Professional Documents
Culture Documents
Technical Background
AISC Design Guide 11
Part 1 / 2
2
1. Introduction
A very realistic limit state
Steel-Framed Structures: floor vibration is
about as likely to control as strength and
deflection limit states.
These are especially vulnerable:
Areas with rhythmic activities.
Areas with sensitive equipment.
Monumental stairs.
3
1. Introduction
Why focus on human induced floor vibration?
Vibration Due to Equipment
Can usually be isolated at the source.
High frequency more difficult to feel.
May still be an issue in some cases.
Wind induced lateral vibration. A different
subject.
Earthquake design has to do with safety. A
different subject.
4
1. Introduction
Floor Vibration Evaluation in a Nutshell
Compare predicted response to the tolerance
limit OK or No Good
Predict Response
Forcing function.
Dynamic properties of the structure.
Response prediction method.
Define Tolerance Limit
Human annoyance.
Equipment problems.
5
1. Introduction
This Presentation
1. Introduction
2. Technical background
3. DG11 Chapter 4
1. Derivation and Implementation
2. Example
3. Accuracy Investigation
Part 2
4. DG11 Chapter 5
5. Remedial Options Case Studies
6
2. Technical Background:
Dynamic Forces
When humans walk, run, bounce, sway, or
jump, inertial forces cause dynamic loads.
Example:
2
Ground Reaction / Weight
Knee bends,
Heel Strike weight shifts
1.5 forward
Push off
1
0.5
0
0 0.2 0.4 0.6 0.8
Time (sec.)
0.4
Left Footstep Summation Fourier
0.35
Series
Ground Reaction (kip)
0.3
or
0.25
0.2
0.15
Need Simple
Mathematical
Individual
0.1
Representations Footstep
0.05
Right Footstep Pulse or
Impulse
0
0 0.5 1 1.5 2
Time (sec.)
8
2. Technical Background:
Dynamic Forces
Fourier Series vs Individual Footstep
Fourier series useful for computing resonant
response.
Pulse or impulse useful for computing
response to one footstep.
9
2. Technical Background:
Dynamic Forces
Review of Fourier Series
Any periodic function is a superposition of the
average plus an infinite number of sinusoids.
Each sinusoid (called a harmonic) has a
different amplitude (Amph), frequency (fh), and
phase lag (h).
F ( t ) Average Value Amph sin(2 fh t h )
h1
N
Average Value Amph sin(2 fh t h )
h1
10
2. Technical Background:
Dynamic Forces
Example Fourier Series
300 100
250
Footstep Force (lbf)
100 -50
50 -100
0 0.2 0.4 0.6 0.8 1 0 0.2 0.4 0.6 0.8 1
Time (sec.) Time (sec.)
F ( t ) 185 lbf
Y: 185
Equivalent Representations 12
2. Technical Background:
Dynamic Forces
Specialize Fourier Series for Human Induced
Forces Weight of Harmonic Phase
Walker (lbf) Number Lag
4
F ( t ) 0 Q h sin(2 h fStep t h )
h1
SCI P354.
14
2. Technical Background:
Dynamic Forces
Sources of Fourier Series Parameters
Stair Ascents and Descents
Davis and Murray (2009) “Slender Monumental
Stair Vibration Serviceability,” Journal of
Architectural Engineering, ASCE.
Miscellaneous
ISO 10137 (2007).
Vibration Problems in Structures by Bachmann
(1995).
15
2. Technical Background:
Dynamic Forces
Example Fourier Series Parameters
AISC DG11 Walking
Bodyweight Q = 157 lbf
= 0.5, 0.2, 0.1, 0.05
Phase lags not given
SCI P354 Walking
Q = 168 lbf
= 0.46, 0.1, 0.08, 0.07 (+/-)
= 0, -/2, , /2
16
2. Technical Background:
Dynamic Forces
Step Frequency
Depends on activity. Use walking for example.
fStep usually between 1.6 Hz and 2.2 Hz
Average is 1.9-2.0 Hz
19
2. Technical Background:
Dynamic Properties of Floors
Natural Modes of Vibration
If a structure is displaced, released, and
allowed to move freely, the motion will:
Be at a specific frequency natural frequency
Be in a specific structural shape mode shape
Decay at a specific rate damping
Natural modes are properties of the structure.
