1 − t T0 ) coos( jω 0 t )dt

T0 2 T0

a j1 =

T0 T0

0 T0 2

(E66)

2

2 48.63 kN

N

a j1 = − p 0 = − ; j = 1, 3, 5, ...

jπ

2

j

b j1 = 0

n of time, thhe sine harmoonics in the Fourier-seriees

expan

nsion of the ho

orizontal load

d vanish.

The

T plot of the Fourier-serries expansion n of the loadding as a funcction of t/T0 iis presented iin

Figure 4 by includiing only threee trigonometrric terms in thhe series.

p of the Fo

ourier-seriess expansion oof the loadinng function bby includingg

three harmonics

h onnly

Comparison

C of Figure 4 with

w Figure 1(b) shows thhat the first ffour terms in the series arre

sufficcient (includin

ng the averagee value a0) to adequately reepresent the ggiven force fuunction.

ment in the dirrection of thee 2nd degree oof

Similarly, the Fourier series coefficientss of the mom

freedo

om can be fo ound by multiplying the ab bove coefficiients by the mmoment arm of 0.35m (seee

Figure 1(a)). As there are no directly app plied forces inn the directioons of the thhird and fourtth

degrees of freedom

m, the corresponding Fourier coefficientts are all zero.

Thus,

T the vecto

ors of the Fou

urier coefficieents of the forrce vector beccome

60 − 48.633

− 21 .02

{a0 } = ; {a j } = 12 17

; {b } = {0}

j

0 j 0

0 0

Vol.

V 17 [2012], Bund.. Z 384

48

The

T vectors off modal Fourieer coefficientts are obtainedd in accordannce with Eq. ((9). Thus,

11.84 − 9.60

− 3.01 44

{a 0m } =

; {a jm } = 12 02..16

(E7)

− 0.19 j

0.09 − 0.08

Next,

N the coeffficient matricces [E]j in Eq. (13) are com

mpiled for j=11, 3, and 5. Thhen, the matriix

operaations of Eq. (12)

( for j=1, 3,

3 5 are performed and thee modal coorddinates determ mined from Eq.

(16). Finally, the displacement

d vector is com

mputed. The pplot of the diisplacement, u1, of the rooof,

against normalized d time, in which

w terms up to and iincluding thee first harmoonics only arre

considdered, is giveen in Figure 5(a).

(a)

( b)

Figu

ure 5: Plots of

o the horizoontal displaceement of thee platform off the flexiblee-base system

m

co

onsidering thhe contributio

ons of the first two moddes only; (a) first three teerms of the

series in

ncluded; (b) all

a the sevenn terms incluuded.

The

T plot of Fig gure 5(b) is th hat of the sam

me displacemeent, where thrree harmoniccs are included.

Both plots

p t the mass oscillates in the range of 15.512mm annd

are pracctically the saame showing that

15.548mm around the static horrizontal displacement of 1 5.53mm. Thiis indicates thhat inclusion oof

the firrst harmonic alone gives sufficiently

s accurate

a resultts. Similar pllots can be obbtained for thhe

Vol.

V 17 [2012], Bund.. Z 384

49

rotatio

on of the ro

oof, the horizzontal displacement of thhe foundationn, and the rrotation of thhe

founddation.

A plot of the displacement,

d u1, of the rooof with the coontribution of all modes aand all Fourieer-

seriess terms, which

h are considerred in the anaalysis, is givenn in Figure 6((a).

( a)

( b)

Figuure 6: Plots of

o the horizo

ontal displaceement of thee platform off the flexiblee-base system

m

as a function of normalized time: (a) byy including thhe contributiions of all m

modes; (b) byy

including

g the contrib

bution of the first mode aalone

The

T plot in FigureF 6(a) iss practically identical to the plot in F Figure 5(b) iindicating that

inclussion of the last

l two mod des in the mode

m superpoosition doesn''t bring abouut any notable

differrence. Figure 6(b) gives the contribution n of the first mode alone tto the displaccement u1. Thhis

plot shows

s he mass oscilllates in the range of 7.5595mm and 77.627mm aroound the static

that th

horizo ontal displaceement of 7.611mm. This is in a significaant deviation from the plots of Figures 5

and 6(a)

6 indicating ng with the fiirst mode alo ne is not suffficient. Hencce, inclusion oof

g that workin

the firrst two modess in the analysis of the probblem under c onsideration is necessary.

