Vibration Design of Floors

RFS2-CT-2007-00033
Human Induced Vibration of Steel Structures
Vibration Design of Floors

Background Document
Background_Floors_EN01.doc 29.10.2008
Background
Table of Content
1.1.
Loadings .......................................................................... 4
1.2.
Perception and perception classes........................................ 4
1.3.
Evaluation ........................................................................ 6
2.1.
Introduction ..................................................................... 7
2.2.
Transfer function method ................................................... 8
2.3.
The one step root mean square value................................... 8
2.3.1.
Standard walking load .................................................. 9
2.3.2.
Weighting ................................................................. 10
2.4.
Obtaining the OS-RMS90 value........................................... 10
2.5.
Hand calculation method .................................................. 12
3.1.
Modal superposition ......................................................... 13
3.1.1.
Steady-state response................................................ 13
3.1.2.
Transient response..................................................... 14
3.1.3.
Weighting factors....................................................... 14
3.1.4.
Contour plots ............................................................ 14
3.1.5.
Detailed procedure..................................................... 14
3.2.
Other vibration considerations........................................... 15
3.3.
Simplified approach ......................................................... 15
4.1.
Increasing the modal mass ............................................... 16
4.2.
Adjusting the frequency ................................................... 17
4.3.
Increasing the damping.................................................... 17
4.4.
Structural means............................................................. 17
4.5.
Retrofit measures............................................................ 17
Background
Summary
This document provides background information to Vibration Design of
Floors Guideline. It presents alternative and more general ways for the
determination of the floor response to dynamic human induced forces.
The theoretical methods presented here and in the guideline document have
been elaborated/investigated in the RFCS-Project Vibration of Floors. The
guideline and background document are here disseminated under the grant
of the Research fund for Coal and Steel within the project HIVOSS.
Background
1. Design Considerations
1.1. Loadings
The mass present in the structure has a very significant effect on both the
frequency of the floor plate and the magnitude of the vibrations. It is
therefore important that the distributed mass used in vibration analysis is
representative of the mass that will be present in service, as a higher mass
will reduce the magnitude of the floor vibration at a given frequency. In
design, the mass per unit area should be taken as the unfactored self-weight
of the structure including superimposed dead loads such as the weight of
ceilings and services. In addition, where the designer can be confident that
such loading will be guaranteed to exist in the finished structure, an
additional allowance may be included for semi-permanent loads. Generally it
is recommended that this allowance should not exceed 10% of the nominal
imposed load. Generally the mass of people present on the floor is not
explicitly considered, but in the case of very light structures this additional
mass is very significant and can be considered.
1.2. Perception and perception classes

In a similar way to human hearing, the human perception of vibration varies
with frequency human ears cannot detect low frequency or high frequency
sounds, and similarly the human body cannot detect very high frequency
vibration. To attenuate a vibration response to take account of this response,
frequency-dependent weighting factors are used. The level of vibration that
can be perceived also depends on the direction of incidence to the human
body, and for this the basicentric coordinate system shown in Figure 1.1 is
used (the z-axis corresponds to the direction of the human spine). The
threshold of perception (the vibration level under which the average human
will not be able to perceive any motion) is higher for z-axis vibration than for
x- or y-axis vibration, indicating that x- or y-axis vibration is more easily
perceived.
Background
z
z
y
y
x
x
Supporting
surface
Supporting
surface
z
Supporting
surface
Figure 1.1 Directions for Vibration defined in ISO 10137
Values of frequency weighting are given in Standards such as ISO 10137[4].

