Nonlinear Vibration Solution For An Inclined Timoshenko Beam Under The Action of A Moving Force With Constant/Nonconstant Velocity
Nonlinear Vibration Solution For An Inclined Timoshenko Beam Under The Action of A Moving Force With Constant/Nonconstant Velocity
3, September, 2014
This study is focused on the nonlinear dynamic response of an inclined Timoshenko beam with different
boundary conditions subjected to a moving force under the influence of three types of motions, including
accelerating, decelerating and constant-velocity types of motion. The nonlinear governing coupled partial
differential equations (PDEs) of motion for the bending rotation of the warped cross section of the beam
and its longitudinal and transverse displacements are derived by using Hamilton’s principle.
To obtain the dynamic response of the beam under the action of a moving force, the derived nonlin-
ear coupled PDEs of motion are solved by applying Galerkin’s method. Then the dynamic response of the
beam is obtained using the mode summation technique. Furthermore, the calculated results are verified
by the results obtained by finite-element method (FEM) analysis. In the next step, a parametric study of
the response of the beam is conducted by changing the magnitude of the traveling concentrated force, its
velocity and boundary conditions for the beam. Similarly, their sensitivity to the dynamic response of the
beam is also studied. It is observed that the existence of quadratic-cubic nonlinearity in the governing
coupled PDEs of motion renders the hardening/softening behavior in the dynamic response of the beam.
Moreover, we note that any restriction imposed on stretching of the mid-plane of the beam introduces a
nonlinear behavior in the PDEs of motion of the beam.
1. Introduction
Under the actual operating conditions, the linear and nonlinear vibration analysis of structural elements, such
as strings, rods, beams, plates, and shells traveling by a moving mass/force is of considerable practical importance
for the structural and railway engineers. For over a century many analytic and numerical methods have been
proposed to investigate the dynamic behavior of different engineering structures. However, until now, almost no
attention has been given to the study of nonlinear dynamic analysis for the coupled bending rotation of the warped
cross section and longitudinal and transverse deflections of an inclined Timoshenko beam subjected to a moving
force. Some practical examples of behavior of this kind are given by a bridge crossed by moving vehicles or trains,
an overhead traveling crane moving on its girder, a beam subjected to pressure waves, as well as by the simulation
of high-axial-speed machining operations and internal two-phase flow in piping systems. Further applications on
the subject of vibrations of inclined beams can be addressed in the aerospace and armed force industries, such as
rocket launcher systems and cannon tubes.
It should be emphasized that from the point of view of design of the mechanical components, the inclusion of
the acceleration/deceleration character of moving masses/forces certainly plays a significant role in the final results.
This becomes more critical when one deals, e.g., with the take-off and landing phases of aircrafts on runways and
1 Parand Branch, Islamic Azad Univ., Tehran, Iran; e-mail: am [email protected].
2 Sharif Univ. Technology, Tehran, Iran.
3 Tarbiat Modares Univ., Tehran, Iran.
Published in Neliniini Kolyvannya, Vol. 16, No. 3, pp. 385–407, July–September, 2013. Original article submitted December 16, 2011;
revision submitted July 10, 2012.
1072-3374/14/2013–0361 c 2014
� Springer Science+Business Media New York 361
362 A. M AMANDI , M. H. K ARGARNOVIN , AND S. FARSI
(a) (b)
Fig. 1. (a) Lateral force traveling on an inclined pinned-pinned beam; (b) forces acting upon the elastic beam at the contact point in the
equivalent moving force model.
angle decks (flight decks) of warship aircraft carriers, with automobiles and locomotives at take off or sudden
braking on the roadways and highway bridges and on the rails and railway bridges, respectively. Furthermore,
frequent braking and accelerating of rail-guided cranes play important role in the stages of design of these types of
structures.
As for the vibration analysis of the Euler–Bernoulli beam, either under motion of a traveling force (or traveling
mass) numerous works are reported [1–12]. Similarly, one can find a number of different studies on the dynamical
behavior of Timoshenko beams subjected to the motion of moving loads and masses [7, 10, 11, 13–18]. The
nonlinear dynamic analyses of Euler–Bernoulli and Timoshenko beams under the action either of moving forces
or of moving masses are specifically performed in [4, 8, 9, 12, 14, 16–18].
From the experimental point of view, it appears that, as the amplitude of vibration increases, the nonlinear
effects come into play; therefore, by considering that the source of nonlinearity may be either inertial, geometric,
or material in its nature, the influence of these terms on the dynamic behavior of the beam should also be taken
into account (see Refs. [8, 14, 19]). In the present paper, our attention is given to the geometric nonlinearity which
may be caused by large curvatures and nonlinear stretching of the mid-plane of the inclined Timoshenko beam.
In general, due to existence of nonlinear terms, the exact analytic (closed form) solutions are not available for the
governing equations of motion.
In the present study, three nonlinear governing coupled PDEs of motion for the bending rotation of warped
cross section and transverse and longitudinal vibrations of an inclined Timoshenko beam subjected to the action of
a moving force are derived using Hamilton’s principle. Then, by applying Galerkin’s method, the three obtained
nonlinear second-order ordinary differential equations (ODEs) governing the modal equations are solved numeri-
cally by using the Adams–Bashforth–Moulton integration method via the MATLAB solver package. It should be
noted that, in the present study, the nonlinear effects of axial strain, bending curvature, and shear strain on the
dynamic responses of the inclined Timoshenko beam are all considered using the von-Karman strain-displacement
relation in combination with the moderately large deflection theory.
In extending the issue of moving mass further to a more applicable study, we believe that the same problem
(but for the motion of a moving force) has its own importance in this field. Based on this postulate and on the
same line to the other studies by the same authors [18], our study is initiated. It should be mentioned that there is
no novelty in the technique of solution used in the present paper as compared with some previous works but what
makes this work new is related to the outcome results very important from the viewpoint of applications. Briefly,
it should be pointed out that the main contribution and significant technical advantages of this paper are to present
some tangible results that have not been reported in the earlier published papers.
