You are on page 1of 14

Journal of Constructional Steel Research 164 (2020) 105764

Contents lists available at ScienceDirect

Journal of Constructional Steel Research


journal homepage: www.elsevier.com/locate/ijcard

Analysis of floor vibration evaluation methods using a large database


of floors framed with W-Shaped members subjected to walking
excitation
Mohammad Royvaran a, Onur Avci b, 1, *, Brad Davis c
a
Graduate Research Assistant Department of Civil Engineering, University of Kentucky, 382 Raymond Building, Lexington, KY, 40506, USA
b
Department of Civil and Architectural Engineering, College of Engineering, Qatar University, Doha, Qatar
c
Department of Civil Engineering, University of Kentucky, 373 Raymond Building, Lexington, KY, 40506, USA

a r t i c l e i n f o a b s t r a c t

Article history: Floor vibration is widely recognized as an important limit state in the design of steel-framed floors, and
Received 12 March 2019 various evaluation methods have been developed over the years. These methods range from simplified
Received in revised form methods that are suitable for routine usage during the design phase to complex computerized or
8 August 2019
probabilistic models. The former are in common usage in the structural design community. It is critically
Accepted 13 September 2019
Available online 15 November 2019
important to structural engineers that the accuracies of the commonly used methods are known;
however, a study of these methods based on a large database of recorded observations is not available in
the literature. Therefore, the authors investigated the evaluation accuracy of four well-known simplified
Keywords:
Steel framed floors
methods by comparing predicted and observed acceptability (whether or not occupants complained) of
Floor vibrations 50 floor bays in real buildings framed with W-shaped members subjected to walking excitations. The
Vibration serviceability evaluation percentage of correct predictions is used to judge the accuracy of each method. It is observed that the
Serviceability methods AISC Design Guide 11 accurately predicted acceptability of floors; two methods from the SCI P354
Walking excitations (Simplified Method and Vibration Dose Value Method) fairly accurately predicted acceptability of floors.
The HIVOSS method provided unconservative predictions. In addition, modified version of the P354
method was investigated in an attempt to improve the accuracy of these methods.
© 2019 Elsevier Ltd. All rights reserved.

1. Introduction is absent when open floor plans are chosen. As a result, modern
floors are often more vulnerable to annoying vibrations due to
With more slender architectural design trends and the high- walking [10e12]. In the last couple of decades, floor vibration due
strength materials made available by modern construction tech- to human activity has become recognized as an important limit
nologies in recent decades, structural floor bays have become state for steel-framed floor systems [13e15].
larger, with longer beams and girders [1e3]. In addition, most Floor vibrations are caused by several types of human activities
building floors often have fewer and lighter furnishings than they such as walking, running, dancing, and aerobics [16e18]. Some
did decades ago, especially in offices as paper offices have been floors known to vibrate at low frequencies may be subject to vandal
phased out in favor of electronic offices with fewer bookshelves and jumping, during which people jump or bounce on the floor in a
filing cabinets [4e6]. Finally, and most importantly, open floor purposeful effort to cause strongly perceptible vibrations [19]. Yet,
plans are often chosen over traditional plans with many full-height the duration of any activity is a significant parameter in service-
partition walls, resulting in much lower superimposed mass and ability assessments [29,30]. While all of these vibration sources
damping [7e9]. It should also be noted that the full-height parti- must be considered by the structural engineer, this paper focuses
tions add very significantly to the stiffness of the bay, and that effect on the most common excitation of floors: human walking. The
typical concern is that one occupant might walk and cause vibra-
tions that annoy another occupant. In steel-framed floors, such
* Corresponding author. problems almost always occur when the frequency of one of the
E-mail addresses: m.royvaran@uky.edu (M. Royvaran), oavci@vt.edu (O. Avci), first four harmonics of the walking force equals a natural frequency,
dbraddavis@uky.edu (B. Davis). resulting in a short resonant build-up. For example, walking at
1
“Former Assistant Professor” at the Department of Civil and Architectural En-
1.5 Hz will cause resonance with natural frequencies around 1.5 Hz,
gineering, Qatar University, Doha, Qatar.

https://doi.org/10.1016/j.jcsr.2019.105764
0143-974X/© 2019 Elsevier Ltd. All rights reserved.
2 M. Royvaran et al. / Journal of Constructional Steel Research 164 (2020) 105764

3 Hz, 4.5 Hz, and 6 Hz. Considering that people routinely walk at 2. Floor vibration serviceability prediction methods
frequencies between approximately 1.6 Hz and 2.2 Hz, a floor with
at least one natural frequency below 9 Hz potentially exhibit to 2.1. AISC design guide 11 chapter 4 method
resonant responses [20].
There are many methods for evaluating floor vibrations, ranging The AISC Design Guide 11 Chapter 4 [20], hereafter referred to as
from probabilistic methods that require complex computerization DG11, contains the most widely used floor vibration evaluation
outside the reach of typical structural engineering offices to method in North America. In this method, the bay is idealized as a
simplified methods that produce a binary “satisfactory” or “un- single degree of freedom (SDOF) system with the fundamental
satisfactory” evaluation of a floor bay. The latter are suitable for frequency and effective mass of the floor bay. The walking-induced
manual calculations and simple computerization and are thus used dynamic load is the product of the bodyweight of the walker and a
in the vast majority of floor vibration evaluations in design offices curve-fit of the Rainer et al. [23] dynamic coefficients, as stated in
today. Ref. [24]. The evaluation criterion is expressed by Equation (1). The
From the perspective of a structural engineer, the most impor- bay evaluation is “satisfactory” if the peak acceleration, ap, does not
tant feature of a method is the accuracy of the final evaluation. In exceed the tolerance limit, ao, which is 0.5%g for low frequency
other words, the method must successfully predict whether the bay floors supporting quiet areas such as offices.
will be satisfactory or unsatisfactory. Predictions of intermediate
results such as natural frequency and acceleration are of lower ap Po e0:35fn ao
practical importance than they are to designers. The consequences ¼  (1)
g bW g
of a false “satisfactory” evaluation range from complaints to liti-
gation and expensive vibration reduction retrofits. The conse- where.
quence of a false “unsatisfactory” evaluation is inefficient usage of
resources and competitive disadvantage. Thus, it is very important fn ¼ bay fundamental frequency, Hz
that structural engineers know the level of accuracy of the b ¼ critical damping ratio
simplified methods that are feasible choices for their design W ¼ bay effective weight, lb
process.
The accuracies of the various methods are largely unknown and As described in Refs. [20,24], Po is a constant 65 lb (289.1 N) force
no comprehensive study is available in the literature. Therefore, the that is the product of 157 lb (698.4 N), part of the dynamic coeffi-
present study intends to fill that gap by comparing predicted cient, and an adjustment factor, 0.5, that accounts for incomplete
evaluations from the following four simplified evaluation methods resonant buildup and the fact that the walker, affected occupant, or
with observed acceptability of each bay in a large database of in- both are probably not at mid-bay.
service floor bays for floors framed with wide flange members The bay natural frequency is computed using Equation (2)
subjected to walking excitation. The percentages of correct pre- (Equations (3) and (4) in DG11); however, according to Pabian
dictions for satisfactory bays, unsatisfactory bays, and all bays are et al. [25], this equation underestimates the bay natural frequency.
used to assess the accuracy of each method.
sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
 American Institute of Steel Construction (AISC) Design Guide 11 g
fn ¼ 0:18 (2)
Chapter 4, by Murray et al. [20]. db þ dg
 Steel Construction Institute (SCI) P354 Simplified Method, by
Smith et al. [21]. where db and dg are beam and girder deflections, respectively, due
 Steel Construction Institute (SCI) P354 Vibration Dose Values, by to the weight supported at midspan, computed using the classical
Smith et al. [21]. equation for a simply supported beam with uniform mass and
 Human Induced Vibration of Steel Structures (HIVOSS) Method flexural rigidity.
[22]. The effective weight, W, is double the weight of the effective
mass, and is a weighted average of the beam mode and girder
Potential improvements to the P354 method were also bending mode effective weights [24].
investigated. In this paper, Equation (1) is used to evaluate each bay for
The authors would like to emphasize that the objective of the comparison with known acceptability. In some cases, Equation (1)
manuscript is to examine the accuracy of four methods on vibra- predicts accelerations very close to the limit. Considering that hu-
tions serviceability of floors. These methods are mostly applicable man perception of floor vibration is very subjective, such pre-
to typical floor bays which comprise the vast majority of floor bays dictions neither indicate that the floor is unsatisfactory or
in typical buildings. The scope of the manuscript does not include satisfactory. When the predicted acceleration is 0.5%g plus or minus
irregular bays that would likely need a finite element analysis. The 0.03%g, which is arbitrarily set, and when the bay did not have
purpose is to evaluate the methods that are in wide usage in many complaints, the evaluation is counted as agreeing with
structural design offices. observed acceptability. Note that a similar tolerance was used for all
It should also be mentioned that all four methods investigated prediction methods.
in this study are based on the SDOF assumption which makes cal-
culations relatively simpler and faster when compared to a po-
tential MDOF procedure. The intention of the research is to evaluate
simplified methods that are well within the reach of typical
structural design offices. With such methods, it is typically only 2.2. SCI P354 simplified method
practical to consider the fundamental mode of a floor. Part of the
research question is whether or not such simplified methods pro- With the Simplified Method in Chapter 7 of the SCI P354, Design
vide accurate evaluations. More sophisticated methods that take of Floors for Vibration: A New Approach [21], the floor bay evaluation
into account multiple modes typically require finite element ana- is satisfactory if the response factor, R, does not exceed the toler-
lyses that are beyond the reach of many structural engineering ance limit, which is 8.0 for quiet spaces such as offices. R is
offices. computed using:
M. Royvaran et al. / Journal of Constructional Steel Research 164 (2020) 105764 3

minus a tolerance of 3% and when the bay did not have many
aw;rms
R¼ (3) complaints, the evaluation is counted as agreeing with observed
0:005 m=sec:2 acceptability.

