Thesis B.Tech

INDIAN INSTITUTE OF TECHNOLOGY, BHUBANESWAR
Computational and Experimental

Investigation of ground Resonance
by
Akshay Ramachandran
A thesis submitted in partial fulfillment for the

degree of Bachelor of Technology
Dr. Satyanarayan Panigrahi Dr. Ranjith Mohan
School of Mechanical Sciences Department of Aerospace Engineering
Indian Institute of Technology Indian Institute of Technology
Bhubaneswar Madras
May 2017
Declaration of Authorship
I, Akshay Ramachandran, declare that this thesis titled, ‘Computational and Experimental In-
vestigation of Ground Resonance’ and the work presented in it are my own. I confirm that:
This work was done wholly or mainly while in candidature for a research degree at this
University.
Where any part of this thesis has previously been submitted for a degree or any other
qualification at this University or any other institution, this has been clearly stated.
Where I have consulted the published work of others, this is always clearly attributed.
Where I have quoted from the work of others, the source is always given. With the exception
of such quotations, this thesis is entirely my own work.
I have acknowledged all main sources of help.
Where the thesis is based on work done by myself jointly with others, I have made clear
exactly what was done by others and what I have contributed myself.
Signed:
Date:
i
Certificate
This is to certify that the thesis entitled “Computational and Experimental Investigation of
Ground Resonance” submitted by Akshay Ramachandran to Indian Institute of Technology,
Bhubaneswar is a record of bona fide research work done under our supervision and we consid-
ered worthy for the award of degree of Bachelor of Technology (Mechanical) of the Institute.
Date:
Dr. Satyanarayan Panigrahi

Associate Professor
Indian Institute of Technology
Bhubaneswar
ii
“Education is what remains after one has forgotten what one has learned in school”
-Albert Einstein
INDIAN INSTITUTE OF TECHNOLOGY, BHUBANESWAR
Abstract
Dr. SATYANARAYAN PANIGRAHI
Dr. RANJITH MOHAN
SCHOOL OF MECHANICAL SCIENCES
Batchelor of Technology Thesis
by Akshay Ramachandran
Helicopter aerodynamic instability is a complex problem and has started receiving attention in
1940s. Ground resonance and air resonance are two examples of aeromechanical instabilities.
Ground resonance happens when the helicopter is in ground whereas air resonance happen when
it is in hover. Ground resonance has been analytically shown to be a purely mechanical instability
wherein the rotor mode frequency coalesce with a fuselage frequency. This project focuses on
analysis of aeromechanical instabilities, specifically ground resonance. First, a simplified model of
coupled rotor-fuselage system with translational fuselage degree of freedom is considered. Later
a similar analysis is extended to a model with fuselage pitch, roll and flap, lag degrees of freedom
and incorporates uniform flow model. An experimental model is designed for ground resonance.
Finite element analysis is performed to check whether the desired translational frequencies are
obtained or not. A smaller helicopter model is fabricated which represents the original model.
Preliminary experiments are done to validate the computational model.
Acknowledgements
I want to express my heartfelt thanks to everybody who have provided support, motivation
and guidance during my undergraduate training at Indian Institute of Technology, Bhubaneswar
without which I would never achieve the desired results of my project.
I would like to express my deep and sincere gratitude to my advisors Dr. Satyanarayan Panigrahi
of School of Mechanical Sciences, IIT Bhubaneswar and Dr. Ranjith Mohan of Department of
Aerospace Engineering, IIT Madras for their endless encouragement and supervision through-
out my research. I would also like to thank Salini Nair and V Ramanujam for their guidance
throughout the project.
A special thanks to V Lokeshwar Rao for his help along the design process of my experimental
setup. I would also like to thank the technical engineering support team of RWRDC, HAL
Bangalore who contributed for the development and accomplishment of my experimental work.
I would like to thank all my family and friends for their constant support and have provided
inspirations throughout the period of my undergraduate stay.
v
Contents
Declaration of Authorship i
Certificate ii
Abstract iv
Acknowledgements v
List of Figures viii
List of Tables ix
Symbols x
1 Introduction 1
1.1 General Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
1.2 Effect of Vibration on Helicopters . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
1.2.1 Ground Resonance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
1.2.2 Cause of Ground Resonance . . . . . . . . . . . . . . . . . . . . . . . . . . 2
1.2.3 Overview of accidents registered due to Ground Resonance . . . . . . . . 2
1.3 Literature Review . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
1.4 Outline of Thesis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
1.4.1 Chapter 2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
1.4.2 Chapter 3 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
1.4.3 Chapter 4 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
1.4.4 Chapter 5 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
1.4.5 Chapter 6 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
2 Formulation using Newtonian Approach 6

2.1 Background Theory . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
2.2 Multiblade Rotor System . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
2.3 Chassis-Rotor System . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
3 Solution Algorithm 15
3.1 Multi-Blade Coordinate Transformation . . . . . . . . . . . . . . . . . . . . . . . 15
3.2 Periodic Coefficient System Analysis . . . . . . . . . . . . . . . . . . . . . . . . . 18
3.2.1 Floquet Theory . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
3.3 Discussion of Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
vi
Contents vii
4 Langrangian Approach 23
4.1 Lagrange Equation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
4.2 Energy terms computation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24
4.3 Coordinate system and transformations . . . . . . . . . . . . . . . . . . . . . . . 25
4.4 Aerodynamic Moment Computation . . . . . . . . . . . . . . . . . . . . . . . . . 27
4.5 Algorithm to Solve the equation . . . . . . . . . . . . . . . . . . . . . . . . . . . 29
4.5.1 Simplification process . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29
4.5.2 Linearization process . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30
4.5.3 Transformation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30
4.5.4 Stability Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30
4.6 Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30
5 Experimental Setup 32
5.1 Design of experimental setup . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33
5.1.1 Mechanical Design of experimental setup . . . . . . . . . . . . . . . . . . 33
5.1.2 Global Assembly . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34
5.2 Equivalent Spring Stiffness Identification . . . . . . . . . . . . . . . . . . . . . . . 34
5.2.1 Analytic Spring Stiffness Determination . . . . . . . . . . . . . . . . . . . 35
5.3 Finite Elemental Analysis of Experimental Setup . . . . . . . . . . . . . . . . . . 38
5.3.1 Modal Analysis of Fuselage . . . . . . . . . . . . . . . . . . . . . . . . . . 38
5.3.2 Modal Analysis of Blade . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40
5.3.3 Calculating stiffness input parameters for the modes . . . . . . . . . . . . 41
6 Fabricated Setup 42
6.1 Obtaining input parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43
6.2 Results and Discussions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45
7 Conclusion 46
Bibliography 47
List of Figures
2.1 Change in center of gravity due to lead lag hinge [1] . . . . . . . . . . . . . . . . 6

2.2 Relation between µb/a and ω 2 k/M . . . . . . . . . . . . . . . . . . . . . . . . . . 8
2.3 Blade displacement in lagging motion . . . . . . . . . . . . . . . . . . . . . . . . 9
2.4 Uncoupled chassis and rotor frequencies articulated rotor, no hinge restraint [2] 10
2.5 Uncoupled chassis and rotor frequencies rotor with drag hinge spring or elastic
element[2] . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
2.6 Mathematical Representation of rotor and hub [3] . . . . . . . . . . . . . . . . . . 11
3.1 Modal damping for the configuration given in Tabel 3.1 . . . . . . . . . . . . . . 21

3.2 Modal frequency for the configuration given in Table 3.1 . . . . . . . . . . . . . . 21
3.3 Modal frequency for the configuration using Floquet technique . . . . . . . . . . 21
4.1 Physical system used for analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . 26

4.2 Imaginary part of Eigenvalues . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31
4.3 Real part of Eigenvalues with Bousman reading . . . . . . . . . . . . . . . . . . . 31
5.1 Isometric view of Global Assembly . . . . . . . . . . . . . . . . . . . . . . . . . . 34

5.2 Modified Swashplate Assembly . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34
5.3 General Representation of Fuselage Lamina . . . . . . . . . . . . . . . . . . . . . 35
5.4 General Representation of Blade Lamina . . . . . . . . . . . . . . . . . . . . . . . 37
5.5 Finite Element Approximation of Fuselage . . . . . . . . . . . . . . . . . . . . . . 38
5.6 Finite Element Model of Fuselage . . . . . . . . . . . . . . . . . . . . . . . . . . . 39
5.7 Modal Analysis of Fuselage . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39
5.8 Finite Element Approximation of Blade . . . . . . . . . . . . . . . . . . . . . . . 40
5.9 Modal Analysis of Blade . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40
6.1 Isometric View of Fabricated Setup . . . . . . . . . . . . . . . . . . . . . . . . . . 42

6.2 Views of Fuselage . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42
6.3 Top View of Rotor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43
6.4 Side View of Rotor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43
6.5 Isometric View of Rotor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43
6.6 Free decay Response for lead lag motion . . . . . . . . . . . . . . . . . . . . . . . 44
6.7 Curve fitted for free decay response . . . . . . . . . . . . . . . . . . . . . . . . . . 44
6.8 FFT analysis for free decay response . . . . . . . . . . . . . . . . . . . . . . . . . 44
6.9 Results obtained from complex-model code . . . . . . . . . . . . . . . . . . . . . 45
viii
List of Tables
3.1 Parameters Used in the Sample Calculations . . . . . . . . . . . . . . . . . . . . 20
4.1 Experimental Configuration for Bousman’s Model . . . . . . . . . . . . . . . . . . 30
5.1 Physical Data for the experimental setup . . . . . . . . . . . . . . . . . . . . . . 33