Various weighting curves are given, depending on the direction of vibration
and the activity. The weighting curves are also specific to the parameter
being considered velocity or acceleration. In most cases, the aim of
vibration analysis is to reduce or remove discomfort, but in special
circumstances, such as operating theatres, the level of vibration will need to
be such that it cannot be perceived and does not affect the steadiness of
hand or vision. Perception and discomfort use the same weightings but
typically perception will have a lower allowable threshold (i.e. a subject can
detect vibration without being discomforted by it), while there are different
weighting curves for considering hand and vision control. The weighting
curves for perception in both the z-axis (Wb) and x- and y-axis directions
(Wd) are shown in Figure 1.2.
0.1
Background
Weighting factor
Weighting factor
0.1
10
100
1
10
100
Frequency (Hz)
Frequency (Hz)
Wb Weighting
Wd Weighting
Figure 1.2 Wb and Wd frequency-acceleration weighting curves
1
To illustrate the use of the curves, for z-axis vibration using curve Wb for
discomfort, a sine wave of 8 Hz has the same feel as a sine wave at 2.5 Hz
or 32 Hz with double the amplitude.
1.3. Evaluation
Acceleration
The response of a system to regular excitation will take the form of one of
the plots shown in Figure 1.3, dependent on the comparison between the
excitation frequency and the natural frequency of the system.
Time
Figure 1.3 Response Envelopes
When the frequency of the excitation (or higher harmonics of the excitation)
is similar to the natural frequency of the floor, the resulting response takes
the form as shown on the left of Figure 1.3 a gradual build up of the
response envelope from zero to a steady-state level. This response is known
as either a resonant response (because the floor is resonating with the
excitation) or a steady-state response. For excitation from walking activities,
this kind of response typically occurs for floors with a fundamental natural
frequency less than 9-10 Hz.
Background
When the frequency of excitation is significantly lower than the natural

frequency of the floor, the response envelope shown on the right of Figure
1.3 is typical, known as a transient response. In this case the floor plate
responds to the excitation as if it is a series of impulses, with the vibration
from one footstep dying away before the next footstep. These excitation
types can be seen on the OS-RMS90 plots as the contour lines become less
dependent on the specific frequency as the floor frequency rises above 9 Hz,
showing that the response is moving from a steady-state response to a
transient response.
2. OS-RMS Method
2.1. Introduction
The one-step root mean square (OS-RMS) method is based on the findings
of a research project funded by the ECSC on floor vibrations, see [1]. This
chapter describes briefly the OS-RMS method which underlies the design
check procedure.
The OS-RMS value represents the response of a floor which is brought into
vibration due to a person walking on that floor. It is obtained from measured
or simulated floor characteristics and a standard walking load function for a
person with given weight and walking pace.
In calculating the OS-RMS value, the excitation point and response points do
not necessarily have to coincide. Further, it is assumed that the excitation
point is kept fixed, that is, the walking path is not taken into consideration.
In general the excitation and response points are selected where the
greatest vibration amplitudes are expected (in regular floors this is usually
the middle of the floor span).
In the design check the 90 percentile of OS-RMS values obtained for
different persons weights and walking speeds (or step frequency) must be
calculated. The 90 percentile is referred to as the OS-RMS90 value and should
subsequently be checked against the recommended values in Table 1 of the
guideline.
A single OS-RMS value for a given step frequency and persons weight can
be obtained from one of the three following methods:
1. Hand calculation method
2. Transfer function method using measurements
3. Transfer function method using finite element analysis of the floor
The hand calculation method is the method which is covered by the guideline
and is applicable to floors which can be easily described as a single degree of
freedom mass-dashpot-spring system. Underlying the hand calculation
Background
method is the transfer function method. An overview of the design check

procedure, comparing the three methods is given in Figure 2.1.
TRANSFER FUNCTION HAND CALCULATION
METHOD
METHOD
Testing
Prediction
Heeldrop
FEA
Calculation
Transfer Function (TF)
TF x Load = Response
Response
Verification
Verification
Figure 2.1:
Simplified overview of design check using the OS-RMS method.
2.2. Transfer function method

In the transfer function method, the floor's characteristics are described in
terms of a frequency response function, FRF, or transfer function. The
transfer function represents the response of a structure when it is subjected
to a harmonic load (a sinusoidal time varying load function) with a given
frequency and amplitude equal to one.
When this function is used in combination with the standard walking load,
the OS-RMS value can be determined.
The transfer function method can be applied where the floor response is
obtained either by measurement or by finite element calculations.
The use of the transfer function method implies that the calculation of the
response of the floor occurs in the frequency domain.
Alternatively, when using a finite element calculation, the response can be
obtained in the time domain. This can be rather time consuming as the
determination of the OS-RMS90 value requires many response calculations.
2.3. The one step root mean square value

The OS-RMS value is obtained from the response of a floor to a standardized
walking load. It is defined as the root mean square value over a given
interval of the weighted velocity response at a point on the floor. The
Background
interval is selected starting from the highest peak in the response and either
the previous or the next peak in the response, see Figure 2.2.
2
v (mm/s)
Original signal
0
-2
0
v (mm/s)
RMS = 0.56 mm/s

