You are on page 1of 53

First Aerospace Engineering Doctoral Students’ Symposium

May 12, 2014,

Department of Aerospace Engineering,

IIT Kanpur, India 208016.

Book of Abstracts

May 2014

Kanpur, India
The Symposium is organized by the doctoral students of the Department of Aerospace
Engineering, IIT Kanpur.

Advisory Committee:

1. Dr. Ashoke De
Convener, Department Post Graduate Committee
2. Dr. Sanjay Mittal
Head of the Department
3. Dr. C. Venkatesan
4. Dr. Debopam Das
Table of contents
Sl. Page
Abstract number Title
no. no.
AERODYNAMICS

1 AEDSS-2014-101 Velocity and Density Measurements in a Circular Buoyant Helium Plume 1

2 AEDSS-2014-102 Velocity and Density field of compressible vortex rings 2


Experimental Aerodynamics of a Butterfly mimicking Flapping Wings at Low
3 AEDSS-2014-103 3
Reynolds Number
4 AEDSS-2014-104 Viscous flow in a twin intake at supersonic speeds 4
Experimental investigation of asymmetric pitching oscillations of a symmetric
5 AEDSS-2014-105 5
airfoil
6 AEDSS-2014-106 Velocity Dynamics of Suddenly Blocked Oscillatory Channel Flow 7
Numerical Investigation of Clap and Fling Motion Using Immersed Boundary
7 AEDSS-2014-107 8
Lattice Boltzmann Method
8 AEDSS-2014-108 Normal interaction of a compressible vortex ring on the wall 10

9 AEDSS-2014-109 Free vibrations of a cylinder beyond the laminar regime 12

10 AEDSS-2014-110 Analysis of Aerodynamic Forces and Inflight Measurements Of An Ornithopter 14

AEROSPACE PROPULSION

11 AEDSS-2014-201 Emission Measurements from Bench Scale Aircraft Combustor Rig 16

12 AEDSS-2014-202 Large Eddy Simulation of Swirling Non Reactive Flow 18

13 AEDSS-2014-203 Effect of transverse periodic loading on an airfoil in a cascade 20

14 AEDSS-2014-204 Computation of Supersonic Flow Past Backward Facing Step in OpenFOAM 22

15 AEDSS-2014-205 Numerical Investigation of Soot Formation in Turbulent Diffusion Flame 24


Effect of air to liquid momentum flux ratio on primary jet breakup in cross flow of
16 AEDSS-2014-206 26
air at atmospheric pressure
17 AEDSS-2014-206 Active Control of Hooting in Gas Turbine Engines 28

AEROSPACE STRUCTURES
Addition of a lead lag damper model and change of root boundry conditions in
18 AEDSS-2014-301 comprehensive aeroelastic code and its effects on structural dynamics and 31
blade loads
Development and Structural Dynamic Analysis of Bio-inspired MAV Flapping
19 AEDSS-2014-302 33
Wings
20 AEDSS-2014-303 A-Posteriori Error Estimation for Non-Linear Problems 34

FLIGHT MECHANICS AND CONTROL


Angle of Attack, Pitch Angle and Glide Angle Modeling at Various Thrust Inputs
21 AEDSS-2014-401 35
for a Powered Parachute Aerial Vehicle
Flight Dynamic Modeling for Mini Helicopter with Stabilizer bar for Trim and
22 AEDSS-2014-402 39
Stability Analysis
Influence of Main Rotor downwash on the Horizontal Tail and its effect on Trim
23 AEDSS-2014-403
of the Helicopter
41

24 AEDSS-2014-404 Estimation of Nonlinear Parameters from Simulated Data of an Aircraft 43

Design, Instrumentation and Data Acquisition of 5 Degree of Freedom (5-


25 AEDSS-2014-405
DOF) Dynamic Test Rig
46
Aerospace Engineering Doctoral Students’ Symposium (AEDSS)
IIT Kanpur, May 12-13, 2014
AEDSS-2014-101

Velocity and Density Measurements in a Circular Buoyant Helium Plume

Kuchimanchi K Bharadwaj1,a, Debopam Das1 and Pavan K Sharma2


1
Department of Aerospace Engineering, Indian Institute of Technology Kanpur, Kanpur
2
Reactor Safety Division, Bhabha Atomic Research Center, Mumbai
(aCorresponding author, e-mail: kkbj@iitk.ac.in)

Medium scale buoyancy dominated helium plumes are experimentally investigated in the present study. It
has been well known that buoyant helium plumes released into air can be used to simulate flames/pool
fires because of the similarity in the flow features and unsteady dynamics exhibited by them in the source
near field1,2. In a certain parameter space (i.e. Reynolds number Re, Froude number Fr and Density ratio
S), these buoyant helium plumes exhibit pulsating instability in the near field, which is similar to the
puffing phenomenon observed in buoyant diffusion flames and pool fires. This instability manifests itself
as the periodic shedding of vortical structures from the plume source. This periodic shedding/puffing
determines the near field entrainment characteristics of these plumes 2. The present study is carried out
with particular emphasis on the instability manifestation and ambient fluid entrainment in circular helium
plumes. The study of entrainment in such flows requires both velocity and helium concentration
measurements. Two-dimensional velocity field has been measured using Particle Image Velocimetry
(PIV). Helium concentration/density field has been measured using Planar Laser Induced Fluorescence
(PLIF), which is developed in-house. Dependence of puffing instability and ambient fluid entrainment on
various parameters will be discussed in the present study.

Figure 1: Velocity, vorticity and density fields at an instant during puffing cycle for helium plume at
Re=117, Fr=0.3 & S=0.14.

References:
1. Cetegen, B.M., Ahmed, T.A.: Experiments on the periodic instability of buoyant plumes and pool
fires. Combust. Flame 93, 157–184 (1993)
2. Cetegen, B.M., Kasper, K.D.: Experiments on the oscillatory behaviour of buoyant plumes of helium
and helium-air mixtures. Phys. Fluids8, 2974–2984 (1996).

1
Aerospace Engineering Doctoral Students’ Symposium (AEDSS)
IIT Kanpur, May 12-13, 2014
AEDSS-2014-102

Velocity and Density field of compressible vortex rings


C. L. Dora and Debopam.Das†

, Department of Aerospace Engineering, Indian Institute of Technology Kanpur-208016, India


 †

Corresponding author‟s email: laxmana@iitk.ac.in


Corresponding author‟s Ph. No.: +91-9956817221

Abstract

The exhaust flow of rocket nozzle start-up, transient supersonic circular jets, volcanic outbursts, and Muzzle blasts are
characterized by shockwaves and compressible vortex rings. In general, a circular vortex ring is produced when an impulsive flow
exits from a nozzle or an orifice. On the other hand, compressible vortex rings are usually generated at the open end of a shock
tube. The present study aims at exploring the flow and density field of compressible vortex rings by experimental and
computational methods with particular emphasis on counter rotating vortex rings (CRVR) formation and their dynamics. In the
earlier studies, the embedded shock strength has been asserted as the primary cause for formation of CRVR. However, in the
present study particle image velocimetry measurements and numerical simulations show that CRVR does not form in the
absence of Mach disk in sonic under-expanded jet behind the primary vortex ring. The Kelvin-Helmholtz type shear flow instability
of the slipstream originated from the triple point of the Mach disk, and subsequent eddy-pairing, as observed by Rikanati et al.[1]
in shock-wave Mach-reflection, is found to be responsible for CRVR formation (see Fig.1). The growth rate of the slipstream in
the present problem follows the model proposed by them. Additionally, the parameters influencing the formation of CRVR as well
as its dynamics are investigated. It is found that the strength of the Mach disk and its duration of persistence results in an exit
impulse that determines the number of CRVRs. Furthermore, the density field of compressible vortex rings is evaluated by
employing Background oriented schlieren Technique (BOS). It has been observed that the core of these compressible vortex
rings are characterized by steep density gradient following a gaussian profile. However, the tangential velocity within the core
follows a linear relation similar to that of a incompressible vortex ring.

Figure.1: Smoke flow visualization showing vortex ring, trailing under-expanded jet, and slipstream at the open-end of a shock
tube.

References:

[1].Rikanati, A and Sadot, O and Ben-Dor, G and Shvarts, D and Kuribayashi, T and Takayama, K.
“Shock-wave Mach-reflection slip-stream instability: a secondary small-scale turbulent mixing phenomenon “ Physical review
letters, 2006, 96, 17450

2
Aerospace Engineering Doctoral Students’ Symposium (AEDSS)
IIT Kanpur, May 12-13, 2014
AEDSS-2014-103

Experimental Aerodynamics of a Butterfly mimicking Flapping Wings at Low


Reynolds Number

Saurav K Ghosha and Debopam Das


Department of Aerospace Engineering, Indian Institute of Technology Kanpur, Kanpur
(aCorresponding author, e-mail: skghosh@iitk.ac.in)

The unsteady aerodynamics of a butterfly mimicking flapping wings under zero and non-zero advance
ratio is investigated experimentally in the Re range of 2000-1000. The present study encompasses two
experimental aspects of unsteady aerodynamics of the flapping wing regime; the force measurement [1]
and the PIV measurement [2]. By the above mentioned approach, the significance of LEV (leading edge
vortex) has been emphasized in the unsteady force generation. The role of the chord-wise flexibility is
substantially studied. The dependence on the kinematics of the wing is also noted by comparing two
flapping models, one with simple flapping and one with lead lag along with flapping motions. The impact
of clap and fling on particularly the lift force is observed in an extensive way. PIV experiments serve both
as a tool of flow visualization as well as a quantitative judgment for studying the nature of flow over a
flapping wing. Furthermore, the observations with the PIV analysis are compared with the force results to
explain the nature of the forces and their patterns. The flow is observed to be more attached and stable for
the lead lag kinematics of the wing. However, the ejection process noted clearly in the normal flapping
isn’t that visible with the lead lag motion. The butterfly shaped wing shows better lift and thrust
generation than a rectangular wing of the same span and same aspect ratio (AR).

