Generation of shape functions for straight beam elements

Charles E. Augarde *
Department of Engineering Science, University of Oxford, Parks Road, Oxford, U.K.
Received 17 June 1997; received in revised form 14 February 1998
Abstract
Straight beam nite elements with greater than two nodes are used for edge stiening in plane stress analyses and
elsewhere. It is often necessary to match the number of nodes on the edge stiener to the number on a whole plane
stress element side. Beam elements employ shape functions which are recognised to be level one Hermitian
polynomials. An alternative to the commonly adopted method for determining these shape functions is given in this
note, using a formula widely reported in mathematical texts which has hitherto not been applied to this task in the
nite element literature. The procedure derives shape functions for beams entirely from the set of Lagrangian
interpolating polynomials. Examples are given for the derivation of functions for a three and four-noded beam
element. # 1998 Elsevier Science Ltd. All rights reserved.
Keywords: Finite elements; Beams; Hermitian interpolation; Shape functions
1. Introduction
Analysis of structures using the nite element
method is well established. Many formulations exist
for complex elements but simple elements remain pop-
ular since they are usually well-tested and easy to im-
plement into an analysis program. Two-dimensional
plane stress analysis, for thin structures subject to in-
plane loading, may employ continuum elements, such
as the fteen-node triangle, having a large number of
nodes along a side. Where edge stiening is required,
beam elements can be connected to continua edge
nodes. There is then a requirement for formulations of
beam elements having more than two nodes.
Conventional two-dimensional beam elements have
two degrees of freedom at each node: one lateral dis-
placement and one rotation. Unless the structure is
loaded entirely laterally, axial stiness must also be in-
corporated, by an additional degree of freedom at each
node. With this amendment, beam elements are usually
referred to as frame elements. The axial eects are
uncoupled for straight beams and the determination of
suitable shape functions is straightforward [1].
The author has recently contributed to the develop-
ment of a complex three-dimensional nite element
model at the Department of Engineering Science,
Oxford University to study the damage accruing to
surface structures from adjacent tunnelling [2].
Modelling was also undertaken in two-dimensions (to
validate the complex model) where tunnel linings were
represented with simple beam elements. The procedure
outlined in this note was used in the implementation of
these beam elements into an existing in-house nite el-
ement code (OXFEM).
In this note we are concerned with the generation of
the shape functions which interpolate the lateral displa-
cements along beam elements having more than two
nodes. Bernoulli-Euler beam theory is assumed where
transverse shear deformation is zero. While most nite
element texts describe the simple two-noded beam [1, 4]
few explain how more complex elements may be
formulated [3].
Computers and Structures 68 (1998) 555560
0045-7949/98/$19.00 # 1998 Elsevier Science Ltd. All rights reserved.
PII: S0045- 7949( 98) 00071- 6
PERGAMON
* Author to whom correspondence should be addressed.
2. Standard Procedure
The beam element of Fig. 1(a) is dened by two
nodes, a distance l apart, along the element local x-
axis. A common convention, and that adopted here, is
to make the nodal coordinates non-dimensional, that
is
" x =
x
l
Y (1)
so that the two-noded beam has nodes at x=0, +1.
Using shape functions, lateral displacement w(x) of
the two-noded beam element of Fig. 1(a) is
w(" x) = Nd (2)
where
N = {N
1
N
2
N
3
N
4
] 8 d
T
= {w
1
y
1
w
2
y
2
]X (3)
N
j
,(j = 1,4) are the bending shape functions and w
i
, y
i
,
(i = 1,2) are the displacements and rotations at the
nodes. Determination of the former, using the method
to be found in many nite element texts [4, 5], proceeds
by rst writing w(x) as an n-termed polynomial with
unknown coecients, n being the number of degrees of
freedom in the element
w(" x) = Xaaa (4)
where,
X = {1 " x " x
2
" x
3
] 8 aaa
T
= {a
1
a
2
l a
3
l
2
a
4
l
3
]X (5)
Taking Eq. (4) and its rst derivative with respect to x,
four further equations can be formed, one for each
degree of freedom:
d = Aaaa (6)
where
A =
1 0 0 0
0 1 0 0
1 1 1 1
0 1 2 3
_
_
_
_
_

_
(7)
Solving for aaa in Eq. (6) and substituting into Eqs. (4)
and (2) gives
N = XA
1
(8)
The three-node beam of Fig. 1(b), of overall length 2 l,
has nodes at x= 1, 0, +1. The same procedure as
above yields
N = {N
1
N
2
N
3
N
4
N
5
N
6
]Y (9)
d
T
= {w
1
y
1
w
2
y
2
w
3
y
3
]Y (10)
X = {1 " x " x
2
" x
3
" x
4
" x
5
] (11)
aaa
T
= {a
1
a
2
l a
3
l
2
a
4
l
3
a
5
l
4
a
6
l
5
] (12)
and to solve for N requires the inversion of
A =
1 1 1 1 1 1
0 1 2 3 4 5
1 0 0 0 0 0
0 1 0 0 0 0
1 1 1 1 1 1
0 1 2 3 4 5
_
_
_
_
_
_
_
_
_

