Large Deflection Dynamic Behavior of Slender

Beams by Method of Characteristics

Raid A. Othman, Baghdad University, Iraq

Samir Z. AL-Sadder, Hashemite University, Jordan
Anis S. Shatnawi, The University of Jordan, Jordan
Mutasim Abdel-Jaber, The University of Jordan, Jordan
Hasan N. Katkhuda, Hashemite University, Jordan

ABSTRACT
The dynamic responses considering large deflection of slender beams are investigated in
this paper. One-dimensional flexural vibration is considered. It is well known that the
flexural wave consists of two components; bending and shear waves. A new formulation
for the flexural wave behavior in geometrically non-linear beams is introduced. The
formulation of the non-linear governing equations is established by considering
elements under the effect of large-deflection and large-rotation while subjected to
dynamic excitation. The governing equations are re-written in a numerical form using
the Method of Characteristics. Various types of support and load conditions are studied.
The results showed that the Method of Characteristics is a suitable method to represent
wave propagation phenomena in large deflection.
INTRODUCTION
Linear behavior can be assumed when material and geometrical changes due to applied
(external) loads are small. Usually the change in material property is due to the change in the
modulus of elasticity, while the change in geometry might be due to a change in cross-
sectional area or length. When these changes are small, the linear system can give accurate
results.

However, there are many situations such as severe earthquake motion with vertical or
horizontal components (or both), suspension beams or bridges of large spans subjected to
pulse, wind or any dynamic loading, in which linear models will not yield acceptable
results.

The most common methods of analysis of structural dynamics are based on finite element
techniques [1, 2] or on analytical solutions of the equations of motion [3-5]. Analytical
methods are difficult to handle for structures with many degrees of freedom or when coupled
action exists between axial and flexural waves. In addition, they are not usually formulated in
a manner that makes it easy to simulate high frequency wave action within the individual
structural members. A number of new studies oriented towards the large deflection and/or
wave propagation are observed [6-7].

There is an obvious advantage in having a single method of analysis capable of simulating all
types of dynamic behavior, whether gradual (slow), more rapid or transient. Such a method is
the Method of Characteristics (MOC) which is considered herein and is shown to be capable
of simulating static as well as dynamic problems. The popularity of the method comes from
three of its attributes:

1
 1. The possibility to reduce the partial differential equations to straightforward
ordinary differential equations.
2. The mathematics of (MOC) underlines the essential physical wave behavior as the
associated characteristics directions are the time-space paths of information flow
in the physical system.
3. The computer resources required are relatively small, thus enabling highly
complex structures to be simulated.

In the theoretical analysis of dynamic elasticity, the (MOC) has received limited attention.
Recently only serious attempts have been made to study the problem of elastic waves
propagation by this method. It was first applied by Pfgiffe [8] who obtained a general
solution of the Timoshenko beam equations. Chou and Mortimer [9] solved the same problem
using (MOC). They also carried out numerical integration for the special case where the shear
and bending wave speeds are equal. Subsequently, Plass [10] extended the numerical method
to involve beams with two different wave speeds. The (MOC) was also applied by Chou and
Koening [11] to several spherical and cylindrical problems. Vardy and Chan [12] used the
method to model waves in skeletal structures comprising prismatic members. Alsarraj [13]
used the method with the presence of coupling between axial and flexural waves. Aldaami
[14] used the (MOC) to modal waves in prismatic and non-prismatic curved members.
Alsarraj et al. [15] applied the (MOC) to problems of non-linear vibration of axially loaded
bars. Fahmi [16] used the method of characteristics to determine the non-linear transient
response of beams with internal resonance.

Large deflection problems of beams was studied in static case by using different methods of
analysis, some of these studies are mentioned herein. Barton [17] and Bishop and Drucker
[18] used the elliptic integrals to find the large deflection behavior of a cantilever beam
subjected to a tip concentrated vertical force. Haisler et al. [19] and Haisler and Stricklin [20]
presented solution techniques which are applicable to the solution of the non-linear algebraic
or differential equations characterizing geometrically non-linear behavior.