where aw,rms is the frequency-weighted root mean square (rms)


2.3. SCI P354 Vibration Dose Values method
acceleration due to walking.
As with DG11, the bay is idealized as an SDOF system with the
Occupants probably accept higher levels of vibrations if the vi-
fundamental frequency and effective mass of the floor bay. The
bration events are short, infrequent, or both. SCI P354 accounts for
dynamic load is the product of bodyweight, Q, and 0.1, a high end
this using the vibration dose value (VDV) approach. If the VDV does
approximation of the dynamic load factors for the second through
not exceed the appropriate limit in P354, then the floor evaluation
fourth harmonics from Kerr [26]. The mid-bay acceleration is scaled
is satisfactory. The general equation for the VDV is:
by the mode shape values at the walker and affected occupant lo-
cations. Human tolerance of vibration depends on the frequency of 2T 31=4
ð
vibrations, so a weighting factor is also included. The predicted 4 4 5
VDV ¼ ½aw ðtÞ dt (8)
frequency-weighted rms acceleration due to walking is:
0
0:1Q
aw;rms ¼ me mr pffiffiffi Wr (4) where.
2 2M b

where. T ¼ exposure duration, taken as 16 h for daytime occupancy of a


building, sec.
me ¼ unity-normalized mode shape value at the excitation aw(t) ¼ acceleration at time, t, m/sec.2 SCI P354 provides the
(walker) following convenient equation for predicting the VDV:
mr ¼ unity-normalized mode shape value at the response p ffiffiffiffiffiffiffiffiffiffi
(affected occupant) VDV ¼ 0:68aw;rms 4
na Ta (9)
where.
b ¼ critical damping ratio
Q ¼ weight of walker, taken as 746 N (168 lb)
na ¼ number of events during the exposure duration
M ¼ modal mass, kg
aw,rms ¼ frequency-weighted rms acceleration (m/sec.2)
W ¼ weighting factor for human perception of vibrations (Fig-
ures 5.3 and 7.5 of P354)
The duration of a walking event, Ta, is computed using the bay
r ¼ partial build-up factor
dimension and velocity of walking, in m/sec., computed using
Equation (7) (Equation 16 of P354) where fp is between 1.7 Hz and
The partial build-up factor, r, is computed using Equation (5).
2.4 Hz.
P354 incorrectly includes the pacing frequency, fp, in the exponent
The difficulty in using Equation (9) is the lack of information on
rather than the fundamental natural frequency, fo,. The fact that fo is
the number of events, na, that occurs during a typical work day. To
correct is evident from the derivation based on an SDOF system
the writers’ knowledge, a recommended value is not available, so a
excited by a sinusoidal force at the fundamental frequency. Lp/v is
designer is unable to evaluate the bay using Equation (9). This
the duration of the resonant buildup, where Lp and v are the
deficiency is addressed for offices in the present research as follows.
walking path length (m) and velocity of walker (m/sec.2)
For a 16 h exposure duration, P354 indicates VDV values between
respectively.
0.2 m/sec.1.75 and 0.4 m/sec.1.75 correspond to “low probability of
2pfo bLp adverse comment” as mentioned in Table 5.4 of P354. If the
r¼1  e v (5) computed VDV does not exceed the upper end, 0.4 m/sec.1.75, then
the evaluation for that bay is satisfactory. An iterative approach was
v ¼ 1:67 f 2p  4:83 fp þ 4:50 (6) used to investigate the effect of selection of na on evaluation pre-
diction accuracy. Note that 3% tolerance was used as with the other
W and M are functions of the fundamental frequency, fo, methods.
computed using Equation (7). fo is computed for the secondary
beam bending mode and each primary beam (girder) bending 2.4. HIVOSS method
mode separately; the minimum is assigned to the bay. Pabian et al.
[25] concluded that the natural frequency is accurately predicted by With the Human Induced Vibration of Steel Structures (here-
choosing the minimum of the secondary and primary beam natural after referred to as HIVOSS [22] “hand calculation method,” toler-
frequencies. ance limits are expressed in terms of floor classes as shown in
Table 1 (Table 1 of the HIVOSS guideline). For example, floors in
18
fo ¼ pffiffiffiffi (7)
d
where d, in mm, is the sum of deflections for the slab and secondary Table 1
Floor recommendations for the classes in HIVOSS guideline.
beam in the beam bending mode and sum of deflections for the
slab, secondary beam, and primary beam for primary beam (girder) Class Function of Floor
bending mode. Different boundary conditions are used in P354 for Healthcare Education Offices
each bending mode in Section 7.1 of this guideline.
A Recommended Recommended Recommended
In this paper, R from Equation (3) was compared to the limit, 8, B Recommended Recommended Recommended
to evaluate each floor bay. If the calculated R does not exceed 8, C Recommended Recommended Recommended
then the floor is predicted to be satisfactory. In some cases, P354 D Critical Critical Recommended
predicts R very close to the limit, so the evaluation is neither E Not Recommended Not Recommended Critical
F Not Recommended Not Recommended Not Recommended
satisfactory nor unsatisfactory. When the predicted R is 8 plus or
4 M. Royvaran et al. / Journal of Constructional Steel Research 164 (2020) 105764

Classes A through D are recommended for offices, floors in Class E  Combination 4: fn is calculated with the Dunkerley approach in
are critical, and floors in Class F are not recommended. The floor HIVOSS Appendix A.5, and Mmod is based on the effective weight
class is determined by the 90th percentile 1-s root mean square from DG11 Chapter 4.
acceleration (OS-RMS90), which is a function of the natural fre-
quency (called eigenfrequency in the guideline), modal mass, and The combinations are summarized in Table 2.
damping ratio of the bay.
The HIVOSS method is based on the response of an SDOF rep-
3. Comprehensive database of floors
resentation of the bay to a single footstep, thus ignoring resonant
buildup. The OS-RMS90 is determined from floor bay characteristics
The database of floor bays was developed by three researchers
and a standard walking load function for a person with 90th
over approximately two decades. The majority of the data were
percentile weight and walking pace [22]. The OS-RMS90 value can
collected by Dr. Thomas M. Murray of Virginia Tech. The remainder
be determined by plotting the natural frequency versus the modal
are from Davis [6] and Pabian [27].
mass of the bay in a series of graphs that vary by the damping ratio.
Each floor bay, which is listed in Table 3, consists of a concrete
The fundamental frequency can be determined by simple
slab on corrugated steel deck and steel I-section beams and girders.
analytical formulas or by finite element analysis. Two analytical
The degree of complaints is specified in the last column of
formula approaches are available in HIVOSS. In the self-weight
Table 3. Bay 1 and Bays 6 through 50 support quiet areas, such as
deflection approach from Appendix A.4, Equation (7) is used. In
hospitals, universities, and mostly offices, and the level of com-
the Dunkerley approach from Appendix A.5, Equation (2) is used. In
plaints was determined by interviewing occupants. “MC”, “SC”, “FC”
both approaches, when the girders are different, the girder with the
and “NC” codes correspond to many, some, few, and no complaints,
highest deflection (corresponding to the lowest girder frequency) is
respectively. In practically all cases with complaints, the building
used. As described by Pabian et al. [25], the self-weight approach
occupants, architect, or engineer-of-record contacted Dr. Murray
which uses the natural frequency equal to the minimum of the
and requested assistance in reducing vibration levels to alleviate
beam and girder natural frequencies is more accurate than Dun-
human discomfort or irritation. The bays with “no complaints” have
kerley approach which underestimates the natural frequency.
been in service for at least five years without any complaints being
The general equation for the modal mass is Equation (10).
issued. For Bays 2 through 5, “CP” indicates that the bay was under
construction and exhibited lively vibration due to walking. Bays
ð with “NC” codes are considered satisfactory; all other bays are
2
Mmod ¼ m d ðx; yÞdxdy (10) considered unsatisfactory. The database contains 22 satisfactory
bays and 28 unsatisfactory bays, for a total of 50 bays.
The damping values were set based on recommendations in the
where. design guides and the judgment of the researchers. The bays with
SDL þ LL ¼ 0 are bare slabs.
m ¼ distributed mass, kg/m2 The following variables are used in Table 3:

The unity-normalized mode shape value, d(x,y) is estimated wc ¼ weight of concrete, pcf
based on engineering judgment, bay geometry, and boundary fc’ ¼ minimum compressive strength of concrete, ksi
conditions. SDL ¼ superimposed dead load, psf
For the bays in this study, Equation (10) becomes Equation (11) LL ¼ live load, psf
(HIVOSS Appendix A.6) using the beam and girder deflections, dx
and dy, respectively.