(b) (c)

(a)

Figure 1: (a) Schematic view of the model (the dimensions are in mm) (b) wing mounted on a linkage (c)
comparison of lift forces in a butterfly shaped wing and a rectangular wing of the same span and same AR
References:
3. Abhijit Banerjee, Saurav K. Ghosh, and Debopam Das (2011) , “Aerodynamics of flapping wing at low
reynolds numbers: force measurement and flow visualization,” ISRN Mechanical Engineering, vol. 2011,
Article ID 162687, 8 pages, 2011. doi:10.5402/2011/162687
4. Ghosh, S.K., Dora, C., and Das, D. (2012). ”Unsteady Wake Characteristics of a Flapping Wing through 3D
TR-PIV.” J. Aerosp. Eng. 25, SPECIAL SECTION: Intelligent Unmanned Systems, 547–558.

3
Aerospace Engineering Doctoral Students’ Symposium (AEDSS)
IIT Kanpur, May 12-13, 2014
AEDSS-2014-104

Viscous flow in a twin intake at supersonic speeds

V M Krushnarao Kotteda and Sanjay Mittal

Department of Aerospace Engineering, IIT Kanpur, Kanpur, Uttar Pradesh-208016


muralik@iitk.ac.in, smittal@iitk.ac.in

Air Intakes are one of the major components of any aircraft engine. Air intakes are essentially a fluid flow
duct whose task is to ensure the engine functions properly to generate thrust. The intake must be designed to provide
the appropriate amount of airflow at low speed with relatively low distortion required by the engine. The design of
the intakes generally aims to provide the shortest possible duct length for a given cross-sectional shape at inlet and
outlet in order to minimize the friction losses and flow distortion at the exit by avoiding separation. The shape of the
intake is generally taken as curved with increasing area from inlet to exit. The curved intakes are generally known as
Y-intake, twin and bifurcated intakes. These are referred to a pair of intakes in the wing roots or on the two sides of a
fuselage, feeding a single engine via a common section of duct or a plenum chamber. With air intakes on the
fuselage, notably the Pitot type, the normal shock positioned at the entrance interacts with the fuselage boundary
layer. This is generally very much attenuated by the presence of a splitter plane just upstream. It should be noted that
this interaction, while limited, may induce a slight compression by oblique shock, which is more favorable than the
nominal normal shock. This interaction favors the starting of air intakes with mixed compression. Y-intakes are
extensively used for engines in fighter aircrafts. These ducts normally operate under steady and symmetric
conditions. Air flows in two intakes join in a common duct are subjected to air-flow instability [1-3] at low mass
flows. A particular type of instability, which is characterized by fluctuations of the quantity of the flow in each duct
and which usually results in reversal of flow in one of the ducts as the mass flow is reduced further. Instability in the
intake adversely affects the mass flow entering the engine and may lead to combustion instability, engine surge and
flame out. It can also lead to deterioration of the performance of propulsion system, thus causing catastrophic loss in
thrust.

The viscous flow in a Y-shaped intake in two dimensions has been studied via a stabilized finite element method.
The free stream Mach number of the flow entering the intake is 1.5 and the Reynolds number, based on the height of
the intake, is 105. Computations have been carried out for various sideslip angles. First, the flow is computed for the
situation when the outflow is supersonic and no condition on back-pressure, pb, exists. The flow is utilized to carry
out computations for various values of back pressure ratio, p b/pi. The strength of the shock increases as pb/pi
increases. It moves upstream of the merger section as p b/pi increases. At the critical operation it is positioned at/near
the leading edge of the cowl. The instability occurs when p b/pi increases beyond critical pb/pi. The mass flow rate at
the merger section decreases when pb/pi increases. The total pressure recovery increases with p b/pi up to the critical
pb/pi and decreases for pb/pi beyond critical pb/pi. As the sideslip angle, β increases the critical pb/pi decreases. The
instability range or flow asymmetry range increases with β. Steady flow range in the intake decreases as β increases.

References

[1] Martin NJ, Holzhauser CA, 1950, Analysis of factors influencing the stability characteristics of symmetrical twin-
intake air-induction systems, NACA TN–2049.

[2] Seddon J, Trebble W, 1955, Model test on the asymmetry of airflow occurring in twin intake systems at subsonic
speeds, Aeronautical Research Council Reports and Memoranda–2910.

[3] Anderson WE, Perkins EW, 1959, Effects of Unsymmetrical Air-flow Characteristics of Twin-intake Air-induction
Systems on Airplane Static Stability at Supersonic Speeds, NACA TM X–94.

4
Aerospace Engineering Doctoral Students’ Symposium (AEDSS)
IIT Kanpur, May 12-13, 2014
AEDSS-2014-105

EXPERIMENTAL INVESTIGATION OF ASYMMETRIC PITCHING OSCILLATIONS


OF A SYMMETRIC AIRFOIL

Anshul Khandelwal1,a, K. Poddar2 and D. Das3,


1
PhD student, IIT Kanpur
2
Professor, IIT Kanpur
3
Associate Professor, IIT Kanpur
(aCorresponding author, e-mail: anshulk@iitk.ac.in)

The aerodynamic characteristics of an airfoil in unsteady motion have many important


implications, and this has been the subject of study of numerous researchers. The present study is
directed towards dynamic effect of asymmetric, unsteady, pitching oscillations of an airfoil on
the flow development and resulting aerodynamic loads on the airfoil at relatively low Reynolds
number which may find application in MAVs and vertical axis wind turbines. Investigation
highlights the results for the conditions when the airfoil is not stalled and dynamically stalled. In
this paper, the effect of asymmetry parameter S, defined as the fraction of the time-period of one
cycle required to reach the minimum amplitude starting from the maximum, on unsteady flow
development on NACA 0012 airfoil at a Reynolds number 8.5*104 has been investigated at
different mean angles and reduced frequencies through unsteady pressure measurements on the
mid-span of airfoil. Pressure data has been ensemble-averaged over 500 cycles to remove the
randomness due to cycle-to-cycle variations in flow and helps to quantify the average load
distribution at different phase angles.

It has been found that introducing asymmetry in oscillation at a fixed value of reduced
frequency has profound effect on unsteady flow development. It significantly affects the
formation, strength and detachment of dynamic stall vortex and consequently the pressure
distribution on airfoil surface, integrated loads and the shape & size of hysteresis loops. The
effect is seen to be most dramatic in the deep dynamic stall regime. Figures 1 to 3 show the effect
of asymmetry on the hysteresis loops at mean angle 10° and reduced frequency 0.1. Higher the
asymmetry, more is the deviation and resulting difference in behavior from symmetric
oscillations; in particular for the case of oscillations with faster upstroke cycle. The reason for
change in shape of hysteresis loops has been discussed.

5
Aerospace Engineering Doctoral Students’ Symposium (AEDSS)
IIT Kanpur, May 12-13, 2014
AEDSS-2014-105

Fig 1(a): κ = 0.1, S =0.5

Fig 1(b): κ = 0.1, S =0.67

Fig 1(c): κ = 0.1, S =0.33

Figure 1: Effect of S on phase-averaged lift & moment coefficients on NACA 0012 airfoil

6
Aerospace Engineering Doctoral Students’ Symposium (AEDSS)
IIT Kanpur, May 12-13, 2014
AEDSS-2014-106

Velocity Dynamics of Suddenly Blocked Oscillatory Channel Flow


Avinash Nayak1,*, Ashok Kannaiyan1 and Debopam Das1

1 Department of Aerospace Engineering, IIT Kanpur, India. 208016.


* Email: nayakav@iitk.ac.in
* Ph. No.: +91-9005813019

ABSTRACT

In internal flow systems, it is frequently required to rest the flow suddenly. Due to the rapid deceleration of the flow
which may be periodic or non-periodic, the flow characteristics such as velocity, pressure and shear stress also
change substantially during this short interval. So, an accurate quantitative knowledge of flow dynamics in this
phase of flow is very important for design and control of internal flow systems and physiological flows.

Much study has been done on the suddenly blocked non-periodic laminar fully developed poiseuille flow. The work
of Weinbaum [1] along with studies of Das [2] and Scandura [3] illustrates the flow dynamics when a fully
developed laminar flow is suddenly blocked by application of end wall or a valve.