_
X (13)
3. Hermitian Interpolation
The shape functions in Eq. (2) are Hermitian poly-
nomials since the displacement w(x) is interpolated
from nodal rotations as well as nodal displacements.
This contrasts with Lagrangian interpolation, used for
continuum elements' shape functions and for the axial
eects in frame elements. Considering small displace-
ments, the nodal rotations are the rst derivatives of
the unknown real displacement function at the nodes
thus fullling the denition of Hermitian interpolation.
This property allows an alternative procedure to be
used to determine the shape functions to that outlined
above. Fig. 1. Beam elements showing bending degrees of freedom.
C.E. Augarde / Computers and Structures 68 (1998) 555560 556
The generation of Hermitian (or Hermite) poly-
nomials from Lagrangian interpolation polynomials is
described in many mathematical texts [68]. Despite
the clear understanding that bending shape functions
are equivalent to Hermitian polynomials the technique
described below has not, to the author's knowledge,
been linked to the generation of those shape functions.
One-dimensional interpolation is required for
straight beam elements. The single dimension is along
the element centreline, dened as the x-axis. One-
dimensional Hermitian interpolation for an unknown,
w(x) proceeds as [6, 7]
w(x) =

nnod
i=1
H
r
0i
w
i
H
r
1i
dw
dx
_ _
i
F F F H
r
ri
d
r
w
dx
r
_ _
i
_ _
(14)
where H
ji
r
is a Hermite polynomial of level r, relating
to node i and to derivative order j of w. The sum is
over the number of nodes, nnod where values of w and
its derivatives are available. The level of the poly-
nomial indicates the highest order derivative used in
the interpolation.
Comparison of Eqs. (2) and (14) reveals that the
bending shape functions are level one Hermitian poly-
nomials as follows:
N
1
= H
1
01
N
2
= H
1
11
N
3
= H
1
02
N
4
= H
1
12
X (15)
Level one Hermitian polynomials are derived from
Lagrangian polynomials by the following formulae:
H
1
0i
= [1 2(x x
i
)L
/
i
(x
i
)][L
i
(x)]
2
(16)
H
1
1i
= (x x
i
)[L
i
(x)]
2
(17)
where L
i
(x) is the one-dimensional Langrangian poly-
nomial of degree (nnod 1) calculated at node i, given
by
L
i
(x) =

nnod
j=1Yj,=i
x x
j
x
i
x
j
(18)
and L'
i
(x) is its rst derivative with respect to x. A
polynomial of order (nnod 2 1) is required to in-
terpolate over nnod points, each contributing two
values. Inspection of Eqs. (16) and (17) shows that the
Hermite polynomials are of the correct order for in-
terpolation.
Re-writing Eqs. (16) and (17) in terms of the non-
dimensional coordinate, x
H
1
0i
= [1 2l(" x " x
i
)L
/
i
(" x
i
)][L
i
(" x)]
2
(19)
H
1
1i
= l(" x " x
i
)[L
i
(" x)]
2
X (20)
The advantage of a derivation based on the
Lagrangian polynomials and their rst derivatives is
that these are already likely to be present in a program
code. The former are required for continuum elements
and the derivatives are required for isoparametric el-
ements. The use of this procedure also provides a sys-
tematic approach to allow simpler coding.
4. Examples
4.1. Three-node beam
A three-node element with six bending degrees of
freedom and a total length of 2 l is shown in Fig. 1(b).
This element has nodes at x=(1, 0, 1). The axial
degrees of freedom are omitted from this element as in
the derivations above. Shape functions for these
degrees of freedom are the Lagrangian polynomials of
order 2. From the preceding section, it is clear that
these are also required for the derivation of the bend-
ing shape functions.
The Lagrangian polynomials are
L
1
(x) =
(x x
2
)(x x
3
)
(x
1
x
2
)(x
1
x
3
)
(21)
L
2
(x) =
(x x
1
)(x x
2
)
(x
2
x
3
)(x
2
x
1
)
(22)
L
3
(x) =
(x x
1
)(x x
2
)
(x
3
x
1
)(x
3
x
2
)
X (23)
Rewriting in terms of the non-dimensional coordinate
x and substituting for values of x at nodes [i.e. (x
1
, x
2
,
x
3
) = (1, 0, 1)], gives
L
1
(" x) =
" x
2
(" x 1) (24)
L
2
(" x) = " x(" x 1) (25)
L
3
(" x) =
" x
2
(" x 1)X (26)
The derivatives are
L
/
1
(" x) =
1
2l
(2" x 1) (27)
L
/
2
(" x) =
1
l
(2" x 1) (28)
L
/
3
(" x) =
1
2l
(2" x 1)X (29)
The bending shape functions are equivalent to the fol-
lowing Hermite polynomials:
N
2
= H
1
01
N
3
= H
1
11
N
5
= H
1
02
(30)
N
6
= H
1
12
N
8
= H
1
03
N
9
= H
1
13
X (31)
C.E. Augarde / Computers and Structures 68 (1998) 555560 557
Note that the rst four are not the same as those in
Eq. (15) since each set is based on dierent order
Lagrangian polynomials.
From Eqs. (24) and (27) and with substitution for x
1
we obtain
H
1
01
= 1 2l(" x 1)
3
2l
_ _ _ _
" x
2
(" x 1)
_ _
2
=" x
2

