Beruflich Dokumente
Kultur Dokumente
sin
_
z
h
_
_
+
w
x
_
z
w
x
(1)
The strain components are
x
=
u
x
=
h
sin
_
z
h
_
_
x
+
2
w
x
2
_
z
2
w
x
2
xz
=
u
z
+
w
x
= cos
_
z
h
_
_
+
w
x
_ (2)
The stress-strain relationships are
x
= E
x
and xz = G
xz
(3)
where E is Youngs modulus, G=E/2(1 +) is the shear modu-
lus, and is Poissons ratio. Here this beam element is assumed
as isotropic.
The equilibriumequations for the beamelement are obtained
using Hamiltons principle [4] based on Eqs. (1) to (3), which
can be expressed as
_
h/2
h/2
_
b
0
_
L
0
(
x
x
+
xz
xz
)dxdydz
_
L
0
qwdx = 0
(4)
where b is the width of the beam and q is the distributed trans-
verse load acting on the beam.
Substituting Eqs. (1) to (3) into Eq. (4), we can have the
equilibrium equations for the beam as
__
2h
3
3
h
3
2
2
_
Eb
_
h
3
2
2
4h
3
3
+
h
3
12
_
Ebw
+
_
0.5Gbh
_
+w
__
+q = 0
__
2h
3
3
h
3
2
2
_
Ebw
h
3
2
2
Eb
+0.5Gbh
_
+w
_
= 0 (5)
Taking the beams cross-sectional area A = bh and the second
moment of the cross-sectional area I = bh
3
/12, Eq. (5) can be
simplied as
_
0.166EI
0.06EIw
+
_
0.5GA
_
+w
__
+q = 0
_
0.166EIw
0.648EI
+0.5GA
_
+w
_
= 0 (6)
where EI is the beams bending rigidity, GA is the beams shear
rigidity, and the prime denotes d/dx.
FromEq. (6), the generalized forces appropriate for this beam
element can be dened as
F
1
= 0.166EI
0.06EIw
+0.5GA
_
+w
_
F
2
= 0.06EIw
0.166EI
F
3
= 0.648EI
0.166EIw
(7)
where F
1
is a generalized shear force, while F
2
and F
3
are
generalized moments.
According to Petrolito [7], in order to fully determine the
boundary conditions for the beam, either quantity in the follow-
ing three pairs needs to be specied
w or F
1
;
w or F
2
;
or F
3
.
So far, the beam theory has been entirely presented, which
includes a sinusoidal displacement eld from Touratier theory
(Eq. (1)), two equilibrium equations (Eq. (6)), and three pairs
of associated boundary conditions at each end of the beam. In
the following sections, the equilibrium equations (Eq. (6)) are
solved to determine the lateral deection of the beam, w, and
the rotation of a normal to the axis of the beam, in a linear
bending problem.
3. GENERAL SOLUTIONS OF EQ. (6)
The coupled differential equations (Eq. (6)) can be solved
and the solutions include two parts: homogeneous solutions and
particular solutions. For the homogeneous solutions, assuming
there is no applied transverse load, q =0, Eq. (6) can be decou-
pled as
w
(6)
2
w
(5)
= 0 and
(5)
(3)
= 0 (8)
with
2
=
46GA
EI
=
276
h
2
(1 +)
(9)
Such a problem (Eq. (8)) has been solved by Petrolito [7] and
Eisenberger [8] and the solution is
w = C
1
+C
2
x +C
3
x
2
+C
4
x
3
+C
5
sinh(x) +C
6
cosh(x)
=
9.768EIC
4
GA
C
2
2C
3
x 3C
4
x
2
+0.226C
6
sinh(x) +0.226C
5
cosh(x) (10)
D
o
w
n
l
o
a
d
e
d
b
y
[
N
a
t
i
o
n
a
l
I
n
s
t
i
t
u
t
e
o
f
T
e
c
h
n
o
l
o
g
y
-
R
o
u
r
k
e
l
a
]
a
t
1
9
:
4
6
1
3
D
e
c
e
m
b
e
r
2
0
1
2
144 Y. LIU
FIG. 2. Presented beam element.