0
-2
0
time (s)
Figure 2.2:
Selection of interval in weighted velocity response for calculation of
the OS-RMS value.
Given this definition, the interval over which the OS-RMS value is obtained
corresponds to the duration of a single step. This ensures a consistent
measure for the vibration level 1 .
2.3.1.
Standard walking load
The standard walking load is taken as a series of consecutive steps whereby

each step (or footfall) load is described by a polynomial. The normalized step
load is given by:
F (t )
= K 1t + K 2 t 2 + K 3 t 3 + K 4 t 4 + K 5 t 5 + K 6 t 6 + K 7 t 7 + K 8 t 8
G
where G is the person's mass. The coefficients K1 to K8 depend on the step
frequency (fs) and are given in Table 2.1. The load duration, ts, is given by
the following formula:
t s = 2.6606 1.757 f s + 0.3844 f s2
For t>ts, F(t)=0.
Table 2.1: Coefficients K1 to K8 for given walking frequency (fs)
K1
K2
K3
K4
K5
K6
K7
K8
fs 1.75
-8 fs + 38
376 fs 844
-2804 fs + 6025
6308 fs 16573
1732 fs + 13619
-24648 fs + 16045
31836 fs 33614
-12948fs + 15532
1.75 < fs < 2

24 fs 18
-404 fs + 521
4224 fs 6274
-29144 fs + 45468
109976 fs 175808
-217424 fs + 353403
212776 fs 350259
-81572fs + 135624
fs 2
75 fs - 120.4
-1720 fs + 3153
17055 fs - 31936
-94265 fs + 175710
298940 fs - 553736
-529390 fs + 977335
481665fs - 888037
-174265fs + 321008
OS-RMS values defined in this way can be unambiguously compared with each other. If on the other hand, a
constant interval greater than the duration of a single step was used, then the rms value over this interval would
depend on the step frequency and interval duration.
Background
The standard walking load function is built from the step load defined above,
by adding the step load to this function repeatedly at intervals of 1/fs.
Examples of the standard step load and walking load functions are given in
Figure 2-3.
Normalized load (-)
1.5
Normalized force (-)
2.5
1.5
1.7
2.0
2.2
0.5
2
1.5
1
0.5
0
0
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
time (s)
time (s)
Figure 2-3:
Step load for four different step frequencies (left) and example of
walking load function (right).
2.3.2.
Weighting
The OS-RMS value is determined from the weighted velocity response at a

point on the floor. The weighted response is obtained by applying the
following weighting function:
1
1
H( f ) =
v0 1 + ( f 0 / f ) 2
where f0=5.6 Hz and v0 is the reference velocity equal 1.0 mm/s. Because of
division by a reference velocity, the weighted response is dimensionless.
2.4. Obtaining the OS-RMS90 value

As stated previously, the final design checked is based on the OS-RMS90
value. This value is obtained by calculating the OS-RMS for all possible
combinations of persons' weights and walking paces defined in Table 2-2.
According to these tables, a total of 35x20=700 OS-RMS values
corresponding to each possible combination must be calculated. From the
relative frequency (probability) of each combination, the cumulative
frequency of the OS-RMS value is obtained. The OS-RMS corresponding to a
cumulative frequency of 90% defines the sought OS-RMS90 value 2 .
In effect, we are treating the OS-RMS value as a random variable and are seeking its 90% upper limit.
10
Background
Table 2-2: Cumulative probability distribution function for persons' walking pace
and persons mass
Cumulative
probability
0.0003
0.0035
0.0164
0.0474
0.1016
0.1776
0.2691
0.3679
0.4663
0.5585
0.6410
0.7122
0.7719
0.8209
0.8604
0.8919
0.9167
0.9360
0.9510
0.9625
0.9714
0.9782
0.9834
0.9873
0.9903
0.9926
0.9944
0.9957
0.9967
0.9975
0.9981
0.9985
0.9988
0.9991
0.9993
Step frequency
fs (Hz)
1.64
1.68
1.72
1.76
1.80
1.84
1.88
1.92
1.96
2.00
2.04
2.08
2.12
2.16
2.20
2.24
2.28
2.32
2.36
2.40
2.44
2.48
2.52
2.56
2.60
2.64
2.68
2.72
2.76
2.80
2.84
2.88
2.92
2.96
3.00
Cumulative
probability
0.0000
0.0002
0.0011
0.0043
0.0146
0.0407
0.0950
0.1882
0.3210
0.4797
0.6402
0.7786
0.8804
0.9440
0.9776
0.9924
0.9978
0.9995
0.9999
1.0000
Mass (kg)
30
35
40
45
50
55
60
65
70
75
80
85
90
95
100
105
110
115
120
125
11
Background
2.5. Hand calculation method

