You are on page 1of 73

DIFFERENTIAL MODEL AND IMPACT RESPONSE OF A FLEXIBLE BEAM

ATTACHED TO A RIGID SUPPORTING STRUCTURE

A Thesis

Presented to

The Graduate Faculty of the University of Akron

In Partial Fulfillment

of the Requirements for the Degree

Master of Science

Harish Chandra

May, 2008
DIFFERENTIAL MODEL AND IMPACT RESPONSE OF A FLEXIBLE BEAM

ATTACHED TO A RIGID SUPPORTING STRUCTURE

Harish Chandra

Thesis

Approved: Accepted:

Advisor Dean of the College


Dr. D. Dane Quinn Dr. George K. Haritos

Faculty Reader Dean of the Graduate School


Dr. T.S.Srivatsan Dr. George R. Newkome

Department Chair Date


Dr. Celal Batur

ii
ABSTRACT

Often electronic components such as laptops and cellular phones are dropped accidentally

during usage and extensive damage is developed due to the impulsive force generated at

the contact point. While external damage is easy to detect, internal damage to the

electronic circuitry go undetected, yet may cause failure of the system. An accurate

description of the impulsive force is necessary to understand the dynamics of the system.

This research study involves development of a differential model of a flexible beam

attached to a rigid supporting structure and studying its response due to impacts. An

Euler Bernoulli beam theory is used to model the beam, and Routh’s graphical method

for two dimensional impacts is used to calculate the impulse at the contact point. The

dynamics of impact at the contact point is used to develop the boundary conditions and

Galerkin’s approach is used to find an approximate solution. An example is presented in

which the response due to drop at different angles of approach is studied. The position of

the beam on the frame, the coefficient of friction (µ ) and the coefficient of restitution (e )

are varied to see their influence on the beam response. Finally the influence of the

boundary conditions on the stresses and strains developed in the beam is discussed.

iii
ACKNOWLEDGEMENTS

First and foremost I would like to thank my advisor Dr.Dane Quinn for his invaluable

guidance, patience and support throughout the course of my study. I remain indebted to

him for awarding me a Research Assistantship during my first semester and helping me

get a Teaching assistantship thereafter till the end of my study.

I would like to thank the Department of Mechanical Engineering at the University of

Akron [Dr. Celal Batur] for awarding me a Teaching Assistantship during my graduate

study for the Master of Science degree. I would like to thank Dr. T.S.Srivatsan and

Dr.Graham Kelly for serving on my thesis committee. I appreciate all the support I got

from the staff members specially Stacy and Stephanie.

I also extend warmest thanks to my parents and friends who have been with me and

encouraged me during my study.

iv
TABLE OF CONTENTS

Page

LIST OF FIGURES …………………………………………..…………….…………viii

CHAPTER

I. INTRODUCTION ………………………………………………………………1

1.1 Overview…………………………………………………………………..1

1.2 Motivation………………………………………………………………....2

1.3 Collisions …………………………………………………………………3

1.3.1 Rigid body collisions……………………………………………...6

1.3.2 Collision law properties…………………………………………...7

1.3.3 Rigid body assumptions…………………………………………...7

1.4 Mathematical modeling…………………………………………………...8

1.4.1 Steps in Mathemaical Modeling…………………………………..9

II. LITERATURE REVIEW …………………………………………………… 12

III. MATHEMATICAL MODELING……………………………………………….17

3.1 Introduction………………………………………………………………17

3.2 Problem formulation……………………………………………………..17

3.3 Boundary conditions……………………………………………………..20

3.3.1 Initial Conditions………………………………………………...20

3.3.2 Transformation of boundary conditions.........................................21

v
3.4 The dynamics of impact………………………………………………….22

3.5 Impact velocities………………………………………………………... 23

3.6 Routh’s method to find the impulse……………………………………...25

3.6.1 Procedure to calculate impulse…………………………………..26

3.6.2 The impact process diagram……………………………………..29

3.7 Galerkin Reduction……………………………………………………....33

3.7.1 Galerkin method as applied to the beam model………………….33

3.8 Stresses and Strains……………………………………………………...36

3.9 Review…………………………………………………………………..36

IV. THE DYNAMICS OF IMPACT AND IMPACT VELOCITIES……………….38

4.1 Overview………………………………………………………………… 38

4.2 An impact problem …………………………………………………….38

4.3 Routh’s graphical method………………………………………………...40

4.4 Using the impulse to find change in velocities…………………………...42

4.5 Mode shapes equations…………………………………………………...43

4.6 Results…………………………………………………………………….45

V. RESULTS AND CONCLUSION………………………………………………..48

5.1 Overview…………………………………………………………………48

5.2 Angle of approach v/s impact……………………………………………48

5.3 Influence of (µ ) on the impulse at contact point………………………...51

5.4 Influence of the coefficient of restitution (e) on the impulse…………….53

5.5 Influence of the angle of impact on the beam velocity…………………..54


vi
5.6 Influence of the angle of impact on the deflection, stress and strain…….56

5.7 Influence of the location of the beam on the boundary conditions………57

5.8 Change in stress and strain with change in position of beam……………59

5.9 Conclusion……………………………………………………………….60

5.10 Underlying simplifications………………………………………………..61

5.11 Recommendations for future work ………………………………………62

REFERENCES…………………………………………………………………………..63

vii
LIST OF FIGURES

Figure Page

1.1 Impact configuration………………………………………………………………1

1.2 Single degree of freedom system………………………………………………….5

3.1 Rigid frame with a flexible beam………………………………………………...18

3.2 Illustration of Impact……………………………………………………………..24

3.3 Position of Mass center…………………………………………………………..27

3.4 Contact velocities………………………………………………………………...27

3.5 Impact process diagram illustrating Slip-Stick…………………………………..30

3.6 Impact process diagram illustrating Stick………………………………………..31

4.1 An Impact Problem………………………………………………………………37

4.2 Impact process diagram………………………………………………………….40

4.3 Maximum Deflection…………………………………………………………….45

4.4 Maximum Stress…………………………………………………………………46

4.5 Maximum Strain…………………………………………………………………46

5.1 Angle of Impact v/s Magnitude of impact……………………………………….48

5.2 Impulse path at θ = 15 o …………………………………………………………..49

5.3 Impulse path at θ = 45o ………………………………………………………….49

5.4 Impulse path at θ = 75o …………………………………………………………..49

5.5 µ v/s Pt …………………………………………………………………………..50

5.6 Impulse path for µ = 0.2 …………………………………………………………51


viii
5.7 Impulse path for µ = 0.8 ………………………………………………………….51

5.8 Impulse path for µ = 1 ……………………………………………………………52

5.9 Impulse path for µ = 2 ……………………………………………………………52

5.10 e v/s P…………………………………………………………………………...53

5.11 Angle of impact v/s Change in magnitude of velocity at end ‘A’………………54

5.12 Angle of impact v/s Change in magnitude of velocity at end ‘B’……………….54

5.13 Angle of impact v/s Deflection………………………………………………....55

5.14 Angle of impact v/s Stress………………………………………………………55

5.15 Angle of Impact v/s Strain……………………………………………………...56

5.16 Frame with beam attached at different locations………………………………...56

5.17 Location of beam v/s change in magnitude of velocity at ‘A’ ………………….57

5.18 Location of beam v/s change in magnitude of velocity at ‘B’ ………………….57

5.19 Stress v/s Location of beam……………………………………………………...58

5.20 Strain v/s Location of beam……………………………………………………...58

ix
CHAPTER 1

INTRODUCTION

1.1 Overview

Impact can be described as the interaction between two or more bodies and plays a

vital role in many mechanical engineering applications. An accurate description of this

interaction is necessary to understand the dynamics of mechanical systems. The objective

of this thesis is to develop a differential model of the response of a flexible beam attached

to rigid supporting structure upon impact. Figure 3.1 illustrates the problem being

studied.

Q ê2 ê1
B

R
A
G ( x, y )
P

iˆ S

Figure 1.1 Impact configuration

This work studies the effect of the impulse produced at the contact point on the

response of the beam. The process of impact is described by Routh’s model, a simplified

description of the post - collision velocity of the body can be predicted, given the pre-

1
collision velocity and geometry. The impact then specifies boundary conditions on the

flexible beam, so that the effect of the impact on the stress, strain and deflection of the

beam are presented.

The first chapter of this thesis introduces the reader to rigid body collisions and

impacts. A few basic terms as applied to rigid body dynamics are explained. The second

chapter discusses the previous work that has been done on collision models and their

results. The main work done in this thesis is explained in chapters three, four and five. In

Chapter 3 the basic equations that form the differential model of the beam are developed.

A simplified model for the approximation of impact is presented. The relationship

between impact and the post-collision velocities is established. In Chapter 4 the equations

developed in the previous chapter are used and an impact problem is explained in detail.

The stresses and strains developed in the beam due to changes in velocity upon impact

are studied. Chapter 5 concludes the thesis with the study of the influence of the angle of

impact of the block on the post-collision velocities and hence the stresses and strains

developed and their location. Finally limitations and recommendations for possible future

work are stated.

1.2 Motivation

Often electronic components are dropped accidentally during usage and can cause

extensive damage. The damage could be exterior in nature such as dents or cracks on the

surface or may lead to failure of the electronic circuits that play a vital role in its

functioning. While external damage may do little to affect the performance of the object,

internal damage may cause the object to cease proper orientation. Therefore the proper

2
functioning of portable electronic products such as mobile phones, laptops, and digital

cameras depends largely on the integrity of the internal components and the electronic

circuitry. Due to the rapid progress made in the technological field in the last few years

electronic devices have become smaller and lighter and the internal components have

become so complex that they are more susceptible to damage due to accidental drop.

