See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/336652621

On the Derivation of Exact Solutions of a Tapered Cantilever Timoshenko

Beam

Article  in  Civil Engineering Dimension · October 2019

DOI: 10.9744/ced.21.2.89-96

CITATIONS READS

3 3,196

5 authors, including:

Foek Tjong Wong

Petra Christian University
38 PUBLICATIONS   120 CITATIONS   

SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Hybrid FE-meshfree with continuous nodal stress View project

Kriging-based Finite Element Method View project

All content following this page was uploaded by Foek Tjong Wong on 25 October 2019.

The user has requested enhancement of the downloaded file.

Civil Engineering Dimension, Vol. 21, No. 2, September 2019, 89-96 DOI: 10.9744/CED.21.2.89-96
ISSN 1410-9530 print / ISSN 1979-570X online

On the Derivation of Exact Solutions of a Tapered Cantilever

Timoshenko Beam

Wong, F.T.1*, Gunawan, J.2, Agusta, K,3, Herryanto4, and Tanaya, L.S.5

Abstract: A tapered beam is a beam that has a linearly varying cross section. This paper
presents an analytical derivation of the solutions to bending of a symmetric tapered cantilever
Timoshenko beam subjected to a bending moment and a concentrated force at the free end and a
uniformly-distributed load along the beam. The governing differential equations of the
Timoshenko beam of a variable cross section are firstly derived from the principle of minimum
potential energy. The differential equations are then solved to obtain the exact deflections and
rotations along the beam. Formulas for computing the beam deflections and rotations at the free
end are presented. Examples of application are given for the cases of a relatively slender beam
and a deep beam. The present solutions can be useful for practical applications as well as for
evaluating the accuracy of a numerical method

Keywords: Timoshenko beam; tapered beam; the principle of minimum potential energy;
slender beam; deep beam.

Introduction Several studies have been carried out to obtain

analytical solutions of non-prismatic beams based on
A tapered beam is a beam with a linearly varying a modified Euler-Bernoulli or Timoshenko beam
cross section and hence has a straight center-line. It model. For example, Fertis and Keene [3] presented
is within the larger class of non-prismatic beams, i.e. an elastic and inelastic bending analysis of non-
“a beam with curvilinear center-line and non- prismatic beams using the so-called method of the
constant cross-section” [1]. It is commonly used in equivalent prismatic system. The advantage of this
various engineering structures such as buildings and approach is that the exact solution of a non-prismatic
bridges to achieve more optimal use of the material Euler Bernoulli beam model can be obtained with
and to satisfy architectural or esthetical needs. There reduced mathematical complexity. Romano and
are several approaches for modeling non-prismatic Zingore [4] obtained analytical solutions of non-
beams for structural analysis and design, that is [1, prismatic Euler-Bernoulli beams with a linearly and
2]: quadratically varying depth or with a linearly vary-
 Using a modified prismatic beam model, i.e., ing width of the rectangular cross section. The
Euler-Bernoulli or Timoshenko beam model with solutions were obtained by solving the fourth-order
varying cross-sectional area and moment of linear differential equation with variable coefficients
inertia, that govern the beam bending. Romaro [5] then
 Using a generalized non-prismatic beam theory extended his previous study [4] to obtain analytical
such as that recently developed by Balduzzi et al. solutions of the same non-prismatic beams based on
[1] and Auricchio et al. [2], a modified Timoshenko beam model including the
 Using 2D or 3D finite elements. effects of tapering slope on shear stress distribution
on the cross sections. The solutions were achieved by
This paper deals with the first approach, in parti- transforming the forth-order differential equation
cular, the Timoshenko beam model. with variable coefficients into that with constant
coefficients. Al-Gahtani and Khan [6] presented an
1, Department of Civil Engineering, Petra Christian University, exact analysis of non-prismatic Euler-Bernoulli
Jl. Siwalankerto 121-131, Surabaya 60236, INDONESIA. beam model with end elastic supports. The boundary
2 PT Waringin Megah, Surabaya, INDONESIA.
integral method was used in their study.
3 Benjamin Gideon and Associates, Surabaya, INDONESIA.
4 PT. Archimetric, Surabaya, INDONESIA.
5 PT. Teno Indonesia, Surabaya, INDONESIA. This paper presents the derivation of analytical
*Corresponding author; email: [email protected] solutions of the Timoshenko beam model of a
Note: Discussion is expected before November, 1st 2019, and will be
symmetric tapered cantilever beam subjected to a
published in the “Civil Engineering Dimension”, volume 22, num- bending moment and a concentrated force at the free
ber 1, March 2020. end, and a uniformly-distributed load (Figure 1). The
Received 17 June 2019; revised 12 September 2019; accepted 19
term „symmetric‟ here means that the longitudinal
September 2019. tapered beam geometry has an axis of symmetry,