N ONLINEAR V IBRATION S OLUTION FOR AN I NCLINED T IMOSHENKO B EAM UNDER THE ACTION OF A M OVING F ORCE 363
2. Mathematical Modeling
We consider an inclined Timoshenko beam with length l and inclination angle ' traveled by a concentrated
force F with a velocity v and a constant acceleration/deceleration a (see Fig. 1a). The longitudinal and lateral
components of the force with which the traveling force acts on the beam are F cos ' and F .sin ' C � cos '/;
respectively, as shown in Fig. 1b. In our upcoming analysis, when the load enters the left end of the beam, the zero
initial conditions are assumed for the beam, i.e., the beam is at rest at time t D 0: It is further assumed that the
external damping is not negligible and the damping behavior follows the viscous nature [2, 8, 10, 11]. Moreover,
the beam deforms within the linear elastic range and, therefore, Hooke’s laws are prevailing. In our study, the
von-Karman’s moderately large strain-displacement relations are used [14, 18]:
1 2
"xx D u;x C w;x ; w;x D C� and D ;x : (1)
2
Here, u D u.x; t / is the axial longitudinal time-dependent in-plane displacement, w D w.x; t / is the time-
dependent transverse deflection of the beam measured upward from its equilibrium position when unloaded, and
D .x; t/ is the time-dependent rotation of the warped cross section of the beam due to bending. The subscripts
( ; t ) and ( ; x ) stand for the derivatives with respect to time ( t ) and spatial coordinate ( x ) in the related order,
respectively. In addition, "; �; and are the longitudinal (or normal) strain, shear strain, and curvature at
the center line of the Timoshenko beam, respectively. To obtain the governing differential equations of motion
by applying Hamilton’s principle, the kinetic energy K and the strain energy U of the beam are given by the
formulas:
0 l 1
Z Zl Zl
1 @
KD 2 2
⇢Aw;t dx C ⇢Au;t dx C ⇢Id ;t
2
dx A ; (2a)
2
0 0 0
0 l 1
Z Zl Zl
1
U D @ EA"2xx dx C EId 2 dx C kGA� 2 dx A : (2b)
2
0 0 0
The total external virtual work done by the traveling force with variable velocity, frictional force, and external
viscous damping forces acting on the beam is:
Zl
� �ˇ
ıWe D � F .sin ' C � cos '/ıu C F cos 'ıw C c1 ;t ı C c2 w;t ıw C c3 u;t ıu ˇxD⇣ .t / dx; (3)
0
where c1 ;t ; c2 w;t ; and c3 u;t are regarded as the external viscous damping forces applied to the beam [2, 8–
11]. We can now establish the Lagrangian function of the system as: L D K � .U � We /: Applying Hamilton’s
principle to L as:
Zt2
ı L dt D 0
t1
364 A. M AMANDI , M. H. K ARGARNOVIN , AND S. FARSI
or
Zt2 Zt2
ı .U � K/ dt D ıWe dt: (4)
t1 t1
By doing some mathematics, one would get the nonlinear governing coupled PDEs of motion (EOMs) as well
as the boundary conditions (BCs) for the analyzed problem in the following form:
the relation for moments in the direction :
⇢Au;t t � EA.u;xx C w;x w;xx / C c3 u;t D �F .sin ' C � cos '/ı .x � ⇣.t // �.t /; (7)
and the boundary conditions at both ends for the Timoshenko beam:
where ⇢ is the density of the beam, A is the cross-sectional area of the beam, Id is the cross-sectional second
moment of inertia of the beam, E is Young’s modulus of elasticity, G is the shear modulus, k is the shear
correction factor, EId is the flexural rigidity of the beam, ⇢A is its mass per unit length, F is the magnitude of
the traveling force, � is the kinetic frictional coefficient, c1 ; c2 ; and c3 are external damping constants related
to the viscous damping of the beam, namely, to ⌘; and also M; Q; and N are the bending (flexural) moment,
shear force, and axial (longitudinal) tension/compression force of the beam, respectively. Furthermore, ı.x � ⇣.t //
is Dirac’s delta function in which ⇣.t / is the instantaneous position of the force moving with the velocity v and
a constant acceleration/deceleration a on the beam such that ⇣.t / D x0 C vt or ⇣.t / D x0 C vt C 1=2at 2 for
describing the constant-velocity or accelerating/decelerating types of motion of the traveling force, respectively,
where x0 is the initial point of application of the force on the beam, v is the entrance/exit velocity of the traveling
force, a is the constant acceleration/deceleration of the traveling force on the beam, and �.t / is the pulse function
equal to one as the force is traveling on the beam and to zero when the traveling force is outside the beam span;
N ONLINEAR V IBRATION S OLUTION FOR AN I NCLINED T IMOSHENKO B EAM UNDER THE ACTION OF A M OVING F ORCE 365
thus, the case of constant-velocity type of motion is described by the formula �.t / D u.t / � u.t � l=v/; where
u.t / represents the unit step function. The derivation of nonlinear governing coupled PDEs of motion is rather
lengthy and, for the sake of brevity, the details are not presented here.
3. Solution Method
In the present study, Galerkin’s method is chosen as a powerful computational tool to analyze the vibrations of
an inclined Timoshenko beam. Based on the technique of separation of variables, the response of the Timoshenko
beam in terms of the linear free-oscillation modes can be represented as follows [18, 21]:
n
X
w.x; t / D �j .x/pj .t / D T .x/P.t /; (9)
j D1
n
X
.x; t / D ⌧j .x/qj .t / D T .x/Q.t /; (10)
j D1
n
X
u.x; t / D ✓j .x/rj .t / D T .x/R.t /; (11)
j D1
where P.t/; Q.t/; and R.t / are vectors of order n listing the generalized coordinates pj .t /; qj .t /; and rj .t /;
respectively, and .x/; .x/ and .x/ are vector functions collecting the first n mode shapes (eigenfunctions) of
�j .x/; ⌧j .x/ and ✓j .x/; respectively.