In the following mathematical formulation, equations of motion in Cartesian coordinates are

obtained for each type of wave and material property relationships are differentiated with
respect to time. This leads to pairs of partial differential equations relating a force and a
velocity or a couple and an angular velocity. Each pair of equations is expressed in ordinary
differential form using the method of characteristics and the explicit finite difference
technique is used to obtain the solution by a process of time marching. The accuracy and
stability of the numerical solution are discussed.

In the present work, the large deflection problem of slender beams is studied in dynamic
state. One-dimensional flexural vibration is considered and a new formulation for the flexural
wave behavior in geometrically non-linear beams is introduced by using the Method of
Characteristics, and the numerical solution of the new equations is done by using the finite
difference techniques.

DERIVATION OF GOVERNING EQUATIONS

A small element of a beam of constant flexural rigidity EI and actual length ds with cross-
section area a is considered as shown in Figure 1. Equilibrium of forces in vertical direction
gives:
dV
V V dx a ds
dx t
(1)

2
or

dV ds
a
dx t dx
(2)

Then, from Figure 2 the relations between the two components and the original vertical force
can be obtained
V Q cos F sin (3)

Q sin F cos (4)

Q
V
cos
(5)

Where
M: bending moment
V: vertical force
: lateral velocity
: density of the material
a: area of cross section
F: axial force
Q: shear force
I: second moment of area
E: modulus of elasticity

,Accordingly

dV d 2 s dQ ds
Q 2 (6)
dx dx dx dx
Substituting Eqn. 6 in Eqn. 2 to get

d 2 s dQ ds d ds
Q 2
a (7)
dx dx dx dt dx

dQ d 2 s dx d
Q 2 a (8)
dx dx ds dt

Equilibrium of moment about point o gives:

dM dV d
M M dx (V dx)dx Ids (9)
dx dx dt

or

dM d ds
V I (10)
dx dt dx
Substituting Eqn. 5 into Eqn. 10 yields

3
dM ds d ds
Q I (11)
dx dx dt dx
where
1
ds dz 2
1 ( ) 2 (12)
dx dx

The material property relationships obtained from Hooks Law are differentiated with respect
to time to give the other two equations.
Q
SHEAR: K 1Ga ( ) (13)
t x

M
MOMENT: EI (14)
t x
Where
K1 = Shear correction factor
G = Shear modulus
=Angular velocity

Eqns. 8, 11, 13 and 14 are pairs of hyperbolic partial differential equations each can be
expressed as an ordinary differential equation provided that their range of applicability is
restricted to certain characteristic directions in the x-t plane. In their characteristic form, the
equations are:
dx
Shear Along C S
dt
dQ d 2 s dx
C S a C S Q 2 K 1Ga (15)
dt t dx ds
dx
Moment Along C S
dt
dM d ds ds
IC C (16)
dt dt dx dx
E K1 E
where C and C S are speeds of transmission of moment and shear waves

along member.
These equations can be integrated along the lines L/P & R/P in the case of shear wave and LP
& RP in the case of bending wave as shown in Figure 3. Using central difference expressions
yields:
Shear:
p L/ d 2 s dx QP QL/
(Q P Q ) aC S ( P ) aC S (
/
L
/
L )x 2 ( )x (17)
2 dx ds 2

p R/ d 2 s dx Q P QR/
(QP Q ) aC S ( P ) aC S (
/
R
/
R ) x 2 ( ) x (18)
2 dx ds 2
Moment:
ds Q Q L ds
( M P M L ) IC ( p L ) ( P ) x (19)
dx 2 dx

4
 ds Q QL ds
( M P M R ) IC ( p R ) ( P ) x (20)
dx 2 dx

NUMERICAL EXAMPLES
The non-linear dynamic governing equations that are represented by Eqns. 17 to 20 are
solved simultaneously by a computer program using FORTRAN language. Several examples
are given to study the dynamic behavior under large deflection. For all examples, the
following numerical data are selected:

L = 1 m, a = 0.01 m2, I = 8.333x10-6 m4, E = 200 GPa, = 8000 kg/m3

Figures 4a and 4b show a comparison between the behavior of a cantilever beam subjected to
a suddenly applied constant bending moment at its free end under large deflection (non-
linear) and under small deflection (linear). The Results show that the shear force and bending
moment-time history have similar shapes with magnifications in values for the case of large
deflection.