89
 Wong, F.T. et al. / On the Derivation of Exact Solutions / CED, Vol. 21, No. 2, September 2019, pp. 89–96

that is, the line passing through the cross section subjected to a bending moment M0 and a concen-
centroids. The effects of the tapering slope on the trated load P0 at the left end, and a uniform trans-
shear stress distribution along the beam is neglected. versal load q (force/length) along the beam.
Thus, the model used is simpler than that considered
in Romano [5] but is finer than the Euler-Bernoulli A right-handed Cartesian coordinate system x-y-z is
beam model [4]. set up with the point of origin O at the left end, as
shown in Figure 1. According to the Timoshenko
beam theory [9: p. 399], the displacement of a
material point at coordinate (x, y, z) of the beam is;
( ) ( )
( ) ( ) (1)
Sect. A-A where u(x, y, z) and w(x, y, z) are the displacement
components of the material point in the x and z
Figure 1. Tapered beam model directions, respectively, w(x) and θ(x) are the
deflection (z-direction displacement) and rotation of
In the derivation, the differential equations govern- the cross section at station x. The positive direction
ing bending of the beam is firstly derived from the for θ(x) is the rotation from the x-axis to the z-axis
total potential energy. The resulting differential (clockwise rotation).
equations are then analytically solved to obtain the
exact solutions. For the sake of comparison and A functional that governs the deformation of the
completeness, the solutions for the case of prismatic beam is
beams are also derived. Lastly, the solutions are
applied to determine the deflections of a slender , - (2a)
beam and a deep tapered beam with different
∫ ∫ ( ) (2b)
tapering angles. In this example, the numerical
solutions are compared to those obtained using the ( ) ( ) ∫ (2c)
analytical solutions of Romano [5] and the finite
element analysis of the plane stress model. where Π[w, θ] is the total potential energy of the
beam, U is the elastic strain energy stored in the
It is worth mentioning here that the use of the Timo- deformed beam, W is the potential work of the
shenko beam model to non-prismatic beams, in applied loads, I is the moment of inertia of the cross
general, introduces a modeling error, which is pro- section about the y axis, and As is the effective shear
portional to the rate of cross section height variation area, i.e.
[1,2]. However, due to its simplicity, it is still worthy (3a)
of using this model in engineering practice to
analyzed a symmetric tapered beam with a small Commas followed by subscripts x indicate differen-
tapering angle (i.e., a small rate of change in the tiation to x. In Equation (3a), A is the area of the
height variation) [6,7]. For non-symmetric tapered cross section, and k is the shear coefficient. This
beams, however, the modeling error may not be coefficient is a correction factor needed to account for
negligible even for a small tapering angle as can be the difference between the assumed constant shear
seen in an example presented in Mercuri et al. [8]. stress in the Timoshenko beam theory and the ac-
The reason for this error is that the Timoshenko tual shear stress distribution. Cowper [10] presented
beam model neglects the coupling between bending a shear coefficient formula based on a derivation
and axial deformations occurred in non-symmetric from the three-dimensional theory of elasticity, that
tapered beams. is,
( )
Governing Equations of the Tapered (3b)
Beam Model for a rectangular cross section.

The beam under consideration is a tapered canti- The deflection and rotation fields in the total poten-
lever beam of the length L, as illustrated in Figure 1. tial energy, Equation (2), are required to be
The cross section is rectangular with the constant sufficiently regular and satisfy the essential boun-
width b, and the height varies linearly from h(0) = h0 dary conditions. These requirements are written as
at the left end to h(L) = hL = αh0 at the right end, follows:
where α ≥ 1 (if α = 1 then the beam is prismatic). The * ( ) ( ) + (4a)
beam is made from a homogeneous, isotropic mate-
rial with the modulus of elasticity E, Poisson‟s ration In this expression, S is the space of admissible
ν, and modulus of shear G = E / (2(1+ν)). It is solutions, H1(0, L) is the Hilbert function space (also