Natural modes do not depend on the dynamic
load.
20
2. Technical Background:
Dynamic Properties of Floors
Natural Modes
Structures have many natural modes.
The lowest frequency mode is referred to as
the fundamental mode.
Some modes are more responsive than
others. Quantify with the Frequency
Response Function (FRF).
21
2. Technical Background:
Dynamic Properties of Floors
FRF Definition: ratio of steady state response to force input.
1 lbf sinusoidal load
Dominant Frequency
0
0 2 4 6 8 10
Steady state Frequency (Hz)
response
amplitude, %g
Sinusoidal Response, Freq. = f
Equal load applied at each frequency,
(Phase not equal to the force’s phase!)
but the structure responded a LOT more
at the natural frequencies.
22
2. Technical Background:
Dynamic Properties of Floors
Variables That Affect the Response
Damping (Response Inversely
Proportional)
Depends mostly on nonstructural components.
Limited experimental data available.
Low end: 0.5% of critical damping for bare
slabs.
High end: 5% of critical damping for floors with
full height drywall partitions. 6% is often used
when crowds are on the floor.
23
2. Technical Background:
Dynamic Properties of Floors
Variables That Affect the Response
Inversely proportional to effective mass (aka modal mass)
which depends on the psf and extent of motion.
24
2. Technical Background:
Dynamic Properties of Floors
Variables That Affect the Response
Location on the floor. Proportional to mode shape
amplitude.
#1
#2
#3
25
2. Technical Background:
Response Prediction
Resonant Responses
4
F (t ) Q h sin( 2 h fStep t h )
h 1
Matches a Natural
Frequency
Causes
Resonance
26
2. Technical Background:
Response Prediction
Low Frequency vs High Frequency Floors
If fundamental frequency exceeds 9 Hz, then
the fourth harmonic can’t reach it.
9 Hz
27
2. Technical Background:
Response Prediction
Low Frequency vs High Frequency Floors
Maximum response for LFF is from resonant
response.
Maximum response for HFF is from a single
footstep.
2
Peak Accel. = 1.18%g 0.4 Peak Accel. = 0.386%g
Meas. Acceleration (%g)
0 0
-0.2
-1
30
2. Technical Background:
Response Prediction
Resonant Response Prediction Options
SDOF Idealization
Driving frequency
a(t ) matches
natural frequency.
K
Psin(2fn t )
M
t
C
P
K aSteadyState
fn (Natural Frequency) 2M
M
C The basis for almost all modern
(Modal Damping Ratio) SDOF approximate methods for
2(2fn )M low frequency floors.
2. Technical Background:
Response Prediction
Resonant Response Prediction Options
SDOF Idealization
Specialize for low frequency floor vibrations.
P = the Fourier series term amplitude that causes
resonance. Recall: F (t ) hQ sin( 2 h fStep t h )
M = an estimate of the effective mass.
Also need to account for incomplete resonant build-up.
Predicted
P acceleration
aSteadyState due to
2M
specialize human
activity at
t resonance.
32
2. Technical Background:
Response Prediction
“Off Resonance” Response of SDOF System
We’re sometimes interested in the steady-state response of
the system to a sinusoid with amplitude P and f ≠ fn.
P a(t )
aSteadyState(fn ) K
2M Psin( 2ft )
M
C
P /M
aSteadyState (f )
2
fn f 2 f
2
n 1 2 n
f f
Frequency, f (Hz)
33
2. Technical Background: Human
Tolerance
Depends on the
setting, what
they’re doing, and
the frequency of
vibration.
Most sensitive in 4
Hz to 8 Hz band.
0.5%g is the limit
for quiet spaces.
36
3. DG11 Chapter 4 Introduction
Criterion in a nutshell
Under no circumstance can the natural frequency
be less than 3.0 Hz vandal jumping.
Predict the peak acceleration due to walking.
Compare that to the human tolerance.
Po e 0.35f
n
ap ao
W
Human tolerance
Predicted Peak limit, g
Acceleration, g
37
3.1 DG11 Ch. 4 Acceleration
Prediction Equation—Derivation
Derivation useful to avoid incorrectly
applying the equation.