By substituting Eqs. (9), (10), and (11) into Eqs. (5), (6), and (7), premultiplying both sides of Eqs. (5),
(6), and (7) by T .x/; T .x/ and T .x/; respectively, integrating over the interval .0; l/ and using the properties
of Dirac’s delta function, we obtain the following resulting nonlinear coupled modal equations of motion in the
matrix form:
n
X n
X n
X ⇥ ⇤
⇢Id Sij qRj .t / C 2⇢Id ⌘!i Sij qPj .t / � EId Kij � kGASij qj .t /
j D1 j D1 j D1
n
X
� kGA Eij pj .t / D 0; (12)
j D1
n
X n
X n
X n
X
⇢A Mij pRj .t / C 2⇢A⌘!i Mij pPj .t / � kGA Hij pj .t / C kGA Fij qj .t /
j D1 j D1 j D1 j D1
n X
X n n X
X n
� EA rj .t /Gij k pk .t / � EA pj .t /Tij k rk .t /
j D1 kD1 j D1 kD1
XX n n
3
� EA pj .t /Iij k pk .t /2 D �F cos '�.t /bi .t /; (13)
2
j D1 kD1
366 A. M AMANDI , M. H. K ARGARNOVIN , AND S. FARSI
and
n
X n
X n
X
⇢A Jij rRj .t / C 2⇢A⌘�li Jij rPj .t / � EA Nij rj .t /
j D1 j D1 j D1
n X
X n
� EA pj .t /Lij k pk .t / D �F Œsin ' C � cos 'ç �.t /di .t / (14)
j D1 kD1
Zl Zl
.M/ij D �i .x/�j .x/ dx; .H/ij D �i .x/�j00 .x/ dx;
0 0
Zl Zl
.F/ij D �i .x/⌧j0 .x/ dx; .S/ij D ⌧i .x/⌧j .x/ dx;
0 0
Zl Zl
.K/ij D ⌧i .x/⌧j00 .x/ dx; .E/ij D ⌧i .x/�j0 .x/ dx;
0 0
(15)
Zl Zl
.J/ij D ✓i .x/✓j .x/ dx; .N/ij D ✓i .x/✓j00 .x/ dx;
0 0
Zl Zl
�i .x/✓j00 .x/�k0 .x/ dx; �i .x/�j00 .x/�k0 .x/ dx;
2
.G/ij k D .I/ij k D
0 0
Zl Zl
.L/ij k D ✓i .x/�j00 .x/�k0 .x/ dx; .T/ij k D �i .x/�j00 .x/✓k0 .x/ dx;
0 0
where i; j; k D 1; 2; 3; : : : ; n:
The primes and dots over any parameter indicate the derivatives with respect to the coordinate .x/ and time
.t / , respectively. Furthermore, the n ⇥ 1 column vectors of b and d are defined as
It is clear that Eqs. (12), (13), and (14) are three nonlinear coupled second-order ordinary differential equations
(ODEs). The boundary conditions for the pinned-pinned Timoshenko beam with fixed end supports are [18, 20,
21]:
º
u.0; t / D u.l; t / D 0 ) ✓j .x/ D 0 for x D 0 and l;
essential BCs W
w.0; t / D w.l; t / D 0 ) �j .x/ D 0 for x D 0 and lI
(17)
natural BCs W M.0; t / D M.l; t / D 0 ) ⌧j;x .x/ D 0 for x D 0 and l:
N ONLINEAR V IBRATION S OLUTION FOR AN I NCLINED T IMOSHENKO B EAM UNDER THE ACTION OF A M OVING F ORCE 367
The boundary conditions for the clamped-pinned Timoshenko beam with immobile end supports are as follows
[18, 20, 21]:
8̂
ˆ u.0; t / D u.l; t / D 0 ) ✓j .x/ D 0 for x D 0 and l;
ˆ
<
essential BCs W w.0; t / D w.l; t / D 0 ) �j .x/ D 0 for x D 0 and l;
ˆ
ˆ
:̂ .0; t / D 0 ) ⌧ .x/ D 0 (18)
j for x D 0I
Finally, the boundary conditions for the clamped-free Timoshenko beam are as follows [18, 20, 21]:
8̂
ˆ u.0; t / D 0 ) ✓j .x/ D 0 for x D 0;
ˆ
<
essential BCs W w.0; t / D 0 ) �j .x/ D 0 for x D 0;
ˆ
ˆ
:̂ .0; t / D 0 ) ⌧ .x/ D 0 for x D 0I
j
8̂ (19)
ˆ M.l; t / D 0 ) ⌧j;x .x/ D 0 for x D l;
ˆ
<
natural BCs W Q.l; t / D 0 ) kGA.w;x � / D 0 for x D l;
ˆ
ˆ
:̂ N.l; t / D 0 ) EA.u
;x / D 0 for x D l:
Moreover, the initial conditions (ICs) for the Timoshenko beam are as follows:
ICs W u.x; 0/ D u;t .x; 0/ D w.x; 0/ D w;t .x; 0/ D .x; 0/ D ;t .x; 0/ D 0: (20)
In Eq. (14), �li denotes the natural angular frequency (rad/sec) of longitudinal vibrations of the beam related
i⇡ p
to its type of boundary conditions. For the pinned-pinned or clamped-pinned beam, it has the form E=⇢ [18,
l
.2i � 1/⇡ p
20, 21]. At the same time, for a cantilever beam, it is given by the formula E=⇢ [18, 20, 21], where
2l
i D 1; 2; 3; : : : ; n:
To solve Eqs. (12)–(14), all entries in the matrices in Eqs. (15) and (16) should be calculated. Thus, by
inspection, it can be seen that the following functions (mode shapes) for �j .x/; ⌧j .x/; and ✓j .x/ satisfy both
the linearized equations of motion of the beam and different types of boundary conditions (see [13, 18, 20, 21]):
(i) For the pinned-pinned Timoshenko beam, the normal modes are expressed as follows:
�j .x/ D D sin.bB⇠/;
x
and ✓j .x/ D sin.j⇡⇠/; ⇠D ; with j D 1; 2; 3; : : : ; n;
l
sin bB D 0; (21b)
368 A. M AMANDI , M. H. K ARGARNOVIN , AND S. FARSI
�j .x/ D DŒcosh bA⇠ � coth bA sinh bA⇠ � cos bB⇠ C cot bB sin bB⇠ç;
h � i
⌧j .x/ D H cosh bA⇠ C sinh bA⇠ � cos bB⇠ C � sin bB⇠ ; (22)
�Z
x
and ✓j .x/ D sin.j⇡⇠/; ⇠D ; with j D 1; 2; 3; : : : ; n;
l
(iii) For the clamped-free Timoshenko beam, the normal modes are
�j .x/ D DŒcos bA⇠ � �ZÅ sinh bA⇠ � cos bB⇠ C Å sin bB⇠ç;
h � i
⌧j .x/ D H cosh bA⇠ C sinh bA⇠ � cos bB⇠ C � sin bB⇠ ; (24)
�Z
x
and ✓j .x/ D sinŒ.j � 1=2/⇡⇠ç; ⇠D ; with j D 1; 2; 3; : : : ; n;
l
b.r 2 C s 2 /
2 C Œb 2 .r 2 � s 2 /2 C 2ç cosh bA cos bB � 1=2
sinh bA sin bB D 0; (25)
.1 � b 2 r 2 s 2 /
⇢Al 4 2 Id EId B 2 � s2 A
b2 D ! ; r2 D ; s2 D ; ZD ; �D ;
EId Al 2 kGAl 2 A2 C s 2 B
º » 12
1 2 2 4 1
2 2 2
A; B D p ⌥.r C s / C Œ.r � s / C 2 ç 2 ; (26)
2 b
1
� sinh bA C sin bB �
sinh bA � sin bB
� D� ; ÅD ;
1
Z cosh bA C cos bB Z cosh bA C cos bB
and Di and Hi ; i D 1; 2; 3; : : : ; n; are normal modal amplitudes which depend on the natural frequencies
of the Timoshenko beam related to the type of BCs. Moreover, it should be pointed out that !i is the natural
angular frequency .rad=s/ of transverse/bending slope warping vibration of the beam, which depends on the type
of boundary condition and can be obtained from Eq. (26) [18, 20].