When an object is accidentally dropped an impulsive force is created when it makes

contact with the ground. This is a potential cause of failure. Under these circumstances

their reliability due to shock and impact becomes a critical issue affecting performance.

They are also exposed to shock during various stages of their production and operation.

When the product is accidentally dropped on the ground, impact forces are transmitted

from the product case to the printed circuit board (PCB) and other components within the

case. This can cause fracture problems. Also sometimes the product may fail due to

motion of the internal components.

With escalating production and manufacturing costs it would help if the damage

caused to these components can be identified or predicted before it is delivered to

customers. This would help in the production of components that are more reliable under

shock and more resistant to damage due to drops. One of the methods used to analyze the

dynamics of the problem is mathematical modeling. It is a powerful and useful approach

to identify the potential weakness of products in the initial stages of fabrication.

1.3 Collisions

Collisions can be defined as the action of two or more bodies coming together or

striking one another for a small amount of time. A collision results in an impulse. In the

3
event of a collision the mass and the change in velocity is easily measured but the impact

force is not. Once the time of collision is known the average impact force can be

calculated. In other words one of the first steps in order to find the impact force is

collision detection. Collisions are classified as single point and multipoint depending on

the way they come into contact. If the bodies come into contact at a single point it’s

called a single point collision. Otherwise it’s called a multipoint collision. During a

collision between two bodies there is large force acting for a very brief period of time.

Integration of that force over that brief time gives the impulse. Collisions can be either

elastic or inelastic. A perfectly elastic collision is defined as one in which there is no loss

of kinetic energy in the collision. Elastic collisions occur only if there is no conversion of

kinetic energy to other forms. Collisions between atoms are perfect examples of elastic

collisions. In an elastic collision the total kinetic energy is the same before and after

collision, hence:

m1u12 m2 u 22 m1v12 m2 v 22
+ = +
2 2 2 2

Momentum balance is maintained in all collisions. Therefore total momentum remains

constant after collision:

m1u1 + m2 u 2 = m1v1 + m2 v 2

An inelastic collision is one in which there is some form of energy conversion during the

collision. During an inelastic collision part of the kinetic energy is converted to internal

energy. Finally, in a plastic collision the objects stick together after collision.

Impact is characterized by a sudden high force or shock applied over a short

interval. Across a collision only the impulse is important. Impact has been studied in

4
detail in a lot of previous works since it forms the subject of a wide variety of

engineering applications. The topic of interest for most design and structural engineers is

the reaction forces that develop during collision and the response of structures to these

forces. There are various methods to determine the response of a system to a shock load.

The two most popular ones are the frequency domain approach or the shock response

spectrum and the time domain approach or the time history of the system. In the

frequency domain approach the steady-state maximum response of the system to a given

shock pulse is determined. This is accomplished by considering the response of a single

degree of freedom model composed of a spring mass and damper. In the time domain

approach the equations of motion governing the system are first written, they are then

integrated with respect to time to determine the response. Mechanical components can

often be modeled with partial differential equations; these equations are discretized to

yield a system of ordinary differential equations in time which can then be integrated.

Figure 1.2: Single degree of freedom system

A shock is defined as a transient physical excitation that causes a sudden jump in

velocity. It can be caused by a drop, collision with another object, sudden disturbance

like an earthquake or explosion. The magnitude of shock is measured using an

5
accelerometer. A shock response spectrum is used for evaluating the response to a

mechanical shock. It is a graph depicting the response of a single degree of freedom

system such as a spring-mass-damper system to an arbitrary transient acceleration as

input. The horizontal coordinate represents the natural frequency of any given single

degree of freedom system and the vertical coordinate represents the acceleration

response. Shocks of high magnitude have the potential of damaging an entire structure.

Also the damage caused depends upon whether the material is brittle or ductile. A brittle

material under shock fractures whereas a ductile material bends. An example of the

damage caused to a brittle material under shock is when a crystal glass is dropped to the

floor it shatters. Whereas when a copper pitcher, a ductile item is dropped on the floor it

bends. Some materials are not damaged by a single shock but experience fatigue failure

under numerous low level shocks.

1.3.1 Rigid Body Collisions

A rigid body is a solid object that has a definite shape and cannot be deformed in

any way. In other words its shape does not change during collision. In order to find the

post collision velocities of two colliding rigid bodies certain laws are used. These laws

are referred to as collision laws. Given the velocities of the centers of mass and the

angular velocities of the bodies at the instant before collision, a collision law is a rule

which predicts the corresponding velocities after the collision. A collision law requires

physical details of the colliding bodies such as material properties, geometric

characteristics of the bodies, friction properties in the region of contact, etc.

6
1.3.2 Collision Law properties

1. A collision law should be consistent with the fundamental laws of mechanics and

dynamics like conservation of mass and momentum, linear and angular momentum

balance.

2. It should consider bodies with arbitrary shape, mass distributions, material and

surface properties, orientations and velocities.

3. It should be able to give results that are in agreement with observed behavior for

simple models.

4. It should be consistent with other simpler laws like the laws of friction and other

phenomena which are modeled using lesser known laws.

5. It should be dependent on reasonably small number of input parameters and

should involve simple calculations.

6. It should be able to capture a wide variety of observed behavior for given input

parameters.

7. The input parameters should have simple physical interpretations.

1.3.3 Rigid Body assumptions

The colliding bodies are treated as rigid before and after collision. This means that

the colliding bodies move almost like rigid bodies before and after collision with

deformation being neglected in the calculation of linear and angular momentum. Any non

rigid behavior taking place during collision causes small deformations for majority of the

body with larger deformations occurring at the contact area. Collision occurs for a very

short duration with displacements and rotations being negligible, accelerations being

large with definite changes in velocities. Impulses other than those occurring at the

7
contact area are neglected with other body forces also being negligible. The mass,

moments of inertia and dimensions of at least one colliding body are finite and known.

Point contact between the bodies is assumed. Even though in reality contact occurs over a

region, this region is assumed to have smaller dimensions when compared to the length

of the smaller colliding body.

There is no kinetic energy created in a collision. The net kinetic energy of rigid

body motion in the bodies after collision is less or equal to the kinetic energy before

collision. The bodies do not pass through each other and there is no interpenetration as

this violates the point contact assumption. These are treated as reasonable assumptions

according to the laws of laws of mechanics.

1.4 Mathematical Modeling

Engineers often use their knowledge of science, mathematics, and appropriate

experience to find suitable solutions to a problem. Creating an appropriate mathematical

model of a problem allows them to analyze it, and to test potential solutions. Usually

multiple reasonable solutions exist, so engineers must evaluate the different design

choices on their merits and choose the solution that best meets their requirements. A

mathematical model usually describes a system by a set of variables and a set of

equations that establish relationships between the variables. The values of the variables

can be real or integer numbers, boolean values or strings. The variables represent some

properties of the system, for example, system outputs in the form of signals, timing data,

counters, event occurrences etc. Mathematical models are of great importance in physics.

It is common to use idealized models in physics to simplify things. It is a effective tool

that engineers use to anlyse, control and optimize physical systems Throughout history,

8
more and more accurate mathematical models have been developed. The laws of physics

are represented with simple equations such as Newton's laws; Maxwell's equations

etc.These laws form a basis for making mathematical models of real situations. Many real

situations are very complex and thus modeled approximately on a computer; a model that

is computationally feasible to compute is made from the basic laws or from approximate

models made from the basic laws. Mathematical modeling problems are often classified

into black box or white box models, according to how much background information is

available of the system. A black-box model is a system of which there is no background

information available. A white-box model is a system where all necessary information is

available. Practically all systems are somewhere between the black-box and white-box

models, so this concept only works as an intuitive guide for approach. Usually it is

preferable to use as much background information as possible to make the model more

accurate. Therefore the white-box models are usually considered easier, because if you

have used the information correctly, then the model will behave correctly.

1.4.1 Steps in Mathematical Modeling

1. Problem identification: The system to be modeled is isolated from its

surroundings and the effects of the surroundings are noted. Known constants and

variables are identified.

2. Assumptions: Assumptions are made to simplify the modeling. Considering all

effects of the system results in complexity and hence a mathematical solution

becomes difficult. Assumptions should only be made if they yield results that are

simpler than those got without the assumptions. Sometimes certain implicit

9
assumptions are made which are taken for granted and are seldom mentioned.

Examples include assuming physical properties as continuous functions,

considering all materials to be linear, homogeneous and isotropic, ignoring

relativistic, chemical, nuclear and other effects.

3. Basic laws of nature: A basic law is a physical law that applies to all physical

systems regardless of the material from which the system is constructed. They are

laws that can be observed but not derived from more fundamental laws. Examples

include conservation of energy, conservation of momentum; the second and a third

law of thermodynamics.Among the above only conservation of momentum plays a

significant role in problems involving vibrating systems.

4. Constitutive equations: They provide information about the materials of which a

system is made. They are used to develop force-displacement relationships for

mechanical components used in vibration problems.

5. Geometric constraints: They are essential in completing the mathematical model

of a system. They can be in the form of kinematic relationships between velocity,

acceleration and dispalcement.Geometric constraints are used to formulate the

boundary conditions and initial conditions once the differential equations are

developed.

6. Mathematical solution: Once the mathematical problem is obtained, the modeling

is not complete till the appropriate mathematics is used to obtain a solution. Exact

analytical solutions, if they exist are preferable to numerical or approximate

10
solutions. Usually exact solutions exist for only for linear problems but for very few

nonlinear problems.

7. Physical interpretation of results: This is the next step after the modeling of the

desired system is complete and the mathematical solution is obtained. In certain

instances it may involve drawing general conclusions from the mathematical

solution or development of design curves or just require simple arithmetic to arrive

at a conclusion.

11
CHAPTER II

LITERATURE REVIEW.

