You are on page 1of 12

INTERNATIONAL JOURNAL FOR NUMERICAL METHODS IN ENGINEERING

Int. J. Numer. Meth. Engng 2003; 57:1603–1614 (DOI: 10.1002/nme.736)

Dynamic stiness vibration analysis using


a high-order beam model

Moshe Eisenberger∗; †
Faculty of Civil Engineering; Technion – Israel Institute of Technology; Technion City 32000; Israel

SUMMARY
This work presents the derivation of the exact dynamic stiness matrix for a high-order beam element.
The terms are found directly from the solutions of the dierential equations that describe the defor-
mations of the cross-section according to the high-order theory, which include cubic variation of the
axial displacements over the cross-section of the beam. The model has six degrees of freedom at the
two ends, one transverse displacement and two rotations, and the end forces are a shear force and two
end moments. Using the dynamic stiness matrix exact vibration frequencies for beams with various
combinations of boundary conditions are tabulated and compared with results from the Bernoulli–Euler
and Timoshenko beam models. Copyright ? 2003 John Wiley & Sons, Ltd.

KEY WORDS: high-order beam theory; vibration; natural frequency; dynamic stiness

1. INTRODUCTION

The analysis of beams has been performed over the years mostly using the Bernoulli–Euler
beam theory. This theory is based on the assumption that plain sections of the cross-section
remain plain and perpendicular to the beam axis. The cross-sectional displacements are shown
is Figure 1, and expressed as

@W (x; t)
U (x; z; t) = U0 (x; t) + z (1)
@x
W (x; z; t) = W0 (x; t) (2)

where U0 (x; t) and W0 (x; t) are the axial and lateral displacements of the beam neutral axis,
z is the distance from the beam neutral axis.

∗ Correspondence to: Moshe Eisenberger, Faculty of Civil Engineering, Technion – Israel Institute of Technology,
Technion City 32000, Israel.
† E-mail: cvrmosh@tx.technion.ac.il

Contract=grant sponsor: Fund for the Promotion of Research


Received 19 November 2001
Revised 14 August 2002
Copyright ? 2003 John Wiley & Sons, Ltd. Accepted 10 October 2002
1604 M. EISENBERGER

Figure 1. Cross-section displacements in dierent beam theories: (a) Bernoulli–Euler beam


theory; (b) Timoshenko beam theory; and (c) high-order beam theory.

A more rened beam theory is the Timoshenko beam theory which relaxes the restriction
on the angle of shearing deformations that exist in the simpler theory. The equations in this
case are

U (x; z; t) = U0 (x; t) + z(x; t) (3)


W (x; z; t) = W0 (x; t) (4)

where (x; t) represents the rotation of a normal to the axis of the beam. Han et al. [1] pre-
sented a comprehensive comparative study of four beam theories (Bernoulli–Euler, Rayleigh,
Shear and Timoshenko).
In recent years, due to the increase in the use of composite materials, it was found that the
Timoshenko shear deformation theory has some major numerical problems such as locking in
the numerical analysis. Another problem was the need to supply an articially derived shear
correction factor. Although some remedies were devised, as a result, several higher-order
theories have emerged.
These theories, with small variations, are due to Levinson [2], Bickford [3], Heyliger and
Reddy [4], Sheinman, Eisenberger and Bernstein [5], and others all relax the restriction on
the warping of the cross-section, and allow variation in the longitudinal direction of the beam
which is cubic. This results, for homogeneous materials and rectangular cross-section with

Copyright ? 2003 John Wiley & Sons, Ltd. Int. J. Numer. Meth. Engng 2003; 57:1603–1614
DYNAMIC STIFFNESS VIBRATION HIGH-ORDER BEAM MODEL 1605

width b and height h in the following set of equations:


  
4  z 2 @W (x; t)
U (x; z; t) = U0 (x; t) + z (x; t) − (x; t) + (5)
3 h @x
W (x; z; t) = W0 (x; t) (6)