By inserting the corresponding normal modes for any type of boundary conditions of the beam in the Eqs.
(15) and (16), it is possible to find all entries in all matrices. In the next step, these evaluated matrices are inserted
N ONLINEAR V IBRATION S OLUTION FOR AN I NCLINED T IMOSHENKO B EAM UNDER THE ACTION OF A M OVING F ORCE 369
in Eqs. (12)–(14) and the indicated set of equations can be solved numerically by using the Adams–Bashforth–
Moulton integration method via the MATLAB solver package out of which the values of pn .t /; qn .t /; and rn .t /
can be obtained. By the backward substitution of pn .t /; qn .t /; and rn .t / in Eqs. (9)–(11), we establish the
quantities u.x; t/; w.x; t /; and .x; t / , respectively [18, 21].
In the next step, based on the obtained values of u.x; t / , w.x; t /; and .x; t / , the dynamic responses of the
inclined Timoshenko beam with three different types of boundary conditions, including pinned-pinned, clamped-
pinned, and clamped-free, are obtained under the influence of three types of force motions: (a) accelerating, (b)
decelerating, and (c) with a constant velocity. The obtained results for the response of the beam under each of these
three types of force motions and boundary conditions are shown separately in Figs. 4–14. The detailed kinematic
discussions of different motions introduced above are described in what follows [18].
a) In the case of a constant accelerating type of motion ( ⇣.t / D x0 C v0 t C 1=2at 2 ; a D const > 0 ), it is
assumed that the beam is at rest when the force F enters the beam at x0 D 0 and t0 D 0 with the initial velocity
v0 D 0 and that the force arrives at the other end of the beam, i.e., x D l with the final velocity v: The total
traveling time in the beam span t1 and the exit velocity v of the force are:
2l
t1 D ; (27a)
v
p
vD 2al: (27b)
b) For the constant decelerating type of motion ( ⇣.t / D x0 C v0 t C 1=2at 2 ; a D const < 0 ), it is also
assumed that the beam is at rest when the force F enters the beam at x0 D 0 and t0 D 0 with an entrance
velocity v0 (nonzero initial velocity); the force finally stops ( v D 0 ) at the other end of the beam, i.e., x D l:
For the total traveling time in the beam span t2 and the entrance velocity of the force v0 ; we get
2l
t2 D ; (28a)
v0
p
v0 D 2ljaj: (28b)
c) For the uniform-velocity type of motion ( ⇣.t / D x0 C v0 t; v0 D const > 0 ), it is also assumed that the
beam is at rest when the force F enters the beam at x0 D 0 and t0 D 0 with a constant velocity v0 I then the
force reaches the other end of the beam, i.e., x D l at time t3 : For the total traveling time in the beam span t3 ;
we obtain
l
t3 D : (29)
v0
As mentioned in the introduction, at present, no specific results are available in the literature for the problem
under consideration. Therefore, to verify the validity of the results obtained in the present study, it is necessary to
consider some special cases of our results and compare them with the data existing in the literature.
370 A. M AMANDI , M. H. K ARGARNOVIN , AND S. FARSI
Fig. 2. Instantaneous normalized vertical displacements under the action of the moving force F : linear analysis performed in the
present study (solid line); linear analysis carried out in Ref [13] (dash-dotted line).
4.1. Verification of the Results in the Linear Case. In the first attempt, we set the higher order terms, i.e.,
✓ ◆
3 2
u;xx w;x C u;x w;xx C w;xx w;x and .u;xx C w;x w;xx /
2
on the left-hand sides of Eqs. (6) and (7), respectively, equal to zero for a constant-velocity type of motion ( a D 0 ).
Furthermore, referring to Eqs. (12)–(14), we set: c1 D c2 D c3 D 0 , i.e., ⌘ D 0; � D 0; and ' D 0: This
leads us to a set of relations for u.x; t /; w.x; t /; and .x; t / corresponding to the case known for the linear
analysis of the horizontal undamped Timoshenko beam. To establish verifications of our analysis, we consider the
data given in [13], namely,
5
Id D 6:236 ⇥ 10 m4 ; A D 0:02736 m2 ; F D 0:2⇢g ⇥ Al.N /; ˛ D 0:11; and ˇ0 D 0:15;
where ˇ0 D ⇡ r0 = l is Rayleigh’s slenderness coefficient with r0 taken as the radius of gyration of the beam.
Based on the above data, the computer code was run for the linear case, the vertical displacements ( w ) of the
instantaneous positions of the moving force were found, and the dimensionless outcome results were depicted
and compared with the other existing results in Fig. 2. The normalization factor for the vertical displacements is
wst D F l 3 =48EId ; which is a mid-point deflection of a simply supported beam under a mid-span concentrated
load F: The close inspection of the curves presented in Fig. 2 reveals good agreement between the results.
4.2. Verification of the Results in the Nonlinear Case. As described earlier, in this study, to extend the
validity of the obtained results, we prepared an appropriate APDL (ANSYS Parametric Design Language) routine
in the environment of the ANSYS software to simulate the response to a moving force on an inclined Timoshenko
beam. Then the linear and nonlinear FEM solutions were compared with the solutions obtained from the linear and
nonlinear analytic solutions by applying the mode-summation technique. In modeling of the Timoshenko beam,
N ONLINEAR V IBRATION S OLUTION FOR AN I NCLINED T IMOSHENKO B EAM UNDER THE ACTION OF A M OVING F ORCE 371
Length (m) l 6
we used BEAM-188 elements defined in this software, which are suitable for the analysis of beam structures. An
element of this sort is a 2D 2-noded second-order beam element having 6 (or 7) DOFs with 3-translational DOF and
3-rotational DOF in each node. Thus, to perform our calculations we consider an inclined undamped Timoshenko
beam with the geometry and mechanical properties listed in Table 1.