But not much work is available in literature for blockage of unsteady flows like an oscillating flow. So, in the current
study, an exact solution to the flow dynamics of laminar pulsatile flow in a channel which encounters a sudden
blockage is attempted. The solution is obtained using Das-Arakeri approach [2], which is comparatively simple but
powerful. Velocity profiles, pressure gradient and shear distribution during decay time are obtained. The solution is
valid for various parameters like Womersley number and the phase of cycle at which the flow is blocked. This
knowledge has an important role in engineering and physiological phenomena, such as an instance of closing,
restarting of fluid control systems, and blood flow in arteries.

Figure 1: Decay of velocity profile with time for initial Figure 2: Pressure gradient variation with time; o: W =
phase = 90o; o: t = 0.0001; : t = 0.1; +: t = 1; ×: t = 2. 4; : W = 10.
References
[1] Weinbaum, S and Parker, K H, 1975, The laminar decay of suddenly blocked channel and pipe flows, Journal of Fluid
Mechanics, 69(04), 729–752.

[2] Das, D and Arakeri, J H, Unsteady laminar duct flow with a given volume flow rate variation, Journal of applied
mechanics, 67(2), 274–281.

[3] Scandura, P, 2003, Two-dimensional perturbations in a suddenly blocked channel flow, European Journal of Mechanics-
B/Fluids, 22(4), 317–329.
7
Aerospace Engineering Doctoral Students’ Symposium (AEDSS)
IIT Kanpur, May 12-13, 2014
AEDSS-2014-107

Numerical Investigation of Clap and Fling Motion Using Immersed Boundary Lattice
Boltzmann Method
Pradeep Kumar S, Ashoke De*and Debopam Das*

*Department of Aerospace Engineering, IIT Kanpur, Kanpur- 208016, INDIA


Corresponding author‟s email: spradeepmahadeek@gmail.com
Corresponding author‟s Ph. No.: +91-8960419188

Extended Abstract
Clap & Fling, a high lift producing mechanism in insects flight is simulated using an
Immersed Boundary Lattice Boltzmann method. Effects of advance ratio and
frequency on flow field at different Reynolds number have been studied. The
numerical results agree well with the existing experimental & numerical data.

1. Introduction
This paper aims at simulating the clap and fling motion first proposed by Weis Fogh in
1973 using immersed boundary Lattice Boltzmann method. LBM has emerged as a powerful
numerical technique since past decade for simulating fluid flows. In this simulation, LBM
with single relaxation BGK model is used to solve Boltzmann transport equation for particle
distribution function from which the macroscopic properties of the fluid can be recovered by
evaluating the moments of the distribution function. The fact that both IBM & LBM are
applied around the Cartesian mesh, makes this combination both attractive and possible.

2. Formulation
In the current simulation, two rigid elliptic airfoils are made to clap and fling using an
implicit velocity correction method introduced by Wu Shu et al.[1] to simulate. A D2Q9
Boltzmann model with forcing scheme as introduced by Guo et al [3] is used.

The Physical velocity at a fluid node is assumed to be sum of intermediate velocity (u*)
(obtained from the moment of distribution function) & correction velocity(due to the discrete
lattice effect).
where &
Unlike conventional IBM methods, like direct forcing or momentum exchange method, the
force density here is considered as unknown & is determined in such a way that the boundary
(Lagrangian) velocities obtained from interpolated corrected velocity field (Eulerian) satisfies
the no slip condition. A dirac delta function interpolation is implemented as proposed by
Peskin.

8
Aerospace Engineering Doctoral Students’ Symposium (AEDSS)
IIT Kanpur, May 12-13, 2014
AEDSS-2014-107

These boundary velocities are used to calculate the velocity correction at the Eulerian
points using a spreading function which then will be introduced in Boltzmann equation which
will make the fluid aware of the boundary through the forcing term.

3. Result
The code is validated with the benchmark problems like flow over a square and circular
cylinder. The fig.1 shows the vorticity contours of four time instances in one complete clap &
fling cycle. The fig 2 shows the corresponding time history of lift and drag coefficients

Fig 1: Vorticity Contour during 2nd Cycle for Re =100,f=0.05Hz

a) b)
Fig:2 Time histories of a) Drag coefficient b) Lift Coefficient at Re = 100, f=0.05 Hz.

4. Conclusion

The Implicit IB LBM method is able to capture the flow physics like wake capture,
dynamic stall, wing - wing interaction, LEV which makes this a robust method to investigate
insect flight phenomenon. This work in future can be extended by introducing flexibility into
the wing structure.

5. References
1. J. Wu, C. Shu: Implicit velocity correction-based immersed boundary-lattice Boltzmann method and its applications. J.
Comput. Physics 228(6): 1963-1979 (2009)
2. Charles S. Peskin (2002). The immersed boundary method. Acta Numerica, 11, pp 479-517.
3. Weis-Fogh, T. (1973). Quick estimates of flight fitness in hovering animals, including novel mechanisms for lift
production. J. Exp. Biol. 59, 169-230.
4. Z. J. Wang, “Two-dimensional mechanism of hovering,” Phys. Rev. Lett. 85, 2216 (2000).
5. S. K. Mishra, A. De (2013), "Coupling of reaction and hydrodynamics around a reacting block modeled by Lattice
Boltzman Method (LBM)", Computers and Fluids, Vol. 71, pp. 91-97

9
Aerospace Engineering Doctoral Students’ Symposium (AEDSS)
IIT Kanpur, May 12-13, 2014
AEDSS-2014-108

NORMAL INTERACTION OF A COMPRESSIBLE VORTEX RING

AND WALL

D.Saravanan, Debopam Das†

*†Department of Aerospace Engineering, IIT Kanpur, Kanpur, UP, 208016,


Corresponding author‟s email:saravana@iitk.ac.in
Corresponding author‟s Ph. No.: +91-9889288665

Extended Abstract

The normal-interaction of a compressible vortex ring with a plane surface is studied both
experimentally and numerically. Isolated vortex rings are generated using an open-ended shock tube in both
experiments and simulation for shock Mach number range of 1.31≤ M ≥1.88 for different wall distance X/Dri=
1.85 and 2.91 where Dri is the initial ring diameter. The trajectory of the vortex ring from the initial to the point
of separation of the vortex and also after separation forms a fundamental study to determine some common
phenomena for all Mach numbers. The shock-vortex interactions and the unsteady separation of the vortex
ring during wall interaction are quite complex and thus plays a major role in altering both the geometric
parameters (ring and core diameters) and the kinematic characteristics (Circulation) of the vortex and hence
these are quantified to know the effects on the vortex.

1. Introduction

Shock waves and vortices are present everywhere on vehicles traveling at supersonic
speeds and hence their interactions also occur in many situations of practical interest like
supersonic aircraft and missiles. This shock associated noise is an important factor in design
of advanced jet engines. Other fields which include these interactions are the case of
helicopter blades operating at super critical speeds and fuel-air mixing in the combustion of a
supersonic combustion ramjet. This normal shock-vortex interaction can be triggered
experimentally by bursting the of a diaphragm in a shock tube resulting with a shock-induced
vortex ring and concentrated vortices at the open ended shock tube after translating and
interacting with the shock finally move onto a wall. The entire sequence can be described as
(i) vortex evolution, (ii) the interactions are accordingly shock-shock (embedded shock),
shock and the vortex interactions depending on the shock Mach number, (iiI) the vortex
stretching (a wall effect), (iv) finally their interactions on the wall (impingement) (v) the
emergence of the separation bubble usually called secondary wall vortex and (vi) finally the
lift-Off. The geometric parameters ring diameter and core diameter and the kinematic
characteristics are calculated as given by Arakeri et al [1]. Murugan and Das [2] also
obtained the entire sequence using smoke flow visualisations using driver section lengths of
Ldrvr/D=1.77 for M=1.31 and 1.55 and Ldrvr/D=4.9 for 1.7 and 1.88 at X/D = 4.7 with the same driven
section length used in this study.

2. Experimental Setup

Experiments are performed using an open ended shock tube, which is shown schematically in figure 1. The
inner and outer diameters of the shock tube are 64mm and 100mm, respectively. Experiments are performed
with a constant driven section length (Ldrvn/D= 18.75, where D is the inner Diameter of the shock tube.) for four
different Mach number, M =1.31, 1.34, 1.55 and 1.88. Mylar sheets are used as diaphragms and they are
ruptured using a thin (0.5mm) electrically heated copper wire. With Helium as driver section gas, the
diaphragm pressure ratios for the respective shock Mach numbers mentioned above are approximately 3.0,
2.7, 5.5 and 11.5, respectively. Particle Image Velocimetry (PIV) technique has been employed for measuring
10
Aerospace Engineering Doctoral Students’ Symposium (AEDSS)
IIT Kanpur, May 12-13, 2014
AEDSS-2014-108

the velocity field of the flow coming out from the shock tube, with a double-pulsed Nd-YAG laser
(100mJ/pulse, 50Hz, thickness of the sheet ' 1mm) as the illumination source and a 12-bit CCD camera of
4MP (Mega Pixel - 2048 pixels X 2048 pixels) resolution to capture the image data, with a frame rate of 4 Hz
in double frame mode. Adaptive cross correlation PIV algorithm was used to analyze the captured flow
images using an interrogation area of 32 pixels X 32 pixels and 50% overlap in the final step of the
processing.