In dynamic analysis the method of characteristics is well suited to determine the fundamental
natural frequencies of the structure by observing the repetitive pattern of the force or velocity
response caused by the dynamic load. In the present work, the software MSC/NASTRAN
[21] is used to predict the natural frequencies of the structure. Then these frequencies are
used by the present FORTRAN computer program to simulate the resonance condition.
Figure 5 shows the dynamic response of a fixed-hinged beam subjected to a periodic load at
its free end and the resonance occurred when the load frequency match the natural frequency
of the beam.

Non-linear systems exhibit phenomena that cannot occur in linear systems. In this
phenomenon, it can be seen that the amplitude of the vibration of the system has been found
to increase or decrease. In the large deformation stage, the fundamental frequency is changed.
This may have an important influence in resonance, as there will be a difference between the
load and structure frequencies, i.e., canceling the resonance condition. When harmonic
loading is applied with a frequency equal to that of the natural frequency of the beam,
resonance will occur. The deflection in this case will increase with time leading to an increase
in the member length and that will decrease the natural frequency of the beam. At a certain
time the frequency of the loading is not matching with the natural frequency of the beam. The
difference between them will be large enough to cause a jump in the vibration of the beam, as
shown in Figure 6a. Figure 6b shows the change in the length of the beam that corresponds to
the change in the response during the resonance case. Jump phenomenon cannot be seen in a
linear analysis.

The propagation of bending waves in beams with difference boundary conditions is

illustrated in Figure 7. Figure 7a show the propagation of bending moment applied at the left
end of simply supported beam, Figure 7b and Figure 7c show bending and shear waves
propagation, respectively, in a cantilever subjected to a constant bending moment at its free
end. Figure 7d explains how the waves are transmitted along a fixed-hinged beam from the
point of applied load to the rest of beam. The interaction of the applied waves that propagate
through the beam length with those reflected at the fixed end, and the differences between the
propagation speeds of the bending and shear waves cause significant changes in the bending
and shear waves shapes as they travel throughout the beam. In addition, the figures show the
dissipative nature of the flexural waves because of the difference between the bending and
shear waves speeds.

Figure 8 shows the dynamic response of a cantilever beam subjected to constant applied
moment at its free end with different lengths. From this Figure, it is clear that the waves

5
travel in constant speed and when the length of beam is increased, the waves need more time
to reach the boundary or any other point. The Figure explains the fact that when the length
increases, the natural frequency will decrease. In linear analysis, the change in length is
assumed to be negligible. The length of the member remains constant during the analysis. In
the present development of the analysis, a new length of the member is calculated every time
step as shown in Figure 9. This leads to a change in natural period and in amplitude of
vibration, thus producing nonlinear behavior of the member.

CONCLUSIONS
A new formulation for the flexural wave behavior in geometrically non-linear beams is
introduced. The formulation of the non-linear governing equations is established by
considering elements under the effect of large-deflection and large-rotation while subjected to
dynamic excitation. The governing equations are re-written in a numerical form using the
Method of Characteristics. Several conclusions can be summarized as:
1- The method of characteristics can be used to simulate the linear and non-linear
dynamic behavior of prismatic members subjected to any type of loading.
2- The interaction between shear and bending waves leads the flexural waves to
propagate in prismatic members in a dispersive (changed) form.
3- The method of characteristics can efficiently predict forces, velocities and
displacements throughout the member at any time.
4- From analysis, convergence of behavior is achieved with 32 grid lengths along the
member.
5- The dynamic load factor is greater than (2) in flexural wave vibration.
6- The large deflection (non-linear) time period and frequency of vibration are
approximately equal to small deflection (linear) values.