4. Comparisons of observed and predicted evaluations


!
d2x 2
þ dy 8 dx dy The 50 floors in the database were evaluated using the four
Mmod ¼ Mtotal 2
þ (11)
2d p2 d2 methods described in Section 2. Table 4 through 12 summarize
comparisons of observed and predicted evaluations. The observed
where d is the sum of beam and girder deflections and Mtotal is the evaluations in the tables are OK, indicating no complaints, or NG
total weight (kg) of the bay. For bays with unequal girders, Mmod (“No Good”), otherwise. The predicted evaluations are “OK,” “B” for
was calculated based on the girder with the lower frequency. borderline, or “NG.” It should be noted that borderline predictions
The guideline provides several methods to calculate the mass are counted as agreeing with observations unless there were many
and frequency of a floor bay, but does not enforce a specific method complaints as explained in Section 2.
for the calculation of them. For example, finite element analysis can
be used to determine these quantities. Therefore, to fully investi- 4.1. AISC design guide 11 chapter 4 method
gate the accuracy of the HIVOSS method, the following four com-
binations of frequency and modal mass calculations were used. DG11 results are summarized in Table 4, which includes the
following variables.
 Combination 1: fn is calculated with the self-weight approach
from HIVOSS Appendix A.4 and Mmod is calculated by Equation
(11), which is from HIVOSS Appendix A.6.
Table 2
 Combination 2: fn is calculated with the Dunkerley approach
Combinations of frequencies and masses used in HIVOSS calculations.
from HIVOSS Appendix A.5 and Mmod from Equation (11), which
is from HIVOSS Appendix A.6. Combination fn Mmod
 Combination 3: natural frequency, fn, is calculated from self- 1 Self-weight (HIVOSS, A.4) (HIVOSS, A.6)
weight approach from HIVOSS Appendix A.4, and modal mass, 2 Dunkerley (HIVOSS, A.5) (HIVOSS, A.6)
Mmod, is based on the effective weight calculated based on DG11 3 Self-weight (HIVOSS, A.4) Effective weight (DG11)
4 Dunkerley (HIVOSS, A.5) Effective weight (DG11)
Chapter 4 (Mmod ¼ W/2g).
Table 3
Database of floor bays.

Bay Length Floor Size Beam Size Beam Girder Size (Left, Right) Slab Depth (Total/ Concrete Concrete SDL þ LL b DG11 & SCI/HIVOSS (%) Complaints
Spacing (if different) Deck) weight (wc) strength (f'c)
Beam Girder Width x Length

(ft) (m) (ft) (m) (ft) x (ft) (m) x (m) (ft) (m) (in/in) (mm/mm) (pcf) (kN/m3) (ksi) (MPa) (psf) (kPa)

1 32.8 10.0 21.3 6.5 120  50.5 36.6  15.4 W16  31 10.7 3.26 W16  31, Wall 5.6/3.0 142/76 115 18.1 3.50 24.1 21 1.01 3.0/3.0 FC
2 30.0 9.1 30.0 9.1 120  90 36.6  27.4 W16  36 10.0 3.05 W21  55 5.5/2.0 140/51 145 22.8 4.00 27.6 0 0.00 1.0/1.0 CP
3 30.0 9.1 30.0 9.1 47  90 14.3  27.4 W16  36 10.0 3.05 W21  55 5.5/2.0 140/51 145 22.8 4.00 27.6 0 0.00 1.0/1.0 CP
4 30.0 9.1 30.0 9.1 120  90 36.6  27.4 W16  36 10.0 3.05 W21  55 5.5/2.0 140/51 145 22.8 4.00 27.6 0 0.00 1.0/1.0 CP
5 30.0 9.1 30.0 9.1 120  60 36.6  18.3 W16  36 10.0 3.05 W21  62 5.5/2.0 140/51 145 22.8 4.00 27.6 0 0.00 1.0/1.0 CP
6 32.8 10.0 21.3 6.5 100  50.5 30.5  15.4 W16  31 10.7 3.26 W16  26, W16  40 5.6/3.0 142/76 115 18.1 3.50 24.1 21 1.01 3.0/3.0 SC
7 40.0 12.2 36.2 11.0 111  120 33.8  36.6 W16  31 9.04 2.76 W21  101 6.0/2.0 152/51 145 22.8 3.00 20.7 12 0.57 2.5/3.0 MC
8 31.1 9.5 36.2 11.0 64.3  111 19.6  33.8 W14  22 9.04 2.76 W21  93, Wall 6.0/2.0 152/51 145 22.8 3.00 20.7 12 0.57 2.53.0 NC
9 36.0 11.0 28.0 8.5 140  140 42.7  42.7 W16  31 7.0 2.13 W24  62, W24  68 4.5/2.0 114/51 110 17.3 4.00 27.6 0 0.00 1.0/1.0 MC

M. Royvaran et al. / Journal of Constructional Steel Research 164 (2020) 105764


10 38.8 11.8 30.0 9.1 120  57 36.6  17.4 W18  40 10.0 3.05 Wall, W21  44 4.5/2.0 114/51 145 22.8 3.50 24.1 15 0.72 3.0/4.0 SC
11 42.0 12.8 30.0 9.1 120  42 36.6  12.8 W21  44 10.0 3.05 Wall, W21  50 4.5/2.0 114/51 145 22.8 3.50 24.1 15 0.72 3.0/4.0 SC
12 30.0 9.1 30.0 9.1 90  81 27.4  24.7 W16  26 10.0 3.05 W18  50, W21  50 5.25/2.0 133/51 110 17.3 4.00 27.6 0 0.00 1.5/1.0 SC
13 30.0 9.1 30.0 9.1 90  71.9 27.4  21.9 W16  26 10.0 3.05 W18  50, W21  50 5.25/2.0 133/51 110 17.3 4.00 27.6 0 0.00 1.5/1.0 SC
14 30.0 9.1 30.0 9.1 90  71.9 27.4  21.9 W16  26 10.0 3.05 W18  50, W21  50 5.25/2.0 133/51 110 17.3 4.00 27.6 0 0.00 1.5/1.0 SC
15 30.0 9.1 30.0 9.1 60  90 18.3  27.4 W16  26 10.0 3.05 W18  50, W21  50 5.25/2.0 133/51 110 17.3 4.00 27.6 0 0.00 1.5/1.0 SC
16 30.0 9.1 30.0 9.1 60  90 18.3  27.4 W16  26 10.0 3.05 W18  50, W21  50 5.25/2.0 133/51 110 17.3 4.00 27.6 0 0.00 1.5/1.0 SC
17 30.0 9.1 30.0 9.1 90  90 27.4  27.4 W16  26 10.0 3.05 W18  50, W21  62 5.25/2.0 133/51 110 17.3 4.00 27.6 0 0.00 1.5/1.0 SC
18 30.0 9.1 40.8 12.4 105  150 32  45.7 W16  26 10.2 3.11 W24  68 5.25/2.0 133/51 115 18.1 3.00 20.7 11 0.53 5.0/4.0 NC
19 28.0 8.5 37.0 11.3 62  150 18.9  45.7 W18  35 12.3 3.75 W24  84 6.25/3.0 159/76 115 18.1 3.00 20.7 11 0.53 5.0/4.0 NC
20 40.0 12.2 32.0 9.8 40  150 12.2  45.7 W21  62 10.7 3.26 W21  62 6.25/3.0 159/76 115 18.1 3.00 20.7 11 0.53 5.0/4.0 NC
21 33.0 10.1 32.0 9.8 96  150 29.3  45.7 W18  40 10.7 3.26 W24  62, W21  44 6.25/3.0 159/76 115 18.1 3.00 20.7 11 0.53 5.0/4.0 NC
22 30.0 9.1 30.0 9.1 75  150 22.9  45.7 W16  26 7.5 2.29 W24  55,W36  135 6.25/2.0 159/51 115 18.1 3.00 20.7 5 0.24 2.0/2.0 NC
23 28.3 8.6 30.0 9.1 36  60 11.2  18.3 W16  26 7.5 2.29 W24  55 6.25/2.0 159/51 115 18.1 3.00 20.7 15 0.72 5.0/4.0 NC
24 30.0 9.1 30.0 9.1 60  60 18.3  18.3 W16  26 7.5 2.29 Wall, W24  55 6.25/2.0 159/51 115 18.1 3.00 20.7 11 0.53 4.0/4.0 NC
25 66.6 20.3 21.3 6.5 64  98.1 19.5  29.9 W36  135 10.7 3.26 W24  55 7.5/3.0 191/76 145 22.8 3.50 24.1 6 0.29 3.0/3.0 NC
26 66.6 20.3 21.3 6.5 64  98.1 19.5  29.9 W36  135 10.7 3.26 W24  55 7.5/3.0 191/76 145 22.8 3.50 24.1 10 0.48 3.5/3.0 NC
27 30.0 9.14 44.5 13.6 71.5  110 21.8  33.5 W14  30 8.9 2.71 W30  90 5.5/3.0 140/76 145 22.8 4.00 27.6 15 0.72 3.0/3.0 NC
28 40.0 12.2 30.0 9.1 90  40 27.4  12.2 W18  40 10.0 3.05 Wall, Wall 5.5/3.0 140/76 145 22.8 4.00 27.6 15 0.72 3.0/3.0 NC
29 40.0 12.2 44.5 13.6 120  110 36.6  33.5 W18  40 8.9 2.71 Wall, W30  90 5.5/3.0 140/76 145 22.8 4.00 27.6 15 0.72 3.0/3.0 NC
30 40.0 12.2 45.0 13.7 120  40 36.6  12.2 W18  40 9.0 2.74 Wall, Wall 5.5/3.0 140/76 145 22.8 4.00 27.6 15 0.72 3.0/3.0 NC
31 40.0 12.2 30.0 9.1 90  40 27.4  12.2 W18  35 10.0 3.05 Walls 6.25/3.0 159/76 110 17.3 3.00 20.7 22 1.05 2.5/3.0 SC
32 30.0 9.1 30.0 9.1 90  30 27.4  9.1 W16  26 10.0 3.05 Walls 5.5/3.0 140/76 145 22.8 3.50 24.1 4.5 0.22 2.5/3.0 NC
33 40.0 12.2 30.0 9.1 90  40 27.4  12.2 W18  40 10.0 3.05 Walls 5.5/3.0 140/76 145 22.8 3.50 24.1 4.5 0.22 2.5/3.0 NC
34 43.0 13.1 30.0 9.1 90  118 27.4  36 W21  50 10.0 3.05 Wall, W2784 5.25/2.0 133/51 115 18.1 3.00 20.7 15 0.72 2.5/2.0 NC
35 43.0 13.1 30.0 9.1 90  119.5 27.4  36.4 W21  50 10.0 3.05 Wall, W30  99 5.25/2.0 133/51 115 18.1 3.00 20.7 15 0.72 2.5/2.0 NC
36 43.0 13.1 30.0 9.1 90  119.5 27.4  36.4 W21  50 10.0 3.05 Wall, W30  99 5.25/2.0 133/51 115 18.1 3.00 20.7 15 0.72 1.0/1.0 SC
37 43.0 13.1 30.0 9.1 90  119.5 27.4  36.4 W21  50 10.0 3.05 Wall, W24  76 5.25/2.0 133/51 115 18.1 3.00 20.7 15 0.72 1.0/1.0 SC
38 45.8 14.0 30.8 9.4 267.2  129.3 81.4  39.4 W21  44 10.3 3.14 Wall, W27  84 5.25/2.0 133/51 115 18.1 3.50 24.1 25 1.17 2.5/3.0 NC
39 32.5 9.9 30.0 9.1 90  81 27.4  24.7 W16  31 7.5 2.29 Wall, W21  44 5.5/2.0 140/51 145 22.8 3.50 24.1 15 0.72 3.0/3.0 NC
40 32.5 9.9 30.0 9.1 90  48.5 27.4  14.8 W16  31 7.5 2.29 Wall, W21  44 5.5/2.0 140/51 145 22.8 3.50 24.1 15 0.72 3.0/3.0 MC
41 28.0 8.53 30.0 9.1 90  87 27.4  26.5 W16  26 10.0 3.05 W21  44, Wall 5.5/3.0 140/76 145 22.8 3.50 24.1 13 0.62 3.0/2.0 SC
42 30.0 9.14 30.0 9.1 90  87 27.4  26.5 W16  26 10.0 3.05 W21  44 5.5/3.0 140/76 145 22.8 3.50 24.1 15 0.72 3.0/3.0 SC
43 30.0 9.14 30.0 9.1 90  87 27.4  26.5 W16  26 10.0 3.05 W21  44 5.5/3.0 140/76 145 22.8 3.50 24.1 13 0.62 3.0/2.0 SC
44 30.0 9.14 30.0 9.1 90  90 27.4  27.4 W16  26 10.0 3.05 W21  62 7.5/2.0 191/51 115 18.1 3.00 20.7 0 0.00 1.5/1.0 FC
45 30.0 9.1 30.0 9.1 80  160 24.4  48.8 W16  26 10.0 3.05 W21  62, W21  57 5.25/3.0 133/76 145 22.8 4.00 27.6 12 0.57 2.5/2.0 SC
46 40.0 12.2 30.0 9.1 120  40 36.6  12.2 W18  40 10.0 3.05 Walls 5.25/3.0 133/76 145 22.8 4.00 27.6 12 0.57 2.5/2.0 SC
47 40.0 12.2 30.0 9.1 120  160 36.6  48.8 W18  40 10.0 3.05 Wall, W21  62 5.25/3.0 133/76 145 22.8 4.00 27.6 12 0.57 2.5/2.0 MC
48 42.0 12.8 30.0 9.1 90  120 27.4  36.6 W21  44 10.0 3.05 W21  50, W21  73 5.63/3.0 143/76 110 17.3 4.00 27.6 15 0.72 3.0/4.0 SC
49 30.0 9.1 52.5 16.0 52.5  85 16  25.9 W16  31 8.75 2.67 Wall, W30  132 4.0/3.0 102/76 145 22.8 4.00 27.6 26 1.24 3.0/3.0 NC
50 25.0 7.6 52.5 16.0 52.5  85 16  25.9 W16  26 8.75 2.67 W30  108 4.0/3.0 102/76 145 22.8 4.00 27.6 26 1.24 3.0/3.0 NC