90
 Wong, F.T. et al. / On the Derivation of Exact Solutions / CED, Vol. 21, No. 2, September 2019, pp. 89–96

called the first Sobolev space) [11,12], which is ( )- ; ( ( ))] (9c)

defined as
( ) ; ( ) (9d)
( ) * ∫ ( ) + (4b)
Equation (9c) are the natural boundary conditions
The principle of stationary potential energy [9: pp. while Equation (9d) are the essential boundary
110-116, 13: pp. 137-140] for the beam can be written conditions.
as
Note that the sign convention for the bending
* ( ) + (5) moment that consistent with the above derivation is
where δ is the variational operator, stands for „for positive when the upper longitudinal fibers are in
all‟, and V is the space of admissible displacement tension (in contrast to the commonly used sign
variations. Applying this principle to Equations (2) convention). For the shear force, it is positive when
and (5) gives; its direction is downward (follows the direction of the
positive z axis) on the cross section with the normal
∫ ∫( ) ( ) vector pointing to the positive x axis.

( ) ( ) ∫ Analytical Solutions of the Beam Govern-

(6) ing Equations

Interpreting δw and δθ as the virtual deflection and The varying height, area, and moment of inertia of
virtual rotation, respectively, Equation (6) is recog- the beam cross sections can be expressed as follows:
nized as of the principle of virtual displacement for
the beam model. ( ) ( ), (10a)

( ) ( ), (10b)
Using integration by parts to the first and second
terms of Equation (6), respectively, and using the ( ) ( ) , (10c)
requirements δw(L) = δθ(L) = 0, these terms can be
expressed as
Integrating Equation (9b) and imposing the shear
∫ ∫ ( ) force boundary condition, i.e., the second equation of
( )( )- (7a) Equation (9c), yield the shear force field
( ) ( ) (11)
∫ ( ) ( )
Subsequently, substituting the shear force, Equation
∫ ( ( )) (11), into Equation (9a), integrating and imposing
the moment boundary condition, i.e., the first equa-
( )( ( ))] tion of Equation (9c), yield the bending moment field
∫ ( ) (7b) ( ) (12)

Now, substituting these expressions into Equation These results, Equations (11) and (12), can be easily
(6) and arranging the resulting terms yields confirmed by using a simple static principle since the
beam is statically determinate.
∫ ,( ) ( )-
Now, substituting the varying moment of inertia,
∫ [( ( )) ] Equation (10c), into Equation (12) and dividing the
result by EI0 give
( ) *( )- +
(13)
( ) ( ) ( )
( )[( ( ))] ] (8)
Integrating Equation (13) (using the partial fraction
Since the variations, δw and δθ, are completely technique for the first and second terms) yields
arbitrary, except at x=L they must be zero, from
( )
Equation (8) one can extract a set of governing ( ) . /
( ) ( )
differential equations and boundary conditions as
(14)
follows: ( )

( ) ( ) (9a)
where Cθ is an integration constant. Imposing the
( ( )) (9b) rotation boundary condition, i.e. the first equation of

91
 Wong, F.T. et al. / On the Derivation of Exact Solutions / CED, Vol. 21, No. 2, September 2019, pp. 89–96

Equation (9d), and substituting the resulting Cθ into ( ) ( ) (19e)

Equation (14) yield the rotation field
( ) . ( )/ (19f)
( ) ( ( ) ) ( ) ( ) ( )
( ) ( )
Here, wbq and wsq are the bending and shear con-
( )
( ) ( ) tributions to the deflection due to the distributed
.( / (15) load q, respectively; wbP and wsP are the bending and
) ( ) shear contributions to the deflection due to the point
load P0, respectively; and wbM is the deflection due to
This solution is not applicable in the case of pris- the moment M0.
matic beams since a = 0 causes a division by zero.
The solution for prismatic beams can be obtained by The deflection field for prismatic beams can be
integrating Equation (13) with a = 0, the result of obtained by substituting θ in Equation (17) by
which is Equation (16), integrating the equation with setting
a = 0, and imposing the deflection boundary
( ) ( ) ( ) ( ) (16) condition. The result is
( ) ( ) ( ) ( ) ( )
Dividing Equation (11) by GAs, substituting As,
( ) (20a)
Equation (10b), and placing θ(x) in the right hand
side yield ( ) . /; ( )
(17) ( ) (20b)