The

T fixed-basee framed foun ndation is con

nsidered nextt. The fixed-bbase assumptiion reduces thhe

numb ber of degreess of freedom to only two and

a fully ignoores the influuence of the ssoil-foundatioon

interaaction.

Vol.

V 17 [2012], Bund.. Z 385

50

The

T natural freequencies are easily determ

mined by solvving the charaacteristics equuation to obtaiin

ω1 = 16.03sec-1 annd ω 2 = 64.222sec-1. These are higher thhan those obtaained earlier on the basis oof

the fleexible-base model

m as expected.

Since the soill-foundation interaction iss neglected, only the prooportional daamping of thhe

superstructure is considered.

c Thus,

T a criticaal damping rratio of 5% iis taken simiilar to the onne

consid

dered for the superstructurre damping suub-matrix of tthe flexible-baase system.

The

T uncoupled

d differential equations

e for the two moddal coordinatees become

zm + 2ξ m ω m z m + ω m2 z m = Pm (t ) M m ; m = 1,2 (E88)

The

T solution to

o this equation

n for the giveen periodic lo ading is givenn by

z m (t ) =

a0m

+

1

(1 − β ) + (2ξ

a jm

[2ξ β jjm sin ( jω 0 t ) + (1 − β jm

2

)coos( jω0t )]

β jm )

2 2 m

Km Km j =1, 3, 5,... jm m

(E99)

wheree a0m and ajmm are elementts of the genneralized forcce vector andd βjm=jω0/ωm are frequenccy

ratios. Eq. (E9) is evaluated fo or the two moodes and thee horizontal ddisplacement, u1, of the toop

floor calculated. A plot of this displacement

d against norm

malized time iss given in Figgure 7.

Figu

ure 7: Plot of the horizon

ntal displaceement of the platform of the fixed-baase system as

a fun

nction of norrmalized time including bboth modal contributionns

The

T plot show ws that that thiis displacemeent time-histoory is significcantly differennt from that oof

the flexible-base model

m given in Figure 6(aa). While, thee fixed-base m model oscillaates with largge

relativve displacemeents about its static deform

mation of 7.4773mm, the fleexible-base m model oscillatees

little about its staatic deformaation of 15.5 52mm. This implies thatt a significannt part of thhe

displaacement of the roof in the flexible-base

f model is cont

ntributed by thhe soil deformmation.

This

T discrepan ncy in both thhe displacemeent amplitude and the oscillation is rem

miniscent of thhe

d the differencce in the enerrgy dissipationn mechanism

soil fllexibility and ms of the two models. In thhe

fixed--base case, thet system depends

d on the

t little dam mping of thee superstructuure alone annd

Vol. 17 [2012], Bund. Z 3851

undergoes no foundation movement. In the flexible-base system the additional damping in the

soil and the additional displacement due to the deformation of the soil play a major role in

modifying the overall response of the system.

CONCLUSIONS

The material presented in this work shows that the method of modal superposition can be

successfully used to transform the coupled differential equations of motion of non-classically

damped systems subjected to periodic loads into a system of coupled algebraic equations. Though

the number of algebraic equations is double that of the number of significant modes included for

the required accuracy, they it is generally much smaller than the total number of degrees of

freedom especially in large and complex structures. The method of solution of the forced

vibration leads to a closed-form solution and thus does not demand any iterative procedure, which

would have generally been the case for irregular excitations `like earthquake ground motions. An

appropriate technique of compiling a non-classical damping matrix has been properly adapted.

