Documente Academic
Documente Profesional
Documente Cultură
Vibration Measurement
Table of Contents
I. Objective
The purpose of this lab is to study and observe free and forced vibrations of simple cantilever
beams.
1. In the first part of the lab, the frequency and amplitude of oscillation will be recorded of a
free vibrating system. Using the acquired data and known information about the
cantilever beam, the characteristics of the vibrating system will be computed and verified.
2. In the second part of the lab experiment, the frequency and amplitude of vibration of a
continuous mass system will be measured. The first three natural frequencies will be
theoretically calculated, experimentally determined using several methods, and observed.
The three mode shapes will be subsequently graphed and comparisons between the
theoretical and the experimentally observed shapes will be discussed.
II. Apparatus
Part 1:
1. Cantilever beam mounted in a clamp support,
2. C-clamp to hold the base to the table,
3. Two strain gages mounted to the beam in an axial direction,
4. Four pound set of weights with pan and hanger,
5. Ruler
6. 24-bit A/D converter and data acquisition software,
Part 2:
8. Cantilever beam with a mounting plate and an adjustable transducer mount,
9. B&K 4818 Mini-Shaker,
10. Bogen 60 watt Audio Amplifier,
System Dynamics
All physical systems have a characteristic response to any forcing function. If you put a
thermometer in the sun, it heats up. If you hit a piece of metal, it makes a noise. These
characteristics can be modeled by differential equations. We can use these system models to
predict how a system will behave even if we have not specifically tested the system response.
This is true of systems we perform measurements on as well as systems we use to take the
measurements.
A general model for an output variable y can be defined as a system of nth order:
n n-1
d y d y dy
an n an-1 n-1 + ..... a1 + a0 y = F(t)
+
dt dt dt
Where F(t) is a forcing function and n is the highest derivative. We will consider cases where
n=0, 1, and 2.
Zero-Order Systems
a0 y = F(t) ; y = KF(t)
where, 1/a0=K is the system gain. From the solution it is clear that the system response is simply
the input multiplied by K. All zero-order systems mimic their input with no time delays,
oscillations or other dynamics. The BAM is an example of a zero-order system (unless the input
frequency becomes high). A lever arm is another one (as long as the motion is slow enough to not
excite the lever arm bending mode).
First-Order Systems
A first-order system has a derivative in the model. The model equation and the solution for the
response to a step change in F(t) of amplitude A is:
dy t
a1 + a0 y = F(t) ; y(t) = KA + ( y0 - KA) e-
dt
where a1/a0=, the system characteristic time constant. A first-order system will respond to an
input with a multiplication by K and an exponential time delay. A thermometer is a first-order
system. Exposed to a temperature, its display exponentially approaches that temperature. One
may think of a first-order system as having the capability to store one form of energy.
Second-Order Systems
2
d y dy
a2 2
+ a1 + a0 y = F(t)
dt dt
1 d 2 y 2 dy
+ + y = KF(t)
2n dt 2 n dt
where n is the natural frequency of the system and is the system damping ratio. The solution
to this equation has three parts because complex numbers occur in the roots.
y(t) = C1 e1 t + C 2 e 2 t
2) If =1, then
y(t) = C 1 e1 t + C 2 t e 2 t
3) If <1, we get
y(t) = C e- n t sin ( d t )
where
1 ,2 = - n + n - 1 ; d = n 1 -
2 2
The first case gives exponential time delays similar to the first-order system and is called
overdamped. The second case is critically damped, and gives the smallest pure time-delayed
response possible for the system. The last case is underdamped. Here the solution is oscillatory
at the damped natural frequency d. A second-order system can thus oscillate. Mechanical
oscillation is called vibration.
Vibrations are an important aspect of mechanical engineering. Vibration is present at some level
in all engineering machines and structures. Vibration can be either helpful, as in the case of
vibrating sifters and mixers, or harmful, such as in the case of the Tacoma Narrows Bridge, which
collapsed due to unstable torsional vibrations. One of the first steps required in the analysis of any
vibration problem is determining the natural frequencies of vibration of a part, or a system
containing the part, in question.
Free vibration is a condition that exists when a system, starting from some initial condition, is
allowed to vibrate without the effect of an external forcing function. In the first part of this lab,
we will study the free vibration of a cantilever beam with a mass on the end, shown in Figure 1.
We will model the freely vibrating mass as a single degree-of-freedom system, as shown in
Figure 2. This is an approximation because the cantilever beam, although stiff and light
compared to the mass, may bend and introduce additional degrees of freedom. The degree of
freedom is equal to the number of independent coordinates that must be used to fully describe the
motion of the system as it vibrates. The spring-mass-damper model in Figure 2 is the simplest
model which can be used to represent all single degree-of-freedom (or point-mass) systems and is
the basis of the modeling equations given earlier.
