Upload
mailson1
View
218
Download
0
Embed Size (px)
Citation preview
8/16/2019 Active Control_floor Vibration
1/5
TA15
= 9~35
Active Control
of
Floor Vibration: Implementation Case Studies
Linda M. Hanagan, Ph.D., P.E.
Departmen t of Civil and Architectural Engineering
University
of
Miami
Coral G ables, FL 33
124
lhanagan@eng rrniami.edu
Thomas M. Murray, PbD ., P.E.
Mo ntague-Betts Professor
of
Structural Steel D esign
Charles E. Via Department of Civil Engineering
Virginia Polytechnic Institute and State University
Blacksburg, VA 24061
Abstract
Progress has been made toward controlling excessive
illustrating the implementation of this control scheme are
presented in this paper.
floor vibration by means of active structural control.
In
this control scheme an electro-magnetic shaker is used to
impart control forces on a floor system, thus, reducing
the floor vibration levels. This paper presents an
overview of the control system setup and describes two
case studies where active control was implemented to
improve floor vibration characteristics.
1.
Introduction
2. Overview of the Active Control System
Control forces are imparted on the floor system by
means of an electro-magnetic shaker. An illustration of
the shaker is presented in Figure
1,
along with the
theoretical model of the shaker. The second order model
of the shaker possesses discrete masses, ma and md,
which represent the reaction (active, 30.4 kg) mass and
the parasitic mass (support frame, etc., 74 kg),
respectively. The spring stiffness, supplied by the
suspension system, is represented by k,. The internal
damping, due to internal motor properties and friction, is
represented by c, and is assumed to be viscous.
The active control of structures is a diverse field of
study, with new applications being developed
continually. One structural system, which is often not
considered a dynamic system, is the floor of a building.
In many cases the dynamics of a floor system are
neglected in the design phase of a building structure.
Occasionally, this omission results in a floor which has
dynamic characteristics found to be unacceptable for the
intended use of the building. Floor motion of very small
amplitudes, often caused by pedestrian movement, is
sometimes found objectionable by occupants of the
building space. Improving an unacceptable floor
system’s dynamic characteristics after construction can
be disruptive, difficult and costly.
In search
of
alternative repair measures, analytical and
experimental research implementing active control
techniques was conducted
to
improve the vibration
characteristics of problem floors. Specifically, a control
scheme was developed utilizing the measured movement
of
the
floor
to
compute
the
input signal
to
an
electromagnetic actuator which, by the movement of the
actuator reaction mass, supplies a force that reduces the
transient and resonant vibration levels. Two case studies
The control law utilized is collocated rate feedback with
a simple command limiter. While adding damping to the
floor system was the key objective, this control law was
selected because it is also robust to system changes and
uncertainties. A presentation of the mathematical model
illustrating the implementation of this control law with
an electro-magnetic shaker is out of the scope of this
paper; however, the formulation is reviewed in
References
[
11and
[2].
The control law is digitally implemented using a 386
computer. With one actuator, the control circuit is
single-inputhingle-output.The input signal, which is a
voltage proportional to the velocity of the floor motion at
the collocated actuator/sensor location, is generated by a
piezoelectric transducer. This input signal is read by an
analog to digital converter housed in the 386 computer.
The output signal is computed by a control program.
The control program implements the control law
discussed previously. Also included in the control
1911
Authorized licensed use limited to: IEEE Xplore. Downloaded on May 6, 2009 at 17:08 from IEEE Xplore. Restrictions apply.
http://rrniami.edu/http://rrniami.edu/
8/16/2019 Active Control_floor Vibration
2/5
algorithm is a nonlinear circuit which limits the output
voltage to the limits of the amplifier and shaker. The
digital to analog converter sends the computed output
voltage to the amplifier which drives the shaker.
The only variable quantity involved in controlling
different floors with different dynamic properties is the
control gain applied in the rate feedback control law.
Because of controYstructure interaction, the selection of
this control gain is a tradeoff between the degree of
control for the floor and the actuator stability. This
concept is illustrated in Reference [3]. The control gain
value can be optimized experimentally by trial
and
error.
An analytical model of the structure to be controlled is
not necessary for the implementation of this control
scheme.
Reaction
Mass
a)
Reaction
Mass
Actuator: Electro-Magnetic Shaker
m a
StructureLZz L
b)
Theoretical Shaker Model
Figure
1.
Illustration of Electro-seis Model
400
shaker
and theoretical shaker model
3. Office Floor Case Study
An office floor in a light manufacturing facility, located
in St. Louis, Missouri, was reported to have annoying
levels of occupant induced floor vibrations. A plan of
this floor is shown in Figure
2.