The illustration on a practical example of a framed machine foundation shows that the method

proposed is straightforward to use. The availability of mathematical software like MATLAB

makes the method even more attractive, because the major work involved is mainly limited to

matrix operations. The method of solution can also be used on spreadsheets especially when the

superstructure is modeled as a lumped-mass system. The solution method can be easily integrated

in existing commercial software for structural analysis. It can also have practical applications in

other kinds of systems subjected to periodic loads like slender structures subjected to vortex

shading of wind excitation and other structures subjected to wind, sea wave and acoustic forces of

periodic nature resulting in deformations within the elastic limit. The method can be easily

extended to deal with non-linear response.

REFERENCES

1. Chopra, A. (1995). Dynamics of Structures: Theory and applications to Earthquake

engineering, Prentice Hall, New Jersey

2. Clough, R.W. and Penzien, J. (1993), Dynamics of Structures, Mc Graw-Hill, 2nd

edition, New York

3. Das, B.M. (1993), Principles of Soil Dynamics, Thomson Engineering, PWS-Kent

Publishing Company

4. Gazetas, G. (1983) “Analysis of Machine Foundation Vibration: State of the Art,”

Soil Dynamics and Earthquake Engineering, 2(1), 2-42

5. Gazetas, G. (1991) “Formulas and Charts for Impedances of Surface and Embedded

Foundations.” Journal of Geotechnical Engineering, ASCE, Vol. 117, No. 9

6. Kreyszig, E. (2006), Advanced Engineering Mathematics, 9th Edition, John Wiley,

New York

7. Luco, J.E. and Westmann, R.A. (1971), “Dynamic Response of Circular Footings,”

ASCE, Engineering Mechanics Division, 97(5), 1381-1395

8. Luco, J.E. (1974), “Impedance functions for a Rigid Foundation on a Layered

medium,” Nuclear Engineering and Design, North-Holland Publishing Company, 31,

204-217

Vol. 17 [2012], Bund. Z 3852

ASCE, Soil Mechanics and Foundation Division, 98(12), 1291-1309

10. Beredugo, Y.O. and Novak, M. (1972), “Coupled Horizontal and Rocking Vibration

of Embedded Footings,” Canadian Geotechnical Journal, 9, 477-497

11. Novak, M. and Sachs, K. (1973), “Torsional and Coupled Vibrations of Embedded

Footings,” Earthquake Engineering and Structural dynamics, 2, 11-33

12. Pais, A. and Kausel, E. (1988), “Approximate Formulas for Dynamic Stiffnesses of

Rigid Foundations.” Soil Dynamics and Earthquake Engineering, Vol. 7, No. 4

13. Paz, M. (1997), Structural Dynamics - Theory and Computation, Chapman & Hall,

fourth edition, Boston

14. Rücker, W (1982), “Dynamic Behaviour of Rigid Foundations of Arbitrary Shape on

a Halfspace,” Journal of Earthquake Engineering and Structural Dynamics, 10, 675-

690

15. Veletsos, A.S. and Wei, W.T. (1971), “Lateral and Rocking Vibrations of Footings,”

ASCE, Soil Mechanics and Foundation Division, 97(9), 1227-1248

16. Villaverde, R. (1988), “Modal superposition method for seismic design of non-linear

multistorey structures,” Earthquake Engineering and Structural dynamics, 16, 691-

704

17. Villaverde, R. and Hanna, M. (1992), “Efficient mode superposition algorithm for

seismic analysis of non-linear structures,” Earthquake Engineering and Structural

dynamics, 21, 849-858

18. Wilson, E. D. (2000), Static and Dynamic Analysis of Structures. Computers and

Structures Inc.

