0% found this document useful (0 votes)
38 views13 pages

Free Vibration Analysis of A Uniform Can

Uploaded by

NİSA AYDIN
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
0% found this document useful (0 votes)
38 views13 pages

Free Vibration Analysis of A Uniform Can

Uploaded by

NİSA AYDIN
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 13

Journal ofSound and Vibration (1990) 136(2), 201-213

FREE VIBRATION ANALYSIS OF A


UNIFORM CANTILEVER BEAM W ITH POINT M A SSES BY A N
A NA LYTICA L-A ND-NUM ERICA L-COM BINED M ETHOD

J.-S. WV
Department of’ Naval Architecture and Marine Engineering, National Cheng-Kung University. Turnan.
Taiwan 70101, Republic of China

AND

T.-L. LIN

Department of Industrial Safety and H.vgiene, Chia-Nan Junior College of Pharmacy, Tainan Hsien,
Republic of China

(Received 10 October 1988, and in revised .form 25 April 1989)

The natural frequencies and mode shapes of a uniform cantilever beam carrying any
number of concentrated masses were determined by using an analytical-and-numerical-
combined method (ANC method). The eigenvalue equation was first derived analytically
by using the expansion theorem and then the eigenvalues and eigenvectors were calculated
numerically. In comparison with the general finite element method (FEM), the ANC
method has the advantage that there is no necessity to derive the property matrices of
each beam element and then to develop the overall ones of the entire beam. In comparison
with the transfer matrix method (TMM) the ANC method requires no tedious matrix
multiplication so that some computer time may be saved. The ANC method is also better
than the pure analytical (closed form) method, since, instead of a few special cases, a
uniform beam carrying any number of point masses of various magnitudes or distributions
along the beam length can be easily solved. Besides, by using the ANC method, there is
no difficulty in taking more modes in the mode superposition equation and hence more
higher mode natural frequencies of better accuracy may be obtained.

1. INTRODUCTION

Before the advent of computers, seeking the analytical (closed form) solution was one
of the most popular ways to solve an engineering problem. In comparison with numerical
methods, the closed form solutions have the advantage of elegance and are time-saving.
However, for most practical engineering problems no closed form solutions can be found,
unless they are significantly simplified. Because of this drawback of the analytical (closed
form solution) method, it has gradually been replaced by numerical methods after the
invention of computer, particularly in recent years. One of the greatest shortcomings of
the numerical approach is that it is time-consuming, but because of its powerful ability
for solving the practical problems this drawback seems allowable.
For various (supporting) boundary conditions the closed form solutions of uniform
beams without concentrated mass loads may be found in many books [l-3]. Most of the
closed form solutions have been obtained under the assumption of negligible shear and
rotatory effects. As for the frequently encountered engineering problem of a uniform
cantilever beam carrying a concentrated mass at the free end, the closed form solution
was presented by Laura and his colleagues in 1974 for a rigidly supported cantilever
201
0022-460X/90/020201 + 13 %03.00/O @ 1990 Academic Press Limited
202 J.-S. WU AND T.-L. LIN

