Impedance Complance
Impedance Complance
Impedance Complance
b) Framed type foundation for machine working at high speed and rotating types.
In the analysis design of machine foundations certain design criteria need [12] to be satisfied. These includes
the following:
1. The foundation has to be safe against shear failure
2. The settlement should not exceed the permissible limit.
3. The combined center of mass of the foundation and machine and the center of gravity of the base
plane of the foundation should lie on a single vertical line.
Use of these charts, as a guide will be made in the selection of an appropriate upper limit for a satisfactory
foundation performance. The limiting displacement amplitudes are typically in the order of a hundredth of a
centimeter compared to the several centimeters that are the usual restriction for foundation settlement under
static load.
2.2 DESIGN PROCEDURES
The design of a machine foundation is a trial-and-error procedure involving some important steps that need
to be satisfied in the design of these foundations which are acted by dynamic loading.
These steps are;
1. Evaluation of the magnitude and characteristics of the dynamic loads.
2. Establishment of proper design criteria in terms of permissible limits of system responses.
3. Simplification and idealization of the physical system and its representation by appropriate
mathematical models guided by relevant mechanism.
4. Determination of relevant system parameters characterizing the elements of the above idealized
model through field and laboratory investigations.
5. Selection of analytical methods and system parameters to compute system responses.
6. Choice of geometric design parameters of the various components of the system which yield
acceptable responses satisfying the design criteria stipulated in step 2.
7. Optimization of the design parameters.
2.3 DESIGN PARAMETERS
The design parameters, which are important for the design of machine foundations can be categorized into
the following subheading groups.
a) The geometric properties of the foundation mat and superstructure.
b) The physical properties of the underlying soil.
c) The machine data.
c
where n lies b/n 0.3 to 0.6 for granular soil and 0.5 to 0.9 for silty and clayey soils
b) the static-stress prehistory expressed for instance through the over consolidation ratio OCR influences the
Gmax of clays mainly.
Method Of Determination Of Shear Modulus Gmax
The soil shear modulus Gmax may be determined from field measurements, laboratory measurements or use
of published empirical correlations.
Field Procedures: two widely used in-situ procedures are the steady-state oscillatory test and the cross-hole
test.
The shear modulus from these methods is obtained as;
Gmax Vs
2
---(2-1)
Gmax f 2 L2R
where, Gmax = shear modulus
= total mass density of the soil
f= frequency of oscillation in cycle per unit of time
LR= measured wave length
Vs = shear wave velocity
Laboratory Procedures: the resonant column test gives an accurate result of Gmax and has got wide
acceptance.
V
fs s
4H
(in the torsional mode)
G/
fs
4H
V
fl l
4H
(in the axial mode)
E/
fs
4H
where H is the height of the sample
E is the Young’s modulus of the soil
= 21 G
2
propagation and particle motion respectively [Knox, et al., 1982]
Alternatively, the following expression can be used for clean sands:
0.20
180
Gmax
2 a
b p a0.60 ---(2-2)
0.3 0.7e
Shear Wave Velocity Vs: a vibrating foundation emits shear and dilatational waves into the supporting
ground. The shear waves denoted by S propagate with a velocity s that is a function of G and the mass
density of the soil.
G
VS ---(2-3)
and the dilatational ones by p propagate with a velocity V p and is given by:
2(1 )
Vp Vs ---(2-4)
(1 2 )
In addition to these there are waves that arise with in the ground under an oscillating foundation (mostly
Rayleigh and Love waves)
Poisson’s Ratio : the Poisson’s ratio to be used is selected based upon the predominant soil type.
Soil Damping
Damping which is dissipation of vibration energy in a soil-foundation system comprises: geometric and
material types of damping.
Geometric damping: the body and Rayleigh waves propagate radially outward from the source. As they
radiate these encounters an increasingly larger volume of material, thus the energy density in each wave
decreases with distance from the source. This decrease in energy density or decrease in displacement
amplitude is called what is termed as geometrical damping. This geometrical damping is a measure of the
energy radiated away from the immediate region of the foundation and is related to the mass.
Material damping is a measure of the energy lost as a result of hysteretic effects. Referring the
accompanying figure the material damping is seen to be proportional to be the ratio of the area of the
hysteresis loop Al (energy lost) to the area enclosed by OAB,A (energy input).
(Shear Stress) B
A (Shear strain)
O
The material damping can be obtained from resonant column testing. After the soil has been vibrating in a
steady state condition, the exciter is stopped and the soil vibrations are monitored as they decay. The
displacement-time relation ship is essentially sinusoidal, but with the amplitude decreasing with time. If two
successive amplitudes are z1 and z2, then the material damping m can be calculated from [1];
m ln( Z1 / Z 2 ) * 4 2 ln Z1 / Z 2 2
0.5
---(2-5)
t
Fig. 3.2 Rigid foundation block with its six degrees of freedom
x2
tan ---(3-6)
x1
The above equations (3-5) and (3-6) can be written equivalently using complex number notation.