Recent developments in the electronics industry over the last few years have a brought

about a huge increase in the usage of portable electronic products like cellular phones,

personal digital assistants and digital cameras. As these products get smaller for ease of

handling, they become more susceptible to drop impacts. Impact loads are the main cause

of mechanical failures in products [1-18]. There can be significant damage caused to the

part as a result of the impact forces developed during drop. Most of the common

electronic devices have a similar structure consisting of an upper structure and lower

structure connected by a hinge device. The upper structure includes the LCD module,

covers, metal frames and the lower structure includes a keyboard, housings, battery,

motherboard and chipsets [3]. There can be serious damage caused to electronic products

when dropped. The impact force produced can cause the internal components to fail or

cracking of the outer cover. Some of the damage can also include housing fracture, joint

breaking, connector disconnection or complete component failure [2]. Of these the one

that is of great concern is the functional failure of the part. The cause of such failure is

normally due to the loss of electrical contacts arising from the breakage of

interconnections between components within the product [1]. The resulting stresses and

strains can cause the housings to deform, cause assemblies to come apart and cause the

12
liquid crystal display (LCD) to crack [2]. As a result, studying the response of these

electronic components subjected to drop impacts is vital in predicting the life of the

component. Over the years various researchers have used different methods to study the

effect of drop on the performance of the system. The traditional method of analysis is

carrying out physical tests using a prototype. But in using such a method it is difficult to

capture the impact event as it happens over such a short span of time and involves a lot of

trial and error [5]. Also it has proven to be expensive, time consuming and requires a lot

of effort [2]. Another method that is very popular and widely used is simulation of the

impact event using a software package such as FEA and comparing it experimental or

analytical results [1, 2]. Tan et al [1] used Finite element (FE) simulation to model a ball

grid array (BGA) package that consists of a integrated circuit(IC) connected to a flexible

printed circuit board (PCB) subjected to drop impact. He used three different kinds of

mesh to model the BGA package and studied the deviation in results with the change in

meshing. The parameters that were compared were the deflection of the PCB and the

stresses that develop within the solder balls. He concluded that the type of mesh used to

model a component has an influence on the stresses and strains developed. Lim et al [2]

studied the feasibility of using FE simulation for the drop impact response of a pager by

performing a numerical simulation and using ABAQUS and verified numerical results

with experiments. He concluded through his experimental and simulation results that the

impact orientation is largely responsible for variation of strains and impact force during

collision. He attributed the variation of strain with time to the difference in local

deformation occurring at different parts of the pager. Low et al [4] studied the impact

effect on mini Hi-Fi audio products He used Pro-E to create his model and used PAM-

13
CRASH to study the effect of material properties on the results. However, using FEA as a

method to simulate the impact event has a few disadvantages. A major difficulty is in

meshing some of the smaller components. Some of these substructures in electronic

components need extremely fine meshes and this causes too much time for the analysis

and requires high end software that could be very expensive [4 - 6].Another problem is

setting up the simulation model accurately including the geometry of the component

being studied, selection of the right material for different components, boundary

conditions etc [5]. Some of the limitations of using software to simulate the impact event

can be overcome by studying the analytical or mathematical models of impact. Goyal et

al [16] used a linear spring-mass system to analytically model the drop impact of a

cellular phone. He concluded that the geometry of the component is an important factor

in analyzing the failure of the component. He suggested that improving the rigidity of the

case may prevent slipping of the cellular phone from its cover upon drop. Goyal [15]

also studied the clattering effects that occur when a component hits the ground at an

angle. He studied the jumps in velocity of the ends of a bar for successive impacts. He

found that when a two dimensional bar was dropped on the floor the second impact could

be twice as large as the first. He concluded that the number of times that the bar would

impact the ground depends on the co efficient of restitution and the mass distribution.

Shan et al [7] also did a comprehensive study on the effect of clattering on three

dimensional bars and found that his results were comparable with Goyal. Xiang [17] did

a study on the effect of a continual shock loads on packaged products during

transportation. Continuous shock loads cause undesirable stresses on packaged material

during transport. Xiang studied the effect of acceleration amplitude on the number of

14
shocks that lead to failure by performing a series of tests in the laboratory. He established

a relation between fragility of the product and the number of continuous shock loads to

failure. Wang et al [9] in his study of two dimensional rigid body collisions found some

methods of rigid body impact could violate the energy conservation principles. He

classified the different modes of impact using Routh’s graphical method to determine the

impulse. He blamed the violation of energy conservation to the use of Newton’s

hypothesis for the coefficient of restitution. Wang proved that using Poisson’s hypothesis

over Newton’s for the co efficient of restitution yielded better results He also stated that

identifying the correct mode of contact is vital in order to satisfy the law of conservation

of energy in an impact process. However Brach [18] blamed the violation of the principle

of energy conservation to the definition of friction. He adopted lower values of

coefficient of friction to satisfy the law of conservation of energy. Keller [8] extended the

Routh’s method to solve three dimensional problems. Barbulescu et al [11] used Kane’s

method to study spatial impact of a slender beam. In his studies he dealt with both slip at

the contact point and no slip. In the former case he assumed tangential velocity at

separation to be zero. He deduced a relationship between the loss of kinetic energy and

the angle of impact. He concluded that for conditions of no slip the energy loss increased

with increase in the angle of impact. Younis et al [13] used Galerkin’s approach to study

the dynamic response of beams to mechanical shock. He investigated the nonlinearity of

the response and attributed it to the effect of mid-plane stretching. He suggested

improving the thickness of the beam to improve shock resistance. He stated that

mechanical shock when combined with electrostatic force could cause a dynamic pull-in

in MEMS devices causing instability. He also studied the effect of packaging on MEMS

15
devices. One of his important findings was that neglecting the effect of packaging could

lead to failure of the device.

16
CHAPTER III

MATHEMATICAL MODELING

3.1 Introduction

In this Chapter we establish an impact model that calculates the impulse and hence the

change in velocities at the ends of the beam upon impact. A rigid frame modeled as the

housing of an electronic component with a flexible beam attached to it modeled as an

internal component is subjected to drop. The impact produces a change in velocity of the

housing that changes the boundary conditions of the beam. The response of the beam is

solved for the resulting boundary conditions. An Euler-Bernoulli beam is used for

analysis. Initially the problem is homogeneous with non homogeneous boundary

conditions. Using a linear function in x we transfer it into a non homogeneous problem

that calculates the change in velocity upon impact. The impulse produced upon impact

produces boundary conditions on the beam. The relation between the impulse and

boundary conditions is shown. Finally the response of the beam with the maximum

amplitude, stress and strain is presented.

3.2 Problem Formulation

The initial configuration of the system is shown in Figure 3.1, illustrating the

collision between a rigid frame with an attached flexible beam and the ground.

17
Q ê2 ê1
B

R
A
G ( x, y )
P

iˆ S

Figure 3.1: Rigid frame with flexible beam

Let PQRS represent the rigid frame with G being the mass center. ‘AB’ represents a

flexible beam attached to the frame with the help of two supports as shown in the figure.

The rigid frame represents the housing of an electronic device while the flexible beam

models an internal component such as a printed circuit board. Let the point of contact S

be the origin. ‘ iˆ ’ and ‘ ĵ ’ are unit vectors in the global coordinate system fixed to the

ground, ‘ ê1 ’ and ‘ ê2 ’ represent unit vectors in the local coordinate system of the beam

relative to the frame. The housing makes direct contact with the ground producing an

impulsive load at the contact point. The impulse produces a change in velocity of the

housing. This change in velocity impulsively loads the beam.

Using Hamilton’s principle, the transverse displacement of the beam w( x, t ) with

respect to the housing is derived as

∂2w ∂4w ∂2w


ρA + EI − P =0 (3.1)
∂t 2 ∂x 4 ∂x 2

18
where E is the elastic modulus of the beam, ρ is its mass density, A(x) is its cross

sectional area, I(x) is its cross section moment of inertia and P is the component of the

axial load in the transverse direction. Equation (3.1) represents the model of an Euler-

Bernoulli beam. It is convenient to nondimensionalize the partial differential equations

governing the vibrations of continuous systems through the introduction of

nondimensional dependent and independent variables. Nondimensional variables are

introduced such as

w
w* = (3.2)
L

x
x* = (3.3)
L

t
t* = (3.4)
T

Where the variable with an * is a nondimensional variable, L is a characteristic length in

the system such as the length of a beam, and T is a characteristic time scale. Equation 3.1

is nondimensionalized using the nondimensional variables of equation 3.2 - 3.4 .The

beam is considered to be of uniform cross section and therefore spatial dependence is not

considered.

ρAL2
β= (3.5)
PT 2

EI
α= (3.6)
PL2

The effect of rotary inertia and shear deflection is ignored and as a result the load in the

axial direction is P = 1 , dropping the *’s and considering α and β to be constants we

get the nondimensional form of equation (3.1) as:


19
∂4w ∂2w ∂2w
α 4 − 2 +β 2 =0 (3.7)
∂x ∂x ∂t

Equation 3.7 represents a homogenous equation governing the vibrations of the beam

‘AB’. The equation is homogeneous because the forcing term is assumed to be equal to

zero.

3.3 Boundary Conditions

The boundary conditions are the displacement and velocities at the two ends of the

∂w
beam at all times. If w( x, t ) is the displacement of a particle ‘x’ at time ‘t’ and, is its
∂t

velocity, the boundary conditions of the non dimensional beam are given by:

w(0, t ) = 0 (3.8)

w(1, t ) = 0 (3.9)

∂w
(0, t ) = v A (t ) (3.10)
∂t

∂w
(1, t ) = v B (t ) (3.11)
∂t

3.3.1 Initial Conditions

The initial conditions represent the displacement and velocities before impact. At time

t = 0 the displacement and velocity of any particle ‘x’ over the entire span of the beam

relative to the rigid frame is zero. This would mean that v A and v B are zero before impact,

so that the initial conditions are:

w( x,0) = 0 (3.12)

∂w( x, t )
=0 (3.13)
∂t t =0

20
It is seen that the differential equation governing the motion of the beam is

homogeneous whereas the boundary conditions are non-homogeneous. We find a

solution by transforming it to a problem with a non-homogeneous differential equation

having homogeneous boundary conditions.