5
6 M. Royvaran et al. / Journal of Constructional Steel Research 164 (2020) 105764

Table 4
Comparisons of observations and predictions e AISC Design Guide 11 Chapter 4.

Bay Observed Evaluation Prediction Correct Prediction?

fn W ap Evaluation

(Hz) (kips) (kN) (%g)

1 NG 4.35 65.5 291.4 0.761 NG Yes


2 NG 4.95 98.8 439.6 1.165 NG Yes
3 NG 4.95 97.9 435.5 1.176 NG Yes
4 NG 4.95 98.8 439.6 1.165 NG Yes
5 NG 5.12 80.3 357.2 1.350 NG Yes
6 NG 4.45 65.8 292.7 0.729 NG Yes
7 NG 2.94 137.0 609.4 0.679 NG Yes
8 OK 3.50 151.7 674.8 0.504 B Yes
9 NG 5.32 60.7 270.0 1.661 NG Yes
10 NG 3.69 75.9 337.5 0.785 NG Yes
11 NG 3.61 75.3 334.9 0.813 NG Yes
12 NG 4.76 67.8 301.5 1.210 NG Yes
13 NG 4.76 63.0 280.4 1.301 NG Yes
14 NG 4.76 63.0 280.4 1.301 NG Yes
15 NG 4.76 71.0 315.9 1.155 NG Yes
16 NG 4.76 71.0 315.9 1.155 NG Yes
17 NG 4.76 71.0 315.9 1.155 NG Yes
18 OK 3.57 131.2 583.5 0.284 OK Yes
19 OK 4.84 119.1 529.6 0.200 OK Yes
20 OK 4.11 136.6 607.8 0.226 OK Yes
21 OK 4.45 116.9 519.9 0.234 OK Yes
22 OK 5.45 92.6 411.8 0.521 B Yes
23 OK 5.06 81.3 361.6 0.272 OK Yes
24 OK 4.93 89.8 399.6 0.323 OK Yes
25 OK 3.83 253.3 1126.9 0.228 OK Yes
26 OK 3.75 264.2 1175.1 0.192 OK Yes
27 OK 3.54 172.0 765.1 0.364 OK Yes
28 OK 4.56 102.4 455.5 0.429 OK Yes
29 OK 3.17 198.6 883.4 0.359 OK Yes
30 OK 4.76 100.8 448.4 0.406 OK Yes
31 NG 4.21 107.4 477.9 0.554 NG Yes
32 OK 6.68 54.2 241.1 0.463 OK Yes
33 OK 4.93 85.7 381.2 0.540 NG No
34 OK 4.12 132.5 589.4 0.463 OK Yes
35 OK 4.43 93.3 415.0 0.591 NG No
36 NG 4.27 132.0 587.2 1.106 NG Yes
37 NG 4.19 97.1 431.9 1.545 NG Yes
38 OK 3.28 177.6 790.2 0.465 OK Yes
39 OK 4.19 103.8 461.8 0.481 B Yes
40 NG 4.19 78.6 349.6 0.635 NG Yes
41 NG 4.44 101.5 451.5 0.452 OK No
42 NG 4.10 107.8 479.6 0.478 B Yes
43 NG 4.16 104.7 465.7 0.482 B Yes
44 NG 4.85 113.9 506.5 1.047 NG Yes
45 NG 4.43 94.7 421.2 0.583 NG Yes
46 NG 4.70 89.8 399.5 0.559 NG Yes
47 NG 3.84 130.5 580.5 0.519 B No
48 NG 3.94 124.6 554.2 0.439 OK No
49 OK 3.47 159.5 709.5 0.403 OK Yes
50 OK 3.44 170.5 758.4 0.382 OK Yes

fn ¼ fundamental natural frequency, Hz bays in the database. Overall, DG11 provided accurate evaluations.
W ¼ effective weight, kips It should be noted that the inaccurate predictions of AISC DG11
b ¼ damping ratio take place when the predicted ap values are close to the limit of 0.5%
ap ¼ peak acceleration due to walking excitation as a percentage g. In Table 4, it is observed that the inaccurate five cases (out of fifty
of the acceleration of gravity bays) are for predicted ap values of 0.540, 0.591, 0.452, 0.519, and
0.439 (respectively for bay 33, bay 35, bay 41, bay 47 and bay 48).
DG11 provided accurate evaluations for 90.9% (20 of 22 bays) of Based on the histogram (Fig. 1), the number of predicted ap values
satisfactory bays and 89.3% (25 of 28 bays) of unsatisfactory bays. that are in the ranges of (0.4e0.5) and (0.5e0.6) are 11 and 8,
DG11 provided accurate predictions for 90% (45 of 50 bays) of the respectively. It is observed that the inaccurately predicted ap values
M. Royvaran et al. / Journal of Constructional Steel Research 164 (2020) 105764 7

satisfactory bays and 100% (28 of 28 bays) of the unsatisfactory


bays. It provided accurate predictions for 64.0% (32 of 50 bays) of
the bays in the database. Overall, this method inaccurately evalu-
ated the bays observed to be satisfactory and very accurately
evaluated the bays observed to be unsatisfactory, indicating that
the method is a bit on the conservative side.
Because the P354 Simplified Method provided conservative
predictions, it was further investigated in an attempt to increase its
accuracy. It is noted that this method uses a fairly high estimate of
the dynamic coefficient, 0.1, which explains why its predictions are
conservative. Calculations performed with dynamic coefficients
ranging from 0.04 to 0.12 (instead of 0.1) in Equation (4) indicate
that the dynamic coefficient 0.06 provides the highest accuracy; see
Table 6 and Fig. 3. A dynamic coefficient of 0.06 is approximately
equal to the average dynamic coefficient for the second through
fourth harmonics of the walking force from the research published
by Kerr [26], summarized in Willford et al. [28].