Fv (t ) F v exp(it )
---(3-7a)
xv (t ) x v exp(it )
F v and x v are complex quantities that can be expressed as
F v F1v iF2v , x v x1v ix 2v ---(3-7b)
Equation (3-7a) and (3-7b) are equivalent to equations (3-5) to (3-6) with the following relations being valid
for the amplitudes.
F v F12v F22v ; x v x12v x22v ---(3-8)
F v and x v being complex numbers equation, the dynamic vertical impedance becomes
F (t ) F v
v v which is a complex number.
xv (t ) x
v
This may be put in the form:
v icv ---(3-9)
Substitution of (3-9) into (3-4) and doing some mathematical manipulations yields
.. .
m x v (t ) cv x v (t ) v xv (t ) Pv (t ) ---(3-10)
Equation (3-10) is the equation of motion for a given mass m.
,,G
kh
v(t) cv
r(t) cr kr
h(t) kv
kv,st(kv+iaocv) 0 0
0 kmh,st(kmh+iaocmh) km,st(km+iaocm)
Where kj,st are the static stiffness of the soil, kj and cj are the dynamic stiffness and damping coefficients
which depend on the frequency parameter ao, poisson’s ratio of the underlying soil, and the foundation
shape. The frequency parameter ao is given by ao =B/Vs where is the excitation frequency; B is half the
least width of the foundation; Vs is the shear wave velocity which is given by Vs=(G/)0.5 , G is the shear
modulus of elasticity of the soil and is the soil mass density. The dynamic stiffness khm and kmh couples the
horizontal and rocking motions. They are equal and relatively small for surface and very shallow foundation
and may be neglected.
The determination of the stiffness and damping parameters (analogy parameters of an equivalent mass-
spring-dashpot model) has been a research work of several researchers and are available in the literature.
Among these are Lysmer and Richart et al, Gazetas and Tassoules, Wolf, Mita and Luco, who have
parameters compiled for the analyses of equivalent SDF model for each modes of vibration, [10].
Among the many researchers who organized the results of their research, Gazetas is one who presented the
results in the form of tables and graphs incorporating the shape of the foundation, effect of embedment of the
foundation and also the effect of coupled oscillation. These tables present comprehensive and easy-to-use
information for dynamic spring and dashpot coefficients. The information is in the form of simple algebraic
formulas and dimensionless graphs pertaining to all possible (translational and rotational) modes of
oscillation and covering a wide range of idealized soil profiles and foundation geometries. Utilization of
these tables and charts will be made in this work. The tables and graphs are given in Appendix A.
These tables and graphs give the dynamic stiffnesses k ( ) as a product of the static stiffness k and the
dynamic stiffness coefficient k ( ) which is k ( ) k k ( ) and the radiation damping (dashpot)
2 k ( ) o
coefficients c( ) . The material damping(dashpot coefficient is) is . The total damping Ctotal is
then given by [7]
2 k ( ) o
Ctotal= c( ) ---(3-12)
ci a o mi
i= i=
2 mi i 2m i i
ao
i=
2 i
ao= 2 i i
Similarly the modal damping ratios for a system with stiffness-proportional damping related to the
coefficient a1 will be as follows,
[k]= i [m]
2
a1
ci= a1 i mi and i= i
2
2
The coefficient a1 can be selected to obtain a specified value of the damping ratio in any one mode, say j for
the jth mode.
a 2 j
From i= 1 i we get a1 .
2 j
Now one can construct the damping matrix, [c]. Such type of damping matrix is called the classical damping
matrix and is given by
[c]=ao[m]+a1[k] ---(3-13)
a a
The total damping n is equal to the sum of 0 and 1 n .
2 n 2
a a
That is n = 0 1 n . The determination of a1 and ao is accomplished by considering two modes i and j
2 n 2
and solving simultaneously. It is done as follows:
Extracted Material to Soil Dynamics Course By Dr. Tensay G - 2012 Page 16
if both modes are assumed to have the same damping ratio , which is reasonably based on experimental data
then;
2 i j 2
ao= , a1= . Hence the damping matrix [c] of the superstructure is determined.
i j i j
In applying this procedure to practical problems, the modes i and j with specified damping ratios should be
chosen to ensure reasonable values for the damping ratios in all the modes contributing significantly to the
response.
Where the natural frequencies i and j are to be determined from the undamped condition.
