100%(1)100% menganggap dokumen ini bermanfaat (1 suara)
1K tayangan8 halaman
Beams represent fundamental structural components in many engineering applications. Shape functions essential for the FINITE ELEMENT DISCRETIZATION of such structures.
Beams represent fundamental structural components in many engineering applications. Shape functions essential for the FINITE ELEMENT DISCRETIZATION of such structures.
Beams represent fundamental structural components in many engineering applications. Shape functions essential for the FINITE ELEMENT DISCRETIZATION of such structures.
Loural of Sound and Vibration (2008) 259(2), 473-480 .
4oi0.1006jsv2002.5122, availabe online at hip idealtbrary.com on IDE De
® @
SHAPE FUNCTIONS OF THREE-DIMENSIONAL
TIMOSHENKO BEAM ELEMENT
A. BAZOUNE AND Y. A. KHULIEF
Department of Mechanical Engineering, King Fahd University of Petroleum and Minerals, KFUPM
BOX # 319, Dhahran 31261, Saudi Arabia, E-mail: abazoune(@kfupm edi.sa
AND
N. G. Srepnen
‘School of Engineering Sciences, Mechanical Engineering, University of Southampton, Southampton
‘SOI IBS, England
(Received 15 March 2002, and in final form 8 April 2002)
L. INTRODUCTION
Beams represent fundamental structural components in many engineering applications,
and shape functions are essential for the finite element discretization of such structures.
Praemeniecki {1} derived explicit expressions for the shape functions of two-dimensional
‘Timoshenko and three-dimensional Euler-Bernoulli (EB) beam elements, Note that for
the three-dimensional EB element presented in reference [1], a change of sign is required in
those entries of the third column of the shape function matrix which correspond to the
twist terms. Since that pioneering work, there does not appear to have been any attempt to
extend these results to a three-dimensional Timoshenko beam element, and it is the
purpose of this note to fill this gap in the literature.
2. FINITE ELEMENT DISCRETIZATION
Consider a typical two-node three-dimensional beam element of length /, where each
node has six degrees of freedom. The nodal displacement vector {e} defined with respect to
the clement axes is denoted by
im yw Or On Or ue tr w2 Ba Op Oa], (1)
{e}irer
where (uj, 1s) are the nodal axial displacements in the x-direction, and (v1, v3) and
(v1, 2) are the translational displacements in the y- and z-directions, respectively,
(Gz1, 8.2) are the torsional displacements about the x-axis, and (01, 8,2) and (0x1, 8.2)
are the rotational displacements in the (xz)- and (xy)-planes, respectively.
According to the standard finite element procedure, the elastic deformation of an
arbitrary point of the beam element can be expressed as
(a) =e), @)
(0022-460X02/$35.00 2002 Elsevier Science Ltd. All rights reserved474 LETTERS TO THE EDITOR
where {d} represents the elastic deformation vector of the beam element and [1] is the
‘matrix of shape functions used to model its deformations. Note that the shape functions
are spatially dependent while the nodal displacement vector is time dependent.
Equation (2) is quite general and is valid for any form of shape functions [1 used to
model the beam elements. The shape functions used for translational and rotational
bending deformation are the conventional cubic Hermitian polynomials that incorporate,
in addition to the continuity and completeness conditions, shear deformation parameters
that account for the effects of shear. The shape functions for torsional and axial
deformation are linear, and are included for completeness.
3. THE DISPLACEMENT FIELD
Timoshenko beam theory (TBT) is applied when the cross-sectional dimensions of the
beam are not small compared to its length and/or when higher bending modes are
required. The kinematic relations for a three-dimensional beam undergoing axial,
torsional and bending deformations in the (xxy)- and (xz)-plane can be expressed as
vs ef) 4B)
Vi = -20+0,
Wo = yOc+w,
@)
where the translations (v, wv) consist of contributions (vp, ws) and (2, 1,) due to bending
and transverse shear, that is
Oy + ty, w= WHE ye (4,3)
‘The relationships between total slope, bending rotation and transverse shear are
Ov _ Buy | dv,
Fen Oe Oe Tot 6)
dw _ Oy Ow
Ber Be ae Ot ae @
Where 7. and 74. are shear strains in the (xy)- and (xz)-planes, respectively.