beam [4] and in 1975 for an elastically supported one [5]. In 1984 Giirgoze presented a
theory for obtaining the closed form solutions of natural frequencies and mode shapes
of a uniform beam or rod carrying any numbers of point masses and springs [6]. In recent
years, various approximate methods have been presented to solve similar problems [7-91.
Because of the complexity of a beam with more than two point masses or springs, only
the closed form solutions of that carrying a single point mass or spring were presented
explicitly in reference [6]. In other words, it is impractical to obtain the closed form
solution for the natural frequencies and mode shapes of a uniform beam with more than
two point masses or springs by using the theory of reference [6], particularly when the
mode number is greater than 2. These difficulties in applying the theory of reference [6]
may be overcome by using the ANC method introduced here.
For any kind of boundary conditions, the natural frequencies and mode shapes of a
uniform beam with or without point masses or springs can be determined by using
numerical approaches such as the transfer matrix method (TMM) or finite element method
(FEM). Since any continuous system must be replaced by an equivalent discrete systems,
and then such numerical methods can be applied, discrete concentrated masses or springs
can be easily accommodated. In the TMM [lo] a concentrated mass is considered as a
part of the mass of the station closest to the concentrated mass, and in the FEM a
concentrated mass is distributed to the diagonal elements of the finite element mass matrix
[ill.
In this paper the analytical process is first used to formulate the problem of a uniform
cantilever beam carrying any number of concentrated masses, and then a purely numerical
method is employed to solve the eigenvalue problem. Because there is no necessity to
apply either TMM or FEM for deriving the eigenvalue equation some computer time
may be saved. Also, since the eigenvalue problem is solved numerically, tedious analytical
derivations for arriving at explicit mathematical expressions such as those given in
reference [6] are avoided.
In order to confirm the reliability of the theory, most of the results obtained with the
ANC method are checked against exact solutions or results obtained from TMM and
good agreement is achieved. In addition, the influence of various factors on the natural
frequencies of the system is also studied.

2. THEORETICAL ANALYSIS
For a uniform beam carrying p concentrated masses (Figure l), the equation of motion
with shear and rotatory effects ignored is [6]
a4w(x,t) + ~ a2wb,t) +i a2w(x,t)
EI mj 6(x-x,)=0, (1)
iJX4 at2 j=l at”
where E is the Young’s modulus, I is the moment of inertia of the cross-sectional area
of the beam, r3 is the beam mass per unit length, mj is the magnitude of the jth concentrated

Figure 1. A uniform cantilever beam carrying p concentrated masses.


BEAM WITH POINT MASSES 203

mass, w(x, I) is the vertical deflection of the beam at location x and time r, and F( x - x, )
is the Dirac delta function.
Equation (1) can be rewritten in the form

EZw”“(X, t)+fiG(x, t) = - $ m,W(x, t) &(x-x,), (2)


j=l

in which the primes (‘) and the overdots (‘) denote the derivatives of M’with respect to
the co-ordinate x and time t, respectively.
Equation (2) is equivalent to the equation of motion of a forced vibrating system with
the inertia1 forces of the p concentrated masses as the excitation forces. According to the
expansion theorem [2] or the mode superposition methodology [3] one may assume that

4x9 r) = i @,(X)%(0, (3)


,=I

where W,(x) are the normal mode shapes of the uniform beam without any concentrated
masses and n,(t) are the generalized coordinates.
Substituting equation (3) into equation (2), multiplying the resulting equation by M’,(x ),
and then integrating over the beam length, one obtains

0’,j, &(x)EZ$“ (x)~,(r) dx+ 1’ i #,(x)fi~tr(x){s( r) dx


I 0 s=t

=- d jg, mj*r(x) .i, %(x)M) 8(x -xi) dx, r=l , . . . , n. (4)


I
By virtue of the orthogonality of the normal mode shapes, one has
I I
$,(x)EZ$“(x) dx =O, ~,(x)mw,(x) dx =O, for r # s. (5)
I0 I0
Upon using the last two relationships and the definition of the Dirac delta function,
equation (4) reduces to

M.,M) + ZLn,(t) = &, r=l,...,n, (6)


where the M,, and K,, are the generalized masses and stiffnesses,
I I
Mrr = ~,(x)rii~,(x) dx, K,, = ~~(x)EZ$“(x) dx, 17a, b)
i0
and N,, are the generalized forces,

N,,= - i mj*r(x,) i ;ir(r)%:(x,). (7c)


j=l C=l

Note that with respect to the concentrated masses mj (j = 1,. . , p) the normal mode
shapes are not orthogonal to each other.
For the mode shapes normalized with respect to the beam mass per unit length, ti,
one has M,, = 1 .O, and equation (6) is reduced to

W)+&,(r) = N,,(r), (8)


where

w, =JK7; (9)
is the rth natural frequency of a uniform beam “without” concentrated masses.
204 J.-S. WU AND T.-L. LIN