Figure 3 illustrates the behavior of the mid-point deflection w .m/ of an inclined undamped pinned-pinned
Timoshenko beam with ' D 36ı vs. the dimensionless time vt = l for the velocity ratio ˛ D 0:25 and the
traveling force F D 2⇢gAl (N) under the influence of constant-velocity motion obtained by using the FEM
method and analytic analysis, respectively. From this figure, we can conclude that the results for the mid-point
lateral dynamic displacements of the beam obtained by using the FEM and analytic solutions either in the nonlinear
or in the linear analysis are almost identical, which reveals very good agreement between these analytic results
obtained by the mode-summation technique and by the FEM analysis. A suitable number of elements which
should be used for the beam to guarantee the convergence of linear/nonlinear results is equal to 80 elements.
4.3. Results and Discussions. In all cases studied in what follows, we use the data given in [18], as well as
the properties of the beam listed in Table 1.
To clarify the results and in order to get a better insight into interpreting the variations of the obtained results,
we try to present the results in the dimensionless form. Thus, we begin with defining the dimensionless dynamic
deflection w.xmax ; t /=w0 and the dimensionless time parameter t =ti ; i D 1; 2; and 3; where w0 and xmax
( D xjwmax ) denote the maximum static deflection and the point on the beam corresponding to this deflection,
respectively. It should be emphasized that w.xmax ; t / is obtained from the dynamic analysis of the governing
equations of motion at xmax : In Table 2, we show the values of w0 and the position of xmax due to the lateral
force, i.e., F; for three different types of boundary conditions.
372 A. M AMANDI , M. H. K ARGARNOVIN , AND S. FARSI
Fig. 3. Variations of mid-point deflection w .m/ vs. the normalized time vt = l for the inclined pinned-pinned Timoshenko beam with
' D 36ı affected by a moving force F D 2⇢gAl.N / in the case of constant-velocity motion for ˛ D 0:25 according to the
analytic and FEM analyses for linear and nonlinear solutions: analytic linear solution (—), analytic nonlinear solution (– –), (—
ı— ) ANSYS linear solution (80 elements), (—O— ) ANSYS nonlinear solution (80 elements).
Table 2. Values of xmax ( D xjwmax ) and w0 Caused by the Applied Lateral Force F for Different
Boundary Conditions on the Beam [2, 18] ( ' is the Inclination Angle of the Beam)
Position of theapplied
Absolute value
Beam geometry BCs lateral force F I
of w0
xmax .D xjwmax /
Fig. 4. Variations of the dimensionless dynamic lateral deflection ( wnl .xmax ; t /=w0 ) for xmax D 0:5l vs. the normalized time
( t=ti ) for the inclined pinned-pinned Timoshenko beam ( ' D 30ı ) traversed by the moving forces F D ⇢gAl.N / with
different velocity ratios ( ˛ D 0:25; 0:5; 0:75; and 1 ) under the influence of three types of motion: (a–c) dimensionless results
for the nonlinear analysis: wnl .xmax ; t /=w0 ; (d–f) dimensionless difference between the nonlinear and linear analyses (in %):
.wnl � wl /=w0 :
Moreover, we define the velocity ratio as ˛ D v=vcr : For the Timoshenko beam, vcr is the critical velocity
of the concentrated moving force acting on this beam. In the general form, it is defined as .vcr /Timo D !1 l=⇡
[10, 11, 18], where !i is the natural angular frequency (rad/sec) of transverse/bending slope warping vibrations
of this beam given by the formula [18, 20]
s
bi EId
.!i /Timo D 2
l ⇢A
with i D 1; 2; : : : ; n (see Eq. (26); please, also see [10, 11, 18]). Note that bi depends on the type of boundary
conditions and, e.g., for a simply supported Timoshenko beam, we get b1 D ⇡=B1 ; bi D i⇡=B pi ; i D 1; 2; : : : ; n:
Consequently, the critical velocity of the first mode in this example is .vcr /Timo D .1= lB1 / EId =⇢A [10, 11,
18]. On the other hand, only for a simply supported kind of boundary conditions, a modified formula for .vcr /Timo
374 A. M AMANDI , M. H. K ARGARNOVIN , AND S. FARSI
Fig. 5. Variations of the dimensionless dynamic lateral deflection ( wnl .xmax ; t /=w0 ) for xmax D 0:55l vs. the normalized time
( t=ti ) for the inclined clamped-pinned Timoshenko beam ( ' D 30ı ) traversed by moving forces F D ⇢gAl.N / with
different velocity ratios ( ˛ D 0:25; 0:5; 0:75; and 1) under the influence of three types of motion: (a–c) dimensionless results
for the nonlinear analysis: wnl .xmax ; t /=w0 , (d–f) dimensionless difference between the nonlinear and linear analyses (in %):
.wnl � wl /=w0 :
is given in [15, 18]. For this case of boundary conditions, the difference between
s
�⇡ � EId
.vcr /Euler D
l ⇢A
[8, 10, 13, 18] and .vcr /Timo [10, 11, 15, 18] is about 0.03%. However, in general, we prefer to use .vcr /Timo [10,
11, 15, 18] in our upcoming calculations for each type of BC.
It should be mentioned that, based on the previous analysis and obtained results which show that the friction
force is very small [6], we neglect the effect of friction in what follows. In Figs. 4–6, we illustrate the variations of
the dimensionless dynamic lateral deflection ( w.xmax ; t /=w0 ) vs. the dimensionless time t =ti ; i D 1; 2; and 3,
at the reference point xmax ( D xjwmax ) for pinned-pinned, clamped-pinned, and clamped-free Timoshenko beams,
respectively, with inclination angle ' D 30ı traversed by moving forces F D ⇢gAl (N) with different velocity
ratios ( ˛ D 0:25; 0:5; 0:75; and 1) under the influence of three types of motion. In Figs. 4–6, the results depicted
in the left, middle, and right columns are related to the cases of accelerating, decelerating, and constant-velocity
motions, respectively. Moreover, in these figures, the first and the second rows show the dimensionless results
N ONLINEAR V IBRATION S OLUTION FOR AN I NCLINED T IMOSHENKO B EAM UNDER THE ACTION OF A M OVING F ORCE 375
Fig. 6. Variations of the dimensionless dynamic lateral deflection ( wnl .xmax ; t /=w0 ) for xmax D l vs. the normalized time ( t=ti )
for the inclined clamped-free Timoshenko beam ( ' D 30ı ) traversed by moving forces F D ⇢gAl.N / with different velocity
ratios ( ˛ D 0:25; 0:5; 0:75; and 1) under the influence of three types of motion: (a–c) dimensionless results for the nonlinear
analysis: wnl .xmax ; t /=w0 I (d–f) dimensionless difference between the nonlinear and linear analyses (in %): .wnl � wl /=w0 .
obtained from the nonlinear analysis ( wnl .xmax ; t /=w0 ) and the percentage of dimensionless difference between
the nonlinear and linear analyses ( .wnl � wl /=w0 ), respectively.