( ) . /; ( )
Substituting θ in Equation (17) by Equation (15) and
integrating the equation (using the integration by ( ) (20c)
parts for the first term of θ) give
( ) ( ) . / (20d)
( )
( )
( )
( ) The deflection and rotation at the free end can be
( ) ( ) obtained by substituting for x = 0 into Equation (19)
( )
( ( ) ) and Equation (15) for the case of tapered beams and
( ) ( )
into Equation (20) and Equation (16) for the case of
(
( ) ( )
) prismatic beams, respectively. The results are, after
simplifying and substituting for α = 1+aL, presented
( ( )) in Tables 1-4.
( ) (18)
Table 1. Tip Deflections of the Tapered Cantilever Beam
Subjected to M0, P0, and q
where e = 2.71828… is the Euler number and Cw is
an integration constant. Imposing the deflection Component
Load
boundary condition, i.e. the second equation of Bending Shear
M0 N.A.
Equation (9d), and substituting the resulting Cw into
Equation (18), yield the deflection field
P0
( )
( ) ( ) ( ) ( ) ( )
q ( )
( ) (19a) (

( ) )
( ) ( )
. . / ( )
( ) ( ) Table 2. Tip Rotations of the Tapered Cantilever Beam
( )/ (19b) Subjected to M0, P0, and q
( ) ( )
Load Rotation
( ) . ( ) . // (19c) M0

P0
( ) . ( ) ( ) ( )

( )/ (19d) q
( ) ( )

92
 Wong, F.T. et al. / On the Derivation of Exact Solutions / CED, Vol. 21, No. 2, September 2019, pp. 89–96

Table 3. Tip Deflections of the Prismatic Cantilever Beam 3. The bending contribution to the tip deflections of
Subjected to M0, P0, and q prismatic beams due to M0, P0, and q (Table 3)
Component are the same as the tip deflection given by the
Load
Bending Shear classical beam theory presented in Ghali and
M0 N.A. Neville [14].
4. For a given constant h0, as the beam becomes
P0 very slender, i.e., as L→∞, the beam approaches
the prismatic beam, i.e., a→0. From Table 3, it is
q evident that for very slender beam the deflection
is dominated by the bending contribution and
hence the deflections converge to the classical
Table 4. Tip Rotations of the Prismatic Cantilever Beam Euler-Bernoulli deflections.
Subjected to M0, P0, and q
Rotation Examples of Application
M0
Consider symmetric cantilever beams of the slender-
P0 ness ratios L/hL = 10 (a relatively slender beam) and
L/hL = 2.5 (a deep beam). The lengths of the beams
q are L = 4 m and L = 1 m for the slender and deep
ones, respectively. The height at the fixed end is hL =
0.4 m; the width is b = 0.2 m. For the present para-
Remarks metric study, the end-thickness ratios α = hL/h0 are
1. Since the deflection and rotation for the case of varied from 1 (prismatic beam), 1.5, 2, and 3. Figure
tapered beams should converge to those of pris- 2 illustrates the slender beam vs. the deep beam for
matic beams when the coefficient a approaches 0, α = 2. Table 5 summarizes the geometric parameters
the results in Tables 3 and 4 can be obtained by of the eight different cases considered. In this table,
taking the limit of the corresponding results in φ is the inclination angle of the top or bottom faces of
Tables 1 and 2 as the coefficient a approaches 0. the beam (tapering angle). The material properties
This is obvious when the coefficient a = 0 in are E = 20000 MPa and ν = 0.2, which are typical
Tables 1 or 2 does not cause a division by zero. properties of normal concrete. The loads are taken as
For the other results, however, one should per- follows: M0 = 10 kN-m, P0 = 10 kN, and q = 10 kN/m
form a limit calculation. For example, according for the slender beam and q = 40 kN/m for the deep
to Table 1, the shear deflection due to P0 is beam (the total load is taken to be equal, that is, 40
kN).
( ) ( ) (21)

Taking the limit as a approaches 0 and applying 0.2 m 0.4 m

the l‟Hospital rule,
( ) ( ) (22a)
( )
( ) 0.2 m 0.4 m

(22b)