19. Wong, H.L. and Luco, J.E (1976), “Dynamic Response of Rigid Foundations of

Arbitrary Shape,” Journal of Earthquake Engineering and Structural Dynamics, 4,

579-587

20. Worku, A. (1996), Vibration of Structures Due to Earthquake Ground Motions with

due Consideration of the Soil-Structure interaction (in German). Doctoral

dissertation, University of Wuppertal, Germany

21. Worku, A. (2005), "A closed-form solution procedure to the vibration of non-

classically damped systems subjected to harmonic loads." Zede, Journal of the

Ethiopian Engineers and Architects, 22, 1-9

© 2012, EJGE

- BACHMANN (1995).pdfЗагружено:Fernanda Lago
- Vibration MonitoringЗагружено:Apurv Khandelwal
- Gas Turbines Materials Modeling and PerformanceЗагружено:kikin_i6364
- Seismic Response of Guyed Telecommunication Towers Literature ReviewЗагружено:zahid nawaz
- A Modal Pushover Analysis Procedure for Estimating Seismic Demands for Buildings - ChopraЗагружено:redpol
- Structural Design and Verification of Locomotive Bogies Using Combined Analytical and Experimental MethodsЗагружено:Prenţa Valeriu Gabriel
- Md Total PapersЗагружено:aeronayak
- Armand_J.Method_for_Evaluatin.1979.TRANS.pdfЗагружено:Ruly Irawan
- Vibration Characteristics of Fluid-filled Cylindrical ShellsЗагружено:ilincast
- Review_Ch1-Fundamental_of_Vibration.pdfЗагружено:Saravanan Sukumaran
- Campbell PDFЗагружено:Anonymous d23gWCRQ
- Technical Vibration 3Загружено:Nicole Mikesell
- Coupled dynamicsЗагружено:kapilkumar18
- Solutions of Circular Membrane With Besse LsЗагружено:spylogo3
- 826785.pdfЗагружено:Jolito Ramos
- Engineering journal ; Analysis of Vibro-Isolated Building Excited by the Technical Seismicity of Traffic EffectsЗагружено:Engineering Journal
- NASA AdministrationsЗагружено:artovolasti
- Paper for StructuresЗагружено:Nono4138
- 2014 - Damage identification-Modal Curvature.pdfЗагружено:paulkohan
- Dynamic Analysis Of A Car Chassis.pdfЗагружено:sultanu89
- Active Vibration Control of Aerospace Structures Using a Modified Positive Position Feedback method.pdfЗагружено:oscar201140
- Suspension Math ModelЗагружено:Santiago Urgiles
- Vibrations of a Taut String With Nonlinear DamperЗагружено:snailbook
- A f 36188193Загружено:Anonymous 7VPPkWS8O
- 226_Shaking Table Testing of a Steel Framce Structure...Загружено:Ing Jaime Espinoza Skinfield
- Tatar 2017Загружено:Francisco Calderon
- AIAAJournalPaperЗагружено:Benjamin Livingston
- Tria Xvi Be Spec DataЗагружено:ho-fa
- 10.1.1.468.2504.pdfЗагружено:huliplay
- 1353303191Загружено:Heidi Webb