For harmonic free vibration one has

nr( t) = 75, eiw’, r= 1,. . . , n, (10)

where W denotes the natural frequency of the uniform beam “with” concentrated masses.
Substituting equation (10) into equation (7~) and then placing the result into equation
(6) gives

(&cj’)?j- f i mjw,(x;)iqxj)w*?y~=o, r=l,...,n. (11)


,=I 7=,

By introducing the relationships

{@(x)1 = i%(x) *2(x) . . . %(x)1, {$=I% f2 . . . %I, (12313)

[B’l=$, mj{*(xj)Hw(xj)IT

(14-16)

equation (11) may be rewritten in the matrix form

C~‘JH~>- u1+w1)~*b71= 0, (17)

or

where

[B]=[Z]+[B’]. (19)

Equation (18) is a standard eigenvalue problem, from which one obtains the frequency
equation

A(W2) = I~(w/~)‘]-[l?]1 =O. (20)

The symbol { } in equations (12), (13), (16), (17) and (18) denotes a column matrix.
The determination of the natural frequencies w, (r = 1,. . . , n) of a uniform beam
“without” concentrated masses is given in the Appendix. Based on this, the natural
frequencies f& (s = 1, . . . , n) of the uniform beam “with” concentrated masses may be
obtained from equation (20). The substitution of the values of W, into equation (18) leads
to the determination of the generalized co-ordinate vector { ?j}(“) and, referring to equation
(3), one obtains the approximate mode shapes of the uniform beam “with” concentrated
masses as

w,(x) % i b3,(X)ij8 = {~(x)}‘{?j}“‘, r=l,...,n, (21)


C=l

or

W,(S) = {*(5)IT{+j)“‘, r=l,...,n, (21)’


BEAM WITH POINT MASSES 205

where 5 = x/ 1. The derivation of the normal mode shapes I?‘,(5) (s = 1, . . , n ) is desired


in the Appendix.
Instead of the analytical method of reference [6], the numerical half-interval method
of reference [12] was used to find the natural frequencies 0, (r = 1,2,. . .) and then the
Gauss elimination method [12] was applied to the determination of the generalized
co-ordinates ii,. (r = 1,2, . . .); the substitution of the latter into equation (21)’ will give
the required mode shapes w,( 5) (r = 1,2, . . .).

3. NUMERICAL ANALYSIS RESULTS AND DISCUSSIONS


The theory presented in the previous section is available for a uniform beam with any
kind of (supporting) boundary conditions. In this section the theory is applied to the
determination of the several lowest natural frequencies and mode shapes of a uniform
clamped-free (cantilever) beam carrying any number of concentrated masses (see Figure
1). The principal physical properties of the cantilever beam are diameter d = 2 in, length
I = 40 in, mass per unit length fi = 0.8891 lbm/in, and Young’s modulus E = 30 x lo6 psi.

3.1. RELIABILITY OF THE THEORY

In order to check the reliability of the theory, the uniform cantilever beam carrying a
single concentrated mass at its free end (see Figure 2) was first investigated. The magnitude
of the concentrated mass was assumed to be WI, = fil = 35564 lbm.
In Table 1 are shown the five lowest natural frequencies W, (i = 1,. . . ,5) obtained by
three different methods, of which those obtained from reference [4] are the exact solutions.
From the table one sees that the values of 0; (i = 1, . . . ,5) obtained by the ANC method
are very close to the exact solutions, particularly those of W, and LQ. Of course, the errors
become larger when the mode number (i) increases. For the three lowest modes, the
values of Gi (i = 1, . . . ,3) obtained from the transfer matrix method (TMM) are also in
close agreement with the exact solutions.

iii,d,E

5
Figure 2. A uniform cantilever beam with a single concentrated
fi = 0.8891 Ibm/in, d = 2 in, E = 30 x 1Ohpsi, I = 40 in, m, = 35.564
mass rn< at the free end. Numerical
Ibm.
values:

TABLE 1

The jive low est natural frequencies W i (i = 1, . . . , 5) of a uniform cantilever beam carry ing
a concentrated mass (m,) at the free end

Natural frequencies (rad/s)

Present paper 5.01108 52.351360 164.10220 342.16890 590.66750


Reference [4] 5.010513 52.283703 163.754434 338.469328 576.669025
Transfer matrix 5.01100 52.200700 163.45510 - -
206 J.-S. WU AND T.-L. LIN

In Table 2 are shown the three mode shapes, Wi([) (i = 1,3,5), corresponding to the
three natural frequencies &ii(i = 1,3,5) of Table 1; each mode shape was normalized so
that the maximum modal displacement is equal to 1*O.The numbers in column A are the
exact mode shape values obtained from reference [4], and those in column B are the
approximate ones obtained by the present method. From Table 2 one sees that the results
of the present paper are also close to the exact ones, particularly the first and the third
mode shapes.
From these results, it is evident that for the several lowest modes the accuracy of the
method introduced in this paper is reasonable.

3.2. INFLUENCE OF M ODE NUM BER

In theory, equation (3) is satisfied only when the mode number (n) approaches infinity
(i.e., n + w). However, in practice equation (3) will give satisfactory results with a small
value of n. In order to realize the effect of the mode number (n) on the accuracy of the
natural frequencies Wi, four cases, each with n = 2,3,4, 5, respectively, were studied. The
numerical analysis results are shown in Table 3; the numbers in the parentheses denote
the percentage errors of the approximate value of Gi comparing with the exact solutions.
From Table 3 one sees that the larger the number of modes (n) the smaller the errors
of the approximate natural frequencies c;ii, which agrees with the theoretical prediction.
It is also seen that the error of the ith natural frequency 6i will be small enough so long
as the number of modes is greater than i + 1. Therefore, it is not necessary to take a very
large number of modes in practice.

3.3. INFLUENCE OF M ASS RATIO (m,/?id)

For a single-degree-of-freedom system the natural frequency is given by w = J-kTm,


o CCl/G, but for a multi-degree-of-freedom system the natural frequencies oi
;E&,... n) are not proportional to I/&, where m, denotes a certain concentrated
mass in the system.
If m, is located at the free end of the cantilever beam and 511 is the total mass of the
beam, then when the value of Erl is kept constant and the mass ratio (m,/r?i1) changes

TABLE 2

The jirst, third and jifth mode shapes corresponding to 6, , W3 and W 5 of Table 1

Mode shapes

w,(5) w,(5) w(5)


Co-ordinate r I Y 3 r .
5 At Bt A B A B

0 (fixed) 0.0 0.0 0.0 0.0 0.0 0.0


0.1 0~0150 0.0149 0.2680 0.2589 0.7620 0.7128
0.2 0.0575 0.0575 0.7359 0.7245 1.0 1.0
0.3 0.1243 0.1243 1.0 1.0 -0.1793 -0.0767
0.4 0.2121 0.2121 0.8635 0.8681 -0.9881 -0.9890
0.5 0.3177 0.3176 0.3501 0.3528 -0.1927 -0.3660
0.6 0.4376 0.4375 -0.3298 -0.3136 0.8660 0.8240
0.7 0.5687 0.5687 -0.8464 -0.8141 0.5257 0.7465
0.8 0.7082 0.7080 -0.9156 -0.9056 -0.5690 -0.4779
0.9 0.8531 0.8527 -0.5098 -0.5381 -0.7368 -0.9560
1.0 (free) 1.0 1.0 0.1242 0.1234 0.0562 o-0774

t A, reference [4]; B, present method.