Fig. 1. Histogram for ap ranges per AISC DG11 predictions of Table 4. 4.3. SCI P354 Vibration Dose Values method

P354 VDV method results are summarized in Table 7, which


includes computed VDV if na is 1400 walking events in the 16 h of
are grouped within these two ranges. exposure duration, as it will be explained in the next paragraph. It
On another note, the estimated ap values with respect to natural should be noted that only the bays used as offices (33 bays consist
frequency (fn) and total weight (w) are displayed in Fig. 2(a) for the of 13 satisfactory bays and 20 unsatisfactory bays) were evaluated
damping ratio of 0.03. The top view of the same plot is shown in with this method.
Fig. 2(b). It is observed that the disagreements between observa- As mentioned in Section 2.3, to the writer's knowledge, there is
tions and predictions are more pronounced in the transition zone no guidance to determine na. Through an iterative process, using
indicated on Fig. 2(b). different values for na in Equation (9) and compare the computed
VDV to the upper end limit which is 0.4 m/sec.1.75, recommended by
4.2. SCI P354 simplified method P354, plus and minus 3%, it was determined that the best prediction
accuracy occurred if na is between 1000 and 1400 walking events
P354 Simplified Method results are summarized in Table 5, for 16 h of exposure (corresponding to less than 1.5 events per
which includes the following variables. minute) for offices. Fig. 4 shows variation of prediction accuracy
versus expected na. As shown, picking any number between 1000
f0 ¼ fundamental frequency, Hz and 1400 events for na provided accurate evaluations for 38.5% (5 of
W ¼ appropriate code-defined weighting factor for human 13 bays) of satisfactory bays and 100% (20 of 20 bays) of unsatis-
perception of vibrations based on the fundamental frequency, f0 factory bays. It provided accurate predictions for 75.8% (25 of 33
M ¼ modal mass, kg bays) of the bays identified as offices in this set of data.
r ¼ resonant build-up factor Regarding Equation (9), it should be emphasized that the cal-
R ¼ response factor culations in VDV method depend on calculated aw,rms from SCI P354
(simplified method). Therefore, any incorrect estimation of aw,rms
P354 provided accurate evaluations for 18.2% (4 of 22 bays) of can affect the results in VDV method significantly.

Fig. 2. ap plot as a function of fn and w.


8 M. Royvaran et al. / Journal of Constructional Steel Research 164 (2020) 105764

Table 5
Comparisons of Observations and Predictions e SCI P354 Simplified Method.

Bay Observed Evaluation Predicted Correct Prediction?

fo W M r R Evaluation

(Hz) (lb) (kg)

1 NG 5.430 1.000 24400 11074 0.987 15.7 NG Yes


2 NG 6.210 1.000 32600 14780 0.904 32.3 NG Yes
3 NG 6.210 1.000 24600 11176 0.904 42.7 NG Yes
4 NG 6.210 1.000 32600 14780 0.904 32.3 NG Yes
5 NG 6.520 1.000 28300 12821 0.915 37.6 NG Yes
6 NG 5.430 1.000 24500 11104 0.987 15.6 NG Yes
7 NG 3.880 0.850 40000 18137 0.988 9.8 NG Yes
8 OK 4.270 0.830 27600 12535 0.985 13.8 NG No
9 NG 7.140 1.000 18900 8577 0.919 56.5 NG Yes
10 NG 5.170 1.000 24100 10925 0.997 16.0 NG Yes
11 NG 5.270 1.000 26500 12040 0.997 14.6 NG Yes
12 NG 5.990 1.000 19500 8840 0.966 38.4 NG Yes
13 NG 5.970 1.000 19300 8745 0.966 38.8 NG Yes
14 NG 5.970 1.000 19300 8745 0.966 38.8 NG Yes
15 NG 5.860 1.000 16600 7526 0.964 45.0 NG Yes
16 NG 5.860 1.000 16600 7526 0.964 45.0 NG Yes
17 NG 5.860 1.000 19100 8654 0.964 39.2 NG Yes
18 OK 4.040 0.850 25300 11498 1.000 7.8 B Yes
19 OK 5.630 1.000 25800 11719 1.000 9.0 NG No
20 OK 5.200 1.000 30800 13965 1.000 7.6 OK Yes
21 OK 5.580 1.000 31800 14441 1.000 7.3 OK Yes
22 OK 7.190 1.000 27200 12358 0.996 21.2 NG No
23 OK 6.050 1.000 24900 11293 1.000 9.3 NG No
24 OK 6.000 1.000 25200 11415 1.000 11.6 NG No
25 OK 4.350 8.500 76600 34752 0.970 41.7 NG No
26 OK 4.260 0.840 79900 36232 0.982 3.4 OK Yes
27 OK 3.950 0.850 26300 11943 0.989 12.4 NG No
28 OK 4.660 0.860 23900 10842 0.995 13.9 NG No
29 OK 3.950 0.860 35700 16215 0.997 9.3 NG No
30 OK 4.860 0.950 23500 10667 0.999 15.6 NG No
31 NG 4.310 0.830 26100 11849 0.983 14.5 NG Yes
32 OK 6.720 1.000 18400 8346 0.998 25.2 NG No
33 OK 5.060 1.000 20700 9381 0.992 22.3 NG No
34 OK 5.420 1.000 27700 12555 0.994 16.7 NG No
35 OK 4.930 0.950 26300 11909 0.990 16.7 NG No
36 NG 5.350 1.000 27200 12321 0.868 37.1 NG Yes
37 NG 4.930 0.950 21500 9763 0.845 43.4 NG Yes
38 OK 4.090 0.850 40400 18331 0.981 9.6 NG No
39 OK 5.320 1.000 30800 13992 0.998 12.5 NG No
40 NG 5.320 1.000 28000 12720 0.998 13.8 NG Yes
41 NG 5.480 1.000 24900 11315 0.997 15.5 NG Yes
42 NG 5.290 1.000 24500 11125 0.998 15.8 NG Yes
43 NG 5.370 1.000 24700 11186 0.998 15.7 NG Yes
44 NG 6.380 1.000 33100 14995 0.973 22.8 NG Yes
45 NG 5.920 1.000 27000 12259 0.996 17.1 NG Yes
46 NG 4.800 0.950 24200 10994 0.989 18.0 NG Yes
47 NG 5.280 1.000 33300 15106 0.993 13.9 NG Yes
48 NG 5.140 1.000 25900 11739 0.997 14.9 NG Yes
49 OK 3.860 0.850 16700 7572 0.987 19.5 NG No
50 OK 3.590 0.850 16900 7656 0.966 18.9 NG No

4.4. HIVOSS method

Table 6 HIVOSS results are summarized in Tables 8e11, which include


Variation of Prediction Accuracy vs Dynamic Coefficients for SCI P354 Simplified
the following variables.
Method.

Dynamic Coefficient Prediction Accuracy (%) fn ¼ fundamental frequency, Hz


All Bays Satisfactory Bays Unsatisfactory Bays Mmod ¼ modal mass, kg
0.12 58 4.6 100 b ¼ damping ratio
0.11 60 9.1 100
0.10 64 18.2 100 Combination 1: fn from the self-weight approach from HIVOSS
0.09 66 22.7 100 Appendix A.4 and Mmod from HIVOSS Appendix A.6. This case
0.08 72 36.4 100
0.07 72 40.9 96.4
correctly predicted that 63.6% of the satisfactory bays would be
0.06 76 50 96.4 satisfactory. It correctly predicted that only 25% of unsatisfactory
0.05 70 63.6 75 bays would be unsatisfactory. It provided accurate predictions for
0.04 64 81.8 50 42% of the bays in the database.
M. Royvaran et al. / Journal of Constructional Steel Research 164 (2020) 105764 9

Fig. 3. Variation of Prediction Accuracy vs Dynamic Coefficients for SCI P354 Simplified Method.

Combination 2: fn from the Dunkerley approach from HIVOSS It could not predict any of unsatisfactory bays. It provided accurate
Appendix A.5 and Mmod from HIVOSS Appendix A.6. This case predictions for 42% of the bays in the database.
correctly predicted that 59.1% of the satisfactory bays would be Combination 4: fn from the Dunkerley approach from HIVOSS
satisfactory. It correctly predicted that only 3.6% of unsatisfactory Appendix A.5 and Mmod from DG11. This case correctly predicted
bays would be unsatisfactory. It provided accurate predictions for that 81.8% of the satisfactory bays would be satisfactory. It could not
28% of the bays in the database. predict any of unsatisfactory bays. It provided accurate predictions
Combination 3: fn from the self-weight approach from HIVOSS for 36% of the bays in the database.
Appendix A.4 and Mmod from DG11, Chapter 4. This case correctly In Table 12, a summary of HIVOSS results are presented. When
predicted that 95.5% of the satisfactory bays would be satisfactory. considering all bays, the prediction accuracies are 42% for

Table 7
Comparisons of observations and predictions e SCI P354 VDV method.

Bay Observed Evaluation Prediction Correct Prediction?