The results showed that the Method of Characteristics is a suitable method to represent wave
propagation phenomena in large deflection.
References
1. Zienkiewicz, O. Z. (1977). The Finite Element Method, (Third edition). McGraw-Hill.
2. Ross, C. T. F. (1985), Finite Element Method in Structural Mechanic, (First edition),
Chichester Ellis Harwood.
3. Clough, R.W. and Penzien, J.(1975). Dynamics of Structures, New York: McGraw-
Hill.
4. Coates, R.C., Couiie M.G. and Kong, F.K. (1986). Structural Analysis, (Second
edition), New York: Van Nostrand Reinhold.
5. Johnson, D. (1986). Advanced Structural Mechanics, London: Collins.
6 Pany, C. and Rao, G. V. (2004). Large amplitude free vibrations of a uniform spring-
hinged beam, Journal of Sound and Vibration, Vol. 271, pp. 11631169.
7. Lande, R.H. and Langley, R.S. (2005), The energetics of cylindrical bending waves
in a thin plate, Journal of Sound and Vibration, Vol. 279, pp. 513518.
8. Pfgiffer, R. (1947), ber Die Differertialbeichung der Trasversalen
Stabschwingunguen, Z Angew. Math. Mech., No. 3, pp. 25-27.
9. Chou, P. C. and Mortimer (1967), Solution of One-dimensional Elastic Wave
Problems by the Method of Characteristics, Journal of Applied Mechanics, ASME,
Vol. 34, pp.745-750.
10. Plass, H.I. (1958), Some Solution of Timoshenko Beam Equation for Short Pulse-
Type Loading, Journal of Applied Mechanics, Vol. 25, pp.379-385.
11. Chou, P. C. and Koening, H. A. (1966), A Unified Approach to Spherical Elastic
Waves by the Method of Characteristics, Journal of Applied Mechanics, Trans.,
ASME, Series 1, 88, pp.159-168.
12. Vardy, A. E. and Chan, L. I. (1988), Transient Analysis by Boundary Characteristics,

6
 Journal of Mechanical Division, ASCE, Vol. 114, No. 3, pp. 420-535.
13. Alsarraj, A. T. (1989), Method of Characteristics Analysis of Vibrating Frames and
Trusses, Ph.D. Thesis, the University of Dundee, Department of civil Engineering,
Dundee, U.K.
14. Aldaami, H. H. (1992), Transient Response of Systems with Non-prismatic Members
by Characteristics, M. Sc. Thesis, Saddam University, Department of Civil
Engineering, Baghdad.
15. Alsarraj, A.T., Othman, R.A. and Jwad, Z.M. (1994), Non-linear Vibration of Axially
Loaded Bars, Journal of Engineering and Technology, Vol.12, No.9, pp.37-45.
16. Fahmi, M.H. (1998), Non-linear Steady State Vibration of Beams with Internal
Resonance, Ph.D. Thesis, the University of Baghdad, department of civil engineering,
Baghdad.
17. Barton, H.J. (1944, 1945), On the Deflection of a Cantilever Beam, Q. Appl. Math.
2, pp. 168-171 and Quart. Appl. Math. 3, pp. 275-276.
18. Bishopp, K. E. and Drucker, D. C. (1945), Large Deflections of Cantilever Beams,
Q. Appl. Math. 3, pp.272-275.
19. Haisler, W. E., Stricklin, J. A. and Stebbins, F. J. (1972), Development and
Evaluation of Solution Procedures for Geometrically Nonlinear Structural Analysis,
AIAA Journal, Vol.10, No. 3, pp. 264-272.
20. Haisler, W.E. and Stricklin, J.A. (1977), Displacement Incrementation in Nonlinear
Structure Analysis by the Self-correcting Method, International Journal for
Numerical Methods in Engineering, Vol.11, pp.3-10.
21. MSC/NASTRAN for Windows 95: McNeal-Schwendler Corporation, CA, USA 1995.

Figure 1. A small beam element under large deflection.

V
Q

F
ds
dz

dx

Figure 2. Relations between forces components.

7
 B P
t=t 2

+
C C
t -

L C+ O C- S R
S
t=t 1

\
t L O\
yradnuoB R\

t=t o

x x

Figure 3. Numerical characteristics scheme.

6000
B e n d in g m o m e n t a t fix e d e n d (N .m )

lin e a r
5000
n o n lin e a r
4000

3000

2000

1000

-1 0 0 0

-2 0 0 0

-3 0 0 0
0 .0 0 0 0 .0 0 5 0 .0 1 0 0 .0 1 5 0 .0 2 0
T im e (s e c )
Figure 4a

8
 25000
20000 lin e a r
15000

S h e a r fo r c e a t f ix e d e n d ( N )
n o n lin e a r
10000
5000
0
-5 0 0 0
-1 0 0 0 0
-1 5 0 0 0
-2 0 0 0 0
-2 5 0 0 0
-3 0 0 0 0
0 .0 0 0 0 .0 0 5 0 .0 1 0 0 .0 1 5 0 .0 2 0
T im e (s e c )
Figure 4b
1000
Mo

Figure 4. Comparison between large deflection and small deflection

dynamic behaviors.