BEAM WITH POINT MASSES 207

TABLE 3

Influence of mode number on the natural frequencies W , (i = 1, . . ,5)

Natural frequencies (rad/s)


Mode
number WI W? WI W? w.
__I__-
2 5.01803 53.25069 -
(0.15%) (1.85%)

3 5.01286 52.58507 167.53370 -


(0.047%) (0.576%) (2.308% )
4 5.01156 52.37130 165.23350 346.06920
(0.021%) (0.168%) (0.9030/o) (2.25%)

5 5.01108 52.35136 164.10220 342.16890 590.66750


(0.011%) (0.129%) (0.212%) (1.09%) (2.43%)

Exact solution [4] 5.010513 52.283703 163.75443 338.46933 576.66903

TABLE 4

Influence of mass ratio (m,/ fil) on the naturalfrequencies Gi (i = 1, . . ,5) w ith concentrated
mass (m,) at free end

Mass Natural frequencies (rad/s)


ratio
m,/FII Method WI 02 W) WLl w,

1 Ai 5*01108 52.35 1360 164.10220 342.16890 590.66750


(0~011%) (0.129%) (0.212%) (1.09%) (2.43% )
B-t 5.010513 52.28370 163.75443 338.46933 576.66903

3 A 3.098224 50.65255 162.65980 340.37030 587.71000


(0.014%) (0.155%) (0.527%) (1.161%) (2.28% )
B 3.097797 50.57744 161.8075 336.4652 574.6313

5 A 2.43574 50.27739 162.25960 339.995 10 587.3901


(0.014%) (0.155%) (0.536%) (1.175%) (2.295% )
B 2.435402 50.19965 161.3944 336.0478 574.2112

7 A 2.071959 50.11084 162.0854 339.8326 587.2599


(0.013%) (0.154%) (0.540%) (1.181%) (2.305%)

B 2.07 1687 50.03361 161.21486 335.8672 574.0297

9 A 1.833996 50.02769 161.9882 339.7426 587.1850


(0.015%) (0.175%) (0.542%) (1.184%) (2.310%)

B 1.833716 49.94026 161.1144 335.7664 573.9286

t A, present method; B, reference [4].


208 J.-S. WU AND T.-L. LIN

1 3 5 7 9
Mass ratio, m,/riIl

Figure 3. The influence of mass ratio m,/riil on the natural frequencies W, (i = 1,. ,5).

160

70 -

40 -
zi,

IO-

I I I I
01
2 4 6 8 to
Station numbor al wtbich point moss m, ia loc0Md

Figure 4. The influence of the location of the concentrated mass m, on the natural frequencies Gi (i = 1,2,3).
BEAM W ITH POINT M ASSES 209

TABLE 5

Influence of location of concentrated mass m, on the natural frequencies (5,


(i=1,2,3) with m,./rirl=l~O

Natural frequencies (rad/s)


Location
of m, Method 0, O? 0,

Station At 11.219830 58.89922 131.87690


No. 2 (0.000% ) (1.073%) (2.380%)
ct 11.21983 58.27417 128.81090

Station A lo.237700 43.82487 171.27850


No. 4 (0.000% ) (1.037%) (3.617% 1
c IO.23767 43.37487 165.29970

Station A 8.274597 53.251670 175.69660


No. 6 (0.000% ) (1.843%) (1.258% )
C 8.27460 52.28824 173.51410

Station A 6.417819 70.665 10 178.13100


No. 8 (0.019%) (1.434%) (4.418%)
c 6.41657 69.66627 170.5935

Station A 5.01108 52.351360 164.10220


No. 10 (0~000%) (0.289%) (0.396% )
C 5.01100 52.20070 163.45510

t A, present method; C, transfer matrix method.

from 1 to 9, the variations of Wj (i = 1,. . . , 5) are shown in Table 4 and Figure 3. Although
thevaluesofC&(i=2,..., 5) decrease slightly in the range of m,/+il= 1-3, the variations
of 6; (i = 2,. . . ,5) are, however, negligible in the range of m,/riil= 3-9.
The numbers in the parentheses of Table 4 show the percentage errors between the
natural frequencies (sj (i = 1, . . . , 5) obtained by the present method (listed in rows A)
and those obtained from reference [4]. It is seen that the mass ratio (m,/tiiZ) does not
affect the accuracy of the present method.