As can be seen from Figs. 4a–c, for the decelerating and uniform-velocity types of motion, as the velocity
ratio ˛ increases to ˛ D 0:75 and ˛ D 0:5; respectively, the value of maximum dynamic deflection increases
and decreases, respectively, whereas for the accelerating type of motion, an increasing trend is observed for the
maximum dynamic deflection of the mid-span of the beam. Moreover, for the decelerating type of motion, the
range of variation of the maximum dynamic deflection is larger as compared with the other two types of motion.
From Figs. 4d–f, it is concluded that the maximum difference for the mid-span deflection of the beam between the
nonlinear and linear analysis occurs mainly for the decelerating motion and with smaller difference in the case of
uniform velocity and then for the accelerating type of motion, respectively.
It can be observed from Figs. 5a–c that, for the accelerating type of motion, by increasing the velocity ratio ˛;
the values of maximum dynamic downward deflection always increase, whereas in the decelerating and uniform
velocity types of motion, the value of maximum dynamic downward deflection increases up to ˛ D 0:5 and
˛ D 0:25; respectively, and decreases after these points. It is also seen from Fig. 5a–c that, for the decelerating
type of motion when the moving force is at the right end of the beam, i.e., x D l , the maximum dynamic upward
376 A. M AMANDI , M. H. K ARGARNOVIN , AND S. FARSI
(a) (b)
(c) (d)
Fig. 7. Variations of the dynamic response of a point at x D 0:5 l of the inclined pinned-pinned Timoshenko beam vs. the moving
force for different velocity ratios in the uniform-velocity motion: (a) ˛ D 0:25; (b) ˛ D 0:5; (c) ˛ D 0:75; (d) ˛ D 1 ; (—)
nonlinear solution, (– � –) linear solution.
(positive) deflection at xmax D 0:55 l which occurs at ˛ D 1 is the greatest value with respect to the other two
types of motion. Moreover, it can be observed from Figs. 5d–f that the maximum difference for xmax D 0:55l for
the deflection of the beam between the nonlinear and linear analysis occurs primarily in the decelerating motion
and with smaller difference in the case of accelerating motion and then in the uniform-velocity type of motion,
respectively.
It is seen from Figs. 6a–c that, for the decelerating type of motion, by increasing the velocity ratio ˛; the
value of the maximum dynamic downward deflection of the beam for x D l almost increases, whereas for the
accelerating type of motion, the reduction trend is always visible for the maximum dynamic downward deflection.
Furthermore, for the uniform-velocity type of motion, as the velocity ratio ˛ increases to ˛ D 0:5; the value of
the maximum dynamic downward deflection increases, and the opposite grand is observed afterward. In addition,
it can be observed from Figs. 6d–f that the maximum difference between the nonlinear and linear analyses occurs
primarily for the decelerating motion and with smaller difference in the case of accelerating motion and then in the
uniform-velocity type of motion, respectively. Moreover, for this type of boundary conditions, the differences are
much lower than the for the other two types of boundary conditions.
In Figs. 7 and 8, the absolute values of the variations of maximum dynamic response ( wmax .x/ ) of a point at
x D 0:5 l and 0:55l; respectively, for the inclined pinned-pinned and clamped-pinned Timoshenko beams with
' D 30ı vs. the moving force are presented for various velocity ratios ˛ D 0:25; 0:5; 0:75; and 1, respectively,
N ONLINEAR V IBRATION S OLUTION FOR AN I NCLINED T IMOSHENKO B EAM UNDER THE ACTION OF A M OVING F ORCE 377
(a) (b)
(c) (d)
Fig. 8. Variations of the dynamic response of a point at x D 0:55 l for the inclined clamped-pinned Timoshenko beam vs. the moving
force for different velocity ratios in the uniform-velocity motion: (a) ˛ D 0:25; (b) ˛ D 0:5; (c) ˛ D 0:75; (d) ˛ D 1 ; (—)
nonlinear solution, (– � –) linear solution.
by using both linear and nonlinear solutions for the uniform-velocity type of motion, respectively. It can be seen
from Fig. 7 that the maximum dynamic x of the nonlinear analysis is always smaller than the of obtained from the
linear solution. The hardening behavior is also seen in this case, as reported in the other works [16, 18]. Moreover,
the dynamic mid-point displacements of this beam obtained form the linear and nonlinear solutions are almost
identical for F 0:5⇢gAl.N /: However, after this point, the magnitude of wmax for the nonlinear and the linear
solutions gradually becomes different, and the difference rapidly grows as the value of F increases. Furthermore,
it follows from Fig. 7 that the difference between the linear and the nonlinear solutions has an increasing trend up
to the load velocity ratio ˛ D 0:5 and a reverse trend afterward. The maximum difference between the linear and
nonlinear solutions for all cases in this figure occurs for F D 3⇢gAl.N / at ˛ D 0:5 (see Fig. 7b). In addition,
the variations of the linear solution mathematically follow the linear trend observed in this figure [8, 16].
It can be observed from Fig. 8 that both the hardening (stiffening) behavior for the lower velocity ratio
( ˛ D 0:25 ) and the softening behavior for higher velocity ratios ( ˛ D 0:5; 0:75 and 1 ) can be predicted for this
type of boundary conditions of the beam under the action of moving forces. Moreover, the absolute values of the
maximum dynamic deflection ( wmax .x D 0:55l/ ) of a point at xmax D 0:55 l for the inclined clamped-pinned
Timoshenko beam obtained from the linear and nonlinear solutions are almost identical independently of the values
of F: In addition, the difference between the linear and nonlinear solutions has a slight increasing trend up to the
378 A. M AMANDI , M. H. K ARGARNOVIN , AND S. FARSI
(a) (b)
(c) (d)
(e)
Fig. 9. Variations of the dimensionless dynamic deflection ( w=w0 ) vs. ( x= l ) for the inclined pinned-pinned Timoshenko beam
with ' D 30ı ; different velocity ratios, and different values of the moving force F in the constant-velocity motion: (a)
F D 0:5⇢gAl.N /; (b) F D ⇢gAl.N /; (c) F D 2⇢gAl.N /; (d) F D 3⇢gAl.N /; (e) F D 4⇢gAl.N /I (—) nonlinear
solution, (– � –) linear solution.