Start with the SDOF steady-state acceleration
equation.
a(t ) P
aSteadyState
K 2M
P sin(2fnt )
M
(at resonance)
C
t
38
3.1 DG11 Ch. 4 Acceleration
Prediction Equation—Derivation
Specialize by defining P for walking.
Murray and Allen (1993) analyzed the footstep
force measurements by Rainer et al. (1988).
P = Amplitude =
Harmonic DLF DLF x Bodyweight
0.5 (lbf)
1 0.5 78.5
2 0.2 31.4
3 0.1 15.7
0.2 4 0.05 7.85
0.1
0.05 Assumed Bodyweight = 157 lbf
Could select P from this table, but
Allen and Murray found a more
Figure by Rainer et al. (1988) elegant method.
39
3.1 DG11 Ch. 4 Acceleration
Prediction Equation—Derivation
Specialize by defining P for walking
Write an equation that works regardless of
which harmonic is applicable.
DLF 0.83e 0.35f n
P Q
(Q )(0.83 )e 0.35fn
41
3.1 DG11 Ch. 4 Acceleration
Prediction Equation—Derivation
Effective Mass
Estimate the weight of plan area participating in
the motion. Call that W.
Effective mass is 0.5W because the mode shape
is similar to a beam (Allen and Murray 1993).
42
3.1 DG11 Ch. 4 Acceleration
Prediction Equation—Derivation
Putting it all together.
P
aSteadyState
2M DLF Call this Po
44
3.2 DG11 Ch. 4 Acceleration Prediction
Equation—Implementation
Natural Frequency
Approach: Compute fn,b for the beam, fn,g for
the girder, and combine to get the bay’s
frequency, fn.
Natural frequency in Hz for a simply supported
beam with uniform mass and EI.
E, ksi What mass is used? Use the
I, in.4 absolute best estimate. It’s
m, kip/in. sec.2/in. E I not conservative to over- or
fn ,b under-estimate the mass.
2 m L4
(over-estimation will be
L, in. conservative for fn, but
unconservative for W.)
45
3.2 DG11 Ch. 4 Acceleration Prediction
Equation—Implementation
Another one of these “don’t be misled by…”
items: m is mass attached to the beam, plus
the beam itself. Don’t think of it as load.
(This will clear up questions such as “Does m
include precomposite DL?”)
E I
fn ,b
2 m L4 also applies to
46
3.2 DG11 Ch. 4 Acceleration Prediction
Equation—Implementation
Natural Frequency
Re-write the equation in a form that’s easier to
remember. fn,b in Hz.
Note how w creeps in. We’re still dealing with
mass, not load. w is just part of m=w/g.
E I g E I g 384 E I 5
fn ,b
2 mL 4
2 w L 4
2 5 w L4 384
5 384 E I g g
g 0.179 0.18
2 384 5w L 4
E I g g
fn,g 0 . 179 0 . 18
2 m L4
48
3.2 DG11 Ch. 4 Acceleration Prediction
Equation—Implementation
Natural Frequency
There are usually two girders. Use the one
with the lowest natural frequency. Discard the
other one.
Spandrel girders supporting cladding are
assumed to be fixed. Modal test data indicates
significant restraint even if cladding has
vertical slip connections.
49
3.2 DG11 Ch. 4 Acceleration Prediction
Equation—Implementation
Natural Frequency
Bay natural frequency, fn, computed as the
combination of fn,b and fn,g. Use Dunkerley’s
Equation.
1 1 1
(DG11 Eq. 3.2)
fn2 fn2,b fn2,g
Similar to before,
g g
fn 0.179 0.18 (DG11 Eq. 3.4)
b g b g
50
3.2 DG11 Ch. 4 Acceleration Prediction
Equation—Implementation
Natural Frequency
Use fully composite transformed MOI.
Amplitude of horizontal shear between beam and
deck is very small, so won’t overcome deck
puddle welds or friction.
Exceptions: noncomposite if actual physical
separation between the beam and deck.
Effective slab width a little different from the AISC
Spec. Ch. I value. See DG11 Section 3.2.