3.4. INFLUENCE OF LOCATION OF THE CONCENTRATED M ASS (In,)

In Table 5 and Figure 4 is shown the influence of the location of a single concentrated
mass (m,) on the three lowest natural frequencies (si (i = 1,2,3). It is seen that the
fundamental natural frequency Oi decreases gradually as the concentrated mass m, is
moved from station No. 2 (near the fixed end) to station No. 10 (at the free end), but
this trend does not apply to the second and third natural frequencies (sz and 6,. The
reason for this is as follows.
For the mode shape corresponding to the first natural frequency (5,, the spring constant
of the cantilever beam is given by k = 3EZ/13; thus the larger the effective length 1 the
smaller the spring constant k and the first natural frequency 6,. The effective length 1
increases when the concentrated mass m, is moved from the fixed end to the free end of
the cantilever beam. For the mode shapes corresponding to the second and third natural
frequencies O2 and G3, however, there are one and two nodes, respectively, between the
fixed end and the free end of the cantilever beam (see Figure Al in the Appendix). Thus
210 J.-S. WU AND T.-L. LIN

the relationship k = 3 EI/ l3 is no longer valid. Of course, no definite trend can be obtained
from the two curves for (3* and W3 as shown in Figure 4.
Allthevaluesof&(i=l,..., 5) shown in Table 1 are for the case of the concentrated
mass m, at the free end of the cantilever beam. For cases in which m, is at other locations
on the cantilever beam, exact values of Wi (i = 1, . . . , n) are not available. Since the values
of (3, (i = 1,2,3) obtained from the transfer matrix method (TMM) are very close to the
exact solutions of reference [4], as one may see from Table 1, the values of Oi (i = 1,2,3)
obtained by using the TMM was used as the basis of the comparisons shown in Table
5. The numbers shown in the parentheses of Table 5 represent the percentage errors
between the values of Wi (i = 1,2,3) obtained by the present method (listed in rows A)
and those obtained by the TMM (listed in rows C). As in the case of the effect of the
mass ratio (m,/fi,I), the locations of the concentrated mass m, do not affect the accuracy
of the present method.

3.5. INFLUENCE OF DISTRIBUTION OF POINT MASSES

With the magnitude of each point mass given by m,i = fil/ii, where ii denotes the total
number of point masses, and each point mass occupies one station, the influence of the
distribution of point masses on the natural frequencies of the cantilever beam is shown
in Table 6 and Figure 5. Note that A = 1 represents a single point mass located at station

TABLE 6

Influence of distribution of point masses on the natural frequencies Wi (i = 1,2,3,4) with


ErnCj = A= constant?

Station Nos. Natural frequencies (rad/s)


at which point
masses located Method (5, w2 (33 “4

t:ee end) AS
cs 5.01108
5.01100 52.35136
52.20070 164.10220
163.45510 342.16890
-

10,9 A 5.30877 54.25649 155.05690 303.7316


C 5.30278 53.98964 150.8027 -

10, 9, 8 A 5.618676 53.05508 141.1548 299.3245


C 5.61209 52.73116 139.35790 -

10, 9, 8, 7 A 5.936067 49.99236 147.5599 301.6688


C 5.94071 49.68168 140.4133 -

10, 9, 8, 7, 6 A 6.254786 47.09436 146.2855 275.4702


C 6.24727 46.77333 144.9746 -

10, 9, 8, 7, 6, 5 A 6.567133 45.26547 143.8442 274.9934


C 6.56297 44.99241 142.3547 -

10, 9, 8, 7, 6, 5, 4 A 6.864671 44.6668 136.1746 279.3388


C 6.85285 44.19888 134.62920 -

10, 9, 8, 7, 6, 5, 4, 3 A 7.13999 45.14380 130.6805 266.8400


C 7.12674 44.87067 129.26210 -

10, 9, 8, 7, 6, 5, 4, 3, 2 A 7.38860 46.30857 130.2567 257.9232


C 7.37392 46.01418 129.0938 -

10, 9, 8, 7, 6, 5, 4, 3, 2, 1 A 7.60987 47.65016 1334415 261.9077


C 7.59385 47.32901 132.0078 -

t m,, = fi// A, where 13 is the total number of point masses.