3.3.2 Transformation of boundary conditions

We first choose a function b( x, t ) that renders the boundary conditions homogeneous.

The solution to the problem is then assumed to be the sum of a variable u ( x, t ) and the

function b( x, t )

w( x, t ) = u ( x, t ) + b( x, t ) (3.14)

Here b( x, t ) is chosen to be a function linear in x.

∂b( x, t )
= [v A (t ) + (v B (t ) − v A (t ) )x ] (3.15)
∂t

This function satisfies the boundary conditions of equations (3.10) and (3.11). In this

manner, we transform the boundary value problem for the variable w( x, t ) into a

boundary value problem for the variable u ( x, t ) . Taking the partial derivative of

equation (3.14), using values of x = 0 and x = 1 we get:

∂w ∂u
(0, t ) = (0, t ) + v A (t ) (3.16)
∂t ∂t

∂w ∂u
(1, t ) = (1, t ) + v B (t ) (3.17)
∂t ∂t

Substituting equations (3.16) and (3.17) in (3.10) and (3.11) we get:

∂u
(0, t ) = 0 (3.18)
∂t

21
∂u
(1, t ) = 0 (3.19)
∂t

Therefore the non homogeneous boundary conditions are made homogeneous by the

choice of a function that satisfies the boundary conditions. The new homogeneous

boundary conditions are written as:

u (0, t ) = 0 (3.20)

u (1, t ) = 0 (3.21)

∂u
(0, t ) = v A (t ) (3.22)
∂t

∂u
(1, t ) = v B (t ) (3.23)
∂t

3.4 The dynamics of impact

When the object hits the ground there is an impulse produced at the contact point. Due

to the sudden impact upon contact with the ground the acceleration at the ends of the

beam are expressed as:

dv A
(t ) = ∆v Aδ (t ) (3.24)
dt

dv B
(t ) = ∆v Bδ (t ) (3.25)
dt

In the above two equations δ (t ) is a unit impulse that has an infinite value when x = 0

and zero at all other places. ∆v A and ∆v B are the changes in velocities at the ends of the

beam. Therefore now substituting equation (3.14) in equation (3.7) and using equation

(3.15) we get the governing equation as:

∂ 4u ∂ 2u ∂ 2u
α 4 − 2 + β 2 = −[∆v A + (∆v B − ∆v A )x ]δ (t ) (3.26)
∂x ∂x ∂t
22
Equation (3.26) represents a non homogeneous boundary value problem for the variable

u ( x, t ) with equations (3.20) - (3.23) representing the homogeneous boundary conditions.

3.5 Impact Velocities

Upon impact there is a change of velocity at the two ends of the beam. These

velocities denoted by ∆v A and ∆v B depend upon the impulse produced upon impact at

the interface between the ground and the rigid block. Using Linear momentum balance,

we get:

∑ F = ma (3.27)

Equation (3.27) can also be written as,

dv
∑ F = m dt (3.28)

Where,

F = force acting on the body

m = mass of the body,

a = acceleration of the mass center of the body.

v = velocity of the mass center of the body

Equation (3.28) is written as,

t
v G (t ) − v G (0) = m −1 ∫ F (τ )dτ (3.29)
0

v G (t ) − v G (0) = m −1 P (t ) (3.30)

Equation (3.30) gives the change in linear velocity at the mass center ‘G’ upon impact.

Using angular momentum balance we get,

∑ M G = I Gα (3.31)
23
Where,

M G = moment acting on the body.

I G = moment of inertia about the mass center

α = acceleration about the mass center of the body.

Using equation (3.31) we get,

I G [ω (t ) − ω (0)] = r BG × P(t ) (3.32)

−1
ω (t ) − ω (0) = I G [r BG × P(t )] (3.33)

Equation (3.33) gives the change in angular velocity at the mass center upon impact. Here

rBG is the position vector of B with respect to the mass center G and P(t ) is the impulse

acting on the body at the point of contact.

rBG R
A
G
P rB
rG
Py

S Px

Figure 3.2: Illustration of Impact

Using basic vector mechanics to express the velocities at the ends of beam with respect to

their mass centers we get,

∆v A = m −1 P(t ) + r AG (t ) × ω (t ) − r AG (0) × ω (0) (3.34)

∆v A = m −1 P(t ) + r AG (t ) × [ω (t ) − ω (0)] (3.35)

24
Similarly we get,

∆v B = m −1 P(t ) + r BG (t ) × [ω (t ) − ω (0)] (3.36)

3.6 Routh’s method to find the impulse

There are several methods to calculate the impulse produced upon impact of a rigid

body. In this section we briefly describe Routh’s graphical procedure for two dimensional

collisions [9]. Routh’s method is a graphical technique for analyzing planar frictional

impact using Coulomb’s law of friction. According to this method an impulse is made of

two parts, a compression impulse, and a restitution impulse. The compression impulse or

the compression phase is from the beginning of collision to the time that their relative

velocity is zero. The restitution impulse or the restitution phase is measured from the time

the relative velocity is zero to the time that objects begin to separate. Accordingly

impulse P is expressed as:

P (t ) = Pt tˆ + Pn nˆ (3.37)

In equation (3.37) subscripts t and n denote the tangential and normal directions

respectively, which are in the iˆ and ĵ directions in the given contact plane. This model

uses the Poisson’s hypothesis to define the coefficient of restitution. Accordingly the

coefficient of restitution denoted by ‘ e ’ is defined by:

(Pn )r
e= (3.38)
(Pn )c

In equation (3.38) (Pn )r is the normal impulse during restitution and (Pn )c is the normal

impulse during compression. By using the Poisson’s hypothesis to define the coefficient

of restitution this method satisfies the basic energy conservation principles as there is no

25
increase in kinetic energy. The Routh-Poisson analysis gives an impulse in accordance

with Coulomb’s law, without an increase in total energy. Also it considers a new type of

impact called the tangential impact, an impact with zero initial approach velocity. It can

be used to distinguish between several types of contact and to identify when sliding

ceases or reverses. In this thesis the Routh’s method is used to find the impulse at the

contact point of the rigid body with the ground. Once the impulse is known the velocities

at the two ends of the beam are calculated. Finally the response of the beam is presented

knowing the velocities.

3.6.1 Procedure to calculate impulse

Once the object hits the ground an impulse P(t ) is produced. This impulse can be

divided into two components, an impulse in the normal direction Pn and an impulse in the

tangential direction Pt .Let the body with mass center ( x, y ) have initial translational and

rotational velocities x& 0 , y& 0 and θ&0 .Using linear and angular impulse-momentum laws we

write the following equations [9]:

m( x& − x& 0 ) = Pt (3.39)

m( y& − y& 0 ) = Pn (3.40)

mρ 2 (θ& − θ&0 ) = Pt y − Pn x (3.41)

Here m is the mass, ρ is the radius of gyration of inertia and x& , y& are the velocities upon

collision. The kinetic energy is given by:

1 1
T=
2
( )
m x& 2 + y& 2 + mρ 2θ& 2
2
(3.41a)

26
m  P   mρ
2 2 2
 P
2
 Pt y − Pn x & 
T =  t + x&0  +  n + y& 0   +  mρ 2 + θ 0  (3.41b)
2  m  m   2  

Expanding and rearranging we get:

m 2 mρ 2 & 2   P 2 P 2 (P y − Pn x)
( 2
T =  x&0 + y&0 + ) [ ]
θ0  + Pt x&0 + Pn y&0 + (Pt y − Pn x)θ&0 +  t + n + t
2mρ 2 
 (3.41c)
2 2   2m 2m

The isoenergetic ellipse satisfies the law of conservation of energy. Hence for this

condition to be satisfied Pt and Pn assume values such that:

 P 2 P 2 (P y − Pn x ) 
[P x&
t 0 ]
+ Pn y& 0 + (Pt y − Pn x )θ&0 +  t + n + t
2mρ 2 
 = 0.
 2m 2m

G
rG rG = xiˆ + yˆj

Figure 3.3: Position of Mass center

The velocity of the object when it comes in contact with the ground is called contact

velocity and is given by:

x& c = x& + θ&y (3.42)

y& c = y& − θ&x (3.43)

27
B

A
G ( x, y )

(x& c , y& c )

Figure 3.4: Contact velocities

Equation (3.42) denotes the tangential component of the relative velocity of the points in

contact. This velocity represented by ‘S’ is called the sliding velocity. Similarly equation

(3.43) denotes the normal component of relative velocity and is denoted by ‘C’. In other

words S = x& c and C = y& c .Substituting equations (3.39)-(3.41) into (3.42) and (3.43) we

get:

S = S 0 + B1 P t − B3 P n (3.44)

C = C 0 − B3 P t + B2 P n (3.45)

Equations (3.44) and (3.45) represent the line of sticking and line of maximum

compression respectively. S 0 and C 0 represent the initial values of sliding and compression

respectively. These depend on the values of the tangential and normal component of the

initial velocity and the initial rotational velocity. They are expressed as:

S 0 = x& 0 +θ& 0y (3.46)

C 0 = y& 0 −θ& 0x (3.47)

B1 , B2 and B3 are constants that depend on the geometry and mass properties of the

system. They are given by:


28
1 y2
B1 = + (3.48)
m mρ 2

1 x2
B2 = + (3.49)
m mρ 2

xy
B3 = (3.50)
mρ 2

The Routh method determines the two components of impulse Pt and Pn using equations

(3.44)-(3.50).This is done graphically and explained in detail in the next section.