The Bernoulli–Euler beam theory does not consider the shear stresses in the cross-section
and the associated strains. Thus, the shear angle is taken as zero through the height of the
cross-section. The Timoshenko beam theory assumes constant shear stress and shear strain
in the cross-section. On the top and bottom edges of the beam the free surface condition
is thus violated. The use of a shear correction factor, in various forms including the eect
of Poisson’s ratio, does not correct this fault of the theory, but rather articially adjusts the
solutions to match the static or dynamic behaviour of the beam. This topic was addressed by
Han et al. [1] and by Soldatos and Sophocleous [6] and the many references they cite. The
high-order theory remedies this physical mismatch at the free edges by assuming variable
shear strain and shear stress along the height of the cross-section. Then there is no need
for the shear correction factor. The theory is more exact and represents much better the
physics of the problem. It results in a sixth-order theory compared to the fourth order of the
other less-accurate theories. This yields a six-degree-of-freedom element with six end forces,
a shear force, bending moment and a high-order bending moment, at the two ends of the beam
element. A complete derivation of the static stiness matrix and the resulting end forces is
given in Reference [7].
In this paper, the vibration frequencies of a beam modelled using the higher-order theory,
with rectangular cross-section, are found using the dynamic stiness approach. The results are
compared with results from other methods for several depth to length ratios, and several types
of boundary conditions.

2. DIFFERENTIAL EQUATIONS OF MOTION

Using this displacement eld in Equations (5) and (6) we arrive at the equations of motion
for a beam segment (linearized from those in Reference [4]) as
 
8GA @(x; t) @2 W (x; t) EI @4 W (x; t) 16EI @3 (x; t)
+ − +
15 @x @x2 21 @x 4 105 @x3

@2 W (x; t) I @4 W (x; t) 16I @3 (x; t)


= A − + (7)
@t 2 21 @x2 @t 2 105 @x@t 2
 
68EI @2 (x; t) 16EI @3 W (x; t) 8GA @W (x; t)
− + + (x; t) +
105 @x2 105 @x3 15 @x

68I @2 (x; t) 16I @3 W (x; t)


=− + (8)
105 @t 2 105 @x@t 2

Copyright ? 2003 John Wiley & Sons, Ltd. Int. J. Numer. Meth. Engng 2003; 57:1603–1614
1606 M. EISENBERGER

Assuming that the motion is harmonic we substitute for W (x; t) and (x; t) the following:

W (x; t) = w(x) sin(!t) (9)


(x; t) = (x) sin(!t) (10)

and obtain a system of two coupled ordinary equations as


 
8GA d(x) d 2 w(x) EI d 4 w(x) 16EI d 3 (x)
+ − +
15 dx dx2 21 dx4 105 dx3
I!2 d 2 w(x) 16I!2 d(x)
= −A!2 w(x) + − (11)
21 dx2 105 dx
 
68EI d 2 (x) 16EI d 3 w(x) 8GA dw(x)
− + + (x) +
105 dx2 105 dx3 15 dx
68I!2 16I!2 dw(x)
= (x) − (12)
105 105 dx
We assume that the solution is

w(x) = W esx (13)


(x) = Pesx (14)

and substitute it into the equations of motion resulting in


8
15 PsGA + 16
105 PsI!2 + 8
15 Ws2 GA − 1
21 Ws2 I!2

− 21
1
EIWs4 + 16
105 EIPs3 + A!2 W = 0 (15)