5.2 Spring stiffness derived from analytic formula . . . . . . . . . . . . . . . . . . . . 36
5.3 Spring stiffness for blade derived from analytic formula . . . . . . . . . . . . . . . 37
5.4 Real Eigenvalues (Obtained from NASTRAN) . . . . . . . . . . . . . . . . . . . . 39
5.5 Frequencis for fuselage derived from Modal Analysis . . . . . . . . . . . . . . . . 39
5.6 Blade Analysis:Real Eigenvalues (Obtained from NASTRAN) . . . . . . . . . . . 40
5.7 Spring stiffness derived from Modal Analysis . . . . . . . . . . . . . . . . . . . . 41
5.8 Spring stiffness for blade derived from modal analysis . . . . . . . . . . . . . . . 41
6.1 Frequencis obtained using free response . . . . . . . . . . . . . . . . . . . . . . . 45
ix
Symbols
Ω Rotor rotational speed Hz

T Time period seconds
Nb Number of blades
ω Modal Frequency Hz
xh Longitudinal displacement of hub m
yh Lateral displacement of hub m
th
ζi Lead-lag displacement of i blade radian
βi Flap displacement of ith blade radian
θ Fuselage pitch angle radian
φ Fuselage roll angle radian
ωx , ωy Fuselage uncoupled frequency in the x and y direction Hz
ωζ Lead-lag uncoupled frequency of ith blade at rest Hz
ψi Blade azimuth angle radian
Sb First mass moment of blade kg-m
R Radius of blade m
e Hinge offset of blade m
x
Dedicated to my loving parents and my brother...
xi
Chapter 1
Introduction
The helicopter is a fine way to travel, but

it induces a view of the world that only
God and CEOs share on a regular basis.
Morley Safer
1.1 General Introduction
Aviation sector has been an essential tool that helps entrepreneurial companies to expand their
business. Private helicopters and jet form a major market share in this sector. Among them,
there are many benefits of the former for it has been the favorite of the corporate executives
since helicopter provides a fast, private and point-to-point travel. Moreover, it avoids the whole
process of checking and eventual delays in airport. Due to such a large fleet of fielded helicopters,
there has been a variety of “lessons learned” and safety over the years in the early 1940s
1.2 Effect of Vibration on Helicopters
Vibrations in helicopters are caused by a variety of sources, such as rotor systems, engine,
transmission etc. Vibratory loads lead to a fatigue damage of structural components, human
discomfort, difficulty in reading instruments and reduced effectiveness of weapon systems. With
increasing demand for high speed and high maneuverability for both civilian and military appli-
cations, vibratory analysis of helicopters plays a significant role in helicopter design.
1.2.1 Ground Resonance
The ground resonance phenomenon consists in a potential instability that occurs in the vicinity
of a frequency coalescence between a rotor cyclic mode and a fuselage mode while helicopter is
on the ground over its landing gear.
1
Experimental and Computational Investigation of Ground Resonance 2
1.2.2 Cause of Ground Resonance
It is imperative that we understand the cause of ground resonance. For that, we are required to
know about the lead-lag motion of the blade. Unlike popular notion, helicopter blades are not
rigidly fixed to the hub, rather a degree of freedom is provided in both in-plane and out-plane
direction. It is done in order to remove the bending stress occurring at the root. There exists a
bending stress as shown below.
If the blades lag in phase, the center of gravity of the rotor system remains on the axis of rotation,
but out-of-phase oscillatory motion of the blades will cause the rotor system center of gravity to
move off the axis of rotation and describe a circle, giving rise to inertia forces which will subject
the fuselage and its chassis to an oscillatory force. If this oscillation frequency matches with
frequency of fuselage, we see phenomenon of ground resonance.
1.2.3 Overview of accidents registered due to Ground Resonance
The ground resonance phenomenon in the helicopter was listed as Miscellaneous Loss of Control
(LOC-MISC) events in a report published by US-JHAT in 2006. Out of the total 59 LOC
accidents registered, 11 were labeled as miscellaneous of which two were of the type Ground
Resonance (GR).
If we closely examine the two instances, the first one transpired when braking while ground-
taxing a wheel-equipped twin turbine aircraft during steady windy winds. Fortunately, this was
stopped by turning into a full blown disaster due to instruction on emergency procedures (which
also included simulator training). With the second one, it occurred during ground operations in
a skid-equipped aircraft.Low pilot time in aircraft (29 hours) and minimal transition training in
the aircraft were identified as problem statements in this accident.
1.3 Literature Review
During the ground resonance, the cyclic mode, excited by the fuselage oscillations creates a
wobble of the rotor center of gravity. The shifting of the effective center of mass of the rotor
couples with the vibration of the fuselage. The continuity of this closed loop process leads to
large displacements and may provoke the total destruction of the helicopter[5];[6].
It is important to remark that Ground Resonance is basically a purely mechanical instability
which can be determined without taking into account the aerodynamical effects. However mainly
in soft in-plane hinge-less rotors, the stability augmentation system can influence the helicopter
dynamics on the ground. By considering the aerodynamic forces, the unstable ground and air
resonance phenomenons are known as aeromechanical instabilities. In this term it is also included
also the air resonance, characterized by the same phenomenon but with helicopter airborne.
The most common ways to overcome the instability are adding mechanism to increase the damp-
ing of the rotor lag mode or the fuselage mode. Traditionally, helicopters have been equipped
with passive elastomeric or hydraulic lag dampers at the rotor hub. Some issues have been
detected with these types of dampers, such as maintanence requirements and hub complexity.
Already applied in automobile suspensions, space structures, the use of semi active techniques
has notably interested aeronautical companies. By combining the positive aspects of passive and
active control devices, the development of semi active devices are in development to suppress
helicopter vibrations.
The earliest research into the ground resonance phenomenon was performed by Coleman[7]. The
periodical equations of motions of the coupled rotor-fuselage system are quite complex, even
for the simplified developed model, and the fundamental theory of air and ground resonance is
reported.
Later a more complete investigation for polar and apolar symmetry is done by Coleman and
Feingold. The aerodynamic efforts are not taken into account, the blades are articulated and
they present identical geometric and mechanical properties (referred as isotropic rotor). This
assumption reduces the complexity on treating the equations of motion and their periodical
characteristics disappears since represented on the fixed frame. The method has been explored
recently on the dynamic analysis of wind turbines[8].
Other early analysis are reported by Hooper and Gabel on ground and air resonance. By con-
sidering aerodynamic effects, Donham et al. reformulated the equations of motion and predicted
the unstable oscillations for ground and air resonance in helicopters. Still considering the aero-
dynamic efforts, Lytwyn et al analyzed the air resonance on hinge-less helicopters. Other contri-
butions towards understanding the ground resonance phenomenon in hinge-less and bearing-less
rotors are made by army researchers, such as Hodges[9].
The investigations mentioned above verified that the occurrence of ground resonance can be
accurately predicted for articulated, hinge-less and bearing-less isotropic rotors. The use of lin-
earized equations of motion provides very accurate frequency prediction.
Hammond[3] uses Floquet analysis for the prediction of mechanical instabilities examining the
case of dissimilar lead-lag damping. Wand and Chopra[10] studied unbalanced rotor with single
dissimilar blade. The results show a shift of boundary speeds of unstable region once a slight
asymmetry of 5 percent is introduced on the lag blade stiffness in the hinged rotors.
The effects of geometric nonlinearities with the use of nonlinear springs or dampers on the pre-
diction of instabilities are verified by Robinson et al.[11][12], Tongue et al. and Kunz. Most
of these works take into account nonlinear effects and treat the equations through the use of
perturbation methods.
1.4 Outline of Thesis
The thesis is organized into seven chapters with each one addressing a specific aspect of the
problem.
1.4.1 Chapter 2
This chapter aims to introducing the fundamental theory used and the fundamental concepts
reached by the earliest researches on the study of helicopter ground resonance. The mechanical
model of the helicopter adopted is illustrated and is termed the “simple model”. It is inspired
on that used by Coleman and the assumptions considered along the modeling process are pre-
sented.It presents the structural without aerodynamics. The equations of motion for the rotor
blade as well as the fuselage are developed using Newton’s principle.
1.4.2 Chapter 3
Chapter 3 describes the solution methodology for solving the obtained equations. In this chap-
ter, we present some mathematics that are useful in the analysis of periodic dynamic systems,
specifically an N bladed rotor rotating at a speed ω. We start our discussion with Coleman’s
method and transform all the periodic terms by appropriate transformations. Later this was
transformed to a matrix form where the eigenvalues were computed. Then we shifted to advan-
tages of periodic coefficient system analysis over multi-blade coordinate transformation in which
we begin our discussion by providing two solutions to a problem posed by US army’s criterion.
The second method is explained in detail and is termed “Floquet Theory”.
1.4.3 Chapter 4
A similar analysis is extended to a model with fuselage pitch, roll and blade flap, lag degrees of
freedom and incorporates uniform inflow model and such a model is termed “Complex Model”.
The equations of motion for a coupled rotor-fuselage system were derived using the Lagrangian
approach.
1.4.4 Chapter 5
This chapter devotes its attention on the experimental setup designed for further understanding
and reproducing the ground resonance in helicopters. A detailed explanation of the experimental
mechanical design are made. Its main components, systems and subsystems, mechanical and
geometrical properties and functioning are described. Two different methods are presented for
determining the equivalent fuselage and blade spring stiffness and the results are discussed.
1.4.5 Chapter 6
In this chapter, the experimental setup that was fabricated was explained. Later, a method was
discussed on how to obtain the physical input parameters to the system. From the physical
parameters obtained, the values were given as input to the code and hence, the unstable regions
were identified. This setup later was later run in that rpm region and results were discussed.
Chapter 2
Formulation using Newtonian