3. Result
The velocity field in the figure 1 (a) and (b) shows the interaction of the reflected shock with the vortex
before and during the impact. The horizontal extraction of velocities U, V and Length (Total Velocity)
along the axial direction at the center of the vortex will show fluctuations in the U velocity (which is not
shown here). This will show that the interaction of the reflected shock with the vortex will affect only the
quantity U, but not the remaining quantities. Thus it also affects the core-diameter and the circulation
and hence the vorticity. Hence the vortex ring suffers a shrink in core-diameter, a loss in translational
velocity, and circulation. After the interaction the vortex ring again regains and these quantities begin to
increase due to stretching and it again falls during the formation of separation bubble.

(a) (b)

Figure 1: Particle Image velocimetry of the vortex ring (a) before (Left) and (b) during
(Right) the reflected shock interaction for Ldrvn/D= 1.02 for X/Dri=1.85, where Dri is the
initial ring diameter.
4. Conclusion
The trajectory of the vortex ring is a fundamental measure of the variation of the geometric parameters such as
ring and core-diameter which have implications on kinematic characteristics such as circulation. The interaction
of the reflected shock with the vortex ring mainly affects the axial velocity (U).

5. References

[1] J. H. Arakeri, D. Das, A. Krothapalli, and L. Lourenco, 2004, Compressible vortex ring: a PIV study, Physics of
Fluids, 16(4), 1008–1019.
[2] Murugan Thangadurai,Debopam Das, 2012,Experimental Study on a compressible vortex ring in collision with
a wall,J. Vis,15, 321-332

11
Aerospace Engineering Doctoral Students’ Symposium (AEDSS)
IIT Kanpur, May 12-13, 2014
AEDSS-2014-109

Free vibrations of a cylinder beyond the laminar regime


Navroseand Sanjay Mittal

Computational Fluid Dynamics lab

Department of Aerospace Engineering, Indian Institute of Technology Kanpur, Kanpur, Uttar Pradesh 208 016

navrose@iitk.ac.in

Extended Abstract

Unsteady aerodynamic forces arising from vortex shedding from a bluff body may lead to its vibration. There is a
strong interdependence between the motion of the body and the wake structure. Free vibrations are often
associated with resonance/synchronization/lock-in, wherein the frequency of oscillation of the structure matches
the frequency of the periodic wake vortex mode (Khalak and Williamson 1999). More often than not, the
synchronization regime is associated with hysteretic and /or intermittent behavior. There have been numerous
efforts in the past to study the lock-in phenomenon associated with a vibrating circular cylinder. For a
comprehensive review the interested reader is referred to review articles by Bearman (1984, 2011), Williamson
and Govardhan (2004), Sarpakaya (2004) and Wu et. al. (2012).

An important parameter that affects the response of the cylinder in the synchronization regime is the Reynolds
number. In the laminar regime, synchronization is associated with two branches of response: initial and lower
(Prasanth and Mittal 2008). At large Reynolds numbers (Re ~ 10, 000) experimental studies indicate presence of
three response branches: initial, upper and lower (Khalak and Williamson 1999, Govardhan and Williamson 2000).
The two kind of response curves are shown in Figure 1. Despite significant amount of work in the two flow
regimes, the exact reason for the difference in the response between the two regimes is unexplored.

Figure 1: Variation of amplitude with reduced velocity. The reduced velocity is defined as the inverse of non-
dimensional natural frequency of the spring mass system. The data for laminar regime is taken from Prasanth
and Mittal (2008), while that for high Re is borrowed from Govardhan and Williamson (2000).

12
Aerospace Engineering Doctoral Students’ Symposium (AEDSS)
IIT Kanpur, May 12-13, 2014
AEDSS-2014-109

In this study, we first present results from 3D computations carried out on a circular cylinder that is allowed to
oscillate for Re = 1000. The aspect ratio of the cylinder is AR = 4.0, and non- dimensional mass in m* = 10.0. The
motion of the cylinder along the cartesian axes is governed by two degree of freedom spring mass system. A
stabilized space- time formulation is utilized to solve the flow equations. Towards the end we will present some
preliminary results from on-going computations for 3000 < Re < 10,000.

References:

1. Khalak, A., Williamson, C. H. K., 1999. Motions, forces and mode transitions in vortex- induced vibrations at
low mass- damping. Journal of fluids and Structures 13, 813-851 .

2. Bearman, P. W., 1984. Vortex shedding from oscillating bluff bodies. Annual Review of Fluid Mechanics 16,
195-222 .

3. Bearman, P. W., 2011. Circular cylinder wakes and vortex- induced vibrations. Journal of Fluids and
Structures 27, 648-658 .

4. Williamson, C. H. K. and Govardhan, R. N., 2004. Vortex- induced vibrations. Annual Review of Fluid
Mechanics 36, 413-455.

5. Govardhan, R. N. and Williamson, C. H. K., 2006. Modes of vortex formation and frequency response of a
freely vibrating cylinder. Journal of Fluid Mechanics 420, 85-130.

6. Prasanth, T. K. and Mittal, S., 2008. Vortex- induced vibrations of a circular cylinder at low Reynolds numbers.
Journal of Fluid Mechanics 594,463-491.

13
Aerospace Engineering Doctoral Students’ Symposium (AEDSS)
IIT Kanpur, May 12-13, 2014
AEDSS-2014-110

Analysis of Aerodynamic Forces and Inflight Measurements Of An


Ornithopter


Joydeep Bhowmik, Nidhish Raj and Debopam Das

*Phd Student, IIT Kanpur, City, UP, 208016.

† Research Associate, IIT Kanpur, City, UP, 208016

Associate professor, IIT Kanpur, City, UP, 208016

Corresponding author’s email: das@iitk.ac.in

Corresponding author’s Ph. No.: +91-9793430022

Extended Abstract

An ornithopter is mechanical replica of a bird which flies by flapping its wings. Lift and thrust is solely
produced from the flapping of the wings but unlike a fixed wing aircraft, the aerodynamic forces and
moments vary with time. The overall effect of these time varying forces affects the kinematics of the vehicle
and essential to understand the same to develop such a vehicle into bio-mimicking stealth surveillance
UAV carrying a camera. Theoretical analysis has been done and compared with some existing methods for
a given wing with a given flapping kinematics. Force measurements on an ornithopter are also carried out
in a wind tunnel but since the model is rigidly fixed to the force transducer, the combined effect of the
aerodynamic forces due to the movement of the vehicle while flight is obscured.

This work presents an experimental approach to determine the aerodynamic forces that act on an
ornithopter in real flying conditions.

In the previous work a theoretical model based on unsteady lifting line theory has been developed and
several ornithopters have been successfully built and tested. Wind tunnel test reveals dominance of the
highly unsteady nature of aerodynamic forces which are not assumed in this model. The model has also
been modified to get the lift and thrust at different position of the wing to get more accurate estimate of the
aerodynamic force during the flapping cycle.

UAV’s now a day have developed a lot, It is not difficult to see an autonomous aircraft of a rotary wing
vehicle accomplishing complicated tasks without human control. But ornithopters on the other hand are at
the developing stage as the unsteady nature of aerodynamic forces offers a challenge in anticipating the
same. The present study focuses mainly on the study of aerodynamic forces and moments that are
generated on an ornithopter and their effect on the flight of the vehicle in a steady level cruise flight.

Figure 3 shows the result of the cyclic variation of aerodynamic forces for a wing flapping at 4 Hz estimated
using an unsteady aerodynamic model using modified Theodorsen’s lift deficiency function C(k).

14
Aerospace Engineering Doctoral Students’ Symposium (AEDSS)
IIT Kanpur, May 12-13, 2014
0.1 40 1.5 30
Thrust Lift
wing position wing position
1 20
0.05 20

Wing position (degrees)


Wing position (degrees)
0.5 10
Thrust (N)

Lift (N)
0 0 0 0

-0.5 -10
-0.05 -20
-1 -20

-0.1 -40 -1.5 -30


0 0.2 0.4 0.6 0.8 1 0 0.2 0.4 0.6 0.8 1
cycle time (t/T) cycle time (t/T)

Figure 3 (a) Variation of Thrust with non dimensionalised cycle time (b) Variation of Lift with non dimensionalised cycle
time.

15
Aerospace Engineering Doctoral Students’ Symposium (AEDSS)
IIT Kanpur, May 12-13, 2014
AEDSS-2014-201

Emission Measurements from Bench Scale Aircraft Combustor Rig


V. Pandey, A. Kushari†, A. De‡, J.S. Dinesh Kumar˚, A. Raut††

* Doctoral Student, Aerospace Engineering, IIT Kanpur, UP, 208016.