− 105
68
EIPs2 + 16
105 EIWs3 + 8
15 PGA − 68
105 PI!2

+ 158 WsGA + 16
105 WsI!2 = 0 (16)

and written in matrix form for the two unknowns P and W


 
2
15 sGA + 105 sI! + 105 EIs 15 s GA − 21 s I! − 21 EIs + A!
2 3 8 2 1 2 2 4
8 16 16 1
P
  =0 (17)
− 105
68
EIs2 + 15
8
GA − 105
68
I!2 8
15 sGA + 16
105 sI! 2
+ 16
105 EIs 3 W

and the non-trivial solution will be when the determinant of the coecient matrix will be
zero, i.e.
4 2 2 6 8 
− 525 E I s + 15 EIGA − 525 8
EI 2 !2 s4
4 2 2 4 
+ − 525  I ! + 158 GAI!2 + 105
68
AEI!2 s2
8 
− 15 GA2 !2 + 105
68
AI2 !4 = 0 (18)

Copyright ? 2003 John Wiley & Sons, Ltd. Int. J. Numer. Meth. Engng 2003; 57:1603–1614
DYNAMIC STIFFNESS VIBRATION HIGH-ORDER BEAM MODEL 1607

Thus, we have a sixth-order equation with the unknowns, resulting in six values and the
general solution will be

w(x) = W1 es1x + W2 es2x + W3 es3x + W4 es4x + W5 es5x + W6 es6x (19)


(x) = P1 es1x + P2 es2x + P3 es3x + P4 es4x + P5 es5x + P6 es6x (20)

The 12 constants, W1 ::: 6 and P1 ::: 6 will be found from Equation (17) and the boundary
conditions.

3. DYNAMIC STIFFNESS MATRIX DERIVATION

The terms in the stiness matrix are the holding actions at the end of the beam element
when the beam is deformed as to have unit displacement in a desired direction and all other
degrees of freedom are restrained. This matrix will be frequency dependent, and thus termed
the dynamic stiness matrix. In the current derivation the degrees of freedom are the transverse
displacement, w(x), its derivative, w (x), and the rotation of the normal (x), at the two ends.
Then the holding end actions will be shear forces, V , bending moment M and a higher-order
moment Mh , at the two ends. The explicit formulas for these are
8 d(x) 1 d 3 w(x)
V (x) = − GA + EI
15 dx 21 dx3
16 d 2 (x) 1 dw(x) 16
− EI 2
− I!2 + I!2 (x) (21)
105 dx 21 dx 105
−1 d 2 w(x) 16 d(x)
M (x) = EI + EI (22)
21 dx2 105 dx
16 d 2 w(x) 68 d(x)
Mh (x) = EI − EI (23)
105 dx2 105 dx

The terms in the stiness matrix for this beam model are found as the forces and moments at
the two ends of the beam (x = 0 and L) when the following six sets of boundary conditions
are enforced:
1. w(0) = 1, w (0) = w(1) = w (1) = (0) = (1) = 0.
2. w (0) = 1, w(0) = w(1) = w (1) = (0) = (1) = 0.
3. w(1) = 1, w(0) = w (0) = w (1) = (0) = (1) = 0.
4. w (1) = 1, w(0) = w (0) = w(1) = (0) = (1) = 0.
5. (0) = 1, w(0) = w (0) = w(1) = w (1) = (1) = 0.
6. (1) = 1, w(0) = w (0) = w(1) = w (1) = (0) = 0.
This will result in a symmetric six by six stiness matrix. Then, the structure dynamic
stiness matrix is assembled as in the nite-element method for the unrestrained degrees of
freedom of the structure.
The natural frequencies are the values of the frequency that will cause the assembled
structure dynamic stiness matrix to become singular. A search procedure is employed to nd

Copyright ? 2003 John Wiley & Sons, Ltd. Int. J. Numer. Meth. Engng 2003; 57:1603–1614
1608 M. EISENBERGER

these values up to the desired accuracy. The Wittrik–Williams [8] algorithm is employed to
ensure that none of the natural frequencies is missed.