Consider the spring-mass-damper in an initial state of equilibrium. When the system is perturbed
by external forces, it absorbs a certain amount of energy. The energy storage may be in the form
of potential energy of the spring, or kinetic energy of the mass. As the perturbing force is
removed, the system tends to return to its equilibrium position and dissipate energy in the
process. Depending on the damping in the system, this process may be characterized by a simple
asymptotic return to its original position (overdamped or critically damped) or by exponentially
decaying vibrations at a single characteristic natural frequency about the equilibrium position
(underdamped).
The underdamped system is characterized by the amplitude of vibration, the damping ratio and
the natural frequency. The amplitude depends on the initial condition and will decrease to zero as
equilibrium is approached, but the frequency and the damping ratio should remain constant. They
are not affected by the nature of the initial disturbance (this is true only for purely linear systems,
non-linear systems may change frequency or damping with amplitude but our experiment behaves
quite linear).
The deflection of the vibrating beam, or the spring-mass-damper system, in our experiment will
be characterized by a damped sinusoidal motion, shown in Figure 3, of the form:
-
d t
1 - 2
x (t) = A e sin ( d t + )
Where :
k eq
d = damped natural frequency = n 1 - 2 ; n=
M eq
: damping ratio
: offset phase angle
keq: equivalent spring rate ( or stiffness) = Ebh3/4l3)
Meq: equivalent mass = mass of weight + 33/140 * mass of the overhung portion of the
beam ( this means that 33/140=23.5% of the beam mass participate in the motion)
Note that as approaches zero (i.e. the system is undamped) d approaches n, the undamped
natural frequency.
The natural log of the amplitude decay of each successive period in Figure 3 is called the
logarithmic decrement, , and it is related to the damping ratio, , by the equation
x (t) 1 x (t) 2
= ln [ ] = ln [ ]=
x (t + T d ) n x (t + n T d ) 1 - 2
Thus, the damping ratio, , can be calculated by measuring the decay in amplitude after any
number of oscillations, n , and plugging it into the above equation. Notice that as the damping
ratio approaches unity, goes to infinity. This corresponds to the critical damping ratio.
Forced vibrations occur when a system is subjected to some type of continuing external excitation
that adds energy to the system. In this lab, we will examine the effect of a steady-state sinusoidal
excitation of a distributed-mass cantilever beam system. With the hanging mass removed, a
cantilever beam will exhibit higher modes of vibration just like the guitar string in our first lab.
The higher modes of vibration had a negligible effect on the initial overhung mass system, and
thus were neglected in the single degree-of-freedom analysis. In the second part of the lab,
however, you will see that these higher modes have a significant effect on the dynamics of the
distributed-mass cantilever beam. Continuous systems like this possess an infinite number of
degrees-of-freedom. In this lab we will examine the characteristics of the first three natural
frequencies. Infinitely many higher modes exist, but their amplitude becomes too small to
measure.
An electrically driven mini-shaker will provide excitation to the system by vibrating the end of
the cantilever beam. The vibration is produced by applying a sinusoidal alternating current to an
inductive coil surrounding a central plunger. By adjusting the frequency of oscillation of this
current with a sine-wave signal generator, different forcing frequencies can be applied to the
beam.
The vibration of the beam used in this lab will be characterized by the first three mode shapes
shown in Table 1, plus an additional spurious mode ???, which may occur (or may not, depending
on boundary condition on your particular system) somewhere between modes 1 and 2. This extra
mode is not predicted by the first-order analysis of the cantilever beam. It is caused by the fact
that the shaker mount does not exhibit perfect translation, but allows for a certain amount of
rotation. The location of the mystery mode will be dependent on the stiffness of the bearing
system - a boundary condition. You may not observe this mode on your particular setup. You
should always think about the effects of system boundaries when designing an experiment as they
can modify what you think you're measuring. Note that the mystery node is not really a
mystery, it is simply not predicted by the theory, because the theory assumes a perfectly
clamped boundary condition.
The primary natural frequencies of the mode shapes (except spurious mode) are determined by
(use consistent units, that means force, mass, time and length unit can not be independently
chosen and are related via Newtons Law)
EI
n = ( n l )
2
A l4
One phenomenon, which you will notice is that the magnitude of the oscillation becomes very
large when the frequency of the excitation matches one of the natural frequencies of vibration of
the beam calculated above. In addition, the cantilever beam will visibly assume each of the above
mode shapes at the respective natural frequencies. This phenomenon can be observed with the
use of the strobe light. Note that in the second and third mode shapes, there are points where
there is no displacement; these points are called nodes (denoted N). The points corresponding to
maximum displacement are likewise called antinodes (denoted A).