VDV (If na ¼ 1400)m/s1.75 Evaluation

1 NG 0.47 NG Yes
2 NG 1.05 NG Yes
3 NG 1.39 NG Yes
4 NG 1.05 NG Yes
5 NG 1.23 NG Yes
6 NG 0.47 NG Yes
9 NG 1.81 NG Yes
10 NG 0.52 NG Yes
11 NG 0.47 NG Yes
18 OK 0.25 OK Yes
27 OK 0.40 OK Yes
28 OK 0.45 NG No
29 OK 0.33 OK Yes
30 OK 0.55 NG No
31 NG 0.47 NG Yes
32 OK 0.82 NG No
33 OK 0.73 NG No
34 OK 0.54 NG No
35 OK 0.54 NG No
36 NG 1.21 NG Yes
37 NG 1.41 NG Yes
38 OK 0.31 OK Yes
39 OK 0.41 OK Yes
40 NG 0.45 NG Yes
41 NG 0.50 NG Yes
42 NG 0.51 NG Yes
43 NG 0.51 NG Yes
45 NG 0.56 NG Yes
46 NG 0.59 NG Yes
47 NG 0.45 NG Yes
48 NG 0.49 NG Yes
49 OK 0.63 NG No
50 OK 0.59 NG No
10 M. Royvaran et al. / Journal of Constructional Steel Research 164 (2020) 105764

Fig. 4. Variation of Prediction Accuracy vs. Expected na.

Table 8
Comparisons of observations and predictions. HIVOSS, combination 1.

Bay Observed b HIVOSS (%) fn from the self-weight approach Mmod from Class Classification Assignment of Evaluation Correct
Evaluation A.4 (Hz) HIVOSS of Floor Class Prediction?
A.6
(lb) (kg)

1 FC 3.0 5.45 21200 9616 D Office Recommended OK No


2 CP 1.0 6.47 25300 11476 D Office Recommended OK No
3 CP 1.0 6.47 25300 11476 D Office Recommended OK No
4 CP 1.0 6.47 25300 11476 D Office Recommended OK No
5 CP 1.0 6.87 25300 11476 D Office Recommended OK No
6 SC 3.0 5.45 21500 9752 D Office Recommended OK No
7 MC 3.0 3.77 52900 23995 D Education Critical FC No
8 NC 3.0 4.89 40500 18370 D Education Critical FC No
9 MC 1.0 6.54 18600 8437 D Office Recommended OK No
10 SC 4.0 4.98 34200 15513 D Office Recommended OK No
11 SC 4.0 4.72 37400 16964 D Office Recommended OK No
12 SC 1.0 6.22 18500 8391 E Health Not Recommended NG Yes
13 SC 1.0 6.22 18500 8391 E Health Not Recommended NG Yes
14 SC 1.0 6.22 18500 8391 E Health Not Recommended NG Yes
15 SC 1.0 6.22 18500 8391 E Health Not Recommended NG Yes
16 SC 1.0 6.22 18500 8391 E Health Not Recommended NG Yes
17 SC 1.0 6.22 18500 8391 E Health Not Recommended NG Yes
18 NC 4.0 4.3 32900 14923 D Office Recommended OK Yes
19 NC 4.0 5.91 30600 13880 D Health Critical FC No
20 NC 4.0 5.45 38400 17418 D Health Critical FC No
21 NC 4.0 5.82 30800 13971 D Health Critical FC No
22 NC 2.0 7.46 25800 11703 C Health Recommended OK Yes
23 NC 4.0 6.66 28000 12701 D Health Critical FC No
24 NC 4.0 6.82 29600 13426 D Health Critical FC No
25 NC 3.0 4.31 62800 28486 D Health Critical FC No
26 NC 3.0 4.22 65400 29665 D Health Critical FC No
27 NC 3.0 4.28 43700 19822 D Office Recommended OK Yes
28 NC 3.0 4.56 41600 18869 D Office Recommended OK Yes
29 NC 3.0 4.24 58400 26490 D Office Recommended OK Yes
30 NC 3.0 4.76 62800 28486 C Office Recommended OK Yes
31 SC 3.0 4.21 42600 19323 D Office Recommended OK No
32 NC 3.0 6.68 25800 11703 D Office Recommended OK Yes
33 NC 3.0 4.93 35300 16012 D Office Recommended OK Yes
34 NC 2.0 4.84 38800 17599 D Office Recommended OK Yes
35 NC 2.0 4.84 41200 18688 D Office Recommended OK Yes
36 SC 1.0 4.84 39400 17872 D Office Recommended OK No
37 SC 1.0 4.84 39700 18008 D Office Recommended OK No
38 NC 3.0 3.75 48400 21954 D Office Recommended OK Yes
39 NC 3.0 5.75 34100 15467 D Office Recommended OK Yes
40 MC 3.0 5.75 34100 15467 D Office Recommended OK No
41 SC 2.0 5.65 25800 11703 D Office Recommended OK No
42 SC 3.0 5.51 28300 12837 D Office Recommended OK No
43 SC 2.0 5.6 27500 12474 D Office Recommended OK No
44 FC 1.0 6.81 28100 12746 D Health Critical FC Yes
45 SC 2.0 6.17 26000 11793 D Office Recommended OK No
46 SC 2.0 4.7 38000 17236 D Office Recommended OK No
47 MC 2.0 4.7 35800 16239 D Office Recommended OK No
48 SC 4.0 5.02 35000 15876 D Office Recommended OK No
49 NC 3.0 4.1 48400 21954 D Office Recommended OK Yes
50 NC 3 3.74 40600 18416 D Office Recommended OK Yes
M. Royvaran et al. / Journal of Constructional Steel Research 164 (2020) 105764 11

Table 9
Comparisons of observations and predictions. HIVOSS, combination 2.

Bay Observed b HIVOSS fn from the Mmod from Class Classification Assignment Evaluation Correct
Evaluation (%) Dunkerley HIVOSS of Floor of Class Prediction?
approach A.6
A.5 (Hz) (lb) (kg)

1 FC 3.0 4.35 21200 9616 D Office Recommended OK No


2 CP 1.0 4.95 25300 11476 D Office Recommended OK No
3 CP 1.0 4.95 25300 11476 D Office Recommended OK No
4 CP 1.0 4.95 25300 11476 D Office Recommended OK No
5 CP 1.0 5.12 25300 11476 D Office Recommended OK No
6 SC 3.0 4.45 21500 9752 D Office Recommended OK No
7 MC 3.0 2.94 52900 23995 D Education Critical FC No
8 NC 3.0 3.50 40500 18370 D Education Critical FC No
9 MC 1.0 5.32 18600 8437 D Office Recommended OK No
10 SC 4.0 3.69 34200 15513 D Office Recommended OK No
11 SC 4.0 3.61 37400 16964 D Office Recommended OK No
12 SC 1.0 4.76 18500 8391 D Health Critical FC No
13 SC 1.0 4.76 18500 8391 D Health Critical FC No
14 SC 1.0 4.76 18500 8391 D Health Critical FC No
15 SC 1.0 4.76 18500 8391 D Health Critical FC No
16 SC 1.0 4.76 18500 8391 D Health Critical FC No
17 SC 1.0 4.76 18500 8391 D Health Critical FC No
18 NC 4.0 3.57 32900 14923 D Office Recommended OK Yes
19 NC 4.0 4.84 30600 13880 D Health Critical FC No
20 NC 4.0 4.11 38400 17418 D Health Critical FC No
21 NC 4.0 4.45 30800 13971 D Health Critical FC No
22 NC 2.0 5.45 25800 11703 D Health Critical FC No
23 NC 4.0 5.06 28000 12701 D Health Critical FC No
24 NC 4.0 4.93 29600 13426 D Health Critical FC No
25 NC 3.0 3.83 62800 28486 D Health Critical FC No
26 NC 3.0 3.75 65400 29665 D Health Critical FC No
27 NC 3.0 3.54 43700 19822 D Office Recommended OK Yes
28 NC 3.0 4.56 41600 18869 D Office Recommended OK Yes
29 NC 3.0 3.17 58400 26490 C Office Recommended OK Yes
30 NC 3.0 4.76 62800 28486 C Office Recommended OK Yes
31 SC 3.0 4.21 42600 19323 D Office Recommended OK No
32 NC 3.0 6.68 25800 11703 D Office Recommended OK Yes
33 NC 3.0 4.93 35300 16012 D Office Recommended OK Yes
34 NC 2.0 4.12 38800 17599 D Office Recommended OK Yes
35 NC 2.0 4.43 41200 18688 D Office Recommended OK Yes
36 SC 1.0 4.27 39400 17872 E Office Critical FC No
37 SC 1.0 4.19 39700 18008 E Office Critical FC No
38 NC 3.0 3.28 48400 21954 D Office Recommended OK Yes
39 NC 3.0 4.19 34100 15467 D Office Recommended OK Yes
40 MC 3.0 4.19 34100 15467 D Office Recommended OK No
41 SC 2.0 4.44 25800 11703 E Office Critical FC No
42 SC 3.0 4.10 28300 12837 D Office Recommended OK No
43 SC 2.0 4.16 27500 12474 E Office Critical FC No
44 FC 1.0 4.85 28100 12746 D Health Critical FC Yes
45 SC 2.0 4.43 26000 11793 E Office Critical FC No
46 SC 2.0 4.70 38000 17236 D Office Recommended OK No
47 MC 2.0 3.84 35800 16239 E Office Critical FC No
48 SC 4.0 3.94 35000 15876 D Office Recommended OK No
49 NC 3.0 3.47 48400 21954 D Office Recommended OK Yes
50 NC 3 3.44 40600 18416 D Office Recommended OK Yes

Combination 1; 28% for Combination 2; 42% for Combination 3 and observed acceptability of each bay in a large database of steel-
36% for Combination 4. framed floors subject to walking excitation. Comparisons were
made on the basis of the final prediction accuracy e satisfactory or
unsatisfactory performance, corresponding to occupant com-
5. Summary and conclusions plaints. In addition, further investigations were performed to
improve the SCI P354 Simplified Method.
In this study, the evaluation accuracies of four well established The floor bay database contains 22 satisfactory bays, based on
floor vibration evaluation methods e DG11 Chapter 4, SCI P354 observations, that is, bays without complaints, and 28 unsatisfac-
Simplified Method, SCI VDV Method, and HIVOSS hand calculation tory bays, for a total of 50 bays. Table 13 summarizes the results of
method e were investigated by comparing predictions with
12 M. Royvaran et al. / Journal of Constructional Steel Research 164 (2020) 105764

Table 10
Comparisons of observations and predictions. HIVOSS, combination 3.