Approach
2.1 Background Theory
About the year 1940, helicopter rotors with the usual hinged-blade construction were observed
to come to a violent critical condition at a speed very much lower than that calculated from
ω 2 = k/M formula. This happens while the aircraft is standing still on the ground prior to
take-off and consequently called the ”ground critical”. The phenomenon was explained by R.P.
Coleman[7] of Langley field in NACA reports of 1942 and 1943, and the simpler portion of his
results are reproduced here[1]. The system is as shown in Figure 2.1.
Consider the three bladed rotor of Figure 2.1 where O is the bottom of the pylon seen from the
above and T is the top of the pylon, displaced to the right through the distance OT = e, the
eccentricity. The pylon is supposed to be bent elastically through the distance e, and the entire
figure as a solid body rotates or whirls at speed ω about the vertical axis O. The blades will turn
Figure 2.1: Change in center of gravity due to lead lag hinge [1]
6
about their hinge axis H through small angles , so that the blade lines up with the centrifugal
field through the center of rotation O. During the whirling motion, these angles are constant
and no relative motion takes place across any of the hinges H. We now calculate the centrifugal
forces of all three blades and of hub and set their sum equal to ke, the elastic homing force of
the pylon. This will give the critical speed.
Figure 2.2[1]
In the triangle OTH the angle OTH is 120 degree, the angle THO= is considered “small”,
the hinge radius TH=a and the eccentricity OT=e is again small with respect to a. From the
geometry of the triangle, we can arrive as follows:
√
3 e
sin ≈ = . (2.1)
2 a
Thus the centrifugal force of blade 2 is mb ω 2 (a + b + 2e ), directed along GH. This force is now
resolved into components parallel and perpendicular to OT. The component parallel to OT(to
the right) is given by
e
mb ω 2 (a + b + ) cos (60 − ) =
2
e
mb ω 2 (a + b + )(cos 60 + sin 60)
2
e 1 3e
= mb ω 2 (a + b + )( + )
2 2 4a
1 3b
= mb ω 2 [ (a + b) + e(1 + )]
2 4a
For blade 3, the result is same for reasons of symmetry, while the components of centrifugal force
perpendicular to OT for blades 2 and 3 cancel each other. The centrifugal force for blade 1 in
the direction of OT(to the left is) mb ω 2 (a + b − e). The centrifugal force of the hub itself (to the
right) is mh ubω 2 e. Thus the total centrifugal force to the right is
3 b
ω 2 e[mhub + mb (3 + . )]
2 a
Let the total mass be

mhub + 3mb = M
and the ratio of hinged mass to total mass be
3mb
µ=
M
µb
Then the total centrifugal force can be written as M ω 2 e(1 + 2a )
Equating this to the elastic force ke, and the critical frequency comes out:
k 1
ω2 = . (2.2)
m 1 + µb
2a
k
It is seen that for the case of no hinged mass, µ = 0,the natural frequency is m :the presence
of the hinged mass diminishes the frequency. The relation is shown graphically by fully drawn
curve of Fig 2.2. Although the above analysis was carried out with a three bladed rotor, the
result is also good for a motor with more than three blades.
Figure 2.2: Relation between µb/a and ω 2 k/M
2.2 Multiblade Rotor System
Let us calculate the displacement of the rotor center of gravity for an arbitrary motion of the
blades. Let xkg and ykg be the coordinates of the center of gravity of the kth blade relative to
the center of the hub, Fig 2.3. This figure can represent either a hinged blade or a hinge-less
blade. The derivation for this has been taken from [2]. We easily see from Fig 2.4 that
Figure 2.3: Blade displacement in lagging motion
xkg = −eR cos(ψk ) − rg cos(ψk + k )

ykg = eR sin(ψk ) + rg sin(ψk + k )
Since k is a small angle, these relationships can be written approximately as
xkg = −(eR + rg ) cos(ψk ) + rg k sin(ψk ) (2.3)

ykg = (eR + rg ) sin(ψk ) + rg k cos(ψk ) (2.4)
Summing over the b blades of the rotor the coordinates of the center of gravity of the rotor are
eR + rg X rg X
x rg = − cos(ψk ) + k sin(ψk )
b b
or r X
g
x rg = k sin(ψk ) (2.5)
b
and similarly r X
g
y rg = k cos(ψk )
b
Now, suppose that the blade oscillate in lagging motion with frequency kω such that
k = o cos(kψk )
Then substituting it in eqn 2.5
rg o X rg o X
x rg = ( cos(kψk ) sin(ψk ) = ( [sin(k + 1)ψk − sin(k − 1)ψk ]
b 2b
Since k is arbitrary we find that
rg o b−1 b−1
x rg = [σ1 sin(k + 1)(ψ + π) − σ−1 sin(k − 1)(ψ + π)] (2.6)
2b b b
sin(k+1)π sin(k−1)π
where σ1 = sin[(k+1)π/b] and σ−1 = sin[(k−1)π/b]
The interpretation is that the center of gravity of the rotor whirls round the hub with a dis-
rg o σ 1
placement consisting of two modes, one with a frequency (k + 1)ω and amplitude 2b and
rg o σ−1
the other with a frequency (k − 1)ω and amplitude 2b . The motion corresponding to the
upper frequency (k + 1)ω is in the same direction as the rotor rotation and is described as a
’progressive’ motion; the motion corresponding to the lower frequency (k − 1)ω is in the same or
opposite direction to the rotor rotation according to when k − 1 is negative or positive. When
negative, the motion is described as ’regressive’.
The whirling center of gravity produces periodic inertia forces which excite motion of the whole
airframe on its undercarriage (’chassis mode’). If one of the frequencies of the oscillating inertia
forces coincides with a chassis frequency, the potential for the occurrence of ground resonance
exists.
This coincidence of frequencies can be represented in diagrammatic form. Fig 2.5, which applies
to an articulated rotor with no lag hinge restraint. The comparable diagram for a rotor with
hinge restraint, or utilizing an elastic hub element is shown in Fig 2.6.
For a blade without hinge restraint, i.e. without a drag-hinge spring or elastic element, the lag
frequency will always be less than the rotor speed and, therefore, k − 1 will always be negative.
In Fig 2.4 and Fig 2.5 the horizontal line represents the (constant) chassis frequency ωc . In Fig
Figure 2.4: Uncoupled chassis and rotor frequencies articulated rotor, no hinge restraint [2]
2.5, the point A represents the blade frequency when the rotor hub is stationary and is assumed
here to be higher than the chassis frequency.
As the rotor speed increases, the branch corresponding to the whirl frequency (k − 1)ω intersects
the chassis line at B and again at C when (k − 1) < 0.
If the chassis frequency were higher than the frequency of the non-rotating blade there would
have been an interaction with the (k + 1) branch. The corresponding intersection for the case of
articulated rotor is point D in Fig 2.6. It can be shown that this intersection and the one corre-
sponding to the point B cannot lead to instability. Ground resonance, if it occurs, is associated
only with point C.
Figure 2.5: Uncoupled chassis and rotor frequencies rotor with drag hinge spring or elastic
element[2]
2.3 Chassis-Rotor System
Now let us derive the equation of motion of the chassis-rotor system. The equations of motion
Figure 2.6: Mathematical Representation of rotor and hub [3]
for the mechanical instability problem would be formulated using an Newtonian approach[3].It
would be assumed that the helicopter on its landing gear can be represented by effective param-
eters applied at the rotor hub. It will be further assumed that only inplane motions of the hub
and blades are important in determining the ground resonance characteristics of the helicopter.
Thus the degrees of freedom to be considered consists of two inplane hub degrees of freedom and
a lead lag degree of freedom for each blade in the rotor. The mathematical model to be used
in this analysis is shown in Figure 2.6. Note that in the figure, a typical blade is shown. The
analysis will be formulated for a rotor having N blades, and each blade is assumed to have a
rotational spring and a damper which act about the lag hinge. Rotor speed is constant.
The blade equations are developed by summing moments about the lag hinge. The coordinates
of the elemental mass dm in the fixed frame are
xi = xh + e cos ψi + ρ cos ψi + ζi (2.7)

yi = yh + e sin sini + ρ sin ψi + ζi (2.8)
where
2π(i − 1)
ψi = ωt + i = 1, 2, 3, ...., N
N
These expressions can be differentiated twice with respect to time to yield the accelerations
experienced by the differential mass
ẍi = x¨h − eω 2 cos ψi − ρ(ω + ζ˙i )2 .cos(ψi + ζi ) − ρζï sin (ψi + ζi )
yï = y¨h − eω 2 sin ψi − ρ(ω + ζ˙i )2 .sin(ψi + ζi ) + ρζï cos (ψi + ζi )
Using D’Alembert’s principle the summation of moments about the lag hinge can be written as
Z
Sb = ρdm
Z
Ib = ρ2 dm
Z Z
ρ sin(ψi + ζi )ẍi dm − ρ cos(ψi + ζi )yï dm − ki ζi − ci ζi = 0
If small displacements are now assumed the blade equation may be linearized to obtain
2
νo
ζï + ηi ζ˙i + (ωoi 2 + Ω2 νo 2 )ζi = ( )[x¨h sin ψi − y¨h cos ψi ] (2.9)
e
where the following parameters have been introduced
eSb
νo2 =
Ib
ki
ωo2i =
Ib
ci
ηi =
Ib
Under the assumptions stated earlier the hub equations of motion can be written as
mx x¨h + cx x˙h + kx xh = Px
my y¨h + cy y˙h + ky yh = Py
where the coefficients on the left hand side of these equations are the effective hub properties
in the x- and y- directions, respectively.The determination of these properties depends on an
extensive knowledge of the helicopter inertial characteristics and the stiffness, damping and
geometrical characteristics of the landing gear system. These properties are determined in detail
in Chapter 4.
The right-hand side of the above equations are the forces acting on the hub due to the fact that
the rotor is experiencing accelerations in the x- and y- directions. If the accelerations of the
rotor center of mass are x¨c and y¨c respectively, the Px and Py are given by
Px = −N mb x¨c
Py = −N mb xÿ
If all the blades in the rotor are assumed to have the same mass distribution, the coordinates
for the total rotor center of mass may be written as
N
1 X
xc = xh + xi
N i=1 c
N
1 X
yc = yh + yi
N i=1 c
where xic and yic are the coordinates of the individual blade center of mass, measured with
respect to the hub. If the center of mass of theith blade is a radial ρc from the lag hinge
xic = e cos ψi + ρc cos(ψi + ζi )
yic = e sin ψi + ρc sin(ψi + ζi )
For N > 1,
N
X N
X
cos ψk = sin ψk = 0
k=1 k=1
the rotor center of mass become