4. EXAMPLES

As the rst example a simply supported beam is analysed. The following numerical values
are used: h = 2, b = 1:5, A = 3,  = 1, E = 1,  = 0:3. The natural frequencies are found for
dierent length to height ratios. In Table I the natural frequencies of a beam with length
L = 10 are given along with the mode shape of vibration. Results are given for the following
four models: Bernoulli–Euler model, Timoshenko model with two values for the k, and the
current high-order model. Many values for the shear correction factor k were suggested, but
here we use the original value suggested by Timoshenko k = 56 , and k = 14 17 [6] that yields the
same results for the pure shear mode as in the high-order theory (shown later in Tables I–III).
Also these are only used for comparison, considering that these are the result of an inferior
theory.
For the Bernoulli–Euler model the frequencies of the shear mode and the second frequency
spectrum of vibration are not available, as the model is not capable of modelling these. Also,
the frequencies obtained from the Bernoulli–Euler model in this case have large errors as the
shear deformations in this case are highly important. The pure shear mode is the same for
the high-order model and the Timoshenko beam model with k = 14 17 , as was shown also in
the work of Soldatos and Sophocleous [6]. In Figures 2 and 3 the mode shapes of vibration
are presented: from these one can easily identify the pure shear mode (mode 4), and the

Table I. Results for the example of a simply supported beam, L = 10.


Frequency Frequency High-order Timoshenko Timoshenko Bernoulli–Euler
number type + number theory theory k = 5=6 theory k = 14=17 theory

1 I-1 0.08390345 0.08387357 0.08376898 0.09869604


2 I-2 0.25417282 0.25345880 0.25276220 0.39478417
3 I-3 0.44476668 0.44126571 0.43962366 0.88826439
4 Shear 0.56279843 0.56613851 0.56279843
5 I-4 0.64077598 0.63094941 0.62820951 1.57913670
6 II-1 0.66285019 0.66618873 0.66308526
7 I-5 0.84010074 0.81920191 0.81531074 2.46740110
8 II-2 0.87849296 0.88181008 0.87902348
9 I-6 1.04332385 1.00562427 1.00057073 3.55305758
10 II-3 1.13645975 1.13963235 1.13714037
11 I-7 1.25145569 1.19045301 1.18424241 4.83610615
12 II-4 1.41398946 1.37400937 1.36665236
13 I-8 1.46539838 1.41692833 1.41471221 6.31654681
14 I-9 1.68583168 1.55657206 1.70321500 7.99437956
15 II-5 1.70250579 1.70518498 1.90880548
16 I-10 1.91320957 1.73835915 1.99851462 9.86960440
17 II-6 1.99784259 1.91953620 2.08839093
18 I-11 2.14778545 2.00027266 2.26759319 11.9422213
19 II-7 2.29768002 2.10022823 2.44648393
20 I-12 2.38964351 2.28053016 2.60120479 14.2122303

Copyright ? 2003 John Wiley & Sons, Ltd. Int. J. Numer. Meth. Engng 2003; 57:1603–1614
DYNAMIC STIFFNESS VIBRATION HIGH-ORDER BEAM MODEL 1609

Table II. Results for the example of a simply supported beam, L = 5.


Frequency Frequency High-order Timoshenko Timoshenko
number type + number theory theory k = 5=6 theory k = 14=17

1 I-1 0.25417282 0.25345880 0.25276220


2 Shear 0.56279843 0.56613851 0.56279843
3 I-2 0.64077598 0.63094941 0.62820951
4 II-1 0.87849296 0.88181008 0.87902348
5 I-3 1.04332385 1.00562427 1.00057073
6 II-2 1.41398946 1.37400937 1.36665236
7 I-4 1.46539838 1.41692833 1.41471221
8 I-5 1.91320957 1.73835915 1.72874312
9 II-3 1.99784259 2.0002726 1.99851462
10 I-6 2.38964351 2.10022828 2.08839093

Table III. Results for the example of a simply supported beam, L = 2:5.
Frequency Frequency High-order Timoshenko Timoshenko
number type + number theory theory k = 5=6 theory k = 14=17

