Bulletin of the Selsmological Society of America. Vol.
54,
No. 1, pp. 233-241. February, 1964
A NEW METHOD FOR THE MEASUREMENT OF THE NATURAL
PERIODS OF BUILDINGS
BY D. E. HUDSON, W. O. I~EIGHTLEY, AND N. N. NIELSEN
ABSTRACT
It is shown that the inertia force obtained by a man moving his body back and forth in syn-
chronism with the natural period of vibration of a large structure is sufficient to build up a
measurable amplitude of motion. By recording such structural vibrations versus time, the
natural period and damping of several of the lower modes of vibration can be determined.
The amplitudes of motion set up in this way are for many structures significantly larger than
can be obtained from wind excitation, which has been used in the past for the measurement of
the period of the fundamental mode.
INTRODUCTION
The natural periods of vibration of a structure are perhaps the most significant
dynamic parameters involved in response analysis. Not only are the numerical
values needed for computations, but a comparison between experimentally deter-
mined values and calculated values is an important check on the validity of the
simplified mathematical models which must of necessity be used in the analysis
(Housner, 1962).
The importance of the fundamental natural period as an indication of the dynamic
behavior of structures is emphasized by the introduction of this parameter into
some modern eaiihquake-resistant building codes. The "Recommended Lateral
Force Requirements" of the Structural Engineers Association of California (1960),
for example, fixes the basic lateral force coefficient to be used in design by a formula
involving the fundamental natural period.
An additional reason for an interest in these natural periods is the belief that
concealed structural damage such as might occur during a strong earthquake might
significantly alter the fundamental period, and hence period measurements before
and after an earthquake might reveal such hidden damage. Although there is some
evidence from past measurements that significant period shifts have been caused by
earthquakes, the presently available data are not sufficient to arrive at any definite
conclusion. It is important that period measurements be made on existing structures
so that, should an earthquake occur, the basic data will be available for comparison.
At present, accurate period data is at hand for only a very few buildings.
EXPERIMENTAL METHODS
The experimental methods that have been used for period measurements of full-
scale structures are: (1) resonance testing, using a variable frequency sinusoidal
vibration generator; (2) free vibration decay tests, excited by initial displacements
or velocities; (3) wind-excited forced vibration tests, using the very small amplitudes
set up by natural gusts.
The first two types of tests require relatively elaborate equipment, and in practice
it is seldom possible to secure permission to make such tests in buildings. The wind-
excited tests can be quickly carried out without the installation of any equipment
233
234 BULLETIN OF THE SEISMOLOGICAL SOCIETY OF AMERICA
in the building, it being only required to temporarily place a portable seismograph
in an upper story position.
The difficulties with the wind-excited tests are: (1) suitable natural gusts may not
always be available; (2) very low amplitude levels are set up, and thus
some structural elements may not be brought into action in a typical way; (3) since
the form of the exciting force is not known, no information on structural damping
can be deduced from the record; and (4) usually only the fundamental mode of
vibration will have an appreciable motion, and thus no information on the higher
modes is obtained. In spite of these defects, the simplicity of the test, and the
fact that it is the only possibility for most buildings, has given it an important place
in structural dynamic investigations. A considerable amount of such data was
FUNDAMENTAL ,MODE 2.25 CYC./SEC. IO.OO, IN. AT X360
2
sE¢. v-x,6o, x~2o
I- "~ HT- XI80
~.OOl IN. AT X4OO ,SECOND MODE 12.,5
CYC./SEC.
HL-X200 ~,,,#,%~
HL- X40O I~ ~1 ~ ~'
V-
XI70, X340
I `SEC.
I z _1
HT-X200 ~l
FIG. 1. Sample records of wind-excited tower vibrations.
collected in the early 1930's by the United States Coast and Geodetic Survey and
reported in the publication "Earthquake Investigations in California" (1936). At
the present time the United States Coast and Geodetic Survey is reactivating and
enlarging this period measuring program with improved instrumentation.
As an example of the results that can be obtained with such wind-excited tests,
figure 1 shows measurements made on a 100 ft high concrete intake tower of a
dam. (Keightley, Housner, and Hudson, 1961). This test was unusual in that a
second mode of vibration was also clearly excited during pai~ of the record, and
hence the first two periods of vibration could be determined. The records of figure 1
were obtained on a photographically recording portable seismograph having a
natural period of about 2 seconds and a magnification of approximately 400.
There is a possibility that some information on structural damping could be
METHOD FOR MEASUREMENT OF NATURAL PERIODS
235
extracted from a wind-excited vibration record such as fig. 1 if some data on the
input force could be obtained. If, for example, the exciting force power spectrum
could be simultaneously measured, something might then be deduced about the
damping. This does not seem to be feasible in practice, however, because of the
large size of the structures involved, which would require that the input wind forces
be simultaneously measured at a number of points to arrive at the integrated excit-
ing force.
A NEW
~ETHOD OF EXCITATION
During some wind-excited tests of a 150 ft high concrete intake tower, it was
noted that a considerable deflection of the seismograph recorder could be set up by
the operator jumping laterally at the top of the tower. This suggested that the
small inertia force generated by the operator himself, if it were properly synchro-
nized with the natural period of the tower, might be sufficient to build up
a measurable resonant vibration. With a little practice it was found that this was
I
0.001 IN.
I I i 8Eo.
V ¥ '¥ '~' ~r~. 1" , ...... ,! ........
