You are on page 1of 8

The 12 International Conference of International Association for Computer Methods and Advances in Geomechanics (IACMAG) 1-6 October, 2008

Goa, India

th

Dynamic Soil-Structure Interaction Analysis of Tall Multi-flue Chimneys under Aerodynamic and Seismic Force
Negar Sadegh Pour
Petrofac International Limited Sharjah, U.A.E

Indrajit Chowdhury
Petrofac International Limited Sharjah, U.A.E Keywords: Soil stiffness, aerodynamic response, chimney, modal analysis, earthquake ABSTRACT: Dynamic response of tall chimneys under wind and earthquake force has been a topic of significant interest to engineers for quite some time in the industry. It is only in recent times it has been acknowledged that the underlying soil affects the dynamic response of the chimneys especially under earthquake force. As far as aerodynamic response is concerned there is yet any consistent study available as to how soil-foundation system affects the across and along-wind response of the chimney which is often the guiding design criteria for a tall chimney. The present paper proposes a semi analytic mathematical model based on which both seismic and aerodynamic response of such tall chimneys are studied for various soil stiffness and are compared with the fixed base conventional method as per UBC 97(for seismic load) and CICIND (for wind loading).Some of the salient feature of the paper is, unlike adding the soil stiffness to the diagonal element of the stiffness matrix (which happens to be the usual practice) it uses a modified mathematical model (considering multi-degree of freedom) for soil coupling as was developed originally by Veletsos and Meek for single degree of freedom. The paper extends this theory to cater to the material and radiation damping of the soil within the modal analysis framework for dynamic soil structure interaction (DSSI) response for the chimney. The method being semi analytic does not require elaborate modelling effort thus could make it computationally attractive for engineers undertaking design of such chimneys in the industry.

1 Introduction
The study of the effects of soil-structure interaction on aerodynamic response of tall RCC chimneys have not received much attention till date , though some attempts have been made for the seismic analysis of tall chimneys with the influence of soil flexibility. While it is an established fact that soil foundation interaction reduces the response under earthquake how it affects the aerodynamics response is still not very clear. The present paper attempts to study the same based on a semi analytic mathematical model considering interaction between the underlying soil foundation and the structure. The soil structure interaction is based on linear behaviour in both structure and soil. The super structure ( the chimney) is assumed as an elastic beam with constant properties along the height. The shear deformation and the effects of the axial forces are neglected. The foundation was assumed as a rigid circular foundation resting on a elastic half space without any embedment with perfect bonding between soil and the foundation.

Formulation for Multi Degree of Freedom including soil-structure interaction

The time period of vibration of a structure increases with decreasing soil stiffness. Based on this phenomena Veletsos and Meek (1974) has proposed an expression for time period for a single degree of freedom system as

T = T 1+

K h2 k (1 + x ) Kx K

(2)

Where

T = Modified period of structure due to DSSI

T = Fundamental period of the fixed base structure

k = Stiffness of fixed base structure

2696

K x = Dynamic Horizontal spring constant of the soil @ K x = 32Grx ( 1 )/( 7 8)


K = Dynamic Rotational spring constant of the soil @ K x = 8Grx / 3( 1 ) h = Inertial centroid of the system
One of the major limitations of eqn (1) is that it is restricted to a model having a single degree freedom, and can neither predict the response of the multi degree of freedom systems for higher modes nor the Eigen vectors for the soil-structure interaction. Chowdhury and Dasgupta(2002) has extended the above equation to cater to multi degree freedom system in three dimension, when eqn.(1) can be written in terms of stiffness as
3

1 1 1 h2 = + + ke k K x K
In which,

(2)

ke

is equivalent stiffness soil structure system.