1 Shear 0.56279843 0.56613851 0.56279843


2 I-1 0.64077598 0.63094941 0.62820951
3 II-1 1.41398946 1.37400937 1.36665236
4 I-2 1.46539838 1.41692833 1.41471221
5 I-3 2.38964351 2.10022828 2.08839093
6 II-2 2.60062253 2.60263176 2.60120479
7 I-4 3.42730393 2.81974484 2.80354058
8 II-3 3.82960415 3.53620588 3.51569480
9 I-5 4.56267312 3.83105554 3.83004010
10 II-4 5.07173436 4.25105161 4.22626750

modes that belong to the second frequency spectrum, as for these the relation of the relative
magnitudes of the rotational component (dotted line) to the vertical displacement component
of the mode (solid line) is higher then the ratio for the rst spectrum modes. In Tables II
and III the results are presented for L = 5 and 2.5, respectively. In these cases, the values
from the simple beam theory are meaningless and are not given. The shear mode frequency is
independent of the beam length as it should be [9, 10]. The dierences are more prominent for
the higher frequencies as these involve a larger amount of warping restraint in the Timoshenko
beam models.
In Tables IV and V the results are presented for a clamped–clamped beam with length
L = 10 and 5, respectively. The dierences between the values of the Timoshenko beam
models with k = 5=6 and 14/17 are small. Here the dierences between the two theories are
more signicant as the clamped restraint at the two ends imposes a stiening eect on the
higher-order model.
Smaller dierences are observed in the results of a clamped–simply supported beam (Ta-
bles VI–VIII), and of a clamped–free beam (Tables IX and X). In general the results given
in all the Tables can be used to assess other approximate methods of analysis.

Copyright ? 2003 John Wiley & Sons, Ltd. Int. J. Numer. Meth. Engng 2003; 57:1603–1614
1610 M. EISENBERGER

.08390345 .25417282
4
4

3 2

2
0
2 4 6 8 10
1 x
-2
0
2 4 6 8 10
x
-4
-1

Mode 1 Mode 2

.44476668 .56279843
1
6
0.8
4

2 0.6

0
2 4 6 8 10 0.4
-2 x

-4 0.2

-6
0 2 4 6 8 10
x
Mode 3 Mode 4
.64077598 .66285019
1
0.8
20
0.6

10 0.4
0.2

0 0
2 4 6 8 10 2 4 6 8 10
x -0.2 x
-10 -0.4
-0.6
-20
-0.8
-1

Mode 5 Mode 6

14
.84010074 .87849296
1
12
10 0.8
8 0.6
6 0.4
4
2 0.2
0 0
-2 2 4 6 8 10 2 4 6 8 10
x -0.2 x
-4
-6 -0.4
-8 -0.6
-10
-0.8
-12
-14 -1

Mode 7 Mode 8

1.04332385 1.13645975
1
0.8
4
0.6

2 0.4
0.2
0 0
2 4 6 8 10 2 4 6 8 10
x -0.2 x
-2 -0.4
-0.6
-4
-0.8
-1

Mode 9 Mode 10

Figure 2. First 10 modes of a simply supported beam.

Copyright ? 2003 John Wiley & Sons, Ltd. Int. J. Numer. Meth. Engng 2003; 57:1603–1614
DYNAMIC STIFFNESS VIBRATION HIGH-ORDER BEAM MODEL 1611

1.25145569 1.41398946
1
3 0.8
2 0.6
0.4
1
0.2
0 0
2 4 6 8 10 2 4 6 8 10
x -0.2 x
-1
-0.4
-2 -0.6
-3 -0.8
-1
Mode 11 Mode 12

1.46539838 1.68583168
2
2
1
1

0 0
2 4 6 8 10 2 4 6 8 10
x x
-1 -1

-2
-2

Mode 13 Mode 14
1.70250579 1.91320957
1
0.8 1.
0.6 1
0.4
0.
0.2
0 0
2 4 6 8 10 2 4 6 8 10
-0.2 x x
-0.
-0.4
-0.6 -1
-0.8 -1.
-1
Mode 15 Mode 16
1.99784259 2.14778545
1 1.
0.8
0.6 1
0.4
0.
0.2
0 0
2 4 6 8 10 2 4 6 8 10
-0.2 x x
-0.
-0.4
-0.6 -1
-0.8
-1 -1.