5
4
" x
3

1
2
" x
4

3
4
" x
5
X (32)
Similarly,
H
1
11
= l(" x 1)
" x
2
(" x 1)
_ _2
=
l
4
[" x
2
" x
3
" x
4
" x
5
] (33)
H
1
02
= 1 2l" x
1
l
_ _ _ _
[" x(" x 1)]
2
= 1 2" x
2
" x
4
(34)
H
1
12
= l" x[" x(" x 1)]
2
= l[" x 2" x
3
" x
5
] (35)
Fig. 2Caption overleaf
C.E. Augarde / Computers and Structures 68 (1998) 555560 558
H
1
03
= 1 2l(" x 1)
3
2l
_ _ _ _
" x
2
(" x 1)
_ _2
=" x
2

5
4
" x
3

1
2
" x
4

3
4
" x
5
(36)
H
1
13
= l(" x 1)
" x
2
(" x 1)
_ _
2
=
l
4
[" x
2
" x
3
" x
4
" x
5
]X
(37)
The suitability of the above as shape functions can be
demonstrated by plotting each polynomial (Fig. 2).
The polynomials related to displacements have value 1
at the associated node and zero elsewhere. Those re-
lated to derivatives have slopes of 1 at the associated
node.
4.2. Four-node beam
Fig. 1c shows a beam with four nodes having the
same overall length as the three-node beam. The non-
dimensional coordinates (x) of the nodes are 1, 1/3,
+1/3 and +1. Eight shape functions are required, de-
rived from the following Lagrangian polynomials and
derivatives:
L
1
(" x) =
9
16

1
9

" x
9
" x
2
" x
3
_ _
(38)
Fig. 2. (a) Shape functions for a three-node beam element. (b) Shape functions for a three-node beam element.
C.E. Augarde / Computers and Structures 68 (1998) 555560 559
L
2
(" x) =
27
16
1
3
" x
" x
2
3
" x
3
_ _
(39)
L
3
(" x) =
27
16
1
3
" x
" x
2
3
" x
3
_ _
(40)
L
4
(" x) =
9
16

1
9

" x
9
" x
2
" x
3
_ _
(41)
L
/
1
(" x) =
9
16l
1
9
2" x 3" x
2
_ _
(42)
L
/
2
(" x) =
27
16l
1
2
3
" x 3" x
2
_ _
(43)
L
/
3
(" x) =
27
16l
1
2
3
" x 3" x
2
_ _
(44)
L
/
4
(" x) =
9
16l

1
9
2" x 3" x
2
_ _
X (45)
The rst two shape functions, relating to lateral displa-
cement and rotation at node 1, can then be derived as
N
1
=
1
512
[13 15" x 243" x
2
281" x
3
1215" x
4
1413" x
5
729" x
6
891" x
7
] (46)
N
2
=
l
256
[1 " x 19" x
2
19" x
3
99" x
4
99" x
5
81" x
6
81" x
7
]X (47)
Plotting these two functions [Fig. 3(a), (b)] shows that
they also satisfy the basic requirement of shape func-
tions as outlined above.
5. Conclusion
This work presents an alternative derivation of
bending shape functions for simple beam elements, for
implementation of many-noded straight beam elements
within a nite element analysis code. While the el-
ements described are simple, the theory will be of inter-
est to developers of other C
1
continuous elements such
as rectangular plates.
Acknowledgements
This work was carried out as part of an EPSRC
funded project examining numerical modelling of tun-
nelling at the Department of Engineering Science,
University of Oxford, UK under the guidance of Dr
H.J. Burd. The author would like to acknowledge the
help and contribution of Dr Burd to this work.
References
[1] Astley RJ. Finite Elements in Solids and Structures.
London: Chapman & Hall, 1992.
[2] Augarde CE. Numerical modelling of tunnelling pro-
cesses for assessment of damage to buildings. D.Phil
thesis. University of Oxford, 1997.
[3] Dawe DJ. Matrix and Finite Element Displacement
Analysis of Structures. Oxford: Clarendon Press, 1984.
[4] Cook RD. Concepts and Applications of Finite Element
Analysis. 2nd ed. Chichester: John Wiley & Sons, 1981.
[5] Mohr GA. Finite Elements for Solids, Fluids and
Optimisation. Oxford: Oxford University Press, 1992.
[6] Jacques I, Judd C. Numerical Analysis. London,
Chapman & Hall, 1987.
[7] Morris JL. Computational Methods in Elementary
Numerical Analysis. Chichester: John Wiley & Sons,
1983.
[8] Spanier J, Oldham KB. An Atlas of Functions. London:
Hemisphere, 1987.
Fig. 3. Shape functions for a four-node beam element.
C.E. Augarde / Computers and Structures 68 (1998) 555560 560