† Associate Professor, Aerospace Engineering, IIT Kanpur, UP, 208016.
‡Assistant Professor, Aerospace Engineering, IIT Kanpur, UP 208016
˚ Doctoral Student, Aerospace Engineering, IIT Kanpur, UP, 208016.
†† Research Associate, Aerospace Engineering, IIT Kanpur, UP, 208016.
Corresponding author‟s email: vivekp@iitk.ac.in
Corresponding author‟s Ph. No.: +91-9005663874

Extended Abstract

Present day global scenario for crude oil production indicates a general overall decline. On the other
hand, there are trends for steady increase in global consumption of crude oil. This would cause a
widening of the demand-supply gap in the foreseeable future. A similar trend is observed for the prices of
aircraft engine fuels as well since aviation turbine fuels (ATF) are derived as a middle distillate from crude
oil. There is therefore a global imperative by most major airlines to find a substitute to ATF that will be
cheaper and also have a small environmental footprint. Extensive experimentation goes into this process
for the purpose of emission evaluation.

1. Introduction

Space and weight are primary constraints in aircraft engine design and much effort is made to reduce combustion
chamber volume by using energy dense liquid fuels that have high specific energy (of the order of 45 MJ/kg). In
addition, storage and transport of liquid fuels are safer and easier compared to gaseous fuels [1]. Hence liquid
fuels are indispensable for aircraft applications. ATF is used for aircraft engines and it a middle distillate of crude
oil. Due to the rapid technological changes in the 20th century and concurrent rapid depletion of fossil fuels, we
have crossed the „peak‟ of production curve for crude oil and by implication, its derivatives. The „peak oil‟ theory
was propounded by various scholars in the first half of the 20th century and it has now found wide credence and
acceptability. The theory suggested that the peak of world oil production would occur in the first decade of the 21st
century. The International Energy Agency reports that world peak for crude oil occurred in 2006. Reports also
predict a continuous, long term and global trend of increase in consumption of these fuels, especially by BRIC
economies, thereby widening the demand-supply gap. At present, there is a global, concerted effort by the aviation
sector to develop and test newer fuels with comparable energy densities and properties so that ready substitutes
to ATF are available. Furthermore, the burning and emission characteristics of the newer fuels should be the same
as the fuel they would replace. Emissions from aircraft engines in particular must conform to stringent cut-offs. The
environmental impact from emissions has become well known in recent years. In addition to the effect on the
global environment, there is a local penalty associated with emissions. Different nations have their own emission
standards, some more strict than others. An aircraft engine must conform to the severest of prevalent emission
norms or else a penalty is levied for excess. Emission characteristics are inexorably linked to combustion
performance and gas turbine/combustor design.

2. Formulation

For the purpose of testing ATF for emissions under varying flight conditions, a bench test rig has been developed
at Combustion and Flame Dynamics Laboratory at IIT Kanpur. Bench tests are necessary prior to full scale, in
flight testing. For this purpose, a can-combustor [2] (single can of a can-annular combustor) is used for testing.
The bench test rig comprises of the following: Air Compressor of 1100 CFM, 300 psi capacity at 2200 rpm, a
16
Aerospace Engineering Doctoral Students’ Symposium (AEDSS)
IIT Kanpur, May 12-13, 2014
AEDSS-2014-201

Vortex Flow Meter, Orifice Flow Meter for air flow measurement, a 200 kW Electric Furnace designed to withstand
20 bar pressure at 1300 K (melting point of steel), closed loop controlled inlet and exit valves, Coriolis Flow Meter
for fuel flow measurement, a high pressure jet pump for after cooling of exhaust prior to exit, an Emission Analyser
with modules for the measurement of NOx, CO, CO2, O2, Total Hydro-Carbons (THC) and lastly a comprehensive
Data Acquisition System by LabView® along with pressure and temperature acquisition devices.

3. Result
Experiments were carried out at various test conditions, one of which is mentioned below in Table1.
The NOx ppm obtained is also depicted in the Figure 1. Experimental results match with computational
results carried out earlier.

P3(bar) dP dP (%) air FAR fuel flow AC Pump T3 T4


flow rate Drive RPM (deg (deg
rate (kg/sec) Freq C) C)
(kg/sec) (hz)

1.34 0.076 0.056 0.059 0.031 0.00186 17 570 194 593

Table 1. Test Condition

Figure 1. NOx ppm from test condition

4. Conclusion
For the purpose of measuring emission characteristics using a gas turbine combustor bench rig, experiments
were done and emissions measured. NOx measurements, pressure drop across combustor and T4 were
measured and these correspond closely with computational results carried out in previous study.

5. Acknowledgement

Support from Pratt and Whitney is acknowledged for this work.

6. References

[1] Weinberg, F., 1974, The first half million years of combustion research and today‟s burning problems, Plenary
Lecture, 15th (Int.) Symposium on Combustion, Tokyo, Japan.
[2] Sampath, P, Shum, F, 1985, Combustion Performance of Hydrogen in a small Gas Turbine Combustor, Int. J.
Hydrogen Energy, 10:12, pp. 829-837.

17
Aerospace Engineering Doctoral Students’ Symposium (AEDSS)
IIT Kanpur, May 12-13, 2014
AEDSS-2014-202

Large Eddy Simulation of Swirling Non Reactive Flow

Suresh Balaji V, Ashoke De†

* M.Tech Student, IIT Kanpur, Kanpur, 208016.


†Assistant Professor, IIT Kanpur, Kanpur, 208016.
Corresponding author‟s email:ashoke@iitk.ac.in.
Corresponding author‟s Ph. No.: +91-512-259-7863

Extended Abstract

In the present work, Large Eddy Simulation technique is used to study the unsteady flow structures of
turbulent swirling non reacting flow. Two non reacting flows from Sydney swirl flame database is used for
validation purpose. The temporal evolution of the flow is studied using the pressure iso-surface. Two
reverse flow zones with a collar-like structure between them is observed in low swirl case. In high swirl
case, a single long reverse flow region is observed. Q-criterion technique is used to study the complex
vortex structures.

1. Introduction

Efficient use of energy and reduction of emissions are the most essential design parameters for engineering
appliances, especially in propulsive devices. This gives rise to the need for constant progress in the research
areas of combustion to meet the stringent emission regulations and environment requirements. Combustion in
swirling flows is one of the most widely used techniques as it shows a significant enhancement in stability of the
flame with the help of recirculation zones. Besides enhancing stability, better mixing is achieved which gives rise
to increase in combustion efficiency thereby reducing the pollutant emissions. In addition to that, it also reduces
the combustion length by increasing the rate of mixing. Hence the modeling of swirling flames is always been an
emphasized task of interest in the combustion community. To have better understanding of these kind of flames,
modeling of non reactive flows is considered to be very vital. In this present study, Large Eddy Simulation (LES)
technique is used to study the unsteady flow structures and the vortex breakdown processes involved in two
isothermal cases (N29S054 & N16S159) from Sydney swirl flame database[1].

2. Formulation

In Large Eddy simulation technique, the large scale energy containing structures are resolved numerically, while
the small scale unresolved structures are modeled. Applying the box filter, the filtered continuity and momentum
equations for the large scale structures are as follows.

 ui
0 (1)
xi
ui (ui u j ) 1  P (2 Sij ) ( ij )
    (2)
t x j  xi x j x j
1  ui u j 
where the strain tensor, Sij    
2  x j xi 
The last term in equation (2) is known as Subgrid scale(SGS) tensor. It represents the contribution of subgrid
 
scale to the momentum equation. For closing the equation, the term,  ij  ui u j  ui u j has to be modeled.
Smagorinsky Eddy Viscosity model[2] is used to model SGS term. Here the eddy viscosity,  sgs is a function of the
filter size and strain rate.
18
Aerospace Engineering Doctoral Students’ Symposium (AEDSS)
IIT Kanpur, May 12-13, 2014
AEDSS-2014-202

2
 sgs  Cs  S (3)

 
1
where Cs is a Smagorinsky model parameter and S  2Sij Sij 2
. The dynamic procedure [3] is used to
calculate the model parameter Cs .

3. Result
The simulations are done using commercial code ANSYS FLUENT- 15.0. The numerical predictions are compared
with experimental results. The predictions show overall good agreement with the experimental measurements.
Figure 1 shows the evolution of unsteady flow structures for high swirl case using pressure iso-surface and axial
velocity contours. The Q criterion technique is also used to study the complex turbulent structures involved in the
flow.

t=0.045 sec t=0.048 sec t=0.066 sec

Figure 1. Spatial-Temporal evolution from Isosurface of Pressure (-462 Pa) and contour axial
velocity from 0.045 sec to 0.066 sec for N16S159

4. Conclusion
Large Eddy simulation technique has been successfully applied to study the unsteady flow structures of two
swirling non reacting jets from Sydney swirl database. The numerical predictions have been compared with the
experimental measurements. The pressure iso-surface in combination with instantaneous axial velocity is used to
study the temporal evolution of the flow. The vortex break down process involved is studied using Q-Criterion
technique. A commercial CFD code is successfully applied in studying the complex flow structures using Large
Eddy Simulation technique.

5. References

[1] http://sydney.edu.au/engineering/aeromech/thermofluids/ swirl.htm


[2] Smagorinsky, J, 1963, General circulation experiments with the primitive equations, M. Weather Review,
Vol.91, pp.99-164
[3] Piomelli, U. and Liu, J, 1995, Large eddy simulation of channel flows using a localized dynamic model, Phy.
Fluids, Vol. 7, pp. 839-848.