Bay Observed b HIVOSS (%) fn from the Mmod from DG11 Class Classification Assignment of Class Evaluation Correct
Evaluation self-weight Ch. 4 of Floor Prediction?
approach (lb) (kg)
A.4 (Hz)

1 FC 3.0 5.45 32700 14832 D Office Recommended OK No


2 CP 1.0 6.47 49400 22407 D Office Recommended OK No
3 CP 1.0 6.47 48900 22181 D Office Recommended OK No
4 CP 1.0 6.47 49400 22407 D Office Recommended OK No
5 CP 1.0 6.87 40100 18189 D Office Recommended OK No
6 SC 3.0 5.45 32900 14923 D Office Recommended OK No
7 MC 3.0 3.77 68500 31071 D Education Critical FC No
8 NC 3.0 4.89 75800 34382 C Education Recommended OK Yes
9 MC 1.0 6.54 30300 13744 D Office Recommended OK No
10 SC 4.0 4.98 37900 17191 D Office Recommended OK No
11 SC 4.0 4.72 37600 17055 D Office Recommended OK No
12 SC 1.0 6.22 33900 15377 D Health Critical FC No
13 SC 1.0 6.22 31500 14288 D Health Critical FC No
14 SC 1.0 6.22 31500 14288 D Health Critical FC No
15 SC 1.0 6.22 35500 16103 D Health Critical FC No
16 SC 1.0 6.22 35500 16103 D Health Critical FC No
17 SC 1.0 6.22 35500 16103 D Health Critical FC No
18 NC 4.0 4.3 65600 29756 D Office Recommended OK Yes
19 NC 4.0 5.91 59500 26989 C Health Recommended OK Yes
20 NC 4.0 5.45 68300 30980 C Health Recommended OK Yes
21 NC 4.0 5.82 58400 26490 C Health Recommended OK Yes
22 NC 2.0 7.46 46300 21001 C Health Recommended OK Yes
23 NC 4.0 6.66 40600 18416 D Health Critical FC No
24 NC 4.0 6.82 44900 20366 C Health Recommended OK Yes
25 NC 3.0 4.31 127000 57606 C Health Recommended OK Yes
26 NC 3.0 4.22 132000 59874 C Health Recommended OK Yes
27 NC 3.0 4.28 86000 39009 C Office Recommended OK Yes
28 NC 3.0 4.56 51200 23224 D Office Recommended OK Yes
29 NC 3.0 4.24 99300 45042 C Office Recommended OK Yes
30 NC 3.0 4.76 50400 22861 D Office Recommended OK Yes
31 SC 3.0 4.21 53700 24358 D Office Recommended OK No
32 NC 3.0 6.68 27100 12292 D Office Recommended OK Yes
33 NC 3.0 4.93 42800 19414 D Office Recommended OK Yes
34 NC 2.0 4.84 66200 30028 C Office Recommended OK Yes
35 NC 2.0 4.84 46600 21137 D Office Recommended OK Yes
36 SC 1.0 4.84 66000 29937 D Office Recommended OK No
37 SC 1.0 4.84 48500 21999 D Office Recommended OK No
38 NC 3.0 3.75 88800 40279 D Office Recommended OK Yes
39 NC 3.0 5.75 51900 23541 C Office Recommended OK Yes
40 MC 3.0 5.75 39300 17826 D Office Recommended OK No
41 SC 2.0 5.65 50700 22997 D Office Recommended OK No
42 SC 3.0 5.51 53900 24449 C Office Recommended OK No
43 SC 2.0 5.6 52300 23723 D Office Recommended OK No
44 FC 1.0 6.81 56900 25809 C Health Recommended OK No
45 SC 2.0 6.17 47300 21455 D Office Recommended OK No
46 SC 2.0 4.7 44900 20366 D Office Recommended OK No
47 MC 2.0 4.7 65200 29574 D Office Recommended OK No
48 SC 4.0 5.02 62300 28259 C Office Recommended OK No
49 NC 3.0 4.1 79700 36151 D Office Recommended OK Yes
50 NC 3 3.74 85200 38646 D Office Recommended OK Yes

the study. It should be noted that only floors used as offices were selected to optimize the prediction accuracy, provided correct
considered in VDV method containing 13 satisfactory bays and 20 predictions for 38.5% of the satisfactory bays and 100% of the un-
unsatisfactory bays, for a total of 33 bays. satisfactory bay. This indicates that it is a slightly conservative
DG11 Chapter 4 accurately predicted the acceptability of 90.9% method. It should be noted that this optimization was performed in
of the satisfactory bays and 89.3% of the unsatisfactory bays in the a way to maximize the prediction accuracy compared to observa-
database, so it is an accurate method. tion in this set of data.
The SCI P354 Simplified Method provided accurate predictions The HIVOSS method provided poor prediction accuracy for all
for 18.2% of the satisfactory bays and accurate predictions for 100% unsatisfactory bays for all combinations. Combination 1 and 2 with
of the unsatisfactory bays, indicating that the method is slightly on modal mass computed by HIVOSS Appendix A.6 could only predict
the conservative side. However, changing the dynamic coefficient 63.4% and 59.1% of satisfactory bays, respectively. However, by use
from 0.1 to 0.06 significantly improved the prediction accuracy for of modal mass computed by DG11 Chapter 4 in HIVOSS method,
satisfactory bays from 18.2% to 50.0% which leads to an increase Combination 3 and 4, this method provided 95.5% and 81.1% correct
from 64.0% to 76.0% in the overall prediction accuracy. prediction for satisfactory bays. Overall, this method is found to be
The SCI P354 VDV approach, with the number of walking events on the unconservative side.
M. Royvaran et al. / Journal of Constructional Steel Research 164 (2020) 105764 13

Table 11
Comparisons of observations and predictions. HIVOSS, combination 4.

Bay Observed b HIVOSS (%) fn from the Mmod from DG11 Class Classification Assignment of Evaluation Correct
Evaluation Dunkerley Ch. 4 of Floor Class Prediction?
approach (lb) (kg)
A.5 (Hz)

1 FC 3.0 4.35 32700 14832 D Office Recommended OK No


2 CP 1.0 4.95 49400 22407 D Office Recommended OK No
3 CP 1.0 4.95 48900 22181 D Office Recommended OK No
4 CP 1.0 4.95 49400 22407 D Office Recommended OK No
5 CP 1.0 5.12 40100 18189 D Office Recommended OK No
6 SC 3.0 4.45 32900 14923 D Office Recommended OK No
7 MC 3.0 2.94 68500 31071 C Education Recommended OK No
8 NC 3.0 3.50 75800 34382 D Education Critical FC No
9 MC 1.0 5.32 30300 13744 D Office Recommended OK No
10 SC 4.0 3.69 37900 17191 D Office Recommended OK No
11 SC 4.0 3.61 37600 17055 D Office Recommended OK No
12 SC 1.0 4.76 33900 15377 D Health Critical FC No
13 SC 1.0 4.76 31500 14288 D Health Critical FC No
14 SC 1.0 4.76 31500 14288 D Health Critical FC No
15 SC 1.0 4.76 35500 16103 D Health Critical FC No
16 SC 1.0 4.76 35500 16103 D Health Critical FC No
17 SC 1.0 4.76 35500 16103 D Health Critical FC No
18 NC 4.0 3.57 65600 29756 D Office Recommended OK Yes
19 NC 4.0 4.84 59500 26989 C Health Recommended OK Yes
20 NC 4.0 4.11 68300 30980 D Health Critical FC No
21 NC 4.0 4.45 58400 26490 D Health Critical FC No
22 NC 2.0 5.45 46300 21001 D Health Critical FC No
23 NC 4.0 5.06 40600 18416 C Health Recommended OK Yes
24 NC 4.0 4.93 44900 20366 C Health Recommended OK Yes
25 NC 3.0 3.83 127000 57606 C Health Recommended OK Yes
26 NC 3.0 3.75 132000 59874 C Health Recommended OK Yes
27 NC 3.0 3.54 86000 39009 C Office Recommended OK Yes
28 NC 3.0 4.56 51200 23224 D Office Recommended OK Yes
29 NC 3.0 3.17 99300 45042 C Office Recommended OK Yes
30 NC 3.0 4.76 50400 22861 D Office Recommended OK Yes
31 SC 3.0 4.21 53700 24358 D Office Recommended OK No
32 NC 3.0 6.68 27100 12292 D Office Recommended OK Yes
33 NC 3.0 4.93 42800 19414 D Office Recommended OK Yes
34 NC 2.0 4.12 66200 30028 D Office Recommended OK Yes
35 NC 2.0 4.43 46600 21137 D Office Recommended OK Yes
36 SC 1.0 4.27 66000 29937 D Office Recommended OK No
37 SC 1.0 4.19 48500 21999 D Office Recommended OK No
38 NC 3.0 3.28 88800 40279 C Office Recommended OK Yes
39 NC 3.0 4.19 51900 23541 D Office Recommended OK Yes
40 MC 3.0 4.19 39300 17826 D Office Recommended OK No
41 SC 2.0 4.44 50700 22997 D Office Recommended OK No
42 SC 3.0 4.10 53900 24449 D Office Recommended OK No
43 SC 2.0 4.16 52300 23723 D Office Recommended OK No
44 FC 1.0 4.85 56900 25809 C Health Recommended OK No
45 SC 2.0 4.43 47300 21455 D Office Recommended OK No
46 SC 2.0 4.70 44900 20366 D Office Recommended OK No
47 MC 2.0 3.84 65200 29574 D Office Recommended OK No
48 SC 4.0 3.94 62300 28259 D Office Recommended OK No
49 NC 3.0 3.47 79700 36151 D Office Recommended OK Yes
50 NC 3 3.44 85200 38646 C Office Recommended OK Yes

Table 12
HIVOSS results summary.