N
ρc X
xc = xh − ( ) ζi sin ψi
N i=1
N
ρc X
yc = yh − ( ) ζi cos ψi
N i=1
These expressions may now be differentiated twice with respect to time and the forces Px and
Py obtained as
N
X
Px = −N mb x¨h + Sb [(ζï − Ω2 ζ) sin ψi + 2Ωζ˙i cos ψi ]
i=1
N
X
Py = −N mb y¨h − Sb [(ζï − Ω2 ζ) cos ψi − 2Ωζ˙i sin ψi ]
i=1
The hub equations thus become
N
X
(mx + N mb )x¨h + cx x˙h + kx xh = Sb [(ζï − Ω2 ζ) sin ψi + 2Ωζ˙i cos ψi ] (2.10)
i=1
N
X
(my + N mb )y¨h + cy y˙h + ky yh = −Sb [(ζï − Ω2 ζ) cos ψi − 2Ωζ˙i sin ψi ] (2.11)
i=1
The equations of motion for the system thus consist of (N + 2) coupled second-order differential
equations with the coupling terms having periodic coefficients. The periodic coefficients arise
because the blade equations are written in a rotating reference system whereas the hub equations
are in a fixed system. It would be later if all the blades have identical lag springs and lag dampers,
the periodic coefficients may be eliminated by transforming the blade equations from the rotating
to the fixed system of reference. The resulting constant coefficient system of equations is the
set normally solved in the classical ground resonance analysis. It should be observed, however,
that if the blades are allowed to have different lag springs and dampers, the periodic coefficients
cannot be eliminated in the usual manner.
Chapter 3
Solution Algorithm
In previous section, it was seen that the equations for the blade is of the form
2
ν
ζ¨k + ηk ζ˙k + (ωo2k + Ω2 νo2 )ζk = ( o )[x¨h sin ψk − y¨h cos ψk ] (3.1)
e
The equation for fuselage is
N
X
(mx + N mb )x¨h + cx x˙h + kx xh = Sb [(ζï − Ω2 ζ) sin ψi + 2Ωζ˙i cos ψi ] (3.2)
i=1
N
X
(my + N mb )y¨h + cy y˙h + ky yh = −Sb [(ζï − Ω2 ζ) cos ψi − 2Ωζ˙i sin ψi ] (3.3)
i=1
In this chapter, we present some mathematics that are useful in the analysis of periodic dynamic
systems, specifically an N bladed rotor rotating at a speed ω. The period for a single blade
is given by T = 2π/ω. For the entire rotor, viewed in the non-rotating frame, the period is
T = 2π/N ω.
3.1 Multi-Blade Coordinate Transformation
For the steady state solution, appropriate representation of the blade motion is a Fourier series,
the harmonics of which describe the motion of the blade motion as a whole. The equation of
motion in the non-rotating frame are simply algebraic equations of harmonics. Now instead of
the steady state behaviour, we are concerned with the general dynamic behaviour, including the
transient response of the rotor. The degrees of freedom that describe the motion of the rotor
in the non-rotating frame are called Multiblade coordinates (MBC). Coleman’s method[7], also
known as “Multi-blade Coordinate Transformation” consists of transforming the equations of
motion by introducing a change of variables which eliminate their periodic characteristic. The
Coleman Transformation is defined as:
15
N N
2 X 2 X
ζs = ζk sin ψk ζc = ζk cos ψk
N N
k=1 k=1
N N
2 X 2 X
ζs = ζk sin ψk ζc = ζk cos ψk
N N
k=1 k=1
where ζc and ζs corresponds to sine and cosine components of a four bladed rotor motor shapes.
These components are called cyclic modes and they correspond to the lateral and longitudinal
shifting of the rotor center of mass. Special names are assigned to the two fundamental coupling
modes between the cosine and sine lag motion. They are called “progressing” and “regressing”
modes, or the high and low frequency modes. ζo and ζd are called collective and differential
component and don’t contribute to ground resonance.
The use of multiblade coordinates is crucial for problems involving the rotor motion coupled with
the fixed frame, such as hub motion, swashplate control, or dynamic inflow. MBC are a physically
relevant, non-rotating frame representation of the rotor motion. Consequently, introduction of
MBC separates the coupling of the rotor and fixed frame into subsets and, most importantly
eliminates periodic coefficients (except for two-bladed rotors).
It can be easily seen that
N
X N
X N
X
sin ψk = cos ψk = sin ψk cos ψk = 0
k=1 k=1 k=1
Also
N
X N
X
cos2 ψk = sin2 ψk = N/2
k1 k1
Here, considering all blades to be symmetric (i.e. isotropic rotors), the stability analysis of
the equations of motions are easily computed since their periodical characteristics is eliminated.
Summing Equation 3.1 over the blades
N N N N N
X X X νo2 X X
ζ¨k + ηk ζ˙k + (ωo2k + Ω2 νo2 ) ζk = ( )[x¨h sin ψk − y¨h cos ψk ]
e
k=1 k=1 k=1 k=1 k=1
Hence
ζö + η ζ˙o + (ωo2 + Ω2 νo2 )ζo = 0
Multiplying the Equation 3.1 by sin ψk and summing over the blades,
N N N N N
X X X νo2 X X
ζ¨k sin ψk +η ζ˙k sin ψk +(ωo2k +Ω2 νo2 ) ζk sin ψk = ( )[x¨h sin2 ψk −y¨h sin ψk cos ψk ]
e
k=1 k=1 k=1 k=1 k=1
Hence
ν 2 x¨h
ζ¨s + η ζ˙s − 2Ωζ˙c + (ωo2 + Ω2 (νo2 − 1))ζs − ηΩζc = o
e
Multiplying the Equation 3.1 by cos ψk and summing over the blades,
N N N N N
X X X νo2 X X
ζ¨k cos ψk +η ζ˙k cos ψk +(ωo2k +Ω2 νo2 ) ζk cos ψk = ( )[x¨h sin ψk cos ψk −y¨h cos2 ψk ]
e
k=1 k=1 k=1 k=1 k=1
Hence
2
ν y¨h
ζ¨c + η ζ˙c + 2Ωζ˙s + (ωo2 + Ω2 (νo2 − 1))ζc + ηΩζs = − o
e
Similarly fuselage equations become
N Sb ζ¨s
(mx + N mb )x¨h + cx x˙h + kx xh =
2
N Sb ζ¨c
(my + N mb )y¨h + cy y˙h + ky yh = −
2
Hence the equations in non-rotating frame are,
ζö + η ζ˙o + (ωo2 + Ω2 νo2 )ζo = 0 (3.4)

νo2 x¨h
ζ¨s + η ζ˙s − 2Ωζ˙c + (ωo2 + Ω2 (νo2 − 1))ζs − ηΩζc = (3.5)
e
2
ν y¨h
ζ¨c + η ζ˙c + 2Ωζ˙s + (ωo + Ω (νo − 1))ζc + ηΩζs = − o
2 2 2
(3.6)
e
N Sb ζ¨s
(mx + N mb )x¨h + cx x˙h + kx xh = (3.7)
2
N Sb ζ¨c
(my + N mb )y¨h + cy y˙h + ky yh = − (3.8)
2
Writing in matrix form, we obtain

    
1 0 0 0 0 η ζ̈o 0 0 0 0 ζ̇o

0 v02     
 1 0 e 0   ζ̈c  0
   η 2Ω 0 0  ζ̇c 
 
v02
−e   ζ̈s  + 0 −2Ω η 0 0   ζ̇s 

0 0 1 0
    
     
N Sb   ÿ  0
1
 2 0 (mx + my ) 0   h  0 0 cy 0    ẏh 
 
0 0 − N2Sb 0 (mx + my ) ẍh 0 0 0 0 cx ẋh
  
ω 2 + Ω2 νo2 0 0 0 0 ζo
 2 2
 

 0 ω +Ω (νo2 − 1) nΩ 0 0   ζc 
 

2 2
+ 0 −nΩ ω +Ω (νo2 − 1) 0 0   ζs  = 0 (3.9)
  
  

 0 0 0 ky 0  yh 
 
0 0 0 0 kx xh
Equation 3.9 can also be written as
[M ][Ẍ] + [C][Ẍ] + [K][X] = 0

where  
1 0 0 0 0

0 v02 
 1 0 e

 0
v02
M = 0 −e

0 1 0


 
1 N Sb
 2 0 (mx + my ) 0 

0 0 − N2Sb 0 (mx + my )
 
η 0 0 0 0
 
0 η 2Ω 0 0 
 
C = 0 −2Ω η 0 0
 
 
0 0 0 cy 0 
 
0 0 0 0 cx
 
ω 2 + Ω2 νo2 0 0 0 0
 

 0 ω 2 + Ω2 (νo2 − 1) nΩ 0 0
K= 0 −nΩ ω 2 + Ω2 (νo2 − 1) 0 0
 
 