This result is identical to the shear deflection of Figure 2. Symmetrically Tapered Beams with the end-
the prismatic beam due to P0 given in Table 3. Thickness Ratios α = 2 (scaled): (a) Slender Beam, (b) Deep
The convergence of the tapered beam solutions Beam
towards prismatic beam solutions as the coeffi-
cient a approaches zero will also be demonstrated Table 5. Geometric Parameters used in the Case Study
in the following numerical examples.
L α hL h0 a φ
2. Equation (15) and (16) and also Tables 2 and 4 (m) (m) (m) (/m) (deg)
show that there is no contribution of the shear 4 1 0.400 0.400 0 0.00
deformation to the cross section rotations. 4 1.5 0.400 0.267 0.125 0.95
Therefore, the rotation formulas are the same as 4 2 0.400 0.200 0.25 1.43
those given by the classical Euler-Bernoulli beam 4 3 0.400 0.133 0.5 1.91
theory. For example, the prismatic beam dis- 1 1 0.400 0.400 0 0.00
placement table in Ghali and Neville [14] gives 1 1.5 0.400 0.267 0.5 3.80
the same results for the free end rotation as those 1 2 0.400 0.200 1 5.65
given in Table 4. 1 3 0.400 0.133 2 7.47

93
 Wong, F.T. et al. / On the Derivation of Exact Solutions / CED, Vol. 21, No. 2, September 2019, pp. 89–96

Numerical Results

Deflections of the slender beam and deep beam due

to end moment M0 for different cases of end-height
ratios α are presented in Figures 3 and 4, res-
pectively, due to end concentrated load P0 in Figures
5 and 6, respectively, and due to distributed load q in
Figures 7 and 8, respectively. It is seen that the
deflections of the slender beam are, as expected,
much larger than those of the deep beam. The tip
deflections of the slender beam subjected to M0 are
16 times larger, while those subjected to P0 and q
(with equal total load) are about 56 to 60 times
larger. As also expected, the thinner the cross-section
at the free end (or the larger the end-thickness ratios
α), the larger the deflections. For the beams Figure 5. Deflection of the Beam of Length 4 m Subjected
subjected to M0, the tip deflections become twice to P0 for Different Values of end Height Ratios α. Values in
larger when the free-end cross sections are twice the Legend Indicate the Deflection at the Left End.
thinner. While for the beams subjected to P0 and q,
the tip deflections are about 1.6 and 1.4 times larger,
respectively when the free-end cross sections are
twice thinner.

Figure 6. Deflection of the Beam of Length 1 m Subjected

to P0 = 10 kN for Different Values of end Height Ratios α.
Values in the Legend Indicate the Deflection at the Left
Figure 3. Deflection of the Beam of Length 4 m Subjected End.
to M0 = 10 kN-m for Different Values of End Height Ratios
α. Values in the Legend Indicate the Deflection at the Left
End.

Figure 4. Deflection of the Beam of Length 1m Subjected Figure 7. Deflection of the Beam of Length 4 m Subjected
to M0 = 10 kN-m for Different Values of End Height Ratios to q = 10 kN/m for Different Values of end Height Ratios α.
α. Values in the Legend Indicate the Deflection at the Left Values in the Legend Indicate the Deflection at the Left
End. End.

94
 Wong, F.T. et al. / On the Derivation of Exact Solutions / CED, Vol. 21, No. 2, September 2019, pp. 89–96

Table 7. Tip Deflections of the Tapered Cantilever Beam

as α→1 for the Case of L = 4 m
α wbM(0) wbP(0) wsP(0) wbq(0) wsq(0)
1.5 5.625 13.43 8.636E-02 18.88 0.1611
1.2 4.500 11.44 7.767E-02 16.69 0.1506
1.1 4.125 10.74 7.444E-02 15.87 0.1465
1.05 3.938 10.37 7.275E-02 15.44 0.1443
1.01 3.788 10.07 7.135E-02 15.09 0.1425
1.001 3.754 10.01 7.104E-02 15.00 0.1420
1 3.750 10.00 7.100E-02 15.00 0.1420