The construction of this
floor consists of a 2 12 in. lightweight concrete slab on
metal deck supported by
joist framing
members as
indicated on the plan. The 28
ft
4 in. span was found
to be the problem area.
In this span, two long rows of
desks are separated by an aisle near the center of the
span. This open of ice area is used primarily for order
processing with personal computers on nearly every
desk. Walking in the aisle causes computer monitors to
rock, thus intensifying the degree of annoyance. One
particularly disturbing characteristic of this floor is that
annoying levels of vibration are felt even when the
occupant movement is several bays away.
An attempt was made to actively control the floor
movemeint at a location where the problem was
particularly acute. The control actuator and sensor were
placed ai: the controller location noted in Figure 2. The
floor response due to a person walking in the aisle
between the desks was measured for the uncontrolled
and controlled system. To provide a valid comparison,
care was taken to keep the walking excitation
as
consistent
as
possible for the two measurements. A
comparison of the results for the uncontrolled and
controlled system is shown in Figure 3. For each
vibration measurement the rms. acceleration was
calculated. The uncontrolled floor system had a
rms.
acceleration level of 0.57%g while the controlled system
had a level of 0.17%g. This represents more than a
300
reduction in the vibration level.
4 Chemistry Laboratory Floor Case Study
Excessivle floor vibration due to occupant movement was
reported to exist in a Vermont university chemistry
laboratory where sensitive microscopes were
in
use.
A
partial pllan of the floor system is shown in Figure 4.
The 7
ft
span is a corridor with laboratory rooms on
either sidle. The floor construction consists of a 31/2 in.
concrete slab on metal deck supported by joist members
as shown in the plan. The problem area, in the
laboratory with the 28 e.- 7 in. span, contains three
island type workbenches where the function has been
severely impaired due to disturbing levels of floor
vibration.
The active control scheme was implemented to reduce
the floor motion. Several tests were performed to assess
the impact of the control. Results from the walking
excitation tests are shown in Figure
5 .
For these test, the
control actuator and sensor were placed between
two
of
the workbenches. This location is noted in Figure 4.
From the data shown in the graphs, the uncontrolled rms.
acceleration was computed to be 0.37%g for the
uncontrolled response and O.O9%g for the controlled
response. This represents over a 400% reduction in the
vibration level.
A Assessment
of
the Control Scheme
The most valuable assessment of the control scheme lies
the human perception of the floor behavior with and
without control. To those present on each of the floor
systems (during testing, the improvements due to the
active control were dramatic. These improvement, with
respect to human perception, can also be illustrated
by
plotting the data from the experimental studies on the
International Standards Organization
ISO)
cale [4] or
assessing floor vibration levels as shown in Figure 6.
The scale: is represented by a baseline acceleration vs.
frequency curve with multipliers for different occupancy
and vibration types. A logarithmic plot of the baseline
1912
Authorized licensed use limited to: IEEE Xplore. Downloaded on May 6, 2009 at 17:08 from IEEE Xplore. Restrictions apply.
8/16/2019 Active Control_floor Vibration
3/5
U
20 -
0
Stair
28' - 4
0.1
0
Figure 2. Off ice Floor Plan
--
Velocity
i n / S e C )
Uncontrolled Time History
0.2
0.1
0
-0.1
-0.2
0
1
2
3
4
5 6
7
8
Time (sec)
Velocity
(in/seC)
Controlled Time History
.2 1
6
I
-0.2
6 7 8
-OS1
t....
0
2
3 4
Time (sec)
Figure 3 Uncontrolled and Controlled Response of an Office Floor
Due to Walking Excitation
1913
Authorized licensed use limited to: IEEE Xplore. Downloaded on May 6, 2009 at 17:08 from IEEE Xplore. Restrictions apply.
8/16/2019 Active Control_floor Vibration
4/5
28
7
21'
-
7
Velocity
(idsec)
Direction of
Walking
Controller
Location
16K7
@24 0 C.
4
~
1 6 K 2 2 2 4 O . C .
Figure 4. Chemistry Laboratory Floor Plan
Uncontrolled Time History
0
1 2 3
4
5 6 7 8
Time (sec)
Velocity
(idsec)
0.1
0.05
0
-0.05
-0.1
Controlled Time History
0
1
2
3
4
5
6
7 8
Time sec)
Figure 5. Uncontrolled and Controlled Response of a Chemistry Laboratory Floor
Due to Walking Excitation
1914
Authorized licensed use limited to: IEEE Xplore. Downloaded on May 6, 2009 at 17:08 from IEEE Xplore. Restrictions apply.