N ONLINEAR V IBRATION S OLUTION FOR AN I NCLINED T IMOSHENKO B EAM UNDER THE ACTION OF A M OVING F ORCE 379
(a) (b)
(c) (d)
Fig. 10. Variations of the dimensionless dynamic deflection ( w=w0 ) vs. ( x= l ) for the inclined pinned-pinned Timoshenko beam with
' D 30ı ; different values of the moving force F; and different velocity ratios in the constant-velocity motion: (a) ˛ D 0:25;
(b) ˛ D 0:5; (c) ˛ D 0:75; (d) ˛ D 1I (—) nonlinear solution, (– � –) linear solution.
load velocity ratio ˛ D 0:75 and the opposite trend afterward. In other words, the comparison of the results
indicates that the trend of decrease for the difference between the linear and nonlinear solutions in the clamped-
pinned beam is much faster than for the pinned-pinned type of BC (cf. Figs. 7 and 8). In addition, it should be
emphasized that the maximum difference between the linear and nonlinear solutions for all cases in this figure
occurs for F D 4⇢gAl.N / and ˛ D 0:75 (see Fig. 8c). The variations of the linear solution also mathematically
follow the linear trend for this type of BC.
It should be mentioned that, for the clamped-free Timoshenko beam, there is similar softening behavior (for
brevity, the results are not presented here) with much less intensity as compared with the clamped-pinned Timo-
shenko beam. It is believed that a more pronounced weakening of the softening/hardening behavior can be seen
when the supports are allowed to slide or to be free.
The variations of normalized lateral dynamic displacements ( w=w0 ) vs. x= l obtained for the inclined
pinned-pinned, clamped-pinned, and clamped-free Timoshenko beams by using nonlinear and linear analyses for
the constant-velocity type of motion by changing the velocity ratios ˛ and the values of moving force F are
shown in Figs. 9–10, 11–12 and 13–14, respectively.
380 A. M AMANDI , M. H. K ARGARNOVIN , AND S. FARSI
Fig. 11. Variations of the dimensionless dynamic deflection .w=w0 / vs. .x= l/ for the inclined clamped-pinned Timoshenko beam with
' D 30ı ; for different velocity ratios of the moving force F D 4⇢gAl.N / in the constant-velocity motion; (—) nonlinear
solution, (– � –) linear solution.
Fig. 12. Variations of the dimensionless dynamic deflection ( w=w0 ) vs. ( x= l ) for the inclined clamped-pinned Timoshenko beam with
' D 30ı ; for different moving forces F and the velocity ratio ˛ D 0:25 in the constant-velocity motion; (—) nonlinear
solution, (– � –) linear solution.
It can be seen from Fig. 9 that the results of nonlinear analysis for w=w0 are almost lower than the results
obtained from the linear analysis. Moreover, as ˛ increases, the difference between the linear and nonlinear
analyses decreases. This difference increases when the value of F increases. The maximum difference between
the linear and nonlinear analyses occurs for ˛ D 0:25 when F D 4⇢gAl (N). The linear solutions always predict
a unique value of w=w0 for the same velocity ratio ˛ independently of the values of F
It can be seen from Fig. 10 that, for all velocity ratios ˛; the linear analysis always gives identical values
of w=w0 independently of the values of F: Moreover, it is seen that the point corresponding to .wmax /linear is
always located to the right of the similar point obtained in the nonlinear analysis.
N ONLINEAR V IBRATION S OLUTION FOR AN I NCLINED T IMOSHENKO B EAM UNDER THE ACTION OF A M OVING F ORCE 381
Fig. 13. Variations of the dimensionless dynamic deflection ( w=w0 ) vs. ( x= l ) for the inclined clamped-free Timoshenko beam with
' D 30ı for different velocity ratios of the moving force F D 4⇢gAl.N / in the constant velocity motion; (—) nonlinear
solution, (– � –) linear solution.
Fig. 14. Variations of the dimensionless dynamic deflection ( w=w0 ) vs. ( x= l ) for the inclined clamped-pinned Timoshenko beam
with ' D 30ı for different moving forces F and the velocity ratio ˛ D 0:25 in the constant velocity motion; (—) nonlinear
solution, (– � –) linear solution.
Further, if the magnitude of the moving force is small, i.e., F D 0:5 ⇢gAl.N /; then the nonlinear and linear
solutions are almost identical independently of the values of ˛: However, after this point .F > 0:5⇢gAl.N // the
difference between the linear and nonlinear solutions becomes more pronounced and, consequently, the difference
between the linear and nonlinear solutions in the maximum value of deflection accordingly increases. Moreover,
as the velocity ratio increases to ˛ D 0:5; the maximum dynamic deflection of linear and nonlinear solutions
increases but the reverse trend prevails afterward. Hence, the maximum dynamic deflection for linear and nonlinear
solutions occurs for F D 4⇢gAl.N / and ˛ D 0:5:
It follows from Fig. 11 that the instantaneous dynamic deflection calculated from the nonlinear analysis is
almost greater than the deflection obtained from the linear analysis. In addition, Fig. 12 shows that the linear anal-
382 A. M AMANDI , M. H. K ARGARNOVIN , AND S. FARSI
ysis always presents a unique value of w=w0 independently of the values of F: Moreover, for small magnitudes
of the moving force F , i.e., for F D 0:5⇢gAl.N /; the nonlinear and linear solutions are almost identical. After
this point .F > 0:5⇢gAl.N //; the difference between the linear and nonlinear solutions slightly increases. It can
be observed from Figs. 13 and 14 that, for the clamped-free beam, the difference between the linear and nonlin-
ear analyses can be graphically negligible. However, the maximum difference between the linear and nonlinear
analyses occurs for F D 4⇢gAl.N / and ˛ D 0:25:
5. Conclusions
Three nonlinear coupled partial differential equations of motion are solved for the rotation of the warped cross
section and longitudinal and transverse displacements of the inclined Timoshenko beam with different boundary
conditions subjected to the action of moving forces under the influence of three types of motions, including the
accelerating, decelerating and constant-velocity motions. The outcome results are as follows:
For the pinned-pinned and clamped-pinned types of boundary conditions, under the influence of the accel-
erating type of motion, the maximum dynamic deflection is reached much later than in the remaining two cases.