3.6.2 The impact process diagram:

When an object deforms, it happens in two phases, compression and restitution.

When compression ends, the normal component of the relative velocity of the points in

contact is zero (C = 0).Therefore from equation (3.45) a linear relationship between the

impulse components at maximum compression is obtained as

C 0 − B3 Pt + B2 Pn = 0 (3.51)

Similarly in the sticking case, the tangential component of relative velocity becomes zero

(S = 0). Therefore a linear relationship between the impulse components at slip is

obtained as

S 0 + B1 Pt − B3 Pn = 0 (3.52)

Now, equations (3.51) and (3.52) represent the line of maximum compression and line of

no sliding respectively. These two lines represent straight lines since they are linear and

are plotted on the impulse plane with Pt representing the horizontal axis and Pn the

vertical axis. Therefore using equations (3.51) and (3.52) we get:

29
B2 C
Pt = Pn + 0 (3.53)
B3 B3

B3 S
Pt = Pn − 0 (3.54)
B1 B1

We also define the line of limiting friction as

Pt = − µsPn (3.55)

In equation (3.55) µ is the coefficient of friction and is a constant and s is the sign of the

S0
initial sliding velocity S 0 , s = if S 0 is not zero. Figures 3.5 and 3.6 give a detailed
S0

explanation of the impact process diagram.

Friction
Compression
Stick
Pn
Pn (t f )

Pn (t c )

Pt

Figure 3.5: Impact process diagram illustrating Slip-Stick

When impact begins the point P which is the total impulse is at the origin and lies on the

line of limiting friction. Assuming initial sliding it increases along the line of limiting

friction. It proceeds along this line until it reaches line sticking or line of maximum

compression. If it reaches the line of maximum compression first, the value of normal

30
impulse Pn (t c )) at that instant is noted. The process of impact will end when the value of

Pn is 1+e times the value of ( Pn (t c )) .This value is denoted by Pn (t f )


Pn (t f ) = (1 + e )Pn (t c ) (3.56)

This is called the termination condition .Then the point continues along the line of

limiting friction until termination is met or P reaches the line of sticking. If P reaches

the line of sticking before termination it continues along this line till termination. This

process of slip-stick is illustrated in Figure 3.5.

Alternatively if it reaches the line of stick first then slipping ends and if the limiting

friction is more than the friction necessary to prevent sliding P will continue to stick till

the process terminates as illustrated in figure 3.6.It will eventually cross the line of

maximum compression and terminate according to the termination condition as stated

earlier. On the other hand if the limiting friction is less than the friction necessary to

prevent sliding P will cross the line of sticking and will travel along the line of reversed

limiting friction and will continue till termination. In such a scenario the object will slip

through out the impact process till termination condition is reached.

31
Friction
Pn Compression
Stick

Pt

Figure 3.6: Impact process diagram illustrating Stick

The isoenergetic ellipse depicted in the above figures should satisfy the energy

conservation principles.Acccordingly if ∆T is the change in kinetic energy then

according to the law of conservation of energy ∆T = 0 .

1 T
∆T =
2
[P (VC + VCO ) ] (3.56a)

Where P = [Pt , Pn ]
T

VC = [S , C ] and VC 0 = [S 0 , C 0 ]
T T

The energy change is given by:

1 T
∆T =
2
[ T
P BP + 2VC 0 P ] (3.56b)

 B − B3 
Where B =  1
− B3 B2 

The energy ellipse requires that ∆T = 0 .

32
3.7 Galerkin Reduction

Galerkin’s method is a method used to find an approximate solution to continuous

systems. Often exact solutions for higher order equations do not exist. Even if they do

they are cumbersome to use, requiring solutions to higher order equations. In such a

scenario Galerkin’s method is a convenient method to use although it yields only an

approximate solution. It is a means for converting a partial differential equation (PDE) to

a system of ordinary differential equations (ODEs), which become easier to handle. It

works on the principle of restricting the possible solutions to a smaller space than the

original. These smaller systems are easier to solve and less time consuming.

Galerkin’s method approximates the solution to a boundary value problem by using a

linear combination of trial functions. In order to solve our problem we choose a trial

functions that satisfy the boundary conditions.

3.7.1 Galerkin’s method as applied to the beam model.

Assume the approximate solution to the problem to be:

N
u ( x, t ) = ∑ Ai (t )φi ( x ) (3.57)
i =1

Where the trial functions φ1 ( x ), φ 2 ( x ),...φ n (x ) are the independent comparison functions

from a complete set and A1 , A2 ,... An are undetermined coefficients. Comparison functions

are trial functions which are differentiable as many times as the order of the system and

satisfy all the boundary conditions. The above solution may not satisfy the exact

differential equation defining the eigenvalue problem, so that some error is incurred. The

error is denoted by R (u (n ) , x ) , known as the residual, and because u (n ) is a linear

combination of comparison functions, the boundary conditions are satisfied exactly. To


33
determine the coefficients A1 , A2 ,.. An we multiply the residual R (u (n ) , x ) by

φ1 ( x ), φ 2 ( x ),..., φ n (x ) in sequence, integrate the result over the domain of the system, and

set equal to zero.

∫ φ (x)R(u )
(n )
i , x dx = 0, i = 1, 2……..n (3.58)
0

N
Now consider equation (3.26), substituting for u ( x, t ) ≈ u ( N ) ( x, t ) = ∑ Ai (t )φi ( x) where
i =1

Ai (t ) is the ith generalized coordinate and φi ( x) is the ith linear undamped mode shape

of the straight beam we get:

∂4  N  ∂2  N  ∂2  N 
α 4 ∑φi (x)Ai (t) − 2 ∑φi (x)Ai (t) + β 2 ∑φi (x)Ai (t) +[∆vA + (∆vB − ∆vA)x]δ (t) = R(u(n) , x) (3.59)
∂x  i=1  ∂x  i=1  ∂t  i=1 

Equation (3.59) represents the residual. It has been shown in previous studies [13] that

four modes are sufficient to capture the dynamic response of a beam pinned at both ends.

Therefore in equation (3.27) substituting N = 4 we get:

u ( N ) ( x, t ) = φ1 ( x) A1 (t ) + φ 2 ( x) A2 (t ) + φ3 ( x) A3 (t ) + φ 4 ( x) A4 (t ) (3.60)

We choose the first four mode shapes to be:

πx 2πx 3πx 4πx


φ1 ( x) = sin , φ 2 ( x) = sin , φ3 ( x) = sin , φ 4 ( x) = sin
l l l l

Therefore equation (3.60) becomes:

πx 2πx 3πx 4πx


u ( x, t ) ≈ u ( N ) ( x, t ) = sin A1 (t ) + sin ( x) A2 (t ) + sin ( x) A3 (t ) + sin A4 (t ) (3.61)
l l l l

34
Equation (3.61) represents the total deflection of the beam. To find the generalized

coordinates in equation (3.61) we substitute φi ( x) = sin kπx . Using values of k = 1 to 4 we

get four sets of equations from which A1 , A2 , A3 and A4 are determined. Therefore we get:

N 
l

∫0 sin (kπx )R ∑ Ai (t )φi ( x), x  dx = 0 (3.62)


 i =1 

For N = 4, k = 1 we get:

β &&  − ∆v B ∆v A 
A1 (t ) + [0.5απ 4 + 0.5π 2 ]A1 (t ) =  − δ (t ) (3.63)
2  π π 

Similarly for k = 2, 3 and 4 we get:

β &&  ∆v ∆v 
A2 (t ) + [8απ 4 + 2π 2 ]A2 (t ) =  B − A δ (t ) (3.64)
2  2π π 

β &&  − ∆v B ∆v A 
A3 (t ) + [40.5απ 4 + 4.5π 2 ]A3 (t ) =  − δ (t ) (3.65)
2  3π 3π 

β &&  ∆v ∆v 
A4 (t ) + [128απ 4 + 8π 2 ]A4 (t ) =  B − A δ (t ) (3.66)
2  4π 4π 

Equations (3.63)-(3.66) represent the first four mode shape equations. δ (t ) is a unit

impulse due to an instantaneous collision with the ground. For an undamped system of

the form &x& + ω 2 n x = δ (t ) , the response to a unit impulse is denoted by h (t) and is given

by:

1
h(t ) = sin ω n t (3.67)
ωn

In the above equations ∆v A and ∆v B are the changes in velocities of the beam upon

impact.

35
3.8 Stresses and strains

If M (x) is the bending moment and y is coordinate measured from the neutral axis in the

cross section of the beam, using elementary beam theory the normal stress at a point

in the cross-section due to bending is expressed as

M ( x) y
σ= (3.68)
I

∂ 2u
M ( x) = EI (3.69)
∂x 2

Equations (3.68) and (3.69) are used to get the stress response of the beam.

Potential energy in the form of strain energy is stored in all deformable systems. For a

loaded member the normal stress (σ ) and normal strain (ε ) follow the Hooke’s law. The

σ
stress -strain curve is linear. Therefore we have = E .Using this we find the maximum
ε

strain.

3.9 Review

The previous sections dealt with calculating the value of the impulse produced

upon drop. Once the impulse is known it is substituted back in equations (3.35 and (3.36)

to get the values of changes in velocities at ‘A’ and ‘B’ respectively. Before this the

geometric dimensions of the problem including the angle of impact is used to calculate

the position vectors of the mass center of the body and the ends of the beam. Once the

change in velocities are known these are substituted back in the mode shape equations

(3.63-3.66).This gives the values of the constants which are then substituted back in the

response on the beam i.e. equation (3.60) to get the deflection of the beam. Upon

36
knowing the response of the beam upon impact, the locations of maximum stress, strain,

deflection can be got.