Particular solutions need to be derived according to the applied
loads or moments. If the applied load q(x) is linearly distributed
along the beam length L with the maximum magnitude Q
q(x) =
x
L
Q
1
(11)
The particular solution of Eq. (6) is obtained as
w
1
(x) =
0.008x
5
EIL
Q
1
1
(x) =
_
0.04x
4
EIL
+
0.78x
2
GAL
+
2.03EI
(GA)
2
L
_
Q
1
(12)
or if the applied load is given by
q(x) =
x
L
Q
1
+
_
1
x
L
_
Q
2
(13)
the particular solution is
w
2
(x) = w
1
(x) +
Q
2
Q
1
w
1
(L x)
2
(x) =
1
(x)
Q
2
Q
1
1
(L x) (14)
Obviously, if the applied load is constant, q(x) = Q, the partic-
ular solution can be obtained from Eq. (14) by assuming Q
1
=
Q
2
= Q.
4. ELEMENT FORMULATION
Based on the above demonstration, a typical beam element,
e, can be presented. Fig. 2 plots this beam element and the ele-
ment has two nodes, each with three degrees of freedom w, w,
and .
The beam element formulation, which is similar to that de-
veloped by Petrolito [7], is represented as
w
e
(x) = A
e
1
(x)(B
1
)
e
_
d
e
d
e
0
_
= N
e
1
(x)d
e
+w
e
f
(x)
e
(x) = A
e
2
(x)(B
1
)
e
_
d
e
d
e
0
_
= N
e
2
(x)d
e
+
e
f
(x) (15)
where the shape function matrices N
1
(x) and N
2
(x) are
N
1
(x) = A
1
(x)B
1
, N
2
(x) = A
2
(x)B
1
(16)
and w
f
(x) and
f
(x) are xed end solutions according to the
applied loading on the element
w
f
(x) = A
1
(x)B
1
d
0
,
f
(x) = A
2
(x)B
1
d
0
(17)
Notations in Eq. (15) to (17) are dened as
A
1
(x) = [1, x, x
2
, x
3
, sinh(x), cosh(x)]
A
2
(x) = [0, 1, 2x,
_
9.768EI
GA
+3x
2
_
,
0.226 cosh(x), 0.226 sinh(x)] (18)
B = [A
1
(0), A
1
(0), A
2
(0), A
1
(L), A
1
(L), A
2
(L)]
d
0
= [w
0
(0), w
0
(0),
0
(0), w
0
(L), w
0
(L),
0
(L)]
d
e
= [w
e
(0), w
e
(0),
e
(0), w
e
(L
e
), w
e
(L
e
),
e
(L
e
)]
where L
e
is the length of the beamelement, d
0
and d
e
are degrees
of freedom of the beam element.
From Eq. (7) and Eq. (15), the generalized forces for the
element are
F
e
1
(x) = S
e
1
(x)d
e
+S
e
f 1
(x)
F
e
2
(x) = S
e
2
(x)d
e
+S
e
f 2
(x)
F
e
3
(x) = S
e
3
(x)d
e
+S
e
f 3
(x)
(19)
where S
1
(x), S
2
(x), and S
3
(x) are generalized force matrices
S
1
(x) = 0.166EIN
2
(x) 0.06EIN
1
(x)
+0.5GA
_
N
2
(x) +N
1
(x)
_
S
2
(x) = 0.06EIN
1
(x) 0.166EIN
2
(x)
S
3
(x) = 0.648EIN
2
(x) 0.166EIN
1
(x)
(20)
while S
f 1
(x), S
f 2
(x), and S
f 3
(x) are generalized xed end forces
S
f 1
(x) = 0.166EI
f
(x) 0.06EIw
f
(x)
+0.5GA
_
(x) +w
f
(x)
_
S
f 2
(x) = 0.06EIw
f
(x) 0.166EI
f
(x)
S
f 3
(x) = 0.648EI
f
(x) 0.166EIw
f
(x)
(21)
Also, from Petrolitos theorem, the stiffness equations for a
beam that is composed of E beam elements are given by
E
e=1
_
K
e
d
e
R
e
_
= 0 (22)
D
o
w
n
l
o
a
d
e
d
b
y
[
N
a
t
i
o
n
a
l
I
n
s
t
i
t
u
t
e
o
f
T
e
c
h
n
o
l
o
g
y
-
R
o
u
r
k
e
l
a
]
a
t
1
9
:
4
6
1
3
D
e
c
e
m
b
e
r
2
0
1
2
A BEAM ELEMENT USING TOURATIER THEORY 145
FIG. 3. Simply supported beam subjected to a uniform load.