 0 0 0 ky 0
0 0 0 0 kx
Hence for us, we need to find the eigenvalue of the matrix A given by
" #
[0] [I]
A= −1 −1 (3.10)
− [M ] [K] − [M ] [C]
Further analysis of the results is given in Section 3.3
3.2 Periodic Coefficient System Analysis
All of the analysis performed before have one assumption in common-all blades are assumed to
have identical properties. This is a reasonable assumption under ordinary circumstances; how-
ever U.S army has a requirement on new helicopters which invalidates this assumption. The
requirement is that the helicopter be free from ground resonance with one blade damper inop-
erative. This one blade damper inoperative requirement has a serious impact on the classical
method of analyzing a helicopter for mechanical instability.
Two methods have been used to circumvent this difficulty. The first involves a physical approx-
imation so that classical analysis becomes applicable. In this approach, all dampers are still
assumed to have identical lag dampers even when one of the dampers is removed, but the value
ci
of each of the damper is reduced by the amount N where N is the number of blades and ci is
the original damper rate. As can be seen, with this approach a system is analyzed which is quite
different from the actual situation of a rotor with no damping on one blade.
The second method which has been used is to reformulate the equations of motion allowing for
differing blade characteristics and to obtain the stability characteristics of the system using a
time history integration of the equations. This second approach has the drawback that inter-
pretation of stability characteristics from time history calculations is often difficult and open to
question. The method will yield correct results, however, provided the equations are integrated
over a sufficiently long time period using a reliable numerical integration procedure.
The equations governing the motion of this system have periodic coefficients. This fact suggests
the use of Floquet theory as the means for determining the stability characteristics of the system.
3.2.1 Floquet Theory
Here the state equation is of the form[13]
Ẋ(t) = A(t)X(t) (3.11)
where A(t) is periodic with period T. The response x(t) at any time t of a periodic system of
the form given above can be expressed as:
x(t) = φ(t) · x(0) (3.12)
where φ(t) represents the state transition matrix and x(0) is the initial condition. Floquet
theorem states that the transition matrix can be written as
φ(t) = P (t) exp Bt,P (0) = I,P (t + T ) = P (t),B-constant Matrix
φ(t) = exp Bt (3.13)
The above equation states that the stability of the system is completely characterized by the
eigenvalues of B. φ(t) is called the Floquet Transition Matrix (FTM) or monodromy matrix.
The eigenvalues of B are called the characteristic exponents(λ) and eigenvalues of FTM, the
characteristics multipliers(ν). From the above equation, eigenvalues of B can be obtained from
the eigenvalues of φ(t) using the relation:
1 1 i Im(ν) 2πn
λ= log [Re(ν)2 + Im(ν)2 ]( + arctan[ ]+i (3.14)
T 2 T Re(ν) T
where n is an integer. The real part of λ gives the modal damping and imaginary part the modal
frequency. For a particular rotor speed, if the real part of any of the λ’s is positive, the system
is unstable.
From the foregoing discussion it can be seen that if either the rotor or the hub is isotropic, the
mechanical stability characteristics of the system may be obtained using conventional techniques.
If both the rotor and hub are non isotropic the equations of motion of the system contain periodic
coefficients and thus the standard eigenvalue techniques cannot be used to determine whether
the system is stable or unstable. It is important to demonstrate that Floquet theory can be used
to analyze this general situation of a non isotropic rotor coupled with a non isotropic hub.
3.3 Discussion of Results
In order to demonstrate the application of Multi-blade transformation and to obtain a general

understanding of Floquet theory, a set of parameters were chosen[3].The parameters in the
mechanical stability analysis were chosen so as to be in the general range of interest for a single
rotor helicopter and were such that the system was stable with all dampers functioning up to
a rotor speed of 400 rpm. The parameter values chosen for the calculations are shown in Table
3.1.
Number of blades, Nb 4
Blade mass, mb 94.9 kg
Blade mass moment, Sb 289.1 kg-m
Blade mass moment of inertia, Ib 1084.7 kg − m2
Lag hinge offset, e 0.3048 m
Lag spring,ki 0.0 N-m/rad
Lag damper,ci 4067.5 N-m-s/rad
Hub mass, mx 8026.6 kg
Hub mass, my 3283.6 kg
Hub spring, kx 1240481.8 N/m
Hub spring, ky 1240481.8 N/m
Hub damper, cx 51078.7 N-s/m
Hub damper, cy 25539.3 N-s/m
Table 3.1: Parameters Used in the Sample Calculations
The parameters presented in Table 3.1 correspond to an isotropic rotor and a non isotropic hub.
In the following discussion results are presented for the case of an isotropic hub coupled with
a non isotropic rotor and a non isotropic hub coupled with an isotropic rotor as well as the
case of interest which involves a non isotropic hub coupled with a non isotropic rotor. When an
isotropic hub is mentioned, this means that the hub parameters in both the x- and y-directions
were assigned the values shown in Table 3.1 for the x-direction.
Figure 3.1 and 3.2 shows the results of the calculations for the isotropic hub with all blade
dampers working. The real parts of the eigenvalues, labeled modal damping, are plotted in the
upper portion of the figure and the imaginary parts of the eigenvalues,labeled modal frequency,
are plotted in the lower portion of the figure.Having examined the case of one blade damper
inoperative and an isotropic hub, the next logical step is to examine the more realistic situation
of a non isotropic hub. Before examining the one damper inoperative situation it was first
desired to confirm that the system was stable with all dampers working. The modal damping
and frequency of the various modes with all dampers working and a non isotropic hub are shown
in Figure 3.1 and 3.2 . As can be seen from the damping plot, all the modes are stable. In this
case the equations of motion are solved in the fixed frame of reference and hence the frequencies
are plotted in this frame. The dashed lines on the frequency plot represent the uncoupled system:
the horizontal dashed lines being the hub modes and the slanted dashed lines being the rotor
modes. Note that because the rotor modes become critically damped at low rotor speeds the
two uncoupled rotor frequencies come together before reaching the origin.
Figure 3.1: Modal damping for the configuration given in Tabel 3.1
Figure 3.2: Modal frequency for the configuration given in Table 3.1
Figure 3.3: Modal frequency for the configuration using Floquet technique
The uncoupled rotor lines also represent the collective modes for the rotor. These modes are
completely uncoupled from the other modes and hence are not included in the eigenvalue analysis
of the non isotropic hub coupled with an isotropic rotor.
The validity of the Floquet analysis was verified by comparing results from this analysis with
results from both the rotating system analysis (isotropic hub) and from the fixed system analysis
(isotropic rotor). In each case the results from the Floquet analysis were identical to results from
the other analyses.
Chapter 4
Langrangian Approach
The methods which I present here do not

require either constructions or reasoning
of geometrical or mechanical nature, but
only algebraic operations proceeding
after a regular and uniform plan.
Lagrange
The equations of motion for a coupled rotor-fuselage system would be derived using the La-
grangian approach. This study is concerned with two models, one being characterized as simple
and the other as complex. The simple model was explained in detail in Chapter 2 where the equa-
tions were formulated using Newtonian approach and Chapter 3 where equations were formulated
equations were solved using MBC and Floquet theory. In complex model, the Lagrangian ap-
proach was selected as opposed to the Newtonian because it is easily implemented with the aid of
a symbolic processor (Mathematica and Matlab).The equations for both models considered were
formulated with Lagrange’s method in their full nonlinear forms, i.e., no ordering scheme, small
angle assumptions, or linearization techniques were applied during derivation and subsequent
coding.
4.1 Lagrange Equation
The Langrange equation can be expressed as follows in Eq 4.1:

d ∂T ∂T ∂U ∂D
− + + = Fi (4.1)
dt ∂ q˙i ∂ q˙i ∂qi ∂ q˙i
where T is the kinetic energy, U is the potential energy, D is the dissipation function, Fi is
a generalized force, and qi , is a generalized displacement. The generalized force term, Fi will
23
describe the aerodynamic forces on the individual rotor blades and will be discussed in a later
section, as a result, this derivation develops only the system of homogeneous equations.
4.2 Energy terms computation
Degrees of Freedom Considered

This model assumes rigid blades and fuselage. The degrees of freedom we have considered in our
analysis are:
• Fuselage degrees of freedom

-Pitch
-Roll
• Rotor degrees of freedom

-Lag motion of each blade
-Flap motion of each blade
Hence the total number of degrees of freedom are 2 + 2 × Nb

Ordering
The roll motion is followed by pitch motion of the fuselage. The flap flexure is assumed to be
outboard of lead lag flexure, implying that lag motion is followed by flap motion.
The various energy terms can be broken down into two categories, terms due to blade motion
and terms due to fuselage motion, to give the following equations:
N
X
T = Tf uselage + (Tblade )k
k=1
N
X
U = Uf uselage + (Ublade )k
k=1
N
X
D = Df uselage + (Dblade )k
k=1
Fuselage Energy
Hence we obtain fuselage energy as
2 2
1 dθ(t) 1 dφ(t)
Tf uselage = Ip + Ir (4.2)
2 dt 2 dt
1 1
Uf uselage = Kp θ(t)2 + Kr φ(t)2 (4.3)
2 2
2 2
1 dθ(t) 1 dφ(t)
Df uselage = Cp + Cr (4.4)
2 dt 2 dt
Rotor Energy
Similarly the rotor energy is given as
Nb
X 1 1
Tblade = mbi vi .vi + ωi . [I] ωi (4.5)
i=1
2 2
Nb
X 1 2 1 2
Ublade = kf lapi βi (t) + klagi ζi (t) (4.6)
i=1
2 2
Nb 2 2
X 1 dβi (t) 1 dζi (t)
Dblade = cf lapi + clagi (4.7)
i=1
2 dt 2 dt
where  
0 0 0
 
I = 0
 Ibf lap 0 

0 0 Iblag
Evaluating Velocity
vi is the velocity of a point on the elastic axis of the ith rotor blade with respect to the inertial
coordinate system at any instant in time.The velocity of a point on the blade is transformed to
the inertial frame using the transformation matrices. It is obtained from the coordinate systems
and transformations.
4.3 Coordinate system and transformations
The following generic transformations are defined

 
1 0 0
 
T 1(α) = 
0 cos (α) − sin (α)
 (4.8)
0 sin (α) cos (α)
 
cos (α) 0 sin (α)
 