37
CHAPTER 1V

THE DYNAMICS OF IMPACT AND IMPACT VELOCITIES

4.1: Overview

In this chapter we apply the basic equations for the dynamics of impact in Chapter

3 to a specific impact problem. We first find the impulse developed because of impact

and hence the change in velocity due to impact. A final solution is then presented using

the Galerkin’s approach. Finally the stress, strain and amplitude responses illustrating

their maximum values are presented.

4.2 An impact problem

Consider a rigid frame (PQRS) of mass m = 2 kg, width W = 0.5m, height H =

0.3m and ρ = 1 with a flexible beam ‘AB’ of length 0.25m falling from a height and

hitting the ground at θ = 45 o . Assume that the angular velocity ω of the block is zero.

B ê2
ê1
rBG R
A
rBS G
P rGS
rAS
ĵ Py

iˆ S Px

Figure 4.1: An Impact Problem


38
Let iˆ and ĵ be unit vectors of the coordinate system fixed to the ground. Let ê1 and

ê2 be unit vectors of the coordinate system of the motion of the frame relative to the

ground. Since there is both translation and rotation of the frame relative to the ground

their unit vectors are expressed as:

eˆ1 = cos θiˆ + sin θˆj (4.1)

eˆ2 = − sin θiˆ + cos θˆj (4.2)

Let r AS , r BS and r GS be the position vectors of points ‘A’, ‘B’, and ‘G’ relative to the

ground and r AG , r BG be the position vectors of ‘A’ relative to ‘G’ and ‘B’ relative to ‘G’

respectively. Using equations (4.1) and (4.2) and the dimensions of the frame and beam

we get the following equations for figure 4.1

rGS = 0.289eˆ1 + 0.0005eˆ2 (4.3)

rAS = 0eˆ1 − 0.398eˆ2 (4.4)

rBS = 0.469eˆ1 − 0.029 × 10 −3 eˆ2 (4.5)

Also using basic vector mechanics we get:

rAS = rGS + rAG (4.6)

rBS = rGS + rBG (4.7)

In equations (4.6) and (4.7) rAG and rBG are the position vectors of ‘A’ relative to ‘G’ and

‘B’ relative to ‘G’ respectively. From equations (4.6) and (4.7) we get:

rAG = −0.289eˆ1 − 0.398eˆ2 (4.8)

rBG = 0.180eˆ1 − 5.29 × 10 −4 eˆ2 (4.9)

39
The frame makes contact with the ground at an angle θ = 45 o .This produces an impulse

at the contact point which affects the velocities at the ends of the beam. We use linear and

angular momentum balance equations to express the relation between the change in

velocity and the impulse produced. Equations (3.35) and (3.36) give the relationship

between the change in velocities and the impulse. We now proceed to find the impulse in

order to calculate the change in velocities.

4.3. Routh’s graphical method

We use Routh’s method for two dimensional impacts to find the value of the

impact P(t ) at the point of contact. Using equation (4.3), we find the mass center(x, y) of

the frame to be (0.289, 0). Considering the geometry of our problem and using equations

(3.48) – (3.50) we get:

1 0.0005 2
B1 = + = 0.50 (4.10)
2 2(1) 2

1 0.289 2
B2 = + = 0.541 (4.11)
2 2(1) 2

0.289 × 0.0005
B3 = 2
= 0.07 × 10 −3 (4.12)
2(1)

At time t = 0 there is an initial tangential and normal component of velocity that is

assumed. These are denoted by S 0 and C 0 respectively. The angular velocity is zero,

therefore the initial values of S 0 and C 0 are given by the initial translational velocities

x& 0 and y& 0 .They are also called the initial stick and compression velocities and their

values are assumed to be 1 and -1 respectively. Using equations (3.53) and (3.54), values

of B1 , B2 , B3 and the initial stick and compression velocities we plot the line of stick and

40
line of maximum compression on an impulse plane with Pt representing the horizontal

axis and Pn the vertical axis. The coefficient of friction µ is assumed to be 0.6. Figure

4.2 illustrates the impact process.

Friction Stick

5
4
P(n)

3 Impulse Path
2
Line of Compression
1
0
-5 -3 -1 1 3 5
P(t)

Figure 4.2: Impact process diagram.

When impact begins P is at the origin, as it progresses Pn increases as it begins to slip

along the line of friction , Pt also accumulates according to the relationship between the

two as expressed in equations (3.53) and (3.54).When Pn reaches the line of maximum

compression the value of the normal impulse is noted. Termination occurs when the value

of Pn reaches 1 + e times the value of Pn obtained at maximum compression. In this case

impact terminates when Pn =2.55.The object continues to slip along the line of limiting

friction till termination is reached.

41
Upon termination the values of Pt and Pn are found to be -1.8 and 2.55 respectively.

Therefore the total impact is given by:

P = −1.8iˆ + 2.55 ˆj (4.13)

P = 0.53eˆ1 + 3.07eˆ2 (4.14)

4.4 Using the impulse to find the change in velocities

The change in angular velocity at the mass center is given by equation (3.33).

Using this we get:

ω (t ) − ω (0) = 0.96kˆ (4.15)

Equations (3.35) and (3.36) give the change in velocities at ‘A’ and ‘B’. Using these

equations we get:

∆v A = 0.09iˆ + 2.13 ˆj (4.16)

∆v B = 0.45iˆ + 1.11 ˆj (4.17)

Since we consider the beam to be pinned at both the ends there is no movement in the

tangential direction and so in further calculations only the normal component of the

velocities are considered. Using equations (4.1) and (4.2) we get the tangential and

normal components of the velocities of the beam relative to the ground to be:

∆v A = 1.44eˆ1 − 1.56eˆ2 (4.18)

∆v B = 0.97eˆ1 − 0.69eˆ2 (4.19)

where ê1 and ê2 are unit vectors in the tangential and normal directions respectively.

42
4.5 Mode shape equations

Consider the first mode as given by equation (3.63).Assuming the values of α and

β to be equal to 1 we get:

&& (t ) + 107.2 A (t ) =  − ∆v B − ∆v A δ (t )
A (4.20)
1 1  π π 

Now substituting the value of the normal component of the change in velocities at ‘A’

and ‘B’ we get:

&& (t ) + 107.2 A (t ) = 0.71δ (t )


A (4.21)
1 1

Equation 4.21 represents the first mode. Similarly for the second, third and fourth modes

we get:

&& (t ) + 1598.02 A (t ) = 0.38δ (t )


A (4.22)
2 2

&& (t ) + 7929.6 A (t ) = 0.56δ (t )


A (4.23)
3 3

&& (t ) + 25094.6 A (t ) = 0.06δ (t )


A (4.24)
4 4

Equation (4.21) represents a second order linear differential equation with f 0δ (t ) being

the forcing function where δ (t ) represents a unit impulse. The standard form of the above

equation for an undamped system is written as:

{&x&(t ) + ω 2 n x(t )} = P(t ) (4.25)

k F (t )
Where ω 2 n = and P(t ) =
m m

ω n = natural frequency and P(t ) is the forcing function.

The solution to equation (4.20) is given by:

43
A1 (t ) = C1 f 1 (t ) + C 2 f 2 (t ) (4.26)

f 1 (t ) and f 2 (t ) represent the solutions to a homogeneous problem, C1 and C 2 represent

constants that are determined using initial conditions. Considering sin ωt and cos ωt to be

the solutions to the homogeneous problem we get:

x(t ) = C1 sin ω n t + C 2 cos ω n t (4.27)

In Equation (4.27), C1 and C 2 are constants that are found using the initial conditions of

displacement and velocity. Equation (4.20) is solved in a similar manner by treating it as

an initial value problem with A1 (t ) being the generalized coordinate representing

displacement. Therefore the solution to equation (4.20) is written as:

A1 (t ) = C1 sin 10.35t + C 2 cos10.35t (4.28)

The initial displacement is taken to be zero. The principle of linear impulse and

momentum is used to determine the velocity upon impact.

Upon impact there is change in momentum of the body. This change in momentum is

the impulse that is produced upon contact with the ground. If 0 + represents the time after

impact the impulse is mathematically expressed as

A&1 (0+ ) − A&1 (0− ) = f 0 (4.29)

The change in velocity before impact is zero, therefore Equation (4.29) becomes

A&1 (0+ ) = f 0 (4.30)

Therefore using initial displacement as zero, i.e. A1 (0) = 0 in Equation (4.28) we get

C 2 = 0 and using Equation (4.30) as the initial velocity we get C1 = −0.005 .Therefore

Equation (4.28) becomes:

44
A1 (t ) = 0.06 sin 10.35t (4.31)

Similarly we get,

A2 (t ) = 0.009 sin 39.97t (4.32)

A3 (t ) = 0.006 sin 89.04t (4.33)

A4 (t ) = 0.0003 sin 158.4t (4.34)

Finally using the above calculated values and the assumed mode shapes in equation

(3.60) we get the total response of the beam to be:

u ( x, t ) = sin πx[0.06 sin 10.35t ] + sin 2πx[0.009 sin 39.97t ]


(4.35)
+ sin 3πx[0.006 sin 89.04t ] + sin 4πx[0.0003 sin 158.4t ]

The stress response is given by:

σ ( x, t ) = − sin πx[6.21 × 10 9 sin 10.35t ] − sin 2πx[3.73 × 10 9 sin 39.97t ]


(4.36)
[ ] [
− sin 3πx 5.59 × 10 9 sin 89.04t − sin 4πx 0.49 × 10 9 sin 158.4t ]
The strain response is given by:

ε ( x, t ) = − sin πx[0.02 sin 10.35t ] − sin 2πx[0.01 sin 39.97t ]


(4.37)
− sin 3πx[0.02 sin 89.04t ] − sin 4πx[0.002 sin 158.4t ]

4.6 Results

Equation (4.35) represents the displacement of the beam upon impact. A time

interval of 0 to 5 seconds is considered and the response of the beam within this interval

is studied to find the maximum amplitude of displacement and its location. A MATLAB

script is written to first find the time at which the maximum deflection occurs. The

response of the beam at this time is plotted to find the value of maximum deflection and

its location. This plot is illustrated in figure 4.3. Similarly maximum stresses and strains

are tabulated and illustrated in figures 4.4 and 4.5 respectively.