For a system having n degree of freedom above equation can be extended to a matrix form as below

[I]nn = [I]nn + [I]nn + [h 2 ]nn [K e ]nn [K ]nn K x K [Fe ]n n = [F]nn + [Fx ]n n + [F ]n n

(3) (4)

Where [F]n n and [K ]n n are flexibility matrix of soil structure system and stiffness matrix with suffix as mentioned earlier, respectively. And [h]nn is radius vector of the lumped masses to the center of foundation springs. The damping matrix for a multi-degree freedom system is similarly given by

[ e ]nn = [K e ]nn ([ ]nn [F]nn + x [Fx ]nn + [F ]nn )

(5)

Here

e = Damping ratio of the equivalent soil structure system = Damping ratio of the equivalent fixed base system usually taken as 0.02 to 0.05 for RCC chimneys.
x = Damping ratio of the soil for horizontal mode x = 0.288/ B x where B x = = Damping ratio of the soil for rocking mode x
32(1 ) s rx 0.375(1 )J x g = 0.15 /[(1 + Bx ) Bx ] where B x = 5 s r x
3

(7 8 )mg

The values Kx, K, x and and their notations are as per Richart et al (1970)

3 Dynamic Analysis of Tall Chimney


Free un-damped vibration of a tall chimney having uniform cross section can be expressed as:

EI(
E = Elastic modulus of the chimney

4y z
4

) + A(

2y t 2

)=0

(8)

I = Moment of Inertia of the chimney cross section = Mass density of the chimney material A = Cross section area y = Displacement of the beam a function of geometry and time, expressed as:

y(z, t) = Y(z)q(t )

(9)

Based on separation of variable technique eqn (8) can be separated into two linear differential equation and one of them is

EI(

d 4 Y A 2 )+ Y=0 EI dz 4

(10)

For a uniform cantilever beam, considering appropriate boundary condition, the shape function solution(Chopra 1995) can be written as:

2697

Ym = sin
Here

z z z mz sinh m m (cos m cosh m ) H H H H

(11)

H = Length of beam or chimney height m = mode number,(1, 2, 3, 4,.)

m = 1.8751, 4.6941, 4.8548, 10.966,


z = Any height from the base of the chimney.

m =

sin m + sinh m cos m + cosh m


2 d 2 Yi (z) d Y j (z)

(12)

Based on Hurty and Rubenstein(1967) it can be shown that

k ij =

EI(z)

dz 2

dz 2

dz

(13)

H m ij = A i (z) j (z)dz 0

(14)

The double derivative of the shape function mentioned in eqn (11) is given by
2 i z z iz i z (15) sin i sinh i + i cos H + cosh H 2 H H H 2 j jz jz jz j z Y j = 2 sin cos sinh + j + cosh (16) H H H H H z dz Now considering = when d = and as z 0, 0 and as z H, 1 we can now express the H H

Yi =

double derivatives as

F ( )i = F ( ) j =

H2 2 j H2

[ sin i sinh i + i (cos i + cosh i )] = i 2

[ sin sinh + (cos + cosh )] =


j j j j j

H 2 j H2

f i ( )
f j( )

(17)

(18)

Thus stiffness of the system can now be expressed as

k ij =
and mass of the system is given by

EI i j
2

H
i

f ( ) f ( )
0
i

(19)

m ij = AH

f ( )
0

f ( ) j d where i=j=1,2,3,m

(20)

Thus for the first three modes the fixed base stiffness matrix is given by
1 4 2 1 f ( )1 d Symmetric 0 1 1 EI 2 2 4 ( )2 d = 3 2 1 f ( )2 f ( )1 d 2 f2 H 0 0 1 1 1 2 2 2 2 4 2 3 1 f ( )3 f ( )1 d 3 2 f ( )3 f ( )2 d 3 f ( )3 d 0 0 0

[K ]ij

(21)

2698

[M]ij

1 2 f ( )1 d 0 1 = AH f ( )2 f ( )1 d 0 1 f ( )3 f ( )1 d 0

Symmetric

f
1 0 0

( )2 d
3 2

f ( ) f ( ) d

1 2 f ( )3 d 0

(22)

Eqn(21) and (22) on numerical integration finally gives

0 0 22.94 EI [ K ]3 x 3 = 0 468 . 04 0 H3 0 3812.81 0 0 0 1.86 [ M ]3 x 3 = 0 0 . 96 0 AH 0 1 0


Based on eigen solution of eqn (23) and (24) the time period can be finally expressed as:

(23)

(24)

0 0 1.787 4 AH [T ]3 x 3 = 0 0 . 285 0 EI 0 0.102 0


To incorporate the soil foundation effect for first three modes, eqn(4) can be rewritten as