T 2(α) = 
 0 
1 0 (4.9)
−sin(α) 0 cos (α)
 
cos (α) − sin (α) 0
 
T 3(α) = 
 sin (α) cos (α) 0
 (4.10)
0 0 1
Where in general, T1, T2, and T3 are rotations about the x, y, and z axes respectively. This
notation can be directly utilized with the symbolic processor and will be used in the following
section where the energy expressions necessary for the Lagrangian derivation are defined. In
developing the equations of motion for the two coupled rotor-fuselage models five coordinate
systems were utilized with transformations between the various systems based on Euler angle
rotations. The five coordinate systems are
• Inertial, fixed relative to earth

• Fuselage, fixed to center of gravity of fuselage
• Hub, parallel to fuselage system but offset a distance h in the positive z direction to the
Earth
• Undeformed blade, fixed to the effective hinge position on the ith blade
The complex model is as shown in figure 4.1

Displacement of a point at a distance r from the hinge on the blade is obtained in the inertial
Figure 4.1: Physical system used for analysis
frame is obtained using the matrices and then the derivative with time is taken to obtain velocity.
We difine transformational matrices with the help of transformation matrices given by 4.8, 4.9
and 4.10.
M1 = T 2 [φ(t)] .T 1 [θ(t)]
M2 = T 2 [φ(t)] .T 1 [θ(t)] .T 3 [−π/2 + ψi (t)]
M3 = T 2 [φ(t)] .T 1 [θ(t)] .T 3 [−π/2 + ψi (t)] .T 3 [ζi (t)]
M4 = T 2 [φ(t)] .T 1 [θ(t)] .T 3 [−π/2 + ψi (t)] .T 3 [ζi (t)] .T 2 [−βi (t)]
With the above equations, we could evaluate vi as

       
0 elag ef lap − elag r
       
= M 1.  0
 
 + M 2.  0  + M 3. 
   0  M 4. 0
  
h 0 0 0
di
vi =
dt
Evaluating ωi
The angular velocity matrix in the inertial frame is obtained using the property
 
0 ω3 −ω2
  d [M3 ] T
 = dt . [M3 ] (4.11)
−ω3 0 ω1 

ω2 ω1 0
The angular velocity vector in the blade frame is:

   
ω ω
 1  1
ω2  = [M3 ]T . ω2 
 (4.12)
  
ω3 ω3
inertial
4.4 Aerodynamic Moment Computation
To compute the aerodynamic moments, the velocity of air incident at a point on the ith blade is
computed.
vaero = vair − vblade
where
   
UR µωR
   
vaero = 
UT 
 vair =
 0 

UP λωR
T
vblade = [M3 ] .vi
U = UR x̂ + UT ŷ + UP ẑ
Negleting the radial component UR ,

q
|U | = UT2 + UR2
UP
φ = tan−1
UT
The lift and drag forces perpendicular and parallel to the direction of net velocity U is,
1 2
L= ρU c × cl
2
1 2
D= ρU c × cd
2
where cl = a(θ − φ)
Resolving the forces in the perpendicular and parallel directions, we get,
dFy = −(L sin (φ) + D cos (φ)) ≈ −(Lφ + D)

dFz = L cos (φ) − D sin (φ) ≈ L − Dφ
Flap and Lag Moments

The lag and flap moments in the deformed frame are given by the equations:
Z R
Mβ = dFz rdr
0
Z R
Mζ = dFy rdr
0
The moments transferred to the undeformed coordinates are obtained as:

    
0 Z R 0
    
= r × dFy  dr
T3 [−β] .  (4.13)
Mβ   
 
0
Mζ dFz
undef ormed
Pitch and Roll Moments

To obtain the pitch and roll moments the aerodynamic moments are transferred to the hub
rst and then transferred to the center of gravity of the fuselage.
• Moment Contribution
   
Mθ 0
   
Mφ 
  = [M4 ] . 
−Mβ 

0 Mζ
moment1 def ormed
• Force Contribution
         
Mθ Z R
e 0 0
        
Mφ  = [T 3(π/2 − ψ(t)] . 0 +  0  × [M4 ] . dFy 
         
0
0 0 h dFz
moment2
• Total Moment
     
Mθ Mθ Mθ
     
Mφ  = Mφ  + (4.14)
Mφ 

   
0 0 0
moment1 moment2
Summation instead of Integration

Since the computation of integration is a time consuming process, we have numerically computed
the value of integral. The procedure we followed is :
• The perturbed and steady state equations of UP and UT are computed.
• The perturbed and steady state equations of |U | and φ are obtained as,
1
|U |pert = (UPs UPP ert + UTs UT )
|U |s
1
φpert = (UTs UPP ert − UPs UTP ert )
|Us |
here q
|U |s = UP2s + UT2s
UPs
φs = tan−1
UTs
• The perturbed equations of Fy and Fz are computed. Using these the perturbed equations
of moments are computed.
• The length of the blade is divided into N number of elements and the moments due to each
element on the blade is computed and summed up to obtain total moment.
The perturbed quantities are as follows after small angle approximations
4.5 Algorithm to Solve the equation
The algorithm to solve computational model is illustrated below.Step 1 and 2 have been explained
in detail in previous sections.
4.5.1 Simplification process
The size of the equations derived using the above procedure is huge. To reduce its size so that
it can be used for further analysis, the following simplifications are done.
• Small angle approximations

cos qi ≈ 1 sin qi ≈ qi
• Products of qi above three terms are negleted

4.5.2 Linearization process
The obtained non-linear equations are linearized by applying perturbation about an equilibrium
position. This is done by replacing qi with qi → δqi + qo . The resulting equations will have two
parts. One part will be linear in terms of qi . The second part will be non-linear in terms of
the equilibrium values qio . The equilibrium equations are solved and the obtained solutions are
applied to the first part to obtain the linearized equations of motion.
4.5.3 Transformation
The obtained equations are periodic. In case of isotropic rotor, they can be converted to constant
coefficient system by using Multi-blade coordinate transformation.
4.5.4 Stability Analysis
The system equation is converted to
rst order form and eigene value analysis is done for each value of rotational frequency Ω. The
modal damping and frequency curves are plotted with respect to the rotational frequency.
4.6 Results
The problem statement was taken from Bousman (1980) described in Tabel ?? The results are
Table 4.1: Experimental Configuration for Bousman’s Model
Stiffness Value
ωf lap 3.13 Hz
ωlag 6.7 Hz
ωpitch 2 Hz
ωroll 4 Hz
shown as in Figure 4.2 and Figure 4.3. As it can be seen the code developed was giving exact
results as same as that given by Bousman’s model. In such a case, the hinge offset difference for
flap as well as lag were kept null. At zero rpm, the flap progressing and regressing, and lead-lag
progressing and regressing mode frequencies are in close proximity due to the matched stiffness
condition. As rotor speed is increased, the progressing and regressing modes separate rapidly.
Figure 4.2: Imaginary part of Eigenvalues
Figure 4.3: Real part of Eigenvalues with Bousman reading

Chapter 5
Experimental Setup
In order to validate the computational complex model developed in the previous chapter and
to help the comprehension of the physical phenomenon, it is imperative for development of an
experimental setup. It is intended to design as a reliable apparatus that reproduces the ground
resonance phenomenon in a safety manner by guaranteeing no dangers for the users and no
damage of the machine.
First, a preliminary and detailed conception are made based, on general specifications attributed
at the begining of the project. The high resemblance with the mechanical model described in
chapter 2 and limiting the maximum allowable revolving speed are some examples of specifica-
tions for this project.
Later at the project test phase, simulations are made in order to verify if the project attempt sat-
isfactorily the following aspects: mechanical resistance, stability simulations and safety systems.
It means:
• Mechanical Resistance: Compute and verify if the maximum stress or deformation of

mechanical elements are within a permissible level, according to the material properties.
• Stability Simulations: Compute the boundary of instabilities by using the approach given
in previous chapter and verify that the maximum critical rotor speed is below than the
maximum allowable adopted.
• Safety System: Optimizing and adapting the mechanical elements in order to minimize the
risk factor of the experimental setup.
A brief description about the experimental setup as well as the main components and functions
are presented in section 5.1. The finite element analysis used to obtain the physical properties
of the system is explained in section 5.2.1.
32
5.1 Design of experimental setup
Before introducing the setup, it is important to discuss about the mechanical design of the ex-
perimental helicopter. Later the global assembly, containing the helicopter and the base system,
is shown in section 5.1.2
5.1.1 Mechanical Design of experimental setup
A CAD model of the design is shown in Figure . For didactic reasons, the helicopter is divided
into two main systems: fuselage and rotor head.
The rotor system is composed of a rigid rotor hub and an assembly of four blades. The angular
stiffness in the blade is obtained through flexible lamina. Each lamina is located over the blades
and links the rotor hub to the blade. At both the joints (rotor head/lamina and lamina/blade),
the kinematic coupling are of type clamped-clamped. Thus, any rotational movement of the
blade deforms the lamina creating then a restoring bending moment, like a spring.
Regarding the fuselage of experimental helicopter, an electric motor is the energy source necessary
to drive the rotor system at constant angular speed ω. The motor is linked to the chassis and
four vertical laminas are linked to it. These laminas would be later linked to rigid transverse
beams and support the whole Fuselage/Rotor Head systems.
The fuselage, suspended by these flexible laminas over an inertial blade (See Figure), can oscillate
in the longitudinal and lateral directions. Similar to those found in the head of rotor system,
these thin blades represent the spring stiffness of the landing skids of an helicopter and their
kinematic coupling at each joints are of the type clamped-clamped.
The spring stiffness of the fuselage and blades are strongly dependent on the laminas geometry,
once the material is fixed (say, made of spring steel with Esteel = 2.2 × 1011 Pa. Intended to
study different helicopter configurations (by altering the spring stiffness), three sets of laminas
have been created for the fuselage and blade. Figure 5.2 shows the general schema of the laminas
with the main geometrical dimensions while Table 5.1 gives the numerical values of all set of
lamina.
Table 5.1: Physical Data for the experimental setup
System Description Units Value