8/16/2019 Active Control_floor Vibration
5/5
curve (satisfactory magnitude curve for critical working
areas) and a satisfactory magnitude curve for office
environments subjected to intermittent and continuous
vibration is shown in the figure.
I S 0
Scale for Limits of Satisfactory
Magnitudes
of
Floor Vibration
with Respect to Human Perception
h4S
Acceleration
10
(%g)
- -
*
_ I _ , - - - - rf - - - I- - , - 1 - 1 - 1
_ _ _
_ _ _
_ _ _ _ _ _
- - _
_ _
_ _ _ _ _ _ _ _ _ _ _ _
+ - 1
4 - I - I ~ I - l i
- - -
- - I - 1 - 1 - 1
~ - - I - ~ - I - I - I - I T
-
I
- 1 I I M
T
_ _
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _
- - l - - l - l - I - l - - - - + - -I 1 - 1 - 1 - 1 +
-
, - I -
- - -
_ _
- - - _
_ _ _
0.1
I I I I
I l l 1
- -
-
+
1
- 1 - I - l - l i - - - +
- - I - - 1 - 1 - 1 - 1 1
I I I I
I l l
Frequency
hz)
Uncontrolled office floor due to walking excitation
A Uncontrolled laboratory floor due to walking excitation
H Controlled office floor due to walking excitation
A
Controlled laboratory floor due to walking excitation
Figure
6.
Evaluation of floor vibration levels using
I S 0
human perception scale [4]
As represented in Figure 6, the active control scheme has
reduced the office floor response to walking excitation to
within an acceptable level, with respect to the I S 0 scale,
for an office environment at the measured location.
While the chemistry laboratory floor response also
shows a significant reduction in vibration levels, its
classification as a “critical work area” has much more
stringent requirements for satisfactory magnitudes
of
vibration and is therefore still unacceptable with respect
to the IS0 scale. Additional damping, beyond what is
provided by the active control scheme, would do little to
further reduce the vibration levels in the laboratory floor.
Alternate repair measures affecting the stiffness of the
system would be necessary to bring the vibration levels
to within satisfactory limits represented by the baseline
curve.
6.
Conclusions
The active control scheme studied in this research
presents several advantages over many of the traditional
methods used in repairing problem floors. An actively
controlled mass provides a larger degree of control than
a passive device with an equivalent reactive mass.
The
active system is also less disruptive to the building
function than most other repair measures. The active
device is rather compact and can be installed with
relative speed and ease in the ceiling cavity present in
most commercial buildings.
There are also disadvantages to the active control
scheme. The cost of the components to provide a single
control circuit are currently very high. The hardware
components alone have a total cost of $21,300 for a
single control circuit. This results in an estimated cost of
$24 per square foot, assuming one actuator is necessary
to control a 30 ft x 30 ft bay. One must keep in mind,
however, that any new technology is expensive and often
becomes more reasonable in time. Maintenance and
reliability issues also detract fiom the attractiveness of a
active system. These issues are not necessarily
prohibitive. Maintenance and repair is necessary for
many building systems. As this technology matures,
maintenance and repair could be considered similar to
changing a filter or overhauling a boiler.
The potential of this application far exceeds the
drawbacks. The results of this research, in addition to
future research, will move this technology toward
acceptance as an alternative
lo
traditional methods in
repairing problem floors and provide desperate building
owners a practicable solution to a very difficult problem.
References
[I]
Hanagan,
L.
M. and Murray, T. M., “Active Control
of Floor Vibrations,” Technical Report CENPI-ST
94/13, Charles E. Via Department of Civil Engineering,
Virginia Polytechnic Institute and State University,
Blacksburg, Virginia, 1994.
[2] Hanagan,
L.
M., and Murray, T. M., “Floor
vibration: A new application for active control,”
Presented at the Fourth Pan American Congress of
Applied Mechanics, Universidad del Salvador, Buenos
Aires, Argentina, January, 1995.
[3] Hanagan, L. M. and Munay, T. M., “Experimental
Results from the Active Control of Floor Motion.”
Proceedings of the
First W orld Conference on Structural
Control, August 3-5, 1994,Los Angeles, CA, 1994.
[4] Intemational Standards I S 0 10137, “Basis for the
design of structures Serviceability of buildings against
vibration,” Intemational Standards Organization,41-43,
1992.
Acknowledgments.
The work described in this paper has been supported in
part by the National Science Foundation
NSF)
Grant
No. MSS-9201944 and by a grant from NUCOR
Research and Development, Norfolk, Nebraska.
1915