1/2 Beam Spacing
be,EachSide min 0.2Lb
Edge Distance 51
3.2 DG11 Ch. 4 Acceleration Prediction
Equation—Implementation
Natural Frequency
Moment of Inertia
Use dynamic elastic modulus for concrete.
Same as in the AISC Specification Ch. I except
35% higher. DG11 Section 3.2.
52
3.2 DG11 Ch. 4 Acceleration Prediction
Equation—Implementation
Natural Frequency
Modified MOI for open-web steel joists.
These have significant shear deformation (unlike
w-shapes) and web eccentricities.
Research by Band and Murray (1996).
2 in. +/-
(Beavers 1998)
53
3.2 DG11 Ch. 4 Acceleration Prediction
Equation—Implementation
Natural Frequency
Modified MOI for open-web steel joists, continued.
1
Ieff (DG11 Eq. 3.18)
1
Ichords Icomp
MOI of chords only Composite transformed MOI
1
1 (DG11 Eq. 3.19)
Cr Single or double angle webs
Cr 0.90 1 e
0.28( L / D ) 2.8
where 6 L / D 24 (DG11 Eq. 3.16)
E Ieff g E Ieff g g
fn ,b 0.179 0.18
2 m L4 2 w L4
55
3.2 DG11 Ch. 4 Acceleration Prediction
Equation—Implementation
Natural Frequency
Modified MOI for joist girders.
Slab is physically separated from the top chord.
Joist seats provide limited horizontal shear
stiffness between the joist girder and the slab.
Slab
56
3.2 DG11 Ch. 4 Acceleration Prediction
Equation—Implementation
Natural Frequency
Modified MOI for joist girders.
Icomp Inc
Ieff Inc (DG11 Eq. 3.14)
4
where
Cr 0.90 1 e
0.28( L / D ) 2.8
where 6 L / D 24 (DG11 Eq. 3.16)
1
1 (DG11 Eq. 3.19)
Cr
57
3.2 DG11 Ch. 4 Acceleration Prediction
Equation—Implementation
Effective Weight
Recall the acceleration prediction equation.
RP RQ0.83e 0.35fn Poe 0.35fn
ap g
2M 20.5W / g W
59
3.2 DG11 Ch. 4 Acceleration Prediction
Equation—Implementation
Effective Weight
Beam Mode Effective Weight
Wb pb Bb Lb (DG11 Eq. 4.2, slightly modified)
Cb Ds / Db Lb
0.25
Bb min , ft (DG11 Eq. 4.3a)
2 / 3 the " floor width"
60
3.2 DG11 Ch. 4 Acceleration Prediction
Equation—Implementation
Effective Weight
Beam Mode Effective Weight
Cb Determination. Default is 2.0. Other option is
1.0 at a free edge “mezzanine condition.”
Interior Opening
(Atrium)
61
3.2 DG11 Ch. 4 Acceleration Prediction
Equation—Implementation
Effective Weight
Beam Mode Effective Weight
Floor width: distance perpendicular to the beams
to discontinuity in the framing such as slab edge,
interior opening, or framing direction change.
Checking
this bay.
Floor
Width
62
3.2 DG11 Ch. 4 Acceleration Prediction
Equation—Implementation
Effective Weight
Beam Mode Effective Weight
Ds and Db are slab and beam transformed flexural
stiffness, respectively, per unit width. See DG11
Section 4.2. Units: in.4/ft
(12 in.)de3 1
Ds de
12 n
64
3.2 DG11 Ch. 4 Acceleration Prediction
Equation—Implementation
Effective Weight
Girder Mode Effective Weight
Beams rigid, girders bend.
Extent of the motion perpendicular to girders
relates to ratio of beam to girder stiffness.
Very flexible beams small Bg.
Very stiff beams large Bg.
Reality: varies, but often a little larger to perhaps
double the beam length.
Calcs almost identical to beam bending mode.
65
3.2 DG11 Ch. 4 Acceleration Prediction
Equation—Implementation
Effective Weight
Girder Mode Effective Weight
Wg pg Bg Lg (DG11 Eq. 4.2, slightly modified)
Lg girder length, ft
Cg Db / Dg Lg
0.25
Bg min , ft (DG11 Eq. 4.3b)
2 / 3 the " floor length"
66
3.2 DG11 Ch. 4 Acceleration Prediction
Equation—Implementation
Effective Weight
Girder Mode Effective Weight
Cg depends on type of connection to girder.