The above method of developing the damping matrix fails when we have two or more parts composing the
system to be analyzed for instance structure-soil system. Modification of the damping matrix will be
important in this case. Because the damping property of such a system is non-proportional type, the
procedure to be followed will be:
[c]s=aos[m]s+a1s[k]s,
[c]f= aof[m]f+a1f[k]f
where the subscripts s and f are to mean superstructure and foundation.
For the determination of aos and a1s, an appropriate damping ratio say =5% can be assumed and for the
coefficients a0f and a1f the damping ratio to be taken depends upon the type of soil in which the structure
rests. Though this is the case, for soils the damping can be fairly easily calculated. Then the two damping
matrices computed will be combined into a single matrix [c] by adding values at the interface where the two
structures interact this is also true for the mass matrix and stiffness matrix of the whole system.
Structure
Interface
Foundation
Soil
Having the mass, stiffness and damping matrix developed, the equation of motion for the whole system can
be written.
.. .
[m]{x} [c]{x} [k ]{x} {P(t )} ---(3-14)
The equations of motion are developed with reference to a discrete planar model of a structure subjected to a
set of harmonic loading that result from machines. The displacement superposition equations yield the
coupled differential equations of motion in the local geometric coordinates. These are transformed into
equations of modal coordinates using the method of modal superposition.
Unfortunately, the equations of motion in the modal coordinates remain coupled with respect to the damping
terms even after transformation. This problem is overcome by adapting a method of solution developed for
structures subjected to seismic loading, [11].
foundations. Although this preliminary phase is often based on experience of designers, suggested guidelines
can be useful in arriving at a satisfactory final design.
These guidelines includes:
All the beams and columns of the foundation should be provided with adequate haunches to
ensure rigidity of joints and to avoid large stress concentrations. [5]
As far as possible, avoid over-hanging cantilevered projections so that rigidity against vibration
will be ensured.[5]
Base slab has to be provided and should be rigid enough to ensure uniform settlement or to
prevent differential settlement. And the depth or thickness of the mat should not be less than
0.07(L)4/3 where L is the average of spans between columns.[4]
As far as possible, the foundation should be so dimensioned that the resultant gravity force due to
the weight of the machine and foundation including the upper deck, intermediate slab if any, the
base slab and the columns passes through the centriod of the base area in contact with the
soil.[4,5]
The following are from {[2]
The analysis should take into account the equipment size and clearance requirements to assure
that sufficient space is allocated to equipment, anchor bolts, piping and clearance for installation,
maintenance and operation.
The columns should be stressed almost equally. The columns should be capable of carrying six
times the vertical load with a maximum spacing of 3.6576m.
The beam depth should be a minimum of one fifth of the clear span, and the beam width is
normally equal to the width of the column.
The flexural stiffness of the beams should be at least twice the flexural stiffness of the column.
The total mass of the structure including the mat should not be less than three times the mass of
the supported machine.
m1
(Damping force, )
(Spring force, )
Fig. 4.1 Figure showing the direction of the drag forces and the direction of motion
Applying Newton’s second law of motion, we have
..
P1 (t ) Fd 1 Fs1 m1 x1 -- -(4-1)
Rearranging equation (4-1)
..
m1 x1 Fd 1 Fs1 P1 (t ) ---(4-2)
These damping force and spring force are from only one degree of freedom. For an n degrees of freedom (x 1,
x2, x3… xn), the drag forces along the degree of freedom x1will be
n .
Fd 1 c1 j x j
j 1
n
---(4-3)
Fs1 k1 j x j
j 1
For any mass mi in any degree of freedom xj, the drag forces becomes
n .
Fdi cij x j
j 1
n
---(4-4)
Fsi k ij x j
j 1
If we consider a system owing mass m resting on a spring of stiffness k and is free to oscillate along only one
direction in the vertical direction,
z m
C K
Extracted Material to Soil Dynamics Course By Dr. Tensay G - 2012 Page 22
Fig. 4.2 Mass-spring-dashpot system.
..
The equation of motion is kz m z (omitting the damping force)
..
m z kz 0 ---(4-7)
The solution of equation (4-7) is a harmonic type given by
z ao cos n t bo sin n t ---(4-8)
where, a o and bo are constants and n is the natural circular frequency.
Equation (4-8) can be rewritten as
z co sin( n t ) ---(4-9)
..
z n2co sin( nt )
..
---(4-10)
z z 2
n
Plugging equation (4-10) into equation (4-7) and solving we obtain
n
n k / m Tn ---(4-11)
2
where Tn is the natural period of oscillation.
z t
z e at .. ---(4-13)
z a 2 e at
Inserting these values into equation (4-12) and simplifying
c k
a2 a 0 ---(4-14)
m m
The solution of equation (4-14) is given by
1
a1, 2 * (c c 2 4km) ---(4-15)
2m
We need to define the following important parameters in the analysis.