Mode 17 Mode 18
2.29768002 2.38964351
1 1.4
1.2
0.8 1
0.6 0.8
0.4 0.6
0.4
0.2 0.2
0 0
2 4 6 8 10 -0.2 2 4 6 8 10
-0.2 x x
-0.4
-0.4 -0.6
-0.6 -0.8
-1
-0.8 -1.2
-1 -1.4

Mode 19 Mode 20

Figure 3. The next 10 modes of a simply supported beam.

Copyright ? 2003 John Wiley & Sons, Ltd. Int. J. Numer. Meth. Engng 2003; 57:1603–1614
1612 M. EISENBERGER

Table IV. Results for the example of a clamped–clamped beam, L = 10.


Frequency High-order Timoshenko Timoshenko
number theory theory k = 5=6 theory k = 14=17

1 0.14150 0.13834 0.13786


2 0.29707 0.28517 0.28399
3 0.48061 0.45665 0.45469
4 0.63803 0.61862 0.61590
5 0.69862 0.68283 0.67957
6 0.84030 0.80412 0.80071
7 0.91416 0.89618 0.89286
8 1.06945 1.14379 1.14090
9 1.29677 1.18856 1.18254
10 1.41810 1.56101 1.55286

Table V. Results for the example of a clamped–clamped beam, L = 5.


Frequency High-order Timoshenko Timoshenko
number theory theory k = 5=6 theory k = 14=17

1 0.35351 0.32950 0.32785


2 0.68353 0.62264 0.61997
3 0.89717 0.89302 0.88987
4 1.12490 0.99965 0.99487
5 1.41765 1.37526 1.36801
6 1.58770 1.41412 1.41163
7 1.99451 1.98230 1.97956
8 2.07048 2.11744 2.10653
9 2.57873 2.45029 2.43738
10 2.59552 2.61095 2.60839

Table VI. Results for the example of a clamped–simply supported beam, L = 10.
Frequency High-order Timoshenko Timoshenko
number theory theory k = 5=6 theory k = 14=17

1 0.11207 0.11082 0.11054


2 0.27722 0.27114 0.27018
3 0.46211 0.44843 0.44662
4 0.58935 0.59203 0.58882
5 0.65646 0.63339 0.63057
6 0.76103 0.76247 0.75951
7 0.86040 0.82338 0.81951
8 0.99903 0.98901 0.98529
9 1.07026 1.02382 1.01982
10 1.25742 1.27695 1.27428

Copyright ? 2003 John Wiley & Sons, Ltd. Int. J. Numer. Meth. Engng 2003; 57:1603–1614
DYNAMIC STIFFNESS VIBRATION HIGH-ORDER BEAM MODEL 1613

Table VII. Results for the example of a clamped–simply supported beam, L = 5.


Frequency High-order Timoshenko Timoshenko
number theory theory k = 5=6 theory k = 14=17

1 0.29707 0.28517 0.28399


2 0.63803 0.61862 0.61590
3 0.69862 0.68283 0.67957
4 1.06945 0.99868 0.99399
5 1.15239 1.14379 1.14090
6 1.52353 1.69062 1.68696
7 1.70165 1.75270 1.74469
8 1.99064 2.09863 2.08703
9 2.29331 2.29923 2.29728
10 2.48389 2.46238 2.44854

Table VIII. Results for the example of a clamped–simply supported beam, L = 2:5
Frequency High-order Timoshenko Timoshenko
number theory theory k = 5=6 theory k = 14=17

1 0.68353 0.62264 0.61997


2 0.89717 0.89302 0.88987
3 1.58770 1.37526 1.36801
4 1.99451 1.98232 1.97956
5 2.57873 2.11744 2.10653
6 3.20737 3.21358 3.21208
7 3.67304 4.24834 4.22398
8 4.44540 4.45291 4.45156
9 4.85590 4.96526 5.63909
10 5.69196 5.66658 5.70112

Table IX. Results for the example of a clamped–free beam, L = 10.