(25)

[Fe ]33 = [K]3x3 1 +


Where

1 [I]33 + 1 h 2 Kx K

[ ]

33

(26)

[h ]33 = H i f( i )d f( i ) 2 d
0
0

when i=1,2,3 and this gives

[h]33
Considering T = 2 M K = 2

0 0 0.726 = 0 0.209 0 H 0 0.127 0

(27)

[M ][F ]1

we finally have

[Te ]3 3

H3 1 H2 1.86 AH( + + 0.5278 ) 22.94 EI K x K = 2 0 0

0 H3 1 H2 + + 0.0438 0.96 AH( ) 468.04 EI K x K


3

0 0 AH( H 1 H2 + + 0.0163 3812.81EI K x K

(28)

Substituting [K] and [F] from eqn(23) and expanding eqn(5) damping ratio matrix can be written as

2699

[ e ]33

H 3 H2 + x + 0.5278 K 22.94EI K x H3 1 H2 + + 0.5278 K 22.94EI K x = 0 0

0 H2 H 3 + x + 0.0438 468.04EI K x K H3 1 H2 + + 0.0438 468.04EI K x K 0

0 0 H 3 x H2 + + 0.0163 3812.81EI K x K 3 2 H 1 H + + 0.0163 3812.81EI K x K

(29)

Eqn (28) and (29) gives the time period and the damping ratio for the first three modes for a multi-flue chimney considering DSSI.

4 Wind Effect on Tall Chimney based on SSI


The wind load acting at any point on a chimney can be considered as a sum of the quasi-static and dynamic load component. The static load component is that force which wind will exert if it blows at a mean steady speed and which will tend to produce steady response in stack. The dynamic component which can cause vibration in stack is generated basically due to gusts, and vortex shedding. This is the concept of wind force calculation in most design codes for chimneys (CCIND, ACI, etc). The static equivalent of the wind load due to gusts in CCIND is assumed to vary linearly with the height based on the following formula:

wg ( z) =
Where as per CICIND

3(G 1) z H2 H

w
0

( z )zdz

(30)

G = Gust factor, and is expressed as G = 1 + 2gi B + ES/ wm ( z ) = Static wind load per unit height
Referring to CICIND it will be observed that in the factor G the terms E,g, and S are all a function of frequency vis-a vis the time period of the chimney. It is thus apparent that as the time period varies due to soil effect so will the gust factor and the aerodynamic response. To assess this effect quantitatively, three different chimneys were analyzed with and without soil effect under wind load. Figs (1), (2) and (3) shows percentage variation in base moment considering, soil- structure interaction (MSSI) and fixed base (Mfx) case. Here t is average thickness of shell along the height of the chimney, and Df is diameter of foundation. The result shows that decrease in shear wave velocity (i.e. softer is the soil), base moment of chimney increases under along wind force and higher is the wind more is the increment. For very stiff soil, M SSI M fx .

Figure 1. Comparison of base moment for longitudinal wind (wind velocity 36m/s)

2700

Figure 2. Comparison of base moment for longitudinal wind (wind velocity 20m/s)
2.5 .10
6

2 .10

Moment(kN.m)

M SSI( z ) M FB( z )

1.5 .10

1 .10

5 .10

50

100

150 z

200

250

300

z(m)

Figure 3. Comparison of moment for longitudinal wind Df=60m Vw=36m/s Height=300m The base moment due to across wind response of tall chimneys as per ACI-307-5 is expressed as

M a = GS s C L ( a 2 )Vcr d(u)h 2 [/{4( s + a }] S p [2L/(h/d(u) + C E )]


2 0.5

(31)

Of all these factors the terms Vcr (The critical wind velocity at which the vortex shedding takes place), ), a,s Sp etc are all functions of the natural frequency of the chimney. It is again evident from eqn(31) that as the frequency of the chimney gets modified due to the soil foundation effect the vortex shedding response also alters from its fixed base response.Fig 4 and Fig 5 shows across wind response attenuates with DSSI.