FIG. 2. "Man-excited" vibrations of 150 ft. high concrete intake tower of dam.
indeed the case. By keeping one eye on the seismograph recorder to observe the
way the vibrations were building up, an operator with an ordinary sense of rhythm
could by periodic motions of his body produce considerably larger amplitudes of
motion at the fundamental period than had been produced by the wind. Figure 2
shows the record obtained in this way on the 150 ft high concrete intake tower using
the same seismograph that was used for figure 1.
It was immediately evident that this method of excitation had three important
advantages over the wind-excited force: (1) larger amplitudes of motion could be
built up at a definite period; (2) the test could be carried out at any time, in the
presence or absence of wind; and (3) in the absence of wind, by stopping the exciting
force after an appreciable motion had been built up, the decay of free vibrations
could be recorded, and hence damping could be measured.
A vibration of the above kind is in one sense a "self-excited" vibration. Since
this phrase, "self-excited", is already used in mechanics for a rather different
phenomenon in which the forces sustaining the motion are derived from the motion
itself, it might be better to refer to this particular type of forced vibration as a
236
BULLETIN OF THE SEISMOLOGICAL SOCIETY OF AMERICA
"man-excited" motion. This method of excitation naturally recalls the old stories
often mentioned in mechanics lectures of bridges destroyed by marching armies,
and the supposedly common rule that soldiers should break step when crossing a
bridge.
At first thought it seemed unlikely that a sufficient exciting force could be
obtained in this very simple way to be useful for large structures such as multistory
buildings. However, it turns out that it is just for such large structures with their
relatively low natural frequencies that the method is most useful. As will be shown
by some specific examples, it is often possible in multistory buildings to achieve
an amplitude far exceeding that obtained from the wind.
The essential feature in producing a significant "man-excited" vibration is to
insure that the center of mass of the body moves with as large as possible amplitude
at a reasonably constant frequency. One effective technique is to stand facing
perpendicular to the direction of excitation and then to sway the body sideways,
shifting the weight from one leg to the other. Another technique is to hold a column
oi" door-jamb, and then move the body backward and forward, transmitting the
force to the structure through the arms and legs. Other interesting possibilities,
perhaps involving the synchronized efforts of several people, will suggest themselves.
An idea as to the magnitude of the force generated in this way can be gained by
supposing that the center of mass of a 150 lb man is moved sinusoidally through a
double amplitude of 6 in. at a frequency of 1 cps. This would produce an inertia
force magnitude of 46 lb. Considering that a structure having 1% of critical damping
has a dynamic amplification factor at resonance of 50, such force magnitudes can
easily produce measurable displacements.
Returning to fig. 2, it is of interest to note how quickly the vibration amplitudes
build up at resonance. Only four cycles of motion were required to bring
the amplitude to a large value, from which a clear free vibration decay record was
obtained. From this decay curve, the damping can be calculated to be less than
1 per cent of critical. The values of period and damping obtained in this way were
subsequently found to be in good agreement with those calculated from resonance
curves determined with a sinusoidal vibration generator installed at the top of the
tower.
At this stage in the investigations, a new portable seismograph became available
which proved to be particularly well suited to such tests. Through the com~esy of
the Seismological Laboratory of the California Institute of Technology, an experi-
mental model of the "lunar seismometer" (Lehner
et al.,
1962), along with a com-
pact recording drum and pen recorder system, was borrowed for a number of build-
ing period tests. This lunar seismometer is a permanent magnet-moving coil type
adjusted to a natural period of about 1 second. By means of a simple transistor
amplifier, a 5 in. ink-recording pen can be driven at a maximum magnification of
about 10,000 at 4 cps and about 1000 at 1 cps. A small recording drum of 3-~ in.
diameter is operated at a recording speed of 1 cm/sec. The combination of large
output, high gain variable over a large range, compactness, and ruggedness, makes
this instrument very suitable for building period tests. Another important advan-
tage is the pen-recorder, which not only gives an immediately visible record, but
which has a relatively large mechanical moving element to watch, which assists in
synchronizing the applied inertia force and the resonant vibrations.
METHOD FOR MEASUREMENT OF NATURAL PERIODS
237
TESTS ON BUILDING FRAMES
The first tests with the lunar seismograph were made on a five-story reinforced
concrete frame building in course of construction. Figure 3 shows typical records
from which both the fundamental natural period and the damping can be obtained
as given. These values were later checked against resonance curve calculations from
steady state sinusoidal vibration generator tests with a satisfactory agreement.
•
I I
ONE SECOND
FIG. 3. "Man-excited" vibration of five-story reinforced concrete frame building. Funda-
mentM mode 2.36 CPS; damping = 1.8% critical.
E-W N -S
220
FT. 1
C
A:
126 FT.
,--40 FT.~
CE
AB
D
~'/4 "//I///////////'2
FIG. 4. Nine-story steel-frame building used for "man-excited" vibration tests. Letters mark
points of excitation and measurement.
A second test of the method was made on the frame of a nine-story steel-frame
building under construction. In this ease it was possible to dearly distinguish be-
tween 8 different modes of vibration, as shown in figures 5, 6, and 7. Figure 4 shows
the general configuration of the building, and indicates the locations of the points
of excitation. These points of excitation should be selected to emphasize the mode
desired, and to suppress other modes. For example, by applying the exciting force at
a node for the 2nd mode of vibration, the 3rd mode can be more easily recognized.
Although these mode shapes will, of course, not be accurately known for actual
structures, it will be possible to make a sufficiently accurate estimate from the gen-
eral configuration of the structure.