Fuselage mf kg 45.2
mb1..4 kg 2.84
a m 0.1
Rotor
b kg 0.22
Izb kg − m2 0.11
5.1.2 Global Assembly
The model is based on conceptual designed proposed by [14]. However changes are proposed
with respect to pitch control given by swashplate.
Figure 5.1: Isometric view of Global Assembly
Figure 5.2: Modified Swashplate Assembly
5.2 Equivalent Spring Stiffness Identification
The present section is aimed to determining the fuselage and blade equivalent spring stiffness of
the experimental setup. The representativeness of these springs are done by laminas that bend
(bending deformation) as the fuselage or blade oscillates.
Two methods are used to identify the equivalent spring stiffness. The first method determines it
analytically by using the theory of strength of materials. The other method employed involves
using Finite Element Techniques to identify the modal frequencies and hence its stiffness.
5.2.1 Analytic Spring Stiffness Determination
The analytical development given here are based on the static bending equilibrium of each lamina
and on the theory of strength of material. It has been referred from [14]. Moreover, the kinematic
coupling are of type coupled-coupled at both extremities of the laminas and small deformations
are assumed:
• Fuselage Spring Stiffness
Figure 5.3 represents a general schema of the fuselage lamina under a lateral shear force Fx . The
z direction is considered parallel to the longitudinal axis of the lamina while x is parallel to the
thickness direction. Also v(z) is the lateral displacement of a given section z where v(lf ) = x(t)
which is coincident with the longitudinal displacement of the fuselage. Accordingly to the theory
Figure 5.3: General Representation of Fuselage Lamina
of strength of material, the following relations are obtained:
∂2v My
v” = =− (5.1)
∂z 2 EIy
∂My
Tx = − (5.2)
∂z
where My and Tx are the static moment and shear force of a given section in z, Iy is the second
moment of inertia area around y-axis and E is the Young’s Modulus of the material.
Regarding the boundary conditions at z = 0 and z = lf (see Figure 5.2), one obtains:
z = 0 → v = v0 = 0 (5.3)
0
z = lf → v = 0, Tx = Fx (5.4)
Developing from Eq 5.1 and 5.3, the analytical equivalent expression of the spring stiffness of
the lamina in the x-direction is obtained and given as,
EIy
Kf bx = 12
lf3
3 (5.5)
ef
= Ewf
lf
Similarly, admitting now a lateral force in the positive y-direction equal to Fy and repeating the
previous steps, the equivalent spring stiffness of the lamina in y direction due to bending load is
EIx
Kf by = 12
lf3
3 (5.6)
wf
= Eef
lf
However the existence of preloaded axial force Faxial on each lamina corresponding to a quarter of
the experimental helicopter weight (in this specific case), increases the equivalent spring stiffness
of the fuselage equally in both directions (i.e. x and y) According to [15], this increment is
determined through the following equation.
12 Faxial
Kfa = (5.7)
10 lf
Finally the equivalent spring stiffness of the fuselage in x- and y- directions provided by the four
laminas are:
Kf x = 4(Kf bx + Kfa ) (5.8)

Kf y = 4(Kf by + Kfa ) (5.9)
Table 5.2: Spring stiffness derived from analytic formula
Stiffness Set 1 Set 2 Set 3

Kf x [kN/m] 17.75 24.06 35.62
Kf y [kN/m] 547.39 2839 16429
• Blade Spring Stiffness
Regarding the blade lamina, Figure 5.4 illustrates it (line AC) over the blade and both are
subjected to a lateral force Fx . The x and y axis are parallel to the longitudinal axis and to the
thickness direction of the lamina, respectively.
The point A (intersection between x and y axis) indicates x = 0 and ebl is the distance between
the point and the blade articulation point B. Once the blade rotates of ϕ around B, the deformed
lamina represents the real dashed line. The following analytic development doesn’t take into
account any longitudinal deformation of the blade lamina. Thus the boundary conditions at
Figure 5.4: General Representation of Blade Lamina
x = 0 and x = lb are:
x = 0 → v = v0 = 0 (5.10)
v(lb )
x = lb → v 0 = ϕ = , Tx = Fx (5.11)
lb + ebl
Following the previously indicated process and using Eq. 5.10, the equivalent spring stiffness of
the lamina in x direction due to bending loads is determined and presented as follows:

lb + 2ebl
EIz
Kbbx = 6
lb3
lb + ebl
3 (5.12)
1 lb + 2ebl eb
= Ewb
2 lb + ebl lb
The prediction of the equivalent spring stiffness for the blades are shown in Table 5.3. Later
these values are compared with those identified experimentally.
Table 5.3: Spring stiffness for blade derived from analytic formula

Kb [Nm/rad] 71.17 139.00 240.20
5.3 Finite Elemental Analysis of Experimental Setup
Since the computational code developed in Chapter 4 requires the input listed below.
As shown of the eight required inputs, we only can physically measure three, namely mass of the
blades, hinge offset and angular velocity of the blades. However we can’t specify mx , my , kx , ky
and kζ . In the previous section, kx , ky and kζ were derived analytically. In order to validate the
analytic expression and also to derive from them mx and my , finite elemental techniques were
used. NASTRAN (NAsa-STRuctural-ANalysis) was used for FEM analysis.
5.3.1 Modal Analysis of Fuselage
Figure 5.5 shows how the fuselage was approximated in the finite elemental model. Its detailed
view is shown below in Figure 5.6. It is to be noted that the entire mass of the rotor was
Figure 5.5: Finite Element Approximation of Fuselage
approximated and concentrated at the hub for analysis of fuselage. It is a fair approximation
since in modal analysis of fuselage, we assume that individual rotor modes do not contribute and
the rotor and fuselage moves as a single unit. Its modal analysis is as shown in Figure 5.7
The modal frequencies obtained from NASTRAN is as shown in Tabel 5.4
From the above table x and y modes correspond to order number 1 and 2 respectively. Hence,
we have obtained the frequencies as given below in Table 5.5
Figure 5.6: Finite Element Model of Fuselage
Figure 5.7: Modal Analysis of Fuselage
Table 5.4: Real Eigenvalues (Obtained from NASTRAN)
MODE EXTRACTION EIGENVALUE CYCLES

NO ORDER
1 1 3.083933E+02 2.794941E+00
2 2 9.435979E+03 1.546015E+01
3 3 2.072787E+05 7.245985E+01
4 4 4.238884E+05 1.036205E+02
5 5 2.195456E+06 2.358210E+02
6 6 2.315225E+06 2.421680E+02
7 7 3.130730E+07 8.905185E+02
8 8 3.130829E+07 8.905325E+02
9 9 3.131254E+07 8.905930E+02
10 10 3.133075E+07 8.908520E+02
Table 5.5: Frequencis for fuselage derived from Modal Analysis
Frequency Value [s−1 ]

wx 2.79
wy 15.46
5.3.2 Modal Analysis of Blade
Figure 5.8 shows how the blade was approximated in the finite elemental model. It is to be noted
that lamina is kept clamped-clamped on both the ends.
The modal frequencies obtained from NASTRAN is as shown in Table 5.6
Figure 5.8: Finite Element Approximation of Blade
Figure 5.9: Modal Analysis of Blade
From the above table lead lag mode correspond to order number 1.
Table 5.6: Blade Analysis:Real Eigenvalues (Obtained from NASTRAN)
MODE EXTRACTION EIGENVALUE CYCLES

NO ORDER
1 1 3.948965E+02 3.162728E+00
2 2 5.264922E+05 1.154825E+02
3 3 3.994273E+06 3.180819E+02
4 4 1.256295E+07 5.641128E+02
5 5 1.538656E+07 6.242966E+02
6 6 4.207798E+07 1.032399E+03
7 7 4.943024E+07 1.118965E+03
8 8 4.952962E+07 1.120089E+03
9 9 9.306334E+07 1.535357E+03
10 10 1.087198E+08 1.659490E+03
Hence, we have obtained the lead lag frequency as 3.163Hz.

5.3.3 Calculating stiffness input parameters for the modes
Once we obtain the frequencies, we could find out kx and ky from the given relations below
FE model of helicopter
dcg(pitch/roll) = r − (hubcentre − C.G) = r − a

where a is the distance between the hub center and C.G.
Moment of Inertial about roll axis is calculated as

Ipitch/roll = Ixx + M × d2cg
Since from modal analysis we know the frequency wpitch/roll . Hence we could obtain the stiffness
value from the relation
s
1 kpitch/roll
wpitch/roll = ×
2π Ipitch/roll
From this we could get the stiffness parameters as listed below

We could see that from Table 5.3 and 5.7, both analytical and Modal analysis gives the same
Table 5.7: Spring stiffness derived from Modal Analysis

Kf x [kN/m] 17.75 24.06 35.62
Kf y [kN/m] 547.39 2839 16429
result. Hence the input to be provided to the code from physical properties of fuselage is deemed
correct. Similarly for blade, stiffness obtained from the above procedure results the following
values of kb as shown in Table 5.8
Table 5.8: Spring stiffness for blade derived from modal analysis