In the hand calculation method, the response calculation and subsequent
statistical process to obtain the OS-RMS90 value has been carried out
beforehand. In this method, the structure is assumed to be a one degree of
freedom system which is easily modelled using a mass-spring-dashpot. OSRMS90 values corresponding to various combinations of floor mass, stiffness
and damping have been obtained and are presented in the graphs given in
the guideline.
In this case it is sufficient to determine the modal parameters (mass,
stiffness and damping) for the structure being investigated and then read off
the corresponding OS-RMS90 value from the graphs.
12
Background
3. Alternative analysis methods

3.1. Modal superposition
As an alternative to the transfer function and probability based design
approach given by the OS-RMS approach, modal superposition techniques
can also be used to determine the response of a floor to human induced
vibration. This vibration can be caused by walking or by more lively activities
such as aerobics or dancing. In this approach the floor is modelled in finite
element software and the modal properties such as the frequencies, modal
masses and mode shapes are extracted.
UZ
(AVG)
RSYS=0
DMX =1
SMN =-.995481
SMX =1
UZ
(AVG)
RSYS=0
DMX =1
SMN =-.005978
SMX =1
MX
MX
MN
MN
Z
Y
Z
Y
UZ
(AVG)
RSYS=0
DMX =1
SMN =-.999972
SMX =1
UZ
(AVG)
RSYS=0
DMX =1
SMN =-.89409
SMX =1
MX
MN
MX
MN
Z
Y
Z
Y
Figure 3.1: Mode shapes for a typical floor
Once the mode shapes have been extracted, a response analysis is

performed to determine the accelerations of the floor. The input forces are
determined by breaking the excitation forces down into frequency
components using Fourier series. By using design values of the Fourier
coefficients, the requirement to consider the variability of pacing frequency
and body mass is removed, and a design acceleration can be calculated
simply by analysing the worst-case pacing frequency.
The accelerations are calculated in two different ways, depending on whether
a steady-state or transient response is expected.
3.1.1.
Steady-state response
For the steady-state or resonant case the accelerations are calculated using
simple dynamic theory, where the acceleration of each mode from each
harmonic of the excitation is equal to the ratio of the applied force to the
modal mass, multiplied by a dynamic magnification factor (DMF). This DMF
takes into account the amount of damping present in the structure and the
ratio between the pacing frequency (or the harmonic that is being
considered) and the modal frequency. At the worst case, the harmonic of the
pacing frequency equals the modal frequency, and the DMF is equal to 1/2,
where is the damping ratio. In typical floors = 3%, and so the
magnification factor is in the region of 17. The calculation also takes into
13
Background
account the magnitude of the mode shape at the excitation and response
points that are being considered (which can either be coincident or
separated). This means the effect of all the mode shapes that are being
considered can be combined without over-estimating the vibration levels.
Once the acceleration levels for each mode shape and excitation harmonic
have been calculated, they are combined to produce a single root-meansquare acceleration that relates the excitation and response points.
3.1.2.
Transient response
For the case of transient vibration, an impulse load is applied to each mode
shape at the excitation point and the corresponding peak acceleration
calculated by comparing this impulse load to the modal mass, the modal
frequency and the mode shape. The decay of the acceleration is governed by
the damping, and by summing the decaying vibration from each mode, the
root-mean-square acceleration can be calculated.
3.1.3.
Weighting factors
During the calculation of the acceleration, the weighting factors as shown in