1.6 if joist seats.
1.8 if typical beam-to-girder connections.
Floor length: same idea as floor width except in
the other direction. Floor length is perpendicular
to the girders.
Continuity effect: same as for beams except only
applies if the girder runs over the column.
Stacked columns, in other words.
67
3.2 DG11 Ch. 4 Acceleration Prediction
Equation—Implementation
Effective Weight
Combined Beam and Girder Mode Effective
Weight
Use weighted average.
Deflections from the natural frequency
calculations.
b g
W Wb Wg (DG11 Eq. 4.4)
b g b g
68
3.2 DG11 Ch. 4 Acceleration Prediction
Equation—Implementation
Damping
Estimate from DG11 Table 4.1, or
Cumulative
Structural system: 0.01
Ceiling and ductwork: 0.02
Paper office: 0.01
Electronic office: 0.005
Churches, schools, malls: 0.0
Drywall partitions: 0.05
69
3.2 DG11 Ch. 4 Acceleration Prediction
Equation—Implementation
Mass (Typical Values)
Member self weight
Slab and deck self weight (nominal)
Superimposed dead load: 4 psf
Paper office live load: 11 psf
Electronic office live load: 8 psf
Residence live load: 6 psf
Assembly or mall live load: 0 psf
70
3.3 DG11 Ch. 4 Acceleration
Limits for Human Comfort
DG11 Fig. 2.1 and Table 4.1.
1.5%g
No Good
0.5%g
3 Hz to 10 Hz
Bandwidth OK
71
References
Allen, D.E., Rainer, J.H., and Pernica, G. (1985). “Vibration Criteria for Assembly Occupancies.” Canadian
Journal of Civil Engineering, 12(3), September.
Allen, D.E., and Murray, T.M. (1993). “Design Criterion for Vibrations Due to Walking.” Engineering Journal, AISC,
4th Qtr, 117-129.
Band, B., and Murray, T.M. (1996). MS Thesis: Vibration Characteristics of Joist and Joist-Girder Members,
Virginia Tech.
Bachmann, Hugo, et al. Vibration Problems in Structures: Practical Guidelines. Basel, Switzerland: Birkhäuser
Verlag, 1995.
Davis, B. and Murray, T.M. (2009). “Slender Monumental Stair Vibration Serviceability.” Journal of Architectural
Engineering, ASCE, 15:4 (111).
Davis, D.B. and Murray, T. M., (May 2010). “Simplified Finite Element Method for Predicting Low Frequency Floor
Vibration due to Walking.” NASCC Presentations. AISC, Chicago, IL.
International Standards Organization, (2007). International Standard ISO 10137, Bases for design of structures –
Serviceability of buildings and walkways against vibrations, 2nd Ed, International Standards Organization,
Geneva, Switzerland.
Murray, T. M. (1975). “Design to Prevent Floor Vibrations.” Engineering Journal, AISC, 3rd Qtr, Chicago, Illinois,
pp. 82-87.
72
References
Murray, T. M. (1981). “Acceptability Criterion for Occupant-Induced Floor Vibrations.” Engineering Journal, AISC, 2nd
Qtr, Chicago, Illinois, pp. 62-70.
Murray, T.M., Allen, D.E., Ungar, E.E. (1997). Steel Design Guide Series 11: Floor Vibrations Due to Human Activity.
American Institute of Steel Construction (AISC), Chicago, Illinois.
Pabian, S., Thomas, A., Davis, B., and Murray, T.M. (2013) “Investigation of Floor Vibration Evaluation Criteria Using
an Extensive Database of Floors,” Proceedings of ASCE Structures Congress.
Rainer, J.H., Pernica, G., and Allen, D.E. (1988). “Dynamic Loading and Response of Footbridges.” Canadian
Journal of Civil Engineering, 15, 66-71.
Smith, A.L., Hicks, S.J., Devine, P.J. (2007). Design of Floors for Vibration: A New Approach. The Steel Construction
Institute (SCI P354), Berkshire, England.
73