Figure 4. Comparison of base moment for across wind (wind velocity 36m/s)

2701

5 .10

4 .10

M SSI( z ) M FB( z)

3 .10

2 .10

1 .10

50

100

150 z

200

250

300

Figure 5. Comparison of moment for across wind Df=60m Vw=36m/s Height=300m

5 Earthquake Effect on Tall Chimney based on SSI


Early chimneys were rarely checked for adequacy under earthquake condition. A first attempt to analyze tall chimneys has done by Hounser, who in 1952 proposed a simplified method. After that some theoretical studies revealed that soil-structure effects on seismic response of tall chimneys for soft soils can not be negligible. In 1986, Luco modeled a chimney as an elastic beam with variable properties along the height and he pointed out that, for vey soft soils( shear wave velocity less than 200m/s), the variation in the chimney seismic response, with respect to the result for a fixed base, could reach a difference of about 200%. In 1998, American Concrete Institute (ACI307) recommended the normalized design response spectrum for stiff soil. In this code, it is emphasized that the response spectrum shall be modified for soft and shallow soil condition by any method that is properly substantiated and complies with basic principle. Following graph presents the effect of soil structure interaction on earthquake loads of three different chimneys , analyzed and compared with the seismic load of the same chimneys with fixed base. Response spectra used in this analysis is acceleration response spectrum given by ACI307-98 for Z=0.40. In this analysis damping of soil and structure assumed as 5%. Shear wave velocity of soil varied from 600m/s to 750m/s. For soft and medium soil, that spectrum can not be used. For this purpose Luco (1985) and Navarro (1992) carried out further investigation, which shows that for soft soils, the direct use of ACI307 without inclusion of soil structure inter action can lead to a significant underestimation of base loads. This seems to be less important when the soil is stiffer.

Figure 6. Comparison of seismic base moment for firm soil spectrum (Z=0.40) (ACI-307)

6 Conclusion
The effect of foundation flexibility on the reinforcement concrete chimneys to wind excitation can be significant and should be addressed in design stage. The presence of soft soils for tall chimney may increase base moment of longitudenal wind up to 10 percent and decrease across wind up to 50 percent that may affect design forces. These changes will be less in the stiffer soils and in the zones with less base wind speed. None of the present codes present any method for estimation of influence of soil structure interaction. It might be advisable to take soil

2702

structure inter action effects into account for calculation of wind load in different codes. Soil structure interaction also has an important effect on seismic forces of tall chimneys. Although for tall chimneys rested on firm soil, earthquake loads decreased as a result of increasing in period values, seismic forces may amplify by using different response spectra in calculation. This means that the soil structure interaction effects are reliant on the characteristic of the seismic excitation in addition to chimneys properties.

7 References
ACI307 1998. Design and Construction of Reinforcement Concrete Chimneys, ACI committee 307 Report ASCE7 2002. Minimum Design Loads for Buildings and other Structures, American Society of Civil Engineers, Virginia Chopra A.K. 1995. Dynamics of Structures, Prentice Hall, Englewood Cliff, New Jersey. Chowdary I., Dasgupta S.P., 2002. Earthquake Response of Soil-Structure System, Indian Geotechnical Journal, 32(2),ppCICIND 2001. Model Code for Concrete Chimneys, Part A. The Shell, Second Edition, Revision 1 Hutry, W.C. and Rubenstein, M. F. 1967, Dynamics of Structures, Prentice-Hall of India Pvt. Ltd, New Delhi IBC 2003. International Building Code, Structural Design, Chapter 16. Luco J. E,1986.Soil-Structure Interaction Effects on the Seismic Response of Tall Chimneys, Soil Dynamics and Earthquake Engineering,Vol.5, pp403-409 Lysmer, J and Richart, F.E. Jr.1966, Dynamic Response of Footing to Vertical Loading. J. Soil Mech. and Found. Div. Proc. ASCE, Vol. 92, No. SMI, January, pp. 65-91 Navaro C. 1992. Influence of Soil Flexibility on the Seismic Behaviour of Chimneys, Soil Dynamics and Earthquake Engineering,Vol.11,No.3., pp170-177 UBC 1997, Uniform building Code, Structural Design Requirement, Chapter 16 Veletsos, A.S. and Meek J. W. 1974, Dynamic Behaviour of Building-Foundation System, Earthquake Engineering and Structural Dynamics 3,No. 2, pp 121-138

2703

You might also like