45
0.06

0.05

0.04
Deflection

0.03

0.02

0.01

0
0 0.05 0.1 0.15 0.2 0.25
Location along length of beam (x)

Figure 4.3: Maximum Deflection

In figure 4.3 it is seen that the amplitude of maximum deflection is 0.05m at x = 0.25m

at time t = 3.18 seconds.

9
x 10
12

10

8
Stress

0
0 0.05 0.1 0.15 0.2 0.25
Location along length of beam (x)

Figure 4.4: Maximum Stress

46
The maximum stress developed is 11927.2 Mpa at x = 0.21m at time t = 2.31 seconds.

0.04

0.035

0.03

0.025
Strain

0.02

0.015

0.01

0.005

0
0 0.05 0.1 0.15 0.2 0.25
Location along length of beam (x)

Figure 4.5: Maximum Strain

The maximum strain developed is 0.03 at x = 0.21m at time t = 2.31seconds.

47
CHAPTER V

RESULTS AND CONCLUSION

5.1 Overview

In this chapter we discuss about how the angle of approach affects the impulse

produced at the point of contact and the velocities at the end of the beam (boundary

conditions). The effect of the coefficient of friction (µ ) and the coefficient of

restitution (e ) on the normal and tangential impulses are explained. The response of the

beam fixed at different positions on the frame is studied to see the location of maximum

stress and strain developed in the beam. The influence of the boundary conditions on the

stresses and strains developed for different angles of impact is discussed.

5.2 Angle of approach v/s impact:

The impact produced upon contact of the rigid frame with the ground depends upon

the material properties of the impacting body and the angle at which it hits the ground.

Observation of the line of sticking and line of maximum compression indicates the

dependency of impact on the geometric constants B1 , B2 and B3 .

S = S 0 + B1 P x − B3 P y (5.1)

C = C 0 − B3 P x + B2 P y (5.2)

These values will change with change in the coordinates of the impacting body which

would in turn change with the angle at which the body makes contact with the ground.

48
In this section we see the change in impact force with the change in angle of approach

and draw suitable conclusions. The coefficient of friction µ remains constant at 0.8 for

all angles of impact with initial conditions of S 0 = 1 , C 0 = −1 and e = 0.8 . The values

of B1 , B2 and B3 are calculated for the different angles of approach with the other

geometric constants being mass m = 2 kg, width W = 0.5m, height H = 0.3m and ρ = 1

with a flexible beam ‘AB’ of length 0.25m

5
Magnitude of Impulse (P)

4.5

3.5

2.5

2
5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85 90
Angle of Impact

Figure 5.1: Angle of Impact v/s Magnitude of impact

Looking at the above figure we find that the impulse starts at a value of 2.95 at θ = 5 o

and increases to a maximum of 4.61 at θ = 15 o .Thereafter it varies continuously till it

reaches a value of 2.95 at θ = 90 o . As stated earlier impulse produced depends on the

material properties of the impacting body in addition to the geometric constraints. Figures

5.2, 5.3 and 5.4 illustrate the impulse paths at a few representative points. These give a

more detailed view of the change in impulse with the orientation angle.

49
Friction Stick

5.5
5
4.5
4
3.5
Impulse Path

P(n)
3
2.5
2
1.5 Line of Compression
1
0.5
0
-5 -3 -1 1 3 5
P(t)

Figure 5.2: Impulse path at θ = 15 o

Friction Stick

5
4
P(n)

3 Impulse Path
2
Line of Compression
1

0
-5 -3 -1 1 3 5
P(t)

Figure 5.3: Impulse path at θ = 45 o

Friction Stick

5.5
5
4.5
4 Impulse Path
3.5
3
2.5 Line of Compression
2
1.5
1
0.5
0
-5 -3 -1 1 3 5

Figure 5.4: Impulse path at θ = 75 o


50
In the next section we see the effect of coefficient of friction (µ ) and coefficient of

restitution (e ) on the normal and tangential impulses.

5.3. Influence of (µ ) on the impulse at contact point.

Consider the collision problem of chapter 4.The coefficient of friction (µ ) causes

an impulsive force in the tangential direction at the point of contact. This section deals

with the change in tangential impulse (Pt ) with µ for the above mentioned collision

problem.

2.5
2
1.5
Tangential impulse

1
0.5
0
-0.5 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2 2.2 2.4
-1
-1.5
-2
-2.5
Co efficient of friction

Figure 5.5: µ v/s Pt

In the above figure it seen that the value of the tangential impulse starts at -0.5

for µ = 0.2 and then increases in the negative direction as µ increases to 0.8 before being

steady for values of µ between 1 and 1.4. For lower values of µ the impact process

terminates before the object reaches the line of stick. Therefore the object slips

throughout the process of impact. For µ varying between 1 and 1.4 the line of friction

meets the line of stick before termination because of this Pt remains constant at -2.For

51
values of µ greater that 1.4 the line of friction meets the line of stick before the line of

maximum compression. This leads to reverse slip. Since the slope of the line of stick is

steeper than the line of reverse limiting friction it continues to stick till termination. As

the value of µ continues to increase the value of tangential impulse will remain constant

at 2, but the object would start to stick sooner. The below figures show the impulse paths

followed for few representative values of friction. This explains the variation in

tangential impulse more clearly.

f=0.2 Stick

6
5
4
Impulse Path
P(n)

3
2
Line of Compression
1
0
-5 -3 -1 1 3 5
P(t)

Figure 5.6: Impulse path for µ = 0.2

Friction Stick

6
5
4
P(n)

3 Impulse Path
2
Line of Compression
1
0
-5 -3 -1 1 3 5
P(t)

Figure 5.7: Impulse path for µ = 0.8


52
f=1 Stick

4 Impulse Path

P(n)
3

2
Line of Compression
1

0
-6 -5 -4 -3 -2 -1 0 1 2 3 4 5

P(t)

Figure 5.8: Impulse path for µ = 1

f=2 Stick

5
4
Impulse Path
P(n)

3
Reverse Friction
2
Line of Compression
1
0
-5 -3 -1 1 3 5
P(t)

Figure 5.9: Impulse path for µ = 2

5.4. Influence of the coefficient of restitution (e) on the impulse.

The coefficient of restitution denoted by ‘e’ is the ratio of speeds of a falling

object from when it hits the surface to when it leaves the surface. It is a measure of the

elasticity of a collision. A perfectly elastic collision has e equal to 1 while a perfectly

plastic collision (object sticks immediately upon impact) has e equal to zero. In the figure

53
below we look at how the magnitude of impulse varies for values of e between 0 and

1.Collision happens at an angle of θ = 45 o .The collision problem is dealt with in detail in

Chapter IV. It is seen from the figure that the magnitude of impulse increases with

increase in e. The figure shows that perfectly elastic collisions produce the highest

impulse. A perfectly plastic collision will produce the least impulse.

4
Magnitude of Impulse(P)

3.5
3
2.5
2
1.5
1
0.5
0
0 0.2 0.4 0.6 0.8 1 1.2
Coefficient of restitution(e)

Figure 5.10: e v/s P

5.5 Influence of the angle of impact on the beam velocity.

The velocity at the ends of the beam (boundary conditions) will change with the

change in impulse and also with the change in position of the object which in turn will

change with the angle of approach. Recalling equations (3.42) and (3.43) we have:

v A (t ) − v A (0) = m −1 P(t ) + r AG (t ) × [ω (t ) − ω (0)] (5.3)

v B (t ) − v B (0) = m −1 P(t ) + r BG (t ) × [ω (t ) − ω (0)] (5.4)

54
C hange in m agnitude of velocity
4
3.5
3
2.5

at 'A'
2
1.5
1
0.5
0
15 30 45 60 75 90
Angle of Impact in degrees

Figure 5.11: Angle of impact v/s Change in magnitude of velocity at end ‘A’
Change in magnitude of velocity at

3.5
3
2.5
2
'B'

1.5
1
0.5
0
15 30 45 60 75 90
Angle of Impact in degrees

Figure 5.12: Angle of impact v/s Change in magnitude of velocity at end ‘B’

Figures 5.5 and 5.6 show the change in magnitude of velocity at the ends of the beam

with the angle of impact. While there is little variation of magnitude of change in velocity

at ‘A’ till θ = 60 o , there is a sudden jump at θ = 60 o .Similarly at end ‘B’ there is a

sudden drop in velocity at θ = 45 o jump at θ = 75 o These sudden changes in velocity

could have an influence on the stress, strain and deflection of the beam. This is

explained in the following section.

55
5.6 Influence of the angle of impact on the deflection, stress and strain

Figures 5.13-5.15 illustrate the continuous variation of the deflection, stress and

strain with the angle of impact. Also the influence of the sudden changes in the

magnitude of velocities at the ends of the beam is investigated. It is seen from the figures

that while there is continuous variation in the deflection, stress, strain with the angle of

impact, there are sudden significant jumps. These could be attributed to the sudden

changes in velocities at the ends of the beam.