$ A, present method; C, transfer matrix method.
BEAM WITH POINT MASSES 211

-.-.

2’ ?? 1
-e-* -e-e-e
1 , 1 I L
1 2 3 4 5 6 7 8 9 1C
Total number of point mosses, 5
Figure 5. The influence of the distribution of point masses m,, (i = 1,. , ri) on the natural frequencies W,
(i = 1,2,3,4).

No. 10, fi = 2 represents two point masses distributed at station Nos. 10 and 9, fi = 3
represents three point masses at station Nos. 10, 9, and 8,. . . . As shown in Figure 2, the
free end and fixed end are designated as station No. 10 and No. 0, respectively.
In Table 6 the results obtained by the present method (A) are compared with those
obtained by the TMM (C), and it is seen that they are in good agreement. From Figure
5 one sees that when the single lumped mass at the free end is sequentially divided into
2,3,4,. .. point masses and distributed to station Nos. 10,9,8,7,. . . respectively, the
fundamental natural frequency W, of the cantilever beam increases gradually, but this is
not true for the other natural frequencies Ui (i = 2,3,4). This phenomenon may have
something to do with the mode shapes of a cantilever beam (see Figure A.l) as indicated
in section 3.4.

4. CONCLUSIONS
1. Usually, the five lowest natural frequencies wi (i = 1, . . . ,5) of a vibration system
are the most important ones. The analytical-and-numerical-combined method (the ANC
method) can determine the values of Wi (i = 1, . . . ,5) of a cantilever beam carrying any
number of concentrated masses with good accuracy.
2. The larger the number of modes (n) the more accurate is the ANC method. The
errors in the ith natural frequency (si will be small enough as long as the number of
modes used is greater than i + 1.
3. For a cantilever beam carrying a single concentrated mass m, at free end, the
variations of natural frequencies W, (i = 1, . . . ,5) are negligible in the mass ratio range
m,/ rid = 3-9.
212 J.-S. WU AND T.-L. LI?J

4. The fundamental natural frequency W, of a cantilever beam carrying a single


concentrated mass m, decreases when m, is moved from the fixed end to the free end of
the cantilever beam, but this trend does not apply to the second and third natural
frequencies (sz and W,.
5. If the total magnitude of all the point masses is kept constant, then when a single
lumped mass at free end of a cantilever beam is divided into a number of point masses
and distributed to a number of stations, from the free end towards the fixed end, the
fundamental natural frequency W, will increase with the total number of point masses,
but there is no particular trend for the other frequencies Wj (i = 2,3,4).

REFERENCES

1. S. P. TIMOSHENKO, D. H. YOUNG and W. WEAVER JR. 1974 Vibration Problems in Engineering


(fourth edition). New York: John Wiley.
2. L. MEIROVITCH 1967 Analytical Method in Vibrations. New York: MacMillan.
3. R. W. CLOUGH and J. PENZIEN 1975 Dynamics ofStructures. New York: McGraw-Hill.
4. P. A. A. LAURA, J. L. POMBO and E. A. SUSEMIHL 1974 Journal of Sound and Vibration 37,
161- 168. A note on the vibration of a clamped-free beam with a mass at the free end.
5. P. A. A. LAURA, M. J. MAURIZI and J. L. POMBO 1975 Journal of Sound and Vibration 41,
397- 405. A note on the dynamic analysis of an elastically restrained-free beam with a mass at
the free end.
6. M. G~RG~~ZE 1984 Journal of Sound and Vibration 96, 461- 468. A note on the vibrations of
restrained beam and rods with point masses.
7. C. N. BAPAT and C. BAPAT 1987 Journal of Sound and Vibration 112, 177- 182. Natural
frequencies of a beam with nonclassical boundary conditions and concentrated masses.
8. P. A. A. LAURA, C. P. FILIPICH and V. H. CORTINEZ 1987 Journal of Sound and Vibration
117, 459-465. Vibrations of beams and plates carrying concentrated masses.
9. C. S. KIM and S. M. DICKINSON 1988 Journal of Sound and Vibration 122, 441- 455. On the
analysis of laterally vibrating slender beams subject to various complicating effects.
10. F. LECKIE and E. PESTEL 1960 ZnternationalJournalofM echanicalSciences 2,137- 167. Transfer
matrix fundamentals.
11. J. S. PRZEMIENIECKI 1968 Theory of M atrix Structural Analysis. New York: McGraw-Hill.
12. B. CARNAHAN, H. A. LUTHER and J. 0. WILKER 1969 Applied Numerical Methods, New
York: John Wiley.