The two rotations (0, 0.) are related to the bending deformations (vp, ws) by the
expressions
og, Om
Ox” Ox”
Note that axial warping displacement uring torsion i ignored
6,
(8,9)
4. DERIVATION OF SHAPE FUNCTIONS
Shape function matrices for axial and torsional deformation, [4
found in any elementary ext, and are given by
Wald] = 00) = (1-9 ah, (10)
where §=.x/l is the dimensionless axial co-ordinate, Shape functions for bending
deformation in the (xp)-plane are derived as follows: the translational deformation v(x) at
an arbitrary location x is expressed as,
0(4) = dy + px + yx? + ax ay
J and [a], can beLETTERS TO THE EDITOR 475
or in matrix form as
v(x) = (allay), (12)
where
b=[ x 2 x] (3)
and
{a} = [a a a as!" (14)
The shear strain is assumed to be independent of the element axial co-ordinate x, in
accordance with reference [2], ie., constant along the finite element
0 (1s)
The bending moment M. and the shearing force Q, are related by
Yay = Const,
(16)
and the moment-curvature relationship is
a7)
where L. is the second moment of area about the z-axis; the shear force is related to the
transverse shear strain by
Q =; G Ary (18)
In the above, x is the shear correction factor that accounts for the non-uniform
distribution of the shear stress over the cross-section A, F is the modulus of elasticity, and
Gis the shear modulus. The slope due to bending can be obtained by using equations (6),
(13) and (15), that is
8, = ay + 2ayx + 3a5x? — Yo, (19)
ith respect to x and substituting it into equation (17) yields
M, = ~El--(2ay + 643%). (20)
Taking the derivative of M. with respect to x and substituting into equation (16) along
with equations (15) and (18) yields
Taking the derivative of 8,
6EL.:03 ~ ky GA, an
from which
(22)
where
(23)
Substitute equation (23) into the expression for 8., to give
0. = ay + 2ayx + (3x? + 6Az)a (2a)476 LETTERS TO THE EDITOR
To express the coefficients a in terms of the bending deformations and slopes, the
following boundary conditions must be satisfied:
w(0) = and (0
600) =. and (0) S
and applying these to equations (11) and (24) gives
20) = a
000) = a +64.a5
w= a tal+anl? + asl (26)
0.) = a, +2ayl + (32 +6A,)ay
In matrix form, this can be written as
o 1000 4%
éa{_]o 10 64, a
af "flree a Cu
2) lo 1 a G46) la
or in more compact form
{a} = [Al{aj}, (28)
from which
fa} = (A) "(a 29)
Solving for {a)} gives
a =r, (30)
GB)
(32)
(33)
where
(34)
and
(35)LETTERS TO THE EDITOR an
is the shear deformation parameter that represents the ratio between bending and shear
stiffnesses. Substituting the values of a; into the expression of o(é = x/I) and simplifying,
one obtains
0G) =8.(1— 367 +28 + (1 ~ gen +h (c-28 +8 + $0-2)ou
48.38 ~ 20 + O.8)02 +I (-2+0+$0e+2))ps (36)
Hence, »() can be written in the form
266) = Wt + Malls +H gta + Mebty (7)
where
My = $138 +22 + O.(1- 6),
We = 16,(¢-28 +8 +$6-2)),
Pee es (38)
No = 632-28 +08)
ee
Wy = 6(-242+$C2+0)
Similarly, substitute a; into the equation for 0,(€) to get
- Pee 2
O48) = FE +P) + (1-48 +32 + 0,1 - 8/0 9)
: 39)
+ Fe Ant B(28+ SP + 2) 02.
Hence, 0.(¢) can be written in the form
HG) = Me, 21 + V ater + Mpg + M pall (40)
where
Mn = Bere),
Vig = 6.0- 2 + 61-8),
(1-45 +32 + 41-9) a
i .
Nig = O26 438 +628)
Shape functions for bending in the (xz)-plane are obtained in a similar manner; the
bending slope 6, is given by equation (7) while the shear deformation parameter is
(42)478 LETTERS TO THE EDITOR
and
(+6) oe;
‘The shape functions corresponding to bending in the (xz)-plane can then be
written as
Ney = (1-38 +28 +014),
He = -He-22 +2 + 2E-2),
As (44)
Vn = 638-28 + 0,2),
ee
Mu = BLP +O + E+ 2)
and the corresponding bending slope shape functions are
Vig =
Le
ig (45)
Mig = AHS 2),
Ny, = —O)(-28 +387 + 0,2).
By virtue of equations (10), (38), (41), (44) and (45), the kinematic relations given by
equation (3) are now expressed as
U = (1-8) — 6b:(-€ +P )noy — 66,(-E + Pow
4 1B, (1 ~ AE +38 + O,(1 ~ E01 ~ 18.(1 — 48 + 32 + 0.1 = 2B
nen ~ 68, (¢ ~ 2 )fy
®,£)6052 ~ 1B,(-25 +32 + 6.2)n82,
W =6,(1-32 +22 +6,(1-)
+In(~ 90a ~ 18, (c-28 +0 +
é-
+ By(36? — 28 + B,2)w2 + nea - 16,(-€LETTERS TO THE EDITOR 479
where =y/I and [= 2/1 are dimensionless co-ordinates in the y- and z-directions
respectively. In matrix form, this can be written as
{@oa=[U VW) =Mseafehioe (47)
These results are summarized in the matrix of the shape functions, [7], shown in
Appendix A. If the shear deformation parameters @, and ®, are neglected, then
[47] reduces to the three-dimensional Euler-Bernoulli beam shape function derived in
reference [1], where a sign change is required in the fourth and tenth entries of the third
column of the shape funetion matrix, which correspond to the twist terms
ACKNOWLEDGMENTS
The support provided by King Fahd University of Petroleum and Minerals is deeply
appreciated. The authors would like to thank Prof. M. H. Balush of the Civil Engineering
Department, KFUPM, for useful discussions on this work.
REFERENCES
1. J.S. PRzeMeNiecKt 1968 Theory of Matrix Structural Analysis, New York: McGraw-Hill p.
293, equation (11.30),
2. R, NARAYANASWAMI and H. M. ADELMAN 1974 American Institute of Aeronautics and
Astronautics Journal 12, \613-1614. Inclusion of transverse shear deformation in finite element
displacement formulations,LETTERS TO THE EDITOR
480
Co + D/L = %@ pur (“+ 1/1 =p asaya
0 (@-22- ete yor N30 + 264 22-)'or~
(@-28-2+2-) er 0 10+ ge4 Yor
up 3- 0
(0+ oe 20% 0 e+>-Y%e9
0 (904 20-00 M54 9-)'69
0 o >
: =r]
o (e-98+ 2tae-2):91 W010 + 204 5— Die
(@ Met e+ ee 2) 0r- ° WG-04 2649-197
Wa-0 We D-
(O-N@+ z+ ge- °
0 (G-D¥0+ get ge-De
0 0
uoUWa[a Mog OYUYSOUNT] JeuOISUOUNP-AaIKy atp LO xINeL oNOUNY adeyg,
V XIGNadaVv