Figure 1.2 are taken into account to ensure that the acceleration calculated
is relevant to human perception. The weighted acceleration can then be
compared to limiting values such as those given in ISO 10137 or other
Standards or guidance.
3.1.4.
Contour plots
By performing this analysis for coincident excitation and response points all
over the floor plate, the vibration performance of different areas of the floor
can be established, as shown in Figure 3.2. This allows the architect or client
to position vibration sensitive areas of a building layout (such as operating
theatres, laboratories, etc.) in areas which are likely to have good vibration
performance, and conversely to locate less sensitive areas (such as
workshops, canteens, etc.) in the more lively areas of the floor.
Figure 3.2: Variation in vibration performance over a typical floor
3.1.5.
Detailed procedure
The exact calculation procedure, including input values such as Fourier

coefficients for walking and dancing activities and Standard defined
acceptability criteria are given in SCI publication 354, entitles Design of
Floors for Vibration: A New Approach[2]. This also includes guidance on the
modelling of floors in finite element software.
14
Background
The P354 method gives similar results to those using the OS-RMS90 method,
but gives a value that is more directly comparable to the limits given in
Standards such as ISO 10137. The effect of vibration travelling across the
floor plate, such as from a busy corridor into a sensitive operating theatre,
can be taken into account, and different weighting factors can be used for
different scenarios. It also allows for different excitations to be considered,
be they from walking or dancing activities or from machinery, and for the
effect of vibration on different receivers (such as sensitive measuring
equipment) to be investigated.
3.2. Other vibration considerations

Another important aspect of vibration design considered in P354 is the
magnification of loads during rhythmic activities. Human activities that can
induce vibration always subject the floor to a load greater than the static
load, but in the case of walking this increase is insignificant and certainly
falls well within the imposed loads considered for design. However, when a
group of people is involved in combined rhythmic activity (usually in
response to music, so aerobics or dancing, for example) the additional load
caused by the activity can exceed the loads that are considered for design,
and must be explicitly taken into account. In extreme cases this can also
cause fatigue problems with the structure.
3.3. Simplified approach

As an alternative to producing a finite element model of the structure, P354
presents a simplified, hand calculation approach. This is based on a
parametric study of a number of models using the modal superposition
approach, and again calculates an acceleration that can be compared to
limiting values in Standards.
15
Background
4. Improving structures
The three most effective ways of improving the response of a floor can be
seen by considering the OS-RMS90 plots. Two of these plots, at different
damping levels, are reproduced in Figure 4.1.
Classification based on a damping ratio of 6%
Classification based on a damping ratio of 3%