TABLE 1
Normalized central displacement, w, normalized maximum
stress,
x
for example
Displacement Stress
L/h Present Elasticity Petrolito Present Elasticity Petrolito
1 3.412 3.399 3.434 1.309 1.271 1.347
2 1.621 1.622 1.620 1.079 1.067 1.087
3 1.276 1.278 1.277 1.036 1.030 1.039
4 1.156 1.157 1.156 1.012 1.017 1.022
5 1.100 1.100 1.100 1.009 1.011 1.014
10 1.025 1.025 1.025 1.003 1.003 1.003
25 1.004 1.004 1.004 1.001 1.000 1.001
50 1.001 1.001 1.001 1.000 1.000 1.000
100 1.000 1.000 1.000 1.000 1.000 1.000
where K
e
denotes the element stiffness matrix, which is sym-
metric, and R
e
is the element load vector
R =
_
L
0
pN
T
1
dx (23)
5. ILLUSTRATIVE EXAMPLE
A beam problem is solved in this section using the present
beam element. The results are compared with the solutions ob-
tained using Petrolitos beam theory [7], elasticity theory [10],
and classical beam theory (the Timoshenko beam). In this ex-
ample, the aspect ratio of the beam L/h varies from 1 to 100 and
Poissons ratio is set as 0.3.
Example: simply supported beam subjected to a uniform load
In this example, a simply supported beam with length L and
depth h (Fig. 1) is subjected to a uniform load q =Q
1
/L (Fig. 3).
This beam is analyzed using the present element and, because
of its symmetry, the beam is meshed with two elements where
each element represents a half beam.
The central displacement w and maximum stress
x
for the
simply supported beams with different L/h are calculated using
the present beamtheory. The results are normalized with respect
to the results obtained from the classical beam theory and are
compared to the normalized elasticity solution and Petrolitos
solution (as listed in Table 1, the normalized elasticity solution
and Petrolitos results were presented in [7]).
The results listed in Table 1 indicate that the present beamfor-
mulation agrees very well to the elasticity theory and Petrolitos
beam model. Evident errors are observed compare to the classi-
cal solution when L/h <5, in which cases the models can not be
considered as beams. As the ratio L/h increases, all the results
tend toward the classical solution. The accuracy of the present
beam formulation is therefore veried.
6. CONCLUSIONS
In this paper, a new beam formulation based on Touratiers
sinusoidal beam theory is presented. The governing differential
equations for such beam are derived and solved. The present
beam formulation can be used to correctly solve regular beam
problems and by using this formulation only the minimumnum-
ber of elements are required to mesh the beam, enabling the
problem to be solved rapidly. Developed based on Touratiers
beam theory, such beam formulation does not require a shear
correction factor and is very suitable for modeling composite
beams.
REFERENCES
1. S. P. Timoshenko, On the Correction for Shear of the Differential Equation
for Transverse Vibrations of Prismatic Bars, Philosophical Magazine vol.
41, pp. 744746, 1921.
2. M. Levinson, A New Rectangular Beam Theory, Journal of Sound and
Vibration, vol. 74(1), pp. 8187, 1981.
3. W. B. Bickford, A Consistent Higher Order Beam Theory, Develop-
ments in Theoretical and Applied Mechanics vol. 11, pp. 137150,
1982.
4. P. R. Heyliger and J. N. Reddy, A Higher Order Beam Finite Element
for Bending and Vibration Problems, Journal of Sound and Vibration, vol.
126(2), pp. 309326, 1988.
5. J. N. Reddy, A Simple Higher-order Theory for Laminated Compos-
ite Plates, ASME Journal of Applied Mechanics, vol. 54, pp. 745752,
1984.
6. J. N. Reddy, A Rened Nonlinear Theory of Plates with Transverse Shear
Deformation, International Journal of Solids and Structures, vol. 20, pp.
881896, 1984.
7. J. Petrolito, Stiffness Analysis of Beams Using a Higher-order Theory,
Computers & Structures, vol. 55(1), pp. 3339, 1995.
8. M. Eisenberger, An Exact High Order Beam element, Computers & Struc-
tures, vol. 81, pp. 147152, 2003.
9. M. Touratier, An Efcient Standard Plate Theory, International Journal of
Engineering Science, vol. 29(8), pp. 901916, 1991.
10. J. Petrolito, AModied ACMElement for Thick Plate Analysis, Computers
& Structures, vol. 31, pp. 553565, 1989.
D
o
w
n
l
o
a
d
e
d
b
y
[
N
a
t
i
o
n
a
l
I
n
s
t
i
t
u
t
e
o
f
T
e
c
h
n
o
l
o
g
y
-
R
o
u
r
k
e
l
a
]
a
t
1
9
:
4
6
1
3
D
e
c
e
m
b
e
r
2
0
1
2