Damping ratio
n k / m
cc 2 km
c
cc
d n 1 2
Depending on the values of the terms in the square root of equation (4-15), we have the following cases.
If c 2 4km, then the system is critically damped.
If c 2 4km, then the system is under damped.
If c 2 4km , then the system is over damped.
If the mass in Fig.4-2 is subjected to a harmonic force f (t) of the form f o Sint , then the equation of motion
will be:
.. .
m z c z kz f (t ) ---(4-16)
The solution of this equation will be
sin(t )
z e wd t c1 sin( d t ) f o ---(4-17)
((k 2 m 2 ) 2 (c ) 2
The first part of the solution is transient and dies out after some time where as the second part is the steady
state response, so
sin(t )
z fo ---(4-18)
((k 2 m 2 ) 2 (c ) 2
sin(t )
z fo ---(4-19)
k (1 r 2 ) 2 (2 rk ) 2
2
where, r is the frequency ratio.
n
The amplitude of displacement z is given by
fo
z k ---(4-20)
(1 r ) (2 r ) 2
2 2
fo
zs is the magnitude of the static displacement under a force f o .
k
z
Define M to be equal to
zs
1
M= ---(4-21)
(1 r 2 ) 2 (2 r ) 2
M is called the magnification factor of the displacement.
The force transmitted to the foundation Ft is
.
Ft c z kz ---(4-22)
.
Plugging the expressions for z and z and after certain mathematical manipulations, we arrive at
Ft q k 2 c 2 2 (Cos(t 1 )) ---(4-23)
In matrix form
m{x} c{x} k x f (t )
.. .
---(4-26)
where n are the natural vibration mode and Z n (t ) are the time variation of the displacements.
The natural vibration mode n is determined from the free-vibration motion of the masses. Accompanying
the mode shapes are the natural vibration frequency n . These terms need to be determined before we go
solving the equation of motion.
Undamped free vibration
The undamped free vibration is given by
m{x} k x 0
..
---(4-28)
The solution for this free vibration equation is a simple harmonic.
{x(t )} {}Z (t ) ---(4-29)
Where Z (t ) a cos t b sin t ---(4-30)
{x(t )} {}(a cos t b sin t ) ---(4-31)
x j 1 mx1 j
j 1
T
---(4-37)
x j 1 m x j 1
3. Check the convergence by comparing two successive values of
j 1 j
tolerance ---(4-38)
j 1
4. If the convergence criterion is not satisfied normalize x j 1 as
x j 1
x
j 1
0.5
---(4-39)
T
x j 1 m x j 1
5. If the convergence criterion is not satisfied then the last values are stored as
Where;
M m { mT }[m]{ m }
Cmn { mT }[c]{ n }
K m { mT }[k ]{ m }
Fm (t ) { mT }{ f (t )}
Equation (4-44) exists for each n=1, to N and the set of N-equations can be written in matrix form;
.
M Z C Z K Z
..
{F (t )} ---(4-45)
M and K are diagonal matrices, where as [C] is a non-diagonal matrix of elements Cmn. The above
equations of motion. Equation (4-45) are coupled through the damping terms. Because of the fact that
damping property of the whole system is non-proportional type, decoupling is not possible. To overcome this
problem the following technique is done. Taking the coupling terms to the right but leaving only the term
corresponding to n=m equation (4-34) can be written as; [11].
n m
.. . .
M m Z m C mm Z m K m Z m Fm (t ) C mn Z n , ---(4-46)
n
Dividing equation (4-22) by Mm yields
C mm . K 1 .
n m
..
Zm Z m m Zm Fm (t ) C mn Z n , ---(4-47)
Mm Mm Mm n
m T m x0
.
z m 0
.
Mm
So for at t =0
Z m (0) Am C m
Am + Cm = Zm(0)
.
Am Bm dm Dm Z m (0)
.
Bm dm Dm Cm Z m (0) Z m (0)
.
B Z m (0) Z m (0) C m Dm
m dm
Am Z m (0) C m
m T M x0
Am Cm
Mm
.
m M x(0)
T
Am Dm
Mm
Bm
D m
For the determination of the peak value f o of the total response, one can combine the peak modal response
vectors f sno by either of the following modal combination rules.
1. The square-root-of- sum-of squares (SRSS) rule developed by E. Rosenblueth’s Ph.D work (1951).
2. The complete quadratic combination (CQC) rule.
As the SRSS method provides an excellent response estimates for structures with well-separated natural
frequencies, I have adopted it for the computation of the peak response amplitude.
According to the rule,
0.5
N
f o f sno ---(4-58)
n 1