5
17
6
12
13
16
13 25
12
13
29
10
45
49 41
9
11
17
2 1.8
2.2
2.4
0.7 0.6
1.2
1
3.2 2.6
2.8
21
25
1.6
0.7 0.6
0.4
1.2 1
1.8
2.2
2.4
0.5
0.8
13
Eigenfrequency of the floor (Hz)
37
116
6
156
29
96
56
76
45
41
49
11
17
136
2.8
2.4
276
4
1.8
0.4
0.70.6 0.5
1 0.8
1.61.4 1.2
3.2
2.8
4
2.2
2
1.8
29
0.7
0.5 0.4
0.3
0.6
0.8 0.7 0.50.4
1
0.6
0.8
2.6 21.8
1.2
1.4
1
2.8 2.2
1.6
3.2
2.4
1.2
1.8
1.4
2
3
4
1.6
2.2
2.6
2.8
2.4
136
96
12
33
21
56
76
49
3.2
156
256236
0.3
0.2
45
41
10
98
13
3.2
5
3 2.6
25
2.8
2.2 1.8
1.6
1.4
1.2
3
2.2
2.6
1.8
2.4 2 1.6
1.4
1.2
1
0.80.7
0.6
1.8
2.6
2.2
2.4
3.22.8
1.2
1
0.8
0.5
2.4
25
3329
0.8
41
0.2
1.4
10
0.2
1.2
2.6
2.2
13
1.6
1.4
1.2
1
0.8
0.6
0.5
0.4
0.5
1.6
0.4
1.8
2.4
2.8
3.2
4
12
256
196
216
236
316
456 356
436
596
636
576
496
336
556
476
536
616
516
416
396
376
0.3
0.8
7
21
0.6
9
176
0.2
0.4
37
4
0.1
0.3
0.5
0.7
11
49
0.2
0.4
17
156
0.1
0.3
0.6
1.6
1.8
3.22.8
45
1.2
2.2
0.1
0.2
0.7
2.6
12
0.7
1.4
3
76
1.8
3.2
45
2.2
11 10
96
33
29
17
6
3
4
7
1.2
1.6
1.4
1.6
1.41.2
1
0.8
200
500
1000
0.4
0.8
2000
5000
10000
20000
50000
100000
200
500
Modal mass of the floor (kg)
1000
2000
5000
10000
0.8
0.7
0.6
0.5
0.4
0.7
0.6
0.5
2.6
2.4
0.3
1
100
0.5
1
2.8
25
56
176
0.3
0.7
1
2.4
2
0.3
1
100
0.2
0.5 0.4
0.7 0.6
1.2 1
0.8
1.6
1.8 1.4
2.2
1.2
2.6
116
0.3
0.6
0.8
136
296
276
0.7
0.4
0.4
0.3
0.6
56
0.5
1.4
21
0.1
0.2
0.8
1.6
1.8
13
96
0.4
0.7
1110
41
0.5
2
2.4
0.6
1.41.2
1.6
11 8
17
116
796
696 576
856
536 476
836
616
676
876
776
756
416336296
816
736
596 496
636
716
396
656
436
556
516 376
0.4
0.50.4
0.3
0.8 0.6
1
9
37
276
316
356
456
0.3
0.5
0.7
2.4
196
216
0.4 0.3
0.6
2.6
13
10
0.3
0.8
17
76
0.2
12
49
0.3
236
256 176
0.1
0.2
0.5
3
7
0.2
1.6
1.8
1.2
37
0.8
25
196
216
25
29
0.4
1.4
13
12
0.1
0.2
0.3
0.4
0.5
2.2
3
21
0.7 0.6
1
1.2
1.6 1.4
3.2 2.6
33
10
0.1
2.2
2.4
3.22.8
0.6
0.7
2.6
0.3
0.8
1
1.4
9
21
10
1110
11
0.3
12
12
17
12
0.2
13
1.2
14
0.3
1.4
16
15
2.6
2.2
2.4 1.6
1.8
3.2 2.8
2
8
17 10
0.2
0.8
18
0.1
0.4
0.5
19
0.2
0.3
13
33
20
0.1
0.4
0.5
0.1
10
11 37
0.8
1.6 1.4
0.70.6
1.2 1
3.2 2.6
2.8
21
12
2 1.8
2.2
2.4
5 4
9
11
17
14
10
15
1.61.4
3.2 2.6
2.8
Eigenfrequency of the floor (Hz)
20 10
9
19
7
11
8
18
20000
50000
100000
Modal mass of the floor (kg)
Figure 4.1 OS-RMS plots for 3% damping (left) and 6% damping (right)
It is clear from these plots that the vibration performance of the floor can be
improved in three different ways: increasing the modal mass (i.e. moving
from left to right within a plot); adjusting the frequency (i.e. moving
vertically within a plot); or increasing the damping (i.e. moving from the
same point on the left plot to the equivalent point on the right plot). When
using more advanced methods of analysis, it is also possible to adjust the
framing to isolate areas of the floor useful when considering the vibration
of operating theatres in hospitals.
4.1. Increasing the modal mass

Usually the most effective way of improving the vibration performance of a
structure is to increase the amount of mass participating in the motion. This
can be done either by increasing the distributed mass of the floor (by
increasing the slab depth, adding a screed, etc.), or by adjusting the
stiffnesses of the supporting steelwork to make a larger area of the floor
participate in the mode shape. A larger floor area will naturally have a larger
modal mass, and this will help to reduce the floor response. However, adding
mass to the structure will also affect the frequency, and so care should be
taken to ensure that the structure does not become more responsive even
with the additional mass as a result of a lower frequency.
16
Background
4.2. Adjusting the frequency

The frequency of the floor plate can easily be increased by increasing the
beam stiffness or, in existing structures, adding plates to the flanges of the
steelwork. The benefits of this are not necessarily significant though, as until
the floor frequency is above 9-10 Hz, there is little frequency dependence in
the vibration performance, and only tuning the structure between the
harmonics of the excitation function will improve the response.
4.3. Increasing the damping

In theory the most effective way of reducing the vibration response of a floor
is to increase the damping that is present, as the magnitude of the floor
response will approximately half if the damping is doubled. However, it is
difficult to add damping to floors, as damping systems generally need to be
connected to points where there is a lot of motion for them to be fully
effective. This would mean attaching dampers between floors in the centre of
beams or slabs, and this is impractical in most circumstances. Significant
damping can be added by the use of tuned mass dampers, but this is
generally not a consideration at the design stage, and is used more as a
remedial measure.
4.4. Structural means