Prediction Accuracy

Combination 1 Combination 2 Combination 3 Combination 4

Observed Satisfactory Bays 63.6% 59.1% 95.5% 81.1%


Observed Unsatisfactory Bays 25.0% 3.6% 0.0% 0.0%
All Bays 42.0% 28.0% 42.0% 36.0%
14 M. Royvaran et al. / Journal of Constructional Steel Research 164 (2020) 105764

Table 13
Summary of accuracies of final evaluation predictions.

Evaluation Method Observed Satisfactory Observed Unsatisfactory All Bays


Bays Bays

Guideline AISC DG11 Chapter 4 90.9% 89.3% 90.0%


SCI P354 Simplified Method 18.2% 100% 64.0%
SCI P354 VDV Method (Only 38.5% 100% 75.8%
Offices)
HIVOSS Combination 1 63.6% 25.0% 42.0%
Combination 2 59.1% 3.6% 28.0%
Combination 3 95.5% 0.0% 42.0%
Combination 4 81.1% 0.0% 36.0%
Investigation to Improve Prediction Accuracy SCI P354 Simplified Method 50.0% 96.4% 76.0%

Acknowledgements laboratory footbridge, J. Archit. Eng. 22 (2016), https://doi.org/10.1061/(ASCE)


AE.1943-5568.0000211.
[13] C.J. Middleton, J.M.W. Brownjohn, Response of high frequency floors: a
The authors would like to express sincere gratitude to Dr. literature review, Eng. Struct. 32 (2010) 337e352, https://doi.org/10.1016/
Thomas M. Murray for providing most of the bays in the floor j.engstruct.2009.11.003.
database. [14] A.V.A. Mello, J.G.S. da Silva, P.C.G.d.S. Vellasco, S.A.L. de Andrade, L.R.O. de
Lima, Dynamic analysis of composite systems made of concrete slabs and steel
The financial support for this research was provided by Qatar beams, J. Constr. Steel Res. (2008), https://doi.org/10.1016/j.jcsr.2007.09.011.
National Research Fund (QNRF; a member of the Qatar Foundation) [15] A.C. Stephenson, A.W. Humphreys, Quantification of steel floor vibration and
via the National Priorities Research Program (NPRP), Project its relationship to human comfort, J. Constr. Steel Res. (1998), https://doi.org/
10.1016/S0143-974X(98)00137-0.
Number NPRP8-836-2-353. The statements made herein are solely [16] M.J. Hudson, P. Reynolds, Implementation considerations for active vibration
the responsibility of the authors. control in the design of floor structures, Eng. Struct. 44 (2012) 334e358,
https://doi.org/10.1016/j.engstruct.2012.05.034.
[17] A. Younis, O. Avci, M. Hussein, B. Davis, P. Reynolds, Dynamic forces induced
References by a single pedestrian: a literature review, Appl. Mech. Rev. 69 (2017), https://
doi.org/10.1115/1.4036327.

[1] S. Zivanovi 
c, A. Pavic, P. Reynolds, Vibration serviceability of footbridges under [18] O. Avci, Nonlinear damping in floor vibrations serviceability: verification on a
human-induced excitation: a literature review, J. Sound Vib. (2005), https:// laboratory structure, in: Conf. Proc. Soc. Exp. Mech. Ser., 2017, https://doi.org/
doi.org/10.1016/j.jsv.2004.01.019. 10.1007/978-3-319-54777-0_18.
[2] I.M. Díaz, P. Reynolds, Acceleration feedback control of human-induced floor [19]  Cunha, C. Moutinho, Vandal loads and induced vibrations on a
E. Caetano, A.
vibrations, Eng. Struct. 32 (2010) 163e173, https://doi.org/10.1016/ footbridge, J. Bridge Eng. (2010), https://doi.org/10.1061/(asce)be.1943-
j.engstruct.2009.09.003. 5592.0000154.
[3] L. Xu, F.M. Tangorra, Experimental investigation of lightweight residential [20] T.M. Murray, D.E. Allen, E.E. Ungar, D.B. Davis, Vibrations of Steel-Framed
floors supported by cold-formed steel C-shape joists, J. Constr. Steel Res. Structural Systems Due to Human Activity, second ed., American Institute of
(2007), https://doi.org/10.1016/j.jcsr.2006.05.010. Steel Construction, 2016.
[4] Z. Muhammad, P. Reynolds, O. Avci, M. Hussein, Review of pedestrian load [21] A.L. Smith, S.J. Hicks, P.J. Devine, Design of floors for vibration - a new
models for vibration serviceability assessment of floor structures, Vibrations approach SCI P354, Revised, in: Steel Constr. Institute, P354SCI, Ascot, Berk-
(2018), https://doi.org/10.3390/vibration2010001. shire, U.K, 2009, pp. 1e128, https://doi.org/10.13140/RG.2.2.29342.95048.
[5] A.R. Barrett, O. Avci, M. Setareh, T.M. Murray, Observations from vibration [22] RFCS, Human Induced Vibrations of Steel Structures Vibration Design of Floors
testing of in-situ structures, in: Proc. Struct. Congr. Expo., 2006, https:// Guideline, 2009. Brussels, Belgium.
doi.org/10.1061/40889(201)65. [23] J.H. Rainer, G. Pernica, D.E. Allen, Dynamic loading and response of foot-
[6] B. Davis, Finite Element Modeling for Prediction of Low Frequency Floor Vi- bridges, Can. J. Civ. Eng. 15 (1988) 66e71, https://doi.org/10.1139/l88-007.
brations Due to Walking, Virginia Polytechnic Institute and State University, [24] D.E. Allen, T.M. Murray, Design criterion for vibrations due to walking, Eng. J.
2008. 30 (1993) 117e129.
[7] Z.O. Muhammad, P. Reynolds, E.J. Hudson, Evaluation of contemporary [25] S. Pabian, A. Thomas, B. Davis, T.M. Murray, Investigation of Floor Vibration
guidelines for floor vibration serviceability assessment, in: Conf. Proc. Soc. Evaluation Criteria Using an Extensive Database of Floors, 2013, https://
Exp. Mech. Ser., 2017, https://doi.org/10.1007/978-3-319-54777-0_42. doi.org/10.1061/9780784412848.216.
[8] A. Bhargava, J. Isenberg, P.H. Feenstra, Y. Al-Smadi, O. Avci, Vibrations [26] S.C. Kerr, Human Induced Loading on Staircases, 1998, https://doi.org/
assessment of a hospital floor for a magnetic resonance imaging unit (MRI) 10.1016/S0141-0296(00)00020-1.
replacement, in: Struct. Congr. 2013 Bridg. Your Passion with Your Prof. - Proc. [27] S.D. Pabian, Investigation of Floor Vibration Evaluation Criteria Using an
2013 Struct. Congr., 2013. Extensive Database of Floors, University of Kentucky, 2011.
[9] O. Avci, Modal parameter variations due to joist bottom chord extension in- [28] M. Wilford, P. Young, C. Field, Predicting footfall-induced vibration: Part I, in:
stallations on laboratory footbridges, J. Perform. Constr. Facil. 29 (2015), Proc. Inst. Civ. Eng. Build., 2007.
https://doi.org/10.1061/(ASCE)CF.1943-5509.0000635. [29] Abdeljaber, et al., A novel video-vibration monitoring system for walking

[10] S. Zivanovi 
c, A. Pavi
c, P. Reynolds, Probability-based prediction of multi-mode pattern identification on floors, Adv. Eng. Software (2019), https://doi.org/
vibration response to walking excitation, Eng. Struct. 29 (2007) 942e954, 10.1016/j.advengsoft.2019.102710. In press.
https://doi.org/10.1016/j.engstruct.2006.07.004. [30] N. Do, et al., Novel Framework for Vibration Serviceability Assessment of
[11] O. Avci, Effects of Bottom Chord Extensions on the Static and Dynamic Per- Stadium Grandstands Considering Durations of Vibrations, J. Struct. Eng.
formance of Steel Joist Supported Floors, Virginia Polytechnic Institute and 144 (2) (2018), https://doi.org/10.1061/(ASCE)ST.1943-541X.0001941. In
State University, 2005. this issue.
[12] O. Avci, Amplitude-dependent damping in vibration serviceability: case of a

You might also like