Table 8. Tip Deflections of the Tapered Cantilever Beam as

α→1 for the Case of L = 1 m

Figure 8. Deflection of the Beam of Length 1 m Subjected α wbM(0) wbP(0) wsP(0) wbq(0) wsq(0)
to q = 40 kN/m for Different Values of End Height Ratios α. 1.5 0.3516 0.2098 2.159E-02 0.2950 4.027E-02
Values in the Legend Indicate the Deflection at the Left 1.2 0.2813 0.1788 1.942E-02 0.2608 3.766E-02
End. 1.1 0.2578 0.1677 1.861E-02 0.2480 3.663E-02
1.05 0.2461 0.1621 1.819E-02 0.2413 3.608E-02
1.01 0.2367 0.1574 1.784E-02 0.2358 3.562E-02
Contribution of Shear Deformation
1.001 0.2346 0.1564 1.776E-02 0.2344 3.551E-02
1 0.2344 0.1563 1.775E-02 0.2344 3.550E-02
For the beams subjected to end moment M0, it is
evident from Equations (19f) and (20d) and from Comparison with the Solutions of Other Beam
Tables 1 and 3 that there is no contribution of shear Models
deformation to the deflections. Contribution of shear
deformation to the tip deflections for the beams To assess the validity of the present Timoshenko
subjected to concentrated load P0 and q is presented beam model, the tip deflections at the free end for
in Table 6. The contribution of shear deformation for the case of the beams with the height ratio of α = 2
the slender beams (L/hL = 10) is, as expected, (see Figures 3-8) are compared to the solutions based
practically insignificant. While for the deep beams on the beam model presented by Romano [5] and the
(L/hL = 2.5), the contribution of shear deformation plane stress model (Table 9). The tip deflections of
cannot be neglected. It is seen that the contribution Romano [5] were obtained by applying the analytical
solution of Romano [5] to the cantilever beam
of shear deformation multiplies about 14 to 15 times
considered in this paper, that is,
when the slenderness, L/hL, decreases four times
(that is, from L/hL = 10 to L/hL = 2.5). This in ( ) ( ) ( )
agreement with the fact that the upper bound of the
shear deformation contribution is proportional to the ( ) ( ) (23a)
square of L/hL, which is 16 for this case.
(23b)
( )
Table 6. Percentages of Shear Deformation Contribution to
the Tip Deflections of Slender and Deep Beams Subjected where C1, C2, C3, and C4 are integration constants
to P0 and q that were determined from the boundary conditions:
P0 q Q(0) = −P0; M(0) = −M0 (following the sign conven-
α tion in Romano [5]); w(L) = 0; θ(L) = 0. While, the tip
L=4m L=1m L=4m L=1m
1 0.70% 10.20% 0.94% 13.15% deflections of the plane stress model were obtained
1.5 0.64% 9.33% 0.85% 12.01% from finite element analyses using the mesh of
2 0.60% 8.78% 0.79% 11.31% 160×8 and 80×16 quadrilateral elements for the
3 0.55% 8.10% 0.72% 10.46% beam of the slenderness ratio L/h = 10 and L/h = 2.5,
respectively. The element used was the four-node
Convergence towards Prismatic Beams quadrilateral element including bending incompa-
tible modes contained in software SAP2000 [15].
To study the convergence of the deflection solutions
of tapered beams, Equation (19), towards the Table 9. Tip Deflections (mm) of the Tapered Cantilever
solutions of the prismatic beam, Equation (20), the Beams for α = 2
values of the tip deflections, Table 1, are numerically
Load Slenderness Romano Plane
examined as the values of α approaches 1 (that is, case ratio
Present
[5] stress
equivalent to coefficient a approaches 0). Tables 7 L/h = 10 7.50 7.43 7.49
and 8 present the tip deflections for the cases of the M0
L/h = 2.5 0.469 0.401 0.465
slender and deep beams, respectively. It is seen that, L/h = 10 16.45 16.51 16.43
P0
as expected, all of the solutions of the tapered beams L/h = 2.5 0.280 0.294 0.276
converge well towards the solutions of the prismatic q
L/h = 10 22.04 22.01 22.00
beams as α approaches 1. L/h = 2.5 0.385 0.376 0.379