For particularly sensitive areas of the floor, such as operating theatres, it
may be preferable to isolate the area from the remainder of the floor rather
than to try and design the entire floor plate to meet the stringent
requirements of the sensitive area. This can be achieved by providing areas
with different framing layouts (different spans, for example), by providing
beams with significantly higher stiffnesses than the typical beams, or by
adding in additional columns around the sensitive areas. The easiest way of
assessing these changes is to model the floor in finite element software, and
adjust the layout until the mode shapes show the isolation of the sensitive
areas.
4.5. Retrofit measures

Remedial action is often expensive and sometimes practicably impossible
within realistic physical constraints. In some situations, it may be feasible to
use measures which will merely reduce the annoyance associated with the
vibration instead of altering the nature or extent of the vibration itself. Such
measures include removing or reducing associated annoyance factors such
as noise caused by vibrating components, altering the timing of the problemcausing activity, or changing the architectural layout to move occupants
away from problem areas.
In general, the methods of improving structures detailed above apply equally
to retrofit measures. Changing the floor mass as a retrofit measure can be
effective, but care must be taken that the frequency of the floor doesnt
reduce such that the beneficial effect of the additional mass is counteracted.
17
Background
The frequency itself can be raised by welding additional steel to the flanges
of the existing beams, and used in combination with an increase in mass
(through a screed, for example) the vibration response can be significantly
improved.
The damping of the structure can also be improved, and the usual methods
of achieving this are:
Changing the placement of non-structural components such as partitions
Provision of tuned mass dampers
Provision of specialist damping materials
Changing the position or increasing the number of non-structural
components will aid the damping of a floor system. Unfortunately, as
damping is an extremely variable characteristic, it is impossible to accurately
quantify the exact improvement which will be provided by increasing such
components. Generally performance testing will be required to establish the
effectiveness of these remedial measures, and trial and error may have to be
used to obtain improvements.
Tuned mass dampers, which exhibit a passive control of floor movement,
may be utilised to reduce the response of the floor to forcing actions such as
footfall. A tuned mass damper (TMD) is a mass attached to the floor
structure through a spring and damping device. A TMD is effective, however,
only if the natural frequency of the TMD closely correlates with that of the
troublesome mode of floor vibration. TMDs which are initially tuned to the
floor vibration modes may become out-of-tune due to changes in the floors
natural frequencies resulting from alterations to the floor characteristics or
movement of materials locally. It should be noted that TMDs have a limited
frequency range where they are effective. As a consequence of this, a floor
with several problematic frequencies may need several TMDs to reduce the
floor response. Typically, the mass of a TMD will be between 2% and 5% of
the modal mass for each mode that needs tuning, and this can result in
problems supporting the additional load.
Specialist materials are generally used in constrained layer damping
systems. Materials with high energy dissipation are sandwiched between the
existing structure and an additional sheet of metal, and the strains that are
subsequently induced in the layer (both direct tension/compression and
shear) dissipate energy by hysteresis. Specialist advice should be sought to
determine whether this method of damping is appropriate and beneficial
(e.g. the effectiveness of the material may be dependent on temperature or
the amplitude of the strains).
18
Background
5. References
[1]
[2]
[3]
[4]
European Commission Technical Steel Research: Generalisation of criteria for

floor vibrations for industrial, office, residential and public building and gymnastic
halls, RFCS Report EUR 21972 EN, ISBN 92-79-01705-5, 2006, http://europa.eu.int
A.L. Smith, S. J. Hicks, P. J. Devine: Design of Floors for Vibration: A New
Approach. SCI 2007, ISBN 1-85942-176-8
Waarts, P. Trillingen van vloeren door lopen: Richtlijn voor het voorspellen, meten
en beoordelen. SBR, September 2005.
ISO 10137 Bases for design of structure Serviceability of buildings and walkways
against vibrations, International Organization for Standardization, 2007.
19

Vibration Design of Floors

Uploaded by

Copyright:

Available Formats

Vibration Design of Floors

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Vibration Design of Floors

Uploaded by

Copyright:

Available Formats

RFS2-CT-2007-00033

Human Induced Vibration of Steel Structures