20000
W p =W n
B e n d in g m o m e n t a t f ix e d e n d )N m. (

n o n l in e a r
10000
l in e a r

-1 0 0 0 0

-2 0 0 0 0
0 .0 0 0 .0 1 0 .0 2 0 .0 3 0 .0 4 0 .0 5

T im e ) s e c (
Figure 5a

9
 40000

30000 W p = W n

S h e a r f o r c e d a t f ix e d )N (
l in e a r
20000
n o n l in e a r
10000

-1 0 0 0 0

-2 0 0 0 0

-3 0 0 0 0
0 .0 0 0 .0 1 0 .0 2 0 .0 3 0 .0 4 0 .0 5
T im e ) s e c (
Figure 5b
Mo Mo=1000 N.m

1 .2 0
W p =W n
L

1 . 1 for
Figure 5. Resonance 5 linear and nonlinear cases.
L e n g th o f b e a m (m )

1 .1 0

10000
8 10 .0 00 5 W p =W n
B e n d in g m o m e n t a t f ix e d e n d )N .m (

6000
4 10 .0 00 0
2000
0 .0 0 0 .0 1 0 .0 2 0 .0 3 0 .0 4 0 .0 5
0
T im e (s e c )
-2 0 0 0
-4 0 0 0
-6 0 0 0
-8 0 0 0
-1 0 0 0 0
0 .0 0 0 .0 1 0 .0 2 0 .0 3 0 .0 4 0 .0 5
T im e ) s e c (
Figure 6a

10
 Figure 6

Figure 6. Jump phenomena in large deflection deflection.

1000

800
M
b e n d in g m o m e n t (N .m )

600
o
L
400
15000
200
L = 1 .0 m
M
0 T im e = 5 0 u s
10000 o
L T im e = 1 5 0 u s
-2 0 0 Mo
s h e a r fo rc e (N )

T im e = 2 5 0 u s
-4 0 0
5000 L
0 .0 0 .2 0 .4 0 .6 0 .8 1 .0
d is ta n c e (m )
Figure 7a

1200 0

1 01 00 00 0

800
M
-5 0 0 0
b e n d in g m o m e n t (N .m )

800 0 .0 0 .2 0 .4 0 .6 0 .8 1 .0
600 M o D i s t a n c Le ( m )
b e n d in g m o m e n t (N .m )

600 Figure 7c
400 o
L
2 04 00 0

0
200
-2 0 0
0
-4 0 0

- 6 -02 00 0
0 .0 0 .2 0 .4 0 .6 0 .8 1 .0
-4 0 0 d is ta n c e (m )
0 .0 0 .2 0 .Figure
4 11 0 . 8
7b 0 . 6 1 .0
d is ta n c e (m )
 Figure 7d

Figure 7. Effect of boundary conditions on wave propagation.

4000
L=1m
b e n d in g m o m e n t a t fix e d e n d (N .m )

L = 1 .5 m
3000 L=2m

2000

1000

-1 0 0 0

-2 0 0 0
0 .0 0 0 .0 1 0 .0 2 0 .0 3 0 .0 4 0 .0 5
T im e (s e c )
Figure 8a

10000
s h e a r fo r c e a t fix e d e n d (N )

5000

-5 0 0 0

L = 1 .0 m
-1 0 0 0 0 L = 1 .5 m
L = 2 .0 m

-1 5 0 0 0
0 .0 0 0 .0 1 0 .0 2 0 .0 3 0 .0 4 0 .0 5
T im e (s e c )
Figure 8b
Mo

L
12
Figure 8. Effect of beam length on wave propagation
 1 .0 2 5
Mo
1 .0 2 0
L
le n g th o f th e b e a m (m )

1 .0 1 5

1 .0 1 0

1 .0 0 5

1 .0 0 0
0 .0 0 0 0 .0 0 5 0 .0 1 0 0 .0 1 5 0 .0 2 0
T im e (s e c )

Figure 9. Actual length of beam during vibration

13