95
 Wong, F.T. et al. / On the Derivation of Exact Solutions / CED, Vol. 21, No. 2, September 2019, pp. 89–96

The table shows that for the slender beam of L/h = 5. Romano, F., Deflections of Timoshenko Beam
10, there is no significant difference among the with Varying Cross-Section, International Jour-
results obtained using different models (the diffe- nal of Mechanical Sciences, Elsevier, 38(84),
rence among the results is less than 1%). For the 1996, pp. 1017–1035.
deep beam of L/h = 2.5 subjected to moment M0 and 6. Al-Gahtani, H.J. and Khan, M.S., Exact Ana-
concentrated load P0, the present solutions are closer lysis of Nonprismatic Beams, Journal of Engi-
to the plane stress results than those of Romano [5]. neering Mechanics, ASCE, 124(11), 1998, pp.
For the deep beam subjected to distributed load q, 1290–1293.
however, Romano‟s solutions are closer. The overall 7. Boley, B.A., On the Accuracy of the Bernoulli-
results of this comparison demonstrate the validity Euler Theory for Beams of Variable Section,
of the present beam model. Journal of Applied Mechanics, ASME, 30(3),
1963, pp. 373–378.
Conclusions 8. Mercuri, V., Balduzzi, G., Asprone, D. and
Auricchio, F., 2D Non-prismatic Beam Model for
Exact solutions of the Timoshenko beam model for Stiffness Matrix Evaluation, Proceedings of the
cantilever tapered and prismatic beams subjected to World Conference on Timber Engineering,
an end moment, an end concentrated force, and a Vienna, Austria, August 22-25, 2016.
uniformly distributed force have been derived by 9. Bathe, K. J., Finite Element Procedures, Pren-
solving the governing differential equations. The tice-Hall, New Jersey, 1996.
solutions include the shear force and bending mo- 10. Cowper, G.R., The Shear Coefficient in Timo-
ment distributions, which are in agreement with the shenko‟s Beam Theory, Journal of Applied
static principle, as well as the rotation and deflection
Mechanics, ASME, 33(2), 1966, pp. 335–340.
fields. Based on these solutions, the formulas for
11. Hughes, T.J.R., The Finite Element Method:
computing the tip deflections and rotations have
Linear Static and Dynamic Finite Element
been presented. The numerical examples demon-
strate the validity of the present solutions. The Analysis, Prentice-Hall, New Jersey,1987.
solutions and formulas may be useful for evaluating 12. Garikipati, K., Introduction to Finite Element
a numerical method such as Timoshenko beam finite Methods, Open Michigan, 2013. Available:
elements [16,17], as well as for evaluating the deflec- https://open.umich.edu/find/open-educational-
tion of a tapered cantilever beam in real engineering resources/engineering/introduction-finite-
practice. element-methods. [Accessed: 23-Jan-2019].
13. Cook, R.D., Malkus, D.S., Plesha, M.E. and Witt,
References R.J., Concepts and Applications of Finite Ele-
ment Analysis, Fourth edition, John Wiley and
1. Balduzzi, G., Aminbaghai, M., Sacco, E., Füssl, J., Sons, New York, 2002.
Eberhardsteiner, J., and Auricchio, F., Non-pris- 14. Ghali, A. and Neville, A., Structural Analysis: A
matic Beams : A Simple and Effective Timoshen- Unified Classical and Matrix Approach, Second
ko-like Model, International Journal of Solids edition, Chapman and Hall, London, 1978.
and Structures, Elsevier, 90, 2016, pp. 236–250. 15. CSI Analysis Reference Manual for SAP2000,
2. Auricchio, F., Balduzzi, G., and Lovadina, C., The ETABS, and SAFE, Computers and Structures
Dimensional Reduction Approach for 2D Non- Berkeley, California, 1995.
prismatic Beam Modelling : A Solution based on 16. Wong, F.T. and Sugianto, S., Study of the Dis-
Hellinger–Reissner Principle, International Jour- crete Shear Gap Technique in Timoshenko
nal of Solids and Structures, Elsevier, 63, 2015, Beam Elements, Civil Engineering Dimension,
pp. 264–276. Petra Christian University, 19(1), 2017, pp. 54–
3. Fertis, D.G. and Keene, M.E., Elastic and in-
62.
elastic Analysis of Nonprismatic Members, Jour-
17. Wong, F.T., Sulistio, A., and Syamsoeyadi, H.,
nal of Structural Engineering, ASCE, 116(2),
Kriging-Based Timoshenko Beam Elements
1990, pp. 475–489.
4. Romano, F. and Zingone, G., Deflections of Beams with the Discrete Shear Gap Technique, Inter-
with Varying Reactangular Cross Section, Jour- national Journal of Computational Methods,
nal of Engineering Mechanics, ASCE, 118(10), World Scientific, 15(07), 2018, pp. 1850064-1–
1992, pp. 2128–2134. 1850064-27.

96

View publication stats