- Applying Bi-directional Evolutionary Structural Optimisation Method for Tunnel Reinforcement Design Considering Nonlinear Material BehaЗагружено:Ionut Patras
- ExpertiЗагружено:Monica Stan
- ASME V NDTЗагружено:zagl660802
- 03. CopyrightЗагружено:Ionut Patras
- 04. JETK Article RequirementsЗагружено:Ionut Patras
- Deep Underground Basements Final With Edits - 10-14-03Загружено:Nadim527
- AaaЗагружено:Ionut Patras
- Geoengineering Considerations in the Optimum Use of Underground SpaceЗагружено:Sam Jandali
- A Comparative Analysis Between Disjunctive Kriging AndЗагружено:Ionut Patras
- Coiled Tube Turbodrilling- A Proposed Technology to Optimise Drilling Deep Hard Rocks for Mineral ExplorationЗагружено:Ionut Patras
- Numerical Analyses of the Hangingwall Failure Due to Sublevel Caving- Study CaseЗагружено:Ionut Patras
- Operating Risk Assessment for Underground Metal Mining Systems- Overview and DiscussionЗагружено:Ionut Patras
- Utilising of Linear and Non-linear Prediction Tools for Evaluation of Penetration Rate of Tunnel Boring Machine in Hard Rock ConditionЗагружено:Ionut Patras
- Jankovic-Validation of a closed circuit ball mill model.pdfЗагружено:rodrigoalcaino
- Tunnel Analysis in Fault Zones and the Effects of Stress Distribution on the SupportЗагружено:Ionut Patras
- Current Condition of Open Mining in KazahstanЗагружено:Ionut Patras
- Baseline Studies of Some Heavy Metals in Top Soils Around the Iron-Ore Mining FieldЗагружено:Ionut Patras
- A Review on Novel Techniques for Chalcopyrite Ore Processing.pdfЗагружено:Ionut Patras
- 10.5923.j.mining.20130202.01Загружено:Ionut Patras
- 10.5923.j.mining.20130202.02.pdfЗагружено:MuhYusuf
- Role of Water Structure-Making Breaking Ions in the Cationic Flotation of Kaolinite- Implications for Iron Ore ProcessingЗагружено:Ionut Patras
- Analysis of Circular Tunnels Due to Seismic P-wave Propagation,Загружено:Ionut Patras
- Assessment of soil–pile–structure interaction influencing seismic response of mid-rise buildings sitting on floating pile foundations.pdfЗагружено:Ionut Patras
- Cylindrical Excavations in Clayey Soils Retained by Jet Grout Walls Numerical Analysis and Parametric Study Considering the Influence of ConsolidationЗагружено:Ionut Patras
- DEM simulation of the behaviour of geogrid stabilised ballast fouled with coal.pdfЗагружено:Ionut Patras
- DEM–SPH simulation of rock blasting.pdfЗагружено:Ionut Patras

- 1Vectores[1]Загружено:Irma Higuera
- Equation EditorЗагружено:nibeer_bhai
- Spring Unit 8Загружено:Bala Subramanian
- Measures of DispersionЗагружено:asdfghjk
- Presentacion Final a3Загружено:sirkoky
- Schaum_s_Outline_of_Differential_Equations.3Ed.pdfЗагружено:Jilberto Cuadra
- What the Difference Between a Matrix and a Tensor DLЗагружено:Duy Nguyen
- Kotz,Etal(2006) Encyclopedia.of.Statistical.sciences.v01,2eЗагружено:Lord Baron
- PartitionЗагружено:Muhammad Andyk Maulana
- 20 GRE Maths Questions -SolvedЗагружено:api-3723544
- Mathematical Náboj 2011-19 EN with solutions.pdfЗагружено:Ilir Hoxha
- Stability Theory of Large-Scale Dynamical Systems.pdfЗагружено:riki
- The Appleton Arithmetics 1000123681Загружено:René Wester
- welcome to gomath-2Загружено:api-279402890
- AMENDMENTS TO PROOF (1)Загружено:Karthi Sundar
- MATH Tangrams ArticleЗагружено:Dave Nichols
- blasiusЗагружено:chengpan4341
- Threshold MathematicsЗагружено:sfhelio
- Assignment2.pdfЗагружено:Abhinav Pimpalgaonkar
- Binary ArithmeticЗагружено:NelsonMoseM
- ProbabilityЗагружено:moiuz
- GRE Numeric Entry Tips & QuestionsЗагружено:Gradestack
- square_of_binomial.pptЗагружено:Jeond Jeff Medina
- 9231_s11_qp_23Загружено:Irtiza Hussain
- PIDЗагружено:mwbmughal
- Solution Manual for Advanced Fluid Mechanics GÇô William GraebelЗагружено:Masoomeh Akbarzadeh
- Vector CalculusЗагружено:Hélio Tanomaru
- IntegralsЗагружено:Osama Adly
- UltimateGuidetoDataScienceInterviews-2Загружено:Anonymous wt1Miztt3F
- Psychology of Learning for instructionЗагружено:Carlos Ramírez

## Гораздо больше, чем просто документы.

Откройте для себя все, что может предложить Scribd, включая книги и аудиокниги от крупных издательств.

Отменить можно в любой момент.