0.14
0.12
Deflection in meters

0.1
0.08
0.06
0.04
0.02
0
15 30 45 60 75 90
Angle of Impact in degrees

Figure 5.13: Angle of impact v/s Deflection

20000
Stress in Mpa

15000

10000

5000

0
15 30 45 60 75 90
Angle of impact in degrees

Figure 5.14: Angle of impact v/s Stress

56
0.08
0.07
0.06
0.05

Strain
0.04
0.03
0.02
0.01
0
15 30 45 60 75 90
Angle of Impact in degrees

Figure 5.15: Angle of Impact v/s Strain

5.7 Influence of location of the beam on the boundary conditions

In this section we look at the change in velocity at the ends of the beam with the

change in location of the beam on the rigid frame. The following figures show the

different positions at which the beam is attached to the frame.

B
B Q
Q Q
B
A
A R R
R
A G G G
P P
P

S S S

(i) (ii) (iii)

Figure 5.16: Frame with beam attached at different locations

Let the frame ‘PQRS’ impact the ground at θ = 45 o . ‘S’ is the point of contact. The

following graphs show the change in velocity at ‘A’ and ‘B’ as their positions change

from (i) to (iii).

57
Change in magnitude of velocity
2.5

1.5

at 'A'
1

0.5

0
(i) (ii) (iii)
Location of beam

Figure 5.17: Location of beam v/s Change in magnitude of velocity at ‘A’


Magnitude of change in velocity at

2.5

1.5
'B'

0.5

0
(i) (ii) (iii)
Location of beam

Figure 5.18: Location of beam v/s Change in magnitude of velocity at ‘B’

It is seen from the above figures that the velocities decrease as the location of the

beam changes from figure (i) to (iii) with the impact configuration shown in (i) having

the highest velocity and, (iii) the lowest velocity. In (i) the points ‘A’ and ‘B’ are closest

to the impact point ‘S’. As a result the velocities are the highest for this configuration and

as they move away from the point of contact their velocities decrease. Thus it can be

concluded that the velocities upon impact depend upon the proximity of the point under

consideration to the point of impact.

58
5.8. Change in stress and strain with change in position of beam

It is seen from preceding sections that a sudden change in velocity has an influence

on the response of the beam. In figures 5.11 and 5.12 there is a sudden drop in velocity

from (i) to (ii).As a result of this there is a sudden change or drop in the values of stress

and strain from (i) to (ii).This is shown in figures 5.13 and 5.14.

12000
10000
Stress in Mpa

8000
6000
4000
2000
0
(i) (ii) (iii)
Location of beam

Figure 5.19: Stress v/s Location of beam

0.05

0.04

0.03
Strain

0.02

0.01

0
(i) (ii) (iii)
Location of beam

Figure 5.20: Strain v/s Location of beam

59
5.9 Conclusion

A simplified differential model of a flexible beam attached to a rigid supporting

block has been developed and its response to an impulse studied. We have used Routh’s

graphical method for two dimensional collisions for an accurate description of the

impulse developed upon impact. The impulse developed for different angles of approach

is calculated. It can be concluded that the impulse produced at the contact point is

independent of the angle at which it strikes the surface. It could depend more on the

material properties of the impacting material such as friction and co- efficient of

restitution (e). For lower values of µ the impact process terminates before the object

reaches the line of stick. For µ varying between 0.8 and 1.4 the line of friction meets the

line of stick before termination, because of this Pt remains constant. For values of µ

greater that 1.4 the line of friction meets the line of stick before the line of maximum

compression. And µ increases the object will begin to stick immediately after impact

and will continue to stick till termination is reached. We also find upon varying e that a

perfectly elastic collision will produce the highest impulse and a perfectly plastic

collision produces the least impulse. We have studied the change in boundary conditions

with change in the angle of impact. It can be concluded that the deflection, stress and

strain developed in the beam depends on the boundary conditions. There is a sudden

jump or drop in these values with corresponding jumps or drops in the velocities. We

have tried to find the influence of the location of the beam on the frame on the velocities.

There is a definite increase in the change in magnitude of velocities for the beam located

at the ends of the block. The effect of the impact force on the beam is more when it is

closest to the impact point. As the beam moves further away from the point of impact the
60
velocities also decrease. The maximum stresses and strains are developed with the beam

located at the corners of the rigid block. In our case considering the frame to be the

housing of an electronic component and the beam to be an internal component it can be

said that the internal components are under maximum stress when placed at the corners of

the housing. This could serve as a pointer in placing the smaller more flexible

components in rigid housings of cellular phones or laptops. Strategic placing of these

components could increase their time to failure and improve performance. It could serve

as a guide for production units before the actual product is manufactured eliminating the

need for a trial and error method in determining failure thereby saving production costs.

In general it can be concluded that there can be several geometric as well as material

parameters or variables that can affect the damage caused to the component upon impact

and knowing some of the critical factors could lead to a better design.

5.10 Underlying simplifications

Impact is an important event in a lot of dynamic mechanical applications and can be

described in variety of ways. It could be treated as a continuum that would make the

process too complex, time consuming and difficult to analyze. However the current topic

simplifies the impact event by a simple differential collision model with the help of

ordinary differential equations. This model predicts the post collision change in velocity

given the pre collision velocity and impact. Using a simple but accurate description of the

impact force helps us understand the influence of the geometric and material parameters

that could affect the post collision state of the component under consideration.

61
5.11 Recommendations for future work.

The results of the current model could be compared against the results got by a simulated

finite element model or against actual physical tests in the laboratory. A different

approach to finding the solution to the beam equation could be attempted and the results

compared. Also a different model of the impact force could give a more accurate

description of the impact event. The current study focuses in single impacts and does not

consider clattering. A new model considering the effects of clattering could be attempted.

Finally the model described could be applied to real world applications and the results

could be verified.

62
REFERENCES

(1) V.B.C.Tan, M.X.Tong, Kian Meng Lim, and Chwee Tech Lim, “Finite element
modeling of electronic packages subjected to drop impact”, IEEE transactions on
Components and Packaging Technologies, Vol.18, No.3, September 2005

(2) C.T.Lim, Y.M.Teo and V.P.W.Shim, “Numerical simulation of the drop impact
response of a portable electronic product”, IEEE transactions on Components and
Packaging Technologies, Vol.25, No.3, pp.478-485, September 2002.

(3) Raymon Ju, Brain Hsiao, “Drop simulation for portable electronic products”, 8th
International LS-DYNA Users Conference, May 2-4, 2004

(4) K.H.Low, Aiqiang Yang, KH Hoon, Xinwei Zhang, Judy K.T.Lim, K.L Lim, “Initial
study of the drop impact behavior of mini Hi-Fi audio products”, Advances in
Engineering Software,Vol 32,Issue 9,September 2001

(5) Y.Y.Wang, C.Lu, J.Li, X.M.Tan, Y.C.Tse, “Simulation of drop/impact reliability for
electronic devices”, Finite Elements in Analysis and Design, Vol 41, Issue 6, March 2005

(6) Low.K.H, Wang Yuqi, Hoon.K.H, Wai.W.K, “A virtual boundary model for a quick
drop impact analysis of electronic components in TV model”, Advances in Engineering
Software, Vol 35, Issue 8-9, Aug-Sept 2004

(7)Hua Shan, Jianzhong Su, Florin Badiu, Jiansen Zhu, Leon Xu, “Modeling and
simulation of multiple impacts of falling rigid bodies”, Mathematical and Computer
Modeling,Vol 43,issue 5-6,March 2006, pp 592-611

(8) J.B.Keller, “Impact with friction”, Journal of Applied Mechanics, Vol 53, March
1986.

(9) Y.U. Wang, Matthew.T.Mason, “Two-Dimensional Rigid-Body collisions with


friction”, Journal of Applied Mechanics, Vol 59, September 1992

(10) V.Bhatt,Jeff Koechling, “Three dimensional rigid body impact”, Journal of Applied
Mechanics,Vol 62,December 1995.

(11) Horaitu Barbulescu, Dan.B.Marghitu, Uday Vaidya, “Spatial impact of a slender


beam”, Journal of Engineering Materials and Technology, Vol-125, Issue 4, pp 368-371
Oct 2003.

63
(12) Mohammad I. Younis, Fadi Alsaleem, Daniel Jordy, “The response of clamped-
clamped microbeams under mechanical shock”, International Journal of Non-linear
Mechanics,Vol 42,Issue 4,pp 643-657,May 2007.

(13) Mohammad I. Younis, Eihab M. Abdel-Rahman, Ali Nayfeh, “A reduced order


model for electrically actuated microbeam based MEMS”, Journal of
Microelectromechanical Systems, Vol 12, No 5, Oct 2003.

(14) E.Suhir, “Could shock tests adequately mimic drop test conditions”, Journal of
Electronic Packaging,Vol 124,Issue 3,pp 170-177,Sept 2002.

(15) S. Goyal, S. Upasani, and D. M. Patel, “Improving impact tolerance of portable


electronic products: Case study of cellular phones”, Journal of Dynamic Systems,
Measurement and Control,Vol 120,Issue 1 pp 83-93,March 1998.

(16) S. Goyal, S. Upasani, and D. M. Patel, “Improving impact tolerance of portable


electronic products: Case study of cellular phones,” Experimental Mechanics,Vol 30,
Issue 1, pp. 43-52,March 1999.

(17) Ming Xiang, R.Eschke, “Modeling of the effects of continual shock loads in the
transport process”,Packaging Technology and Science,Vol 17,Issue 1,pp 31-35,March
2004.

(18) Brach.R. M. 1989, “Rigid Body Collisions”, ASME Journal of Applied Mechanics,
Vol. 56, pp. 133-138.

(19) Abhishek Dubey, “Tangential restitution in three dimensional collisions”, M.S


thesis, December 2004

(20) Kelly.S.G, “Fundamentals of mechanical vibrations”, McGraw Hill publications,


Boston, MA, 2000.

(21) Chatterjee.A, Ruina.A, “Two interpretations of rigidity in rigid body collisions”,


Journal of Applied Mechanics, Vol 65, No 4, pp 894-900, December1998.

64

You might also like