19
Aerospace Engineering Doctoral Students’ Symposium (AEDSS)
IIT Kanpur, May 12-13, 2014
AEDSS-2014-203

Effect of transverse periodic loading on an airfoil in a cascade


M. C. Keerthi, Abhijit Kushari†

*Graduate Student
†AssociateProfessor
Department of Aerospace Engineering, IIT Kanpur, UP, India 208016.
Corresponding author‟s email: mckee@iitk.ac.in
Corresponding author‟s Ph. No.: +91-512 259 7851

Extended Abstract

Aircraft gas turbine engines are being subjected to increasingly severe requirements in terms of thrust to
weight ratio. The blades of axial turbomachines are therefore subjected to higher loading while the blade
thickness is getting smaller. As a result, strong fluid-structure interaction results between the unsteady
air stream and the vibrating blades. Due to the effect of adjacent blades, the aeroelastic phenomena in
turbomachines establishes in a manner different from an isolated aircraft wing aeroelasticity. The present
study is a preliminary CFD study of damping performance for a compressor cascade over a range of
relevant parameters.

1. Introduction

Present-day aircraft gas turbine designs are tending towards maximize the performance at lower costs, without
compromising on safety. In the context of axial turbomachines, this entails having thinner blades while being
subjected to the same or higher aerodynamic loading. Due to the high relative velocities encountered, this can
lead to a self-starting and sustaining blade vibratory phenomena called flutter. The prevailing unsteadiness in the
air stream, due to upstream rotor wakes or gusts, can lead to blades vibrating at their natural frequencies, known
as forced response. Both flutter and forced response cause high-cycle fatigue leading to catastrophic blade
failures. The current predictive tools for aeroelasticity in turbomachines are not adequate possibly resulting in a
costly redesign of the blading in the later stage of engine design.

Some of the critical parameters in turbomachine aeroelasticity are the interblade phase angle and the reduced
frequency. The present study is a fundamental CFD study on the effect of interblade phase angle on the damping
characteristics of a reference stationary airfoil.

2. Formulation

In an actual turbomachine, the blades vibrate with a non-zero phase difference with each other. This is enforced
by the nodal diameter of the rotor disc, resulting in a constant phase difference between any two adjacent blades.
The phase difference between the forces acting on a vibrating blade and the motion of the blade is a critical
parameter that decides whether the amplitude of the blade vibration can grow in time. The net work done on the
structure is a function of this parameter, and its sign is a direct indicator of instability [1]. In the present study, only
the effect of the oscillating neighboring blades on a stationary airfoil is studied.
The cascade under study consists of five airfoils with a spacing of 60.5 mm. The airfoil is a standard test
configuration 1, recommended for low-speed compressor aeroelastic studies Bölcs and Fransson [2]. The camber
angle of the airfoil is 10° and the chord is 121.7 mm. The stagger of the blades is set at 0°. In the present study,
the second and fourth blades were oscillated about the mid-chord, while the third blade is chosen as the
reference. ANSYS CFX is used for computing the transient solution.
The effect of the phase difference between the neighboring blades is referred to as interblade phase angle (IBPA).
The reduced frequency is defined as , where b is the semi-chord, ω the blade oscillation angular
frequency and U is the inlet velocity. The damping parameter is a function of the phase difference between the
blade force component (lift or drag) and the blade oscillation motion.

20
Aerospace Engineering Doctoral Students’ Symposium (AEDSS)
IIT Kanpur, May 12-13, 2014
AEDSS-2014-203

3. Results

Figure 1 shows the plot of various parameters as a function of IBPA. The phase angle determines whether the
conditions lead to instability, while the corresponding amplitude indicated its extent. It can be seen that the
amplitudes and phase differences of the forces depend strongly on IBPA while the dependence is minimal for the
presently considered range of reduced frequency.

0.14 180

0.12

Lift phase difference, °


90
0.1
Lift Amplitude, (N)

0.08
0
0.06

0.04
k = 0.255 -90 k = 0.255
k = 0.382 k = 0.382
0.02
k = 0.510 k = 0.510
0 -180
-150 -100 -50 0 50 100 150 -150 -100 -50 0 50 100 150
IBPA, ° IBPA, °
(a) (b)
-3
x 10
3.5 180

3
Drag phase difference, °

90
Drag Amplitude, (N)

2.5

2
0
1.5

1
k = 0.255 -90 k = 0.255
0.5 k = 0.382 k = 0.382
k = 0.510 k = 0.510
0 -180
-150 -100 -50 0 50 100 150 -150 -100 -50 0 50 100 150
IBPA, ° IBPA, °
(c) (d)

Figure 1: Variation of the following parameters with IBPA for different reduced frequencies (a) lift
amplitude, (b) lift phase difference, (c) drag amplitude, and (d) drag phase difference.

For the case of the constant blade-to-blade phase angle, instability occurs when the phase difference is between
0° and 180°. In the present case, both lift and drag show a reversal in stability around IBPA = 0°.

4. Conclusion

A moving boundary model has been implemented and tested in ANSYS CFX. Simulations have been
conducted for three values of reduced frequencies and four IBPAs. Regions corresponding to high and low
damping for axial (drag) and transverse loading (lift) have been identified using the phase difference information
of the periodic loads.

5. References

[1] Carta, F. O. (1982). An experimental investigation of gap-wise periodicity and unsteady aerodynamic
response in an oscillating cascade. 1: Experimental and theoretical results (turbine blades). NASA
Contractor Rep., CR-3513.
[2] Bölcs, A., & Fransson, T. H. (1986). Aeroelasticity in turbomachines: Comparison of theoretical and
experimental cascade results. EPFL.

21
Aerospace Engineering Doctoral Students’ Symposium (AEDSS)
IIT Kanpur, May 12-13, 2014
AEDSS-2014-204

Computation of Supersonic Flow Past Backward Facing Step in OpenFOAM


Rahul Kumar Soni, Nitish Arya† and Ashoke De

*Phd Student, IIT Kanpur, Kanpur, 208016.


†M.Tech Student, IIT Kanpur, Kanpur, 208016.
Assistant Professor, IIT Kanpur, Kanpur, 208016.
Corresponding author‟s email: ashoke@iitk.ac.in
Corresponding author‟s Ph. No.: +91-512-2597863

Extended Abstract

The understanding of flow field associated with Scramjet engine is crucial in the efficient design of high
speed propulsion system. The flow physics is often complex and experimental study with prototypes are
quite expensive. This immediately poses the need for robust and customizable solver that could be
utilized to investigate detailed flow structure. In the present work, systematic validation of density based
solver (rhoCentralFoam) available in OpenFOAM, for supersonic flow, is carried out. The test case
investigated here is backward-facing step at Mach 2, reasonable agreement was observed in numerical
and experimental results.

1. Introduction

Scramjet engines are found efficient air-breathing propulsion system in high speed flow regime. However, due to
intrinsic difficulties associated with the combustion mechanism, there exists need for development in the area of
fuel mixing and flame holding. At higher supersonic speed, due to very small residence time (O~ms),
achievement of efficient fuel mixing and flame holding is of paramount importance. In recent years backward-
facing step being conventional geometric configuration to establish subsonic recirculation zone, has been studied
widely by many researchers [1-3]. They also found out that low recirculation zone established behind the step
prolongs the residence time. Huang et al [3] put forward that vortices generated at the step corner enhance the
fuel and air mixing which is reflected in the improved combustion and mixing efficiency.

The objective of the present work is to evaluate the capability of OpenFOAM framework, an open source
computational fluid dynamics class library based on C++ [4], for high speed computation. The density based
solver, that utilizes Kurganov and Tadmor schemes is chosen and modified as per our needs to simulate the flow
physics over backward-facing step at Mach 2. The test case investigated here represents the experimental
investigation of McDaniel et al [2].

2. Formulation

Numerical results were obtained by employing density based solver (rhoCentralFoam) available in OpenFOAM.
Compressible, unsteady mass averaged, Reynolds-averaged Navier-Stokes equation are solved with cell-
centered finite volume scheme [5]. Turbulence is represented through one equation Spalart-Allmaras (SA) and
two equation models like, k-ε, RNG k-ε and SST k-ω. A comparative study of these turbulence models is also
performed as part of current work. The molecular dynamic viscosity was evaluated by Sutherland's law, turbulent
and molecular Prandlt number is 1 and 0.7 respectively. The convection terms are discretized using monotone
preserving schemes and diffusion terms are discretized using central difference scheme, temporal discretization
is obtained through second order backward scheme. At inlet boundary uniform flow properties, i.e. Mach number
(M∞= 3), free stream velocity (V=520 m/s), static pressure (P∞ = 35 KPa) and static temperature (T∞ = 167 K) are
specified. Adiabatic, no-slip boundary conditions were enforced at the top and bottom wall along with condition
that noraml pressure gradient vanishes at wall

3. Result
The numerical results computed through various RANS models are validated against the experimental results.
The streamwise velocity, static temperature and pressure profiles at two locations, namely, x/h=1.75 and 6.66,
22
Aerospace Engineering Doctoral Students’ Symposium (AEDSS)
IIT Kanpur, May 12-13, 2014
AEDSS-2014-204

are compared with experimental data, presented in Figure 1. It is observed that numerically predicted results
match well with experimental data at both the locations for almost all the turbulence models, except in the region
of y/h < 1, which is consistent with the observations of Huang et al. [3].