1 1.8751
????? ??????
2 4.6941
3 7.8548
Table 1. A schematic of the mode shapes for a cantilever beam subjected to forced
vibration with the corresponding nodal and anti-nodal positions.
IV. Procedure
1. Connect axial half bridge strain gage wires to the data acquisition system (refer back to
strain lab)
6. Cleanup experiment A. Carefully dismantle and store all equipment. Close the VI.
The transducer signal will be recorded with the data acquisition system, and the fundamental
vibration frequency and any harmonics will be evaluated with the FFT function of LABVIEW as
in the first lab. Since the beam is in simple harmonic motion, dividing the magnitude of the
beam's maximum velocity at each point by the fundamental frequency will yield the maximum
displacement of the beam at that point. The conversion from measured voltage to velocity
depends on the gap between beam and transducer and is not linear, so the calibration curve given
in Figure 4 must be used. The transducer will be moved along the length of the beam in one inch
increments so that data can be taken along the length of the beam at each of the three lowest
natural frequencies of vibration of the beam. This data can then be used to plot the mode shapes
of vibration at these three natural frequencies. In addition, the strobe light will be used to view
the beam's vibration at these three resonant conditions in a darkened laboratory. The frequency
of the beam vibration can be found by "stopping" the motion with the stroboscopic illumination.
a. Refer to figure 5 for a schematic of the experiment. Connect the function generator to
the Bogen Audio Amplifier (AUX1). Connect its output to the mini-shaker. The
magnetic pickup connects back to the data acquisition board.
b. Remove the cardboard shim on the shaker.
c. Position the magnetic transducer at the end of the beam and use the gap gauge to set
the transducer-to-beam spacing. Measure and record the gap gauge thickness for use
in Data Analysis Question 3.
d. On the amplifier, turn bass to cut and turn treble to zero. Maximize the signal from
signal generator so you dont have to increase the amplification too much.
New
velocity amplitude measurement for a total of seven to eight positions along the beam.
At each position verify that the gap is maintained with the test gauge.
7. Clean up
a. Replace the cardboard shim on the shaker.
b. Carefully dismantle the equipment and return it to its original state for use by the next
laboratory group
1 a. Plot the time-series strain gage output data stored in Step A4. Measure the amplitude
of the first two successive peaks of the decaying vibration of the approximate point-
mass system. From the amplitude of the peaks find the logarithmic decrement and
thus the damping ratio . Then calculate the undamped natural frequency using the
damped natural frequency from the FFT measurement and the damping ratio just
found.
b. Compare the experimental natural frequency to the theoretical natural frequency of the
beam.
c. Repeat the above calculation for two other successive peaks near the end of oscillation.
d. Evaluate the uncertainty in the damped natural frequency d, resulting from
uncertainties in the beam dimensions and measurements. Don't forget boundary
conditions that affect d and .
2 a. Calculate the first two theoretical mode resonance frequencies of the continuous mass
system using Table 1. Assume that = 0.283 lbm/in3 and modulus is 29x106 psi. Use
consistent units in your calculation. Compare these values to the measured resonant
frequencies.
b. The actual resonant frequencies can be evaluated from the readout of the signal
generator, from the FFT analysis of the magnetic reluctance transducer measurement,
and by stopping the beam motion with the strobe light. Discuss the effectiveness of
these measurements by listing the potential errors in each. Consider how each device
is calibrated, the effect each device has on the experiment itself, and any boundary
conditions. Evaluate these errors from your data if possible. Evaluate which of the
readings is most trusted.
VI. References
1. Theory and Design for Mechanical Measurements, R.S. Figliola and D.E. Beasley,
Wiley, (1991)
2. Theory of Vibration with Applications, W. T. Thomson, Prentice Hall, 2nd edition,
1981, Appendix D, pp. 464 - 474.
3. Vibration of Mechanical and Structural Systems, M. L. James, Harper & Row, New
York, (1989).
4. Vibration Problems in Engineering, S. Timoshenko, Wiley, New York, (1974).
VII. Figures
Figure 3. Free vibrating oscillation of a damped single degree of freedom (SDOF) system.
Figure 4. Conversion chart from voltage to velocity as a function of spacing for the
magnetic velocity transducer.
LabVIEW VI
Part 2