Moreover, for the decelerating type of motion, the range of variation of the maximum dynamic deflection is greater
than the ranges obtained for the other two types of motion for all types of boundary conditions of the beam.
It is concluded that the maximum difference between the nonlinear and linear analyses for the deflection of the
beam at a reference point, i.e., xmax .D xjwmax / occurs primarily in the decelerating motion, with consecutively
smaller values of the difference in the cases of accelerating and uniform-velocity types of motion, respectively.
The maximum dynamic displacements of the moving force obtained by using the linear and nonlinear solutions
are almost identical for lower values of F . However, for higher values of F , the magnitude of the beam deflection
for the nonlinear and linear solutions gradually becomes different, and the difference rapidly grows as the value of
F increases. Moreover, as the velocity ratio increases, the difference between the linear and nonlinear solutions
for the maximum dynamic deflection of the beam becomes negligible. In addition, for pinned-pinned, clamped-
pinned, and clamped-free types of boundary condition, the variations of the linear solution mathematically follow
the linear trend in the moving-force problem.
From the viewpoint of nonlinear analysis, in view of the existence of quadratic-cubic nonlinearity of the
coupled governing PDEs of motion, for the pinned-pinned Timoshenko beam, the system behaves like a hard
spring. This means that, as the magnitude of the moving force increases, the dynamic deflections become smaller
than those from the solution of the linear system, whereas for the clamped-pinned and clamped-free types of
boundary conditions, the system behaves as a hard/soft or soft spring, respectively. For the soft system, as the
magnitude of the moving force increases, the nonlinear dynamic deflection becomes larger than the deflection
obtained from the linear solution.
As ˛ increases, the value of the maximum instantaneous dynamic deflection decreases and, hence, for higher
velocity ratios, i.e., for ˛ D 0:75 and 1; the values of linear and nonlinear solutions are almost identical indepen-
dently of the values of F: Furthermore, as the magnitude of the moving force F increases, the difference between
the maximum values of the linear and nonlinear solutions increases.
It can be observed that the instantaneous dynamic deflection obtained from the linear solution under the action
of a moving force always gives a unique value for any analyzed value of the velocity ratio ˛ independently of the
values of F: At the same time, this behavior cannot be seen in the nonlinear analysis.
REFERENCES
1. G. T. Michaltsos, “Dynamic behavior of a single-span beam subjected to loads moving with variable speeds,” J. Sound Vibrat., 258,
No. 2, 359–372 (2002).
2. M.Abu Hilal and H. S. Zibdeh, “Vibration analysis of beams with general boundary conditions traversed by a moving force ,” J. Sound
Vibrat., 229, No. 2, 377–388 (2000).
N ONLINEAR V IBRATION S OLUTION FOR AN I NCLINED T IMOSHENKO B EAM UNDER THE ACTION OF A M OVING F ORCE 383
3. Y. H. Lin and M. W. Trethewey, “Finite element analysis of elastic beams subjected to moving dynamic loads,” J. Sound Vibrat., 136,
No. 2, 223–242 (1990).
4. X. Xu, W. Xu, and J. Genin, “A nonlinear moving mass problem,” J. Sound Vibrat., 204, No. 3, 495–504 (1997).
5. M. Olsson, “On the fundamental moving load problem,” J. Sound Vibrat., 145, No. 2, 299–307 (1991).
6. J. -J. Wu, “Dynamic analysis of an inclined beam due to moving loads,” J. Sound Vibrat., 288, 107–133 (2005).
7. K. Kiani, A. Nikkhoo, and B. Mehri, “Prediction capabilities of classical and shear deformable beam models excited by a moving
mass,” J. Sound Vibrat., 320, 632–648 (2009).
8. A. Mamandi, M. H. Kargarnovin, and D. Younesian, “Nonlinear dynamics of an inclined beam subjected to a moving load,” Nonlin.
Dynamics., 60, 277–293 (2010).
9. A. N. Yanmeni Wayou, R.Tchoukuegno, and P. Woafo, “Nonlinear dynamics of an elastic beam under moving loads,” J. Sound Vibrat.,
273, 1101–1108 (2004).
10. L. Fryba, Vibration of solids and structures under moving loads, Thomas Telford Publ., London (1999).
11. L. Fryba, Dynamics of railway bridges, Thomas Telford Publ., London (1996).
12. A. Mamandi, M. H. Kargarnovin, and D. Younesian, “Nonlinear vibrations of an inclined beam subjected to a moving load,” J. Phys.
Conf. Ser., 181, 012094, doi: 10.1088/1742-6596/181/1/012094 (2009).
13. H. P. Lee, “The dynamic response of a Timoshenko beam subjected to a moving mass,” J. Sound Vibrat., 198, No. 2, 249–256 (1996).
14. R. T. Wang and T. H. Chou, “Nonlinear vibration of Timoshenko beam due to moving force and the weight of beam,” J. Sound Vibrat.,
218, No. 1, 117–131 (1998).
15. P. Sniady, “Dynamic response of a Timoshenko beam to a moving force,” J. Appl. Mech. Trans. ASME, 75, 024503-1–024503-4
(2008).
16. A. Mamandi and M. H. Kargarnovin, “Dynamic analysis of an inclined Timoshenko beam traveled by successive moving masses/forces
with inclusion of geometric nonlinearities,” Acta Mech., 218, No. 1, 9–29 (2011).
17. A. Mamandi and M. H. Kargarnovin, “Nonlinear dynamic analysis of an inclined Timoshenko beam subjected to a moving mass/force
with beam’s weight inclined,” Shock Vibrat., 18, No. 6, 875–891 (2011).
18. A. Mamandi, M. H. Kargarnovin, and S. Farsi, “An investigation on effects of traveling mass with variable velocity on nonlinear
dynamic response of an inclined Timoshenko beam with different boundary conditions,” Int. J. Mech. Sci., 52, No. 12, 1694–1708
(2010).
19. A. H. Nayfeh and D. T. Mook, Nonlinear oscillations, Wiley, New York (1995).
20. I. Karrnovsky and O. I. Lebed, Formulas for structural dynamics, McGraw-Hill, New York (2001).
21. S. S. Rao, Mechanical vibrations, Addison-Wesley, New York (2000).