APPENDIX: CLOSED FORM SOLUTION FOR THE NATURAL FREQUENCIES AND


MODE SHAPES OF A UNIFORM CANTILEVER BEAM
With the effects of shear deformation and rotatory inertia neglected, the equation of
motion of a uniform beam with no external loads is [l-3]
IV”(X) -p%(x) = 0, (Al)

where

p4 = fitw*/( EZ). (AZ)

From equation (Al) one obtains the natural frequencies OJ, and the corresponding
normal mode shapes #,,(&) as
CO,= (P,l)‘JEZ/ rit14, r= 1,2,3, . . . , (A3)

G,(5) = (l/m)[(cos I%&- cash pJ&) + D,(sinh &I&- sin &&)], (A4)
where
0, = (cos &l+ cash p,I)/(sin &l+ sinh &l), 5=x1& (A% A6)
BEAM WITH POINT MASSES 213

and the values of&l (called the frequency coefficients) are determined from the frequency
equation

1 + cos PI1 cash &l = 0, r=1,2,... . 1.47)

Based on these expressions, the lowest five natural frequencies and the corresponding
normal mode shapes of a uniform cantilever beam may be determined. The results are
shown in Table Al and Figure Al, the latter being plotted from the data of Table Al.
Conversion factors for the English units used in this paper are as follows: 1 ft = 0.3048 m;
1 in = 25.4 mm; 1 lbm = 0.45 kg; 1 lbf = 4.45 N; 1 psi = 6894.76 N/m’.

TABLE Al

Coejicients for the low est jive natural frequencies and normal mode shapes sf‘a uniform
cantilever beam

‘r Natural frequency coefficients PJ = \lw,./a

1.875097 4.694087 7.854762 IO.99550


Normal mode shape coefficients hf, (5)
Co-ordinate
5(=x/O $ b%(5) b%(t) h(S) b*,,(5)
0.0 0~000000 0~000000 0~000000 0~000000 0~000000
0.1 0.033547 0.185258 0.456138 0.770017 1.074493
0.2 0.127741 0.602109 1.20912 1.507581 1.312949
0.3 0.272964 1.052265 1.512478 0.867737 -0.422575
0.4 0.459766 1.366938 1.051845 -0.63 1120 - 1.393088
0.5 0.679042 1.427332 0.039370 -1.414237 0.001692
0.6 0.922264 1.178955 -0.947538 -0.652987 1.400254
0.7 1.181746 0.634108 -1.314869 0.79488 1 0.451216
0.8 1.450947 -0.140064 -0.789746 1.286106 -1.201698
0.9 1.724790 - 1.047493 0.456963 0.104557 -0.591270
1.0 1.999990 - 1.999992 1.999971 - 1.998376 1.986626
,-
L
1 a=JEI/(td).
tb=m.

0
/ I 1 I 1 1 I J
0.1 02 0.3 0.4 0.5 O-6 07 00 09 10
(=x/l

Figure Al. The five lowest mode shapes of a uniform cantilever beam (b = J’?d). Mode numbers: -. I;
---. 2; --( 3; ---, 4;. ,5.

You might also like