Kb [Nm/rad] 71.17 139.00 240.20
Chapter 6
Fabricated Setup
There were several reasons that the experimental setup built in previous section could not be
completed within time frame.
Figure 6.1: Isometric View of Fabricated Setup
(a) Side View (b) Front View
Figure 6.2: Views of Fuselage
Hence we decided to fabricate a mini model that resembled ground resonance setup in Depart-
ment of Aerospace Engineering, IIT Madras. The fabricated model is shown below with the
fuselage and rotor parts separately. The fuselage part is shown in Figure 6.1 and rotor part is
shown in Figure 6.3,6.4 and 6.5.
42
Figure 6.3: Top View of Rotor
Figure 6.4: Side View of Rotor
Figure 6.5: Isometric View of Rotor
6.1 Obtaining input parameters
As discussed in Chapter 5, we need to find ωpitch , ωroll and ωlead−lag . This is generally obtained
by taking the natural response of the system by free vibration. For example, to obtain ωlead−lag ,
the blades are given excitation in lead lag direction and as a result, its response was noted in
IMU. Then, a fourier analysis was done and with the help of curve-fitting tool in Matlab, the
frequency of excitation as well as damping were plotted down.
Figure 6.6: Free decay Response for lead lag motion
Figure 6.7: Curve fitted for free decay response
Figure 6.8: FFT analysis for free decay response
A similar analysis was performed for obtainin roll as well as pitch frequencies and they are
summarized below in Table 6.1
Table 6.1: Frequencis obtained using free response
Frequency Value
wleadl ag 7.617 Hz
wpitch 6.7 Hz
wroll 4.5 Hz
6.2 Results and Discussions
The obtained frequencies were taken as input to the code developed in Chapter 4. It was run
through the code and the output generated is plotted in Figure 6.9
From the results obtained above, the rotor was run at those rpm and vibrations were observed
Figure 6.9: Results obtained from complex-model code
in those ranges corresponding to pitch and roll modes. This obtained results matched with the
results given by the computational model. The vibrations were not validated to be that caused
due to ground resonance, however at other rpms other than the two values obtained above, the
setup was not shaking.
Chapter 7
Conclusion
We first explained the mechanical model of the helicopter adopted and termed it as the “simple
model”. It is inspired by the one used by Coleman and the assumptions considered along the
modeling process were presented. It presented a structural model without aerodynamics. Later,
the equations of motion for the rotor blade as well as the fuselage were developed using Newton’s
principle. The solution methodology was described for solving the obtained equations.
With respect to the mathematics involved, we started from Coleman’s method and with that,
transformed all the periodic terms by appropriate transformations. Later this was transformed to
a matrix form where the eigenvalues were computed. Then we shifted to advantages of periodic
coefficient system analysis over multi-blade coordinate transformation in which we discussed by
providing two solutions to a problem posed by US army’s criterion. The second method was
explained in detail and is termed “Floquet Theory”.
A similar analysis is extended to a model with fuselage pitch, roll and blade flap, lag degrees of
freedom and incorporates uniform inflow model and such a model is termed “Complex Model”.
The equations of motion for a coupled rotor-fuselage system were derived using the Lagrangian
approach. Then we shifted our attention on the experimental setup designed for further under-
standing and for validation of the ground resonance model in helicopters. A detailed explanation
of the experimental mechanical design was given. Its main components, systems and subsystems,
mechanical and geometrical properties and functioning were described. Two different methods
were presented for determining the equivalent fuselage and blade spring stiffness and the results
were discussed.Finally, the experimental setup which was fabricated was explained. Later, a
method was discussed on how to obtain the physical input parameters to the system. From
the physical parameters obtained, the values were given as input to the code and hence, the
unstable regions were identified. This setup later was later run in that rpm region and results
were discussed.
46
Bibliography
[1] Jacob Pieter Den Hartog. Mechanical vibrations. Courier Corporation, 1985.
[2] Anthony RS Bramwell, David Balmford, and George Done. Bramwell’s helicopter dynamics.
Butterworth-Heinemann, 2001.
[3] CE Hammond. An application of floquet theory to prediction of mechanical instability.

Journal of the American Helicopter Society, 19(4):14–23, 1974.
[4] US JHSAT. Calendar year 2000 report to the international helicopter safety team. 2007.
US Joint Helicopter Safety Analysis Team, USA.
[5] RF Ganiev and IG Pavlov. The theory of ground resonance of helicopters. International
Applied Mechanics, 9(5):505–510, 1973.
[6] Nitin K Garg, Brian P Mann, Nam H Kim, and Mohammad H Kurdi. Stability of a time-
delayed system with parametric excitation. Journal of Dynamic Systems, Measurement,
and Control, 129(2):125–135, 2007.
[7] Robert P Coleman and Arnold M Feingold. Theory of self-excited mechanical oscillations
of helicopter rotors with hinged blades. 1958.
[8] Gunjit Bir. Multi-blade coordinate transformation and its application to wind turbine
analysis. In 46th AIAA aerospace sciences meeting and exhibit, page 1300, 2008.
[9] Dewey H Hodges. An aeromechanical stability analysis for bearingless rotor helicopters.
Journal of the American Helicopter Society, 24(1):2–9, 1979.
[10] JAMES WANG and Inderjit Chopra. Dynamics of helicopters in ground resonance with
and without blade dissimilarities. In Dynamics Specialists Conference, page 2108, 1992.
[11] CS Robinson, RL King, and ER Wood. Non-linear simulation of coupled rotor-fuselage

response using symbolically derived equations of motion. part 1: derivation. Proceedings of
the Institution of Mechanical Engineers, Part G: Journal of Aerospace Engineering, 218(1):
1–9, 2004.
[12] CS Robinson, ER Wood, and RL King. Full nonlinear simulation of coupled rotor/fuselage
response using symbolically derived equations of motion. In AHS International, Annual
Forum, 54 th, Washington, DC, pages 785–799, 1998.
47
Bibliography 48
[13] Salini S Nair and Ranjith Mohan. Analysis and control of aeromechanical instabilities in
helicopters. In ASME 2015 International Mechanical Engineering Congress and Exposition,
pages V04BT04A005–V04BT04A005. American Society of Mechanical Engineers, 2015.
[14] Leonardo Sanches, Daniel Alazard, Guilhem Michon, and Alain Berlioz. Robustness analysis
of helicopter ground resonance with parametric uncertainties in blade properties. Journal
of Guidance, Control, and Dynamics, 36(1):272–281, 2012.
[15] Michel Lalanne and Guy Ferraris. Rotordynamics prediction in engineering. Wiley, 1998.

Thesis B.Tech

Uploaded by

Copyright:

Available Formats

Thesis B.Tech

Uploaded by

Document Information

Original Description:

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Thesis B.Tech

Uploaded by

Copyright:

Available Formats

INDIAN INSTITUTE OF TECHNOLOGY, BHUBANESWAR

Computational and Experimental

A thesis submitted in partial fulfillment for the

Dr. Satyanarayan Panigrahi Dr. Ranjith Mohan

School of Mechanical Sciences Department of Aerospace Engineering

Indian Institute of Technology Indian Institute of Technology

I have acknowledged all main sources of help.

Dr. Satyanarayan Panigrahi

Batchelor of Technology Thesis

List of Figures viii

2 Formulation using Newtonian Approach 6

2.1 Change in center of gravity due to lead lag hinge [1] . . . . . . . . . . . . . . . . 6

3.1 Modal damping for the configuration given in Tabel 3.1 . . . . . . . . . . . . . . 21

4.1 Physical system used for analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . 26

5.1 Isometric view of Global Assembly . . . . . . . . . . . . . . . . . . . . . . . . . . 34

6.1 Isometric View of Fabricated Setup . . . . . . . . . . . . . . . . . . . . . . . . . . 42

3.1 Parameters Used in the Sample Calculations . . . . . . . . . . . . . . . . . . . . 20

4.1 Experimental Configuration for Bousman’s Model . . . . . . . . . . . . . . . . . . 30

5.1 Physical Data for the experimental setup . . . . . . . . . . . . . . . . . . . . . . 33

6.1 Frequencis obtained using free response . . . . . . . . . . . . . . . . . . . . . . . 45

Ω Rotor rotational speed Hz

The helicopter is a fine way to travel, but

1.1 General Introduction

1.2 Effect of Vibration on Helicopters

1.2.1 Ground Resonance

1.2.2 Cause of Ground Resonance

1.2.3 Overview of accidents registered due to Ground Resonance

1.3 Literature Review

1.4 Outline of Thesis

Formulation using Newtonian

2.1 Background Theory

Let the total mass be

and the ratio of hinged mass to total mass be

Figure 2.2: Relation between µb/a and ω 2 k/M

2.2 Multiblade Rotor System

Figure 2.3: Blade displacement in lagging motion

xkg = −eR cos(ψk ) − rg cos(ψk + k )

Since k is a small angle, these relationships can be written approximately as

xkg = −(eR + rg ) cos(ψk ) + rg k sin(ψk ) (2.3)

Then substituting it in eqn 2.5

Since k is arbitrary we find that

2.3 Chassis-Rotor System

Figure 2.6: Mathematical Representation of rotor and hub [3]

xi = xh + e cos ψi + ρ cos ψi + ζi (2.7)

ẍi = x¨h − eω 2 cos ψi − ρ(ω + ζ˙i )2 .cos(ψi + ζi ) − ρζ¨i sin (ψi + ζi )

y¨i = y¨h − eω 2 sin ψi − ρ(ω + ζ˙i )2 .sin(ψi + ζi ) + ρζ¨i cos (ψi + ζi )

where the following parameters have been introduced

xic = e cos ψi + ρc cos(ψi + ζi )

yic = e sin ψi + ρc sin(ψi + ζi )

the rotor center of mass become

The hub equations thus become

The equation for fuselage is

3.1 Multi-Blade Coordinate Transformation

It can be easily seen that

ζ¨o + η ζ˙o + (ωo2 + Ω2 νo2 )ζo = 0 (3.4)

Writing in matrix form, we obtain

Equation 3.9 can also be written as

[M ][Ẍ] + [C][Ẍ] + [K][X] = 0

Further analysis of the results is given in Section 3.3

xkg = −eR cos(ψk ) − rg cos(ψk + k )

Since k is a small angle, these relationships can be written approximately as

xkg = −(eR + rg ) cos(ψk ) + rg k sin(ψk ) (2.3)