Frequency High-order Timoshenko Timoshenko
number theory theory k = 5=6 theory k = 14=17

1 0.01976 0.01968 0.01968


2 0.10748 0.10602 0.10587
3 0.25766 0.25235 0.25179
4 0.43039 0.41896 0.41776
5 0.61509 0.59578 0.59380
6 0.80102 0.77315 0.77024
7 0.97002 0.94332 0.93929
8 1.02824 1.02889 1.02348
9 1.13220 1.12378 1.11856
10 1.21020 1.18689 1.18168

Copyright ? 2003 John Wiley & Sons, Ltd. Int. J. Numer. Meth. Engng 2003; 57:1603–1614
1614 M. EISENBERGER

Table X. Results for the example of a clamped–free beam, L = 5.


Frequency High-order Timoshenko Timoshenko
number theory theory k = 5=6 theory k = 14=17

1 0.07375 0.07269 0.07262


2 0.32004 0.30594 0.30507
3 0.69443 0.65451 0.65229
4 0.98308 0.95054 0.94633
5 1.19881 1.18556 1.17973
6 1.36527 1.32511 1.32021
7 1.59643 1.56951 1.56219
8 1.87475 1.77190 1.76549
9 2.00582 1.99224 1.98464
10 2.43207 2.19680 2.18680

5. SUMMARY AND CONCLUSIONS

The exact stiness terms for the high-order beam theory are given explicitly in this work
and can be readily used in computer programs. As was shown, the dierences between the
Timoshenko model and the current model for isotropic beams are for most cases insigni-
cant. For very short beams the dierences become more prominent, but beam theory is not
capable of accurately modelling beams in that range. However, to be on the safe side it is
recommended to use the higher-order theory for all cases.

ACKNOWLEDGEMENT
This research was supported by the Fund for the Promotion of Research at the Technion.

REFERENCES
1. Han SM, Benaroya H, Wei T. Dynamics of transversely vibrating beams using four engineering theories. Journal
of Sound and Vibration 1999; 225:935– 988.
2. Levinson M. A new rectangular beam theory. Journal of Sound and Vibration 1981; 74:81–87.
3. Bickford WB. A consistent higher order beam theory. Developments in Theoretical and Applied Mechanics
1982; 11:137–150.
4. Heyliger PR, Reddy JN. A higher order beam nite element for bending and vibration problems. Journal of
Sound and Vibration 1988; 126(2):309 –329.
5. Sheinman I, Eisenberger M, Bernstein Y. High order element for prebuckling and buckling analysis of laminated
plane frames. International Journal for Numerical Methods in Engineering 1996; 39:2155–2168.
6. Soldatos KP, Sophocleous C. On shear deformable beam theories: the frequency and normal modes of the
homogeneous orthotropic bickford beam. Journal of Sound and Vibration 2001; 242:215–245.
7. Eisenberger M. An exact high order beam element. Computers and Structures 2003; 81:147–152.
8. Wittrik WH, Williams FW. A general algorithm for computing natural frequencies of elastic structures. The
Quarterly Journal of Mechanics and Applied Mathematics 1971; 24:263–284.
9. Huang TC. The eect of rotary inertia and of shear deformation on the frequency and normal modes equations
of uniform beams with simple end conditions. Journal of Applied Mechanics (ASME) 1961; 28:579–584.
10. Downs B. Transverse vibration of a uniform, simply supported Timoshenko beam without transverse deection.
Journal of Applied Mechanics (ASME) 1976; 76:671–674.

Copyright ? 2003 John Wiley & Sons, Ltd. Int. J. Numer. Meth. Engng 2003; 57:1603–1614

You might also like