Figure 1 : Comparison of numerical results with experimental observations at various


streamwise locations

1. Conclusion
Systematic validation of OpenFOAM for supersonic flow has been successfully attempted. Flow at Mach 2 over
backward-facing step was studied and validated. Reasonable accuracy was observed, except for pressure profile
in the region y/h < 1, due poor performance of most of the RANS based turbulence models in the separation
region. Minor oscillation after the reattachment shock was observed which needs to be addressed. Overall
OpenFOAM, appear to be competitive computational fluid dynamics tool for the investigation of supersonic flow
physic

2. References

[1] Takahashi S, Yamano G, Wakai K, Tsue M, Kono M (2000) Self-ignition and transition to flame-holding in a
rectangular scramjet combustor with a backward step. Proc Combust Inst 28:705–712
[2] McDaniel JC, Fletcher DG, Hartfield RJ (1992) Staged transverse injection into Mach 2 flow behind a
rearward-facing-step: a 3D compressible flow test case for hypersonic combustor CFD validation: AIAA Paper
1992-0827
[3] W. Huang, M. Pourkashanian, L. Ma, D. B. Ingham, S. B. Luo, and Z. G. Wang, “Investigation on the
flameholding mechanisms in supersonic flows: backward-facing step and cavity flameholder,” Journal of
Visualization, vol. 14, no. 1, pp. 63–74, 2011.
[4] Weller, H. G., Tabor, G., Jasak, H., and Fureby, C., "A tensorial approach to computational continuum
mechanics using object-oriented techniques," Comput. Physics, Vol. 12, No. 6, 1998, pp.620–631.
[5] Christopher JG, Henry GW, Luca G, Reese JM. Implementation of semi-discrete, non-staggered central
schemes in a collocated, polyhedral, finite volume framework, for high-speed viscous flows. Int. J. Numer.
Methods Fluid 2010;63(1):1–21.

23
Aerospace Engineering Doctoral Students’ Symposium (AEDSS)
IIT Kanpur, May 12-13, 2014
AEDSS-2014-205

Numerical Investigation of Soot Formation in Turbulent Diffusion Flame


Manedhar Reddy B and Ashoke De

* M. Tech. Student, IIT Kanpur, Kanpur, 208016.


 Assistant Professor, IIT Kanpur, Kanpur, 208016.
Corresponding author‟s email: ashoke@iitk.ac.in
Corresponding author‟s Ph. No.: +91-512-259-7863

Abstract

The present work is aimed at examining the ability of semi-empirical soot models in predicting soot
formation in 'Delft flame III', which is a non-premixed pilot stabilized natural gas flame. The two-step
model is a semi empirical model, where the soot formation is modelled by solving the governing transport
equations for the soot mass fraction and normalized radical nuclei concentration. The turbulence-
chemistry interaction effects on soot formation are studied using a single variable probability density
function (PDF) in terms of a normalized temperature. The results shown in this work clearly elucidate the
effect of radiation and turbulence-chemistry interaction on soot formation. The soot volume fraction
decreases with the introduction of radiation interactions, which is consistence with the theoretical
predictions. It has also been observed in the current work that the soot volume fraction is sensitive to the
variable used in the PDF to incorporate the turbulence interactions.

1. Introduction

Pollutants are not the unburned hydrocarbon, they are the hydrocarbon produced during combustion and not
consumed by the flame [1]. Soot is an impure form of carbon particles resulting from the incomplete combustion of
hydrocarbons which indicates poor utilization of fuel. Nucleation is the starting step of soot formation. It is followed
by surface growth of polycyclic aromatic hydrocarbon (PAH) species and results in formation of small particles.
The growth of these small particles depends on the heterogeneous surface reactions with acetylene, which is one
of the primary growth species. Modeling of these reactions is done by H-Abstraction-Carbon-Addition (HACA)
mechanism [2]. The collisions between this small particles result in coagulation. Heterogeneous reactions with
molecular oxygen and OH radicals result in the oxidation of soot. In addition to the type of fuel, soot formation also
depends on the operating and prevailing flow conditions. Therefore, the soot formation can be reduced by
controlling operating parameters like residence time, temperature and the turbulence.

2. Formulation

The turbulence-chemistry interactions are modeled using a steady Laminar flamelet method (SLFM) and
presumed shape Eulerian PDF Transport method. The radiative transfer equation is simplified based on the
optically thick medium approximation into a truncated series expansion in spherical harmonics (P1 approximation).
Soot Modeling
In the one-step model, a single transport equation is solved for the soot mass fraction Eqn (1)
Ysoot 
 ·(  )Ysoot  ·( t Ysoot )  Rsoot (1)
t  soot
Where Rsoot is the net rate of soot generation, calculated using the balance of soot formation and combustion. In
the Two- Step model, in addition to the above Eqn 2, the normalized radical nuclei concentration is also solved
 bnuc
*

 ·(  )bnuc
*
 ·( t bnuc
*
)  Rnuc
*
(2)
t  nuc
In the Moss-Brookes model, the soot mass concentration is calculated instead of Rsoot in Eqn 1 and the
instantaneous production rate of soot particles is calculated instead of R*nuc in Eqn 2.
24
Aerospace Engineering Doctoral Students’ Symposium (AEDSS)
IIT Kanpur, May 12-13, 2014
AEDSS-2014-205

Burner Description
The burner consists of an axisymmetric jet centred in an annulus in which a number of premixed flames are used
to stabilise the flame. The fuel is natural gas and is assumed to consist of 81% CH4, 4% C2H6 and 15% N2 by
volume.
Numerical Details
In the current work, all computations are performed using ANSYS FLUENT 13.0. For all equations, a second order
discretization scheme is used. SIMPLE algorithm is used for pressure-velocity coupling.

3. Result
Contours of mean temperature indicate that with the inclusion of gray radiation the peak centerline
temperature drops by ~100K and with the non-gray radiation the peak centerline temperature dropped
by another ~350K and the peak temperature now shifted by ~12D in the upstream direction. Radiative
transfer of thermal energy from the flame is substantial increased by soot which has high emissivity.
This has lowered the flame temperature. The soot volume fraction has reduced by a factor of ~2 with
inclusion of gray radiation and by a factor of ~10 with inclusion of non-gray radiation.

Figure 1: Contours of mean temperature and soot volume fraction with (a) no radiation, (b) gray radiation
and (c) non-gray radiation using EPDF for TCI.

4. Conclusion

a. The gray radiation was not sufficient to account for the radiation losses and the non-gray model has
significantly lowered the global maximum temperature by ~300K and temperature was found to be
~1900K.
b. The soot volume fraction was observed to be very sensitive with temperature and reduced significantly
with the gray radiation. But still over predicted the experimental value by a factor of ~10.
c. The soot nucleation is highly sensitive to the radial maximum temperature and reduced by a factor of 4
with the inclusion of non-gray radiation. The soot surface growth and oxidation also show a similar trend
with the inclusion of non-gray radiation.

5. References

[1] Haynes, B. S., and Wagner, H. G., 1981. Soot Formation. Progress in Energy and Combustion Science, 7,
pp. 229-273.
[2] Frenklach, M., and Wang, H., 1990. Detailed Modeling of Soot Particle Nucleation and Growth. 23 rd
Symposium (International) on Combustion, The Combustion Institute, pp. 1559–1566.
[3] Bart Merci, Bertrand Naud, and Dirk Roekaerts, 2005. Flow and Mixing Fields for Transported Scalar PDF
Simulations of a Piloted Jet Diffusion Flame („Delft Flame III‟), Flow, Turbulence and Combustion, 74, pp. 239–
272.
25
Aerospace Engineering Doctoral Students’ Symposium (AEDSS)
IIT Kanpur, May 12-13, 2014
AEDSS-2014-206

Effect of air to liquid momentum flux ratio on primary jet breakup in cross flow of air
at atmospheric pressure
Deepak Kumar and A. Kushari†

*M. Tech Student of Aerospace Department, Kanpur, Uttar Pradesh, 208016.


† Associate Professor of Aerospace Department, Kanpur, Uttar Pradesh, 208016.
Corresponding author‟s email:akushari@iitk.ac.in
Corresponding author‟s Ph. No.: +91-941505712

Extended Abstract

This work represent experimental study of primary breakup of liquid jet in an annular passage in a cross
flow of air at a fixed mach number 0.12 at ambient pressure. The experiment is conducted for various
velocities of liquid jet from 1.417 m/s to 7.084 m/s (based on orifice diameter = 1mm). The droplet
distribution (diameter and velocity) and volume flux is measured by Phase Doppler Particle Analyzer in
three mutually perpendicular direction. High speed images of spray are captured at the frame rate of 1000
fps. Observed droplet sizes and velocities distribution show that there are three distinct zones. However,
different trend is observed in all zones. Furthermore, the dimensionless number (Weber number and
Reynolds number) contribute only during breakup process and becomes independent later.

1. Introduction

Liquid jet atomization in cross flow of air has major applications in propulsion systems such as gas turbine
engines, ramjet and scramjet. The combustion efficiency and NOx emissions from these engines are strongly
dependent on fuel/air mixing ratio, spatial distribution of liquid drops and drop vaporization rate. This requires
detail understanding of jet breakup. The breakup of liquid jet in cross flow of air is caused by the multiphase force
interactions like inertial force of air, viscous force of air and liquid and surface tension force of liquid. These
processes affect the atomization process in a complicated way. Present work study, experimentally, the effect of
jet momentum ratio on jet breakup process.

2. Experimental setup

Experimental set up (see Figure 1) consists of a pressurized cylindrical chamber containing test section leading to
air- liquid cross flow. Inlet air velocity was measured using a pitot static probe. Phase Doppler Particle Analyzer
(PDPA) was used to measure drop sizes, axial and radial velocities for entire spray. High speed camera was
used to capture images at 1000 fps and LED light for illuminating the region of interest. Water flow rate was
measured using a turbine flow meter. Centrifugal blower was used to create air flow rate.

Figure 1. Experimental set up

26
Aerospace Engineering Doctoral Students’ Symposium (AEDSS)
IIT Kanpur, May 12-13, 2014
AEDSS-2014-206

3. Result
The dimensionless number (Weber number and Reynolds number), volume flux and jet velocity at various
momentum flux used during experiments are tabulated in Table 1. Figure 2 shows sauter mean diameter variation
at various momentum flux along axial direction. Three distinctive zones (Zone I:4-9mm, Zone II: 9-19mm, Zone III:
19-29mm) are recognized where sauter mean diameter decrease with axial direction. Small size drops are present
in zone III, where the deviation more for different momentum flux. In zone II (medium size drops) the change is
moderate at various momentum flux. However at zone I relatively large size drops are present which are less
affected by momentum flux.

Sauter mean diameter Vs axial distance


500 q=4,y=2mm
q=6,y=3mm
450
q=9,y=5mm
D32(µm)

400
q=11,y=5mm
350 q=13,y=5mm

300 q=16,y=6mm
q=20,y=7mm
250
q=23,y=7.35
200
4 9 14 19 24

Z(mm)
Figure 2. variation of Sauter mean diameter with axial distance

q 5.00 7.00 10.00 12.00 15.00 16.00 18.00 20.00 23.00 25.00
Vj (m/s) 3.17 3.75 4.48 4.91 5.49 5.67 6.01 6.34 6.79 7.08
Re 3168.15 3748.60 4480.44 4908.07 5487.39 5667.35 6011.14 6336.30 6794.92 7084.19
We 139.60 195.44 279.20 335.04 418.80 446.72 502.56 558.40 642.15 697.99
Q 149.30 176.65 211.14 231.29 258.59 267.07 283.27 298.59 320.20 333.83

Table 1. various air to liquid momentum flux ratios and corresponds to dimensionless numbers.

4. Conclusion
Penetration height increases as q increases. Droplet size is constant along radial direction and reduces in axial
direction after jet breakup. Axial velocity increases with axial direction at given radial distance and increases with
lower axial distance with radial direction. Radial velocity decreases along axial direction and increases along
radial direction. Weber number and Reynolds number of drops contribute only in breakup process.

5. Acknowledgement

Authors would like to thank Pratt and Whitney, USA for sponsoring this research activity.

6. References

[1] Wu PK, Kirkendall KA and Fuller RP, 1997, Breakup processes of liquid jet in subsonic crossflows, Journal of
Propulsion and Power, 13(1), 173-182.

[2] Santolaya JL, Aisa LA, Calvo E,Garcia I, Garcia JA,2010, Analysis by droplet size classes of liquid flow
structure in a pressure swirl hollow cone spray, Chemical Engineering and Processing, 49(1 ), 125-131.

[3] Elshamy OM, Tambe SB, Cai J, Jeng SM, 2007, PIV and LDV measurements for liquid jets in cross flow, AIAA
2007-1338, 45th AIAA Aerospace Sciences Meeting and Exhibit, Reno, Nevada
27
Aerospace Engineering Doctoral Students’ Symposium (AEDSS)
IIT Kanpur, May 12-13, 2014
AEDSS-2014-207
Active Control of Hooting in Gas Turbine Engines

Dinesh Kumar.S.J. and A. Kushari†

*Ph.D Scholar of Aerospace Department, Kanpur, Uttar Pradesh, 208016.

† Associate Professor of Aerospace Department, Kanpur, Uttar Pradesh, 208016.

Corresponding author‟s email:akushari@iitk.ac.in

Corresponding author‟s Ph. No.: +91-941505712

Extended Abstract

The present experimental study investigates the combustion instability phenomenon in a Gas turbine
combustor which employs swirling flow of air and an Air-blast atomizer to inject both fuel and Atomizing
air. The Combustor employs four air supplies (Atomizing air, Swirling air, Quenching air and Secondary
air). For this present study swirling air, secondary air and quenching air flow rate is fixed and only the
atomizing air flow rate is varied, Quenching air was reduced in phases and the reduction of quenching
air was compensated by primary air. The data indicated that the predominant frequency of oscillations is
around 300Hz which is twice that of designed value (150Hz). Combustion instability is predominant with
45 degree swirler and the magnitudes of oscillations in the combustor are matched with that of acoustic
measurements.

1. Introduction

Combustion instability remains one of the critical issues in developments of any combustion systems. The
phenomenon is observed in all types of propulsion systems. Stringent emission levels have restricted the
operation of Diffusion controlled Gas turbine combustors. So to control the NOx formation, development of Lean
premixed combustor has started. But these combustors are prone to get combustion instability phenomenon as
they operate in very lean conditions. Hence, lean combustion process generally exhibits disturbances in flow
parameters such as pressure, temperature, velocity, and species concentration. If these disturbances interact
with the heat release rates then the disturbances grow with time and they form limit-cycle oscillations having
large amplitude. This growth of disturbances due to their interaction with the combustion process (heat release
rates) is referred as Combustion Instability.

2. Experimental setup

 Experimental set up (see Figure 1) Air blast Atomizer shown in figure 1(a), shows that air enters the atomizer
tangentially and moves into a manifold (which consists of 1mm hole of eight numbers) and from that manifold it
enters into the main manifold of 11 mm inside diameter. Fuel enters axially and leaves the fuel pipe radially. Air
which comes out of the manifold pushes the fuel flow out from the Air blast Atomizer. A 400 notch angle is
provided at the exit so that it helps in better mixing of the air and fuel with the swirling air.

28
Aerospace Engineering Doctoral Students’ Symposium (AEDSS)
IIT Kanpur, May 12-13, 2014
AEDSS-2014-207

Figure 1. Schematic of Experimental set up (Left Air Blast Atomizer)

3. Result

Quenching air was found out to be one of the dominating sources for hooting in gas turbine, Hence by
decreasing the quenching air, It was found that the SPL values were decreased upto 25db even at lean
combustion process. By pulsing down the atomizing air at 15hz with an solenoid valve it was found out that
the instability decreases due to change in the droplet diameter with the apparent changes in the coherent
structures across the dump combustor.

Pressure and Heat Release Fluctuations FFT‟s of Corresponding Pressure and HR

Fig.1 Details of Pressure and Heat Release at FAR 0.0015

29
Aerospace Engineering Doctoral Students’ Symposium (AEDSS)
IIT Kanpur, May 12-13, 2014
AEDSS-2014-207
4. Conclusion

Most of the tests showed frequency of oscillations as 360Hz, which was supported by the acoustic
measurements. With 45 swirler a sound pressure level of 145db was achieved and was reduced by 26db by
decreasing the quenching air.

5. Acknowledgement

The authors would like to acknowledge the sponsors Pratt & Whitney (P&W) Cannada, under the project of
Active and Passive Control of Hooting in Gas Turbine Engines.

6. References

1. Sreenivasan, K. R., and Raghu, S., " The control of combustion instability: A perspective ", Vol. 79,

September 2000.

2. Jakob J.Keller., “Thermo-acoustic oscillations in Combustion chambers of Gas Turbines”. AIAA , Vol.33,

No.12 , December 1995.

30
Aerospace Engineering Doctoral Students’ Symposium (AEDSS)
IIT Kanpur, May 12-13, 2014
AEDSS-2014-301

ADDITION OF A LEAD LAG DAMPER MODEL AND CHANGE OF ROOT BOUNDRY


CONDITIONS IN COMPREHENSIVE AEROELASTIC CODE AND ITS EFFECTS ON
STRUCTURAL DYNAMICS AND BLADE LOADS

Parwez Alam, Dr. C Venkatesan†

*Graduate student, parwez@iitk.ac.in


† Professor and former Head, cven@iitk.ac.in
Department of Aerospace Engineering
Indian Institute of Technology, Kanpur, India, 208016

Abstract

An Analytical linear model of lead lag damper is developed by accounting the flap, lag and torsion
kinematic coupling of damper attachment point of blade. The lead lag damper is mounted between the
rotor blade and rotor hub. This damper model is added to the comprehensive aeroelastic code for finding
the natural frequency and mode shape of hingeless rotor blade which is very much useful to calculate the
loads on the blades. In structural modeling rotor blades are treated as long slender beams undergoing
deformation in flap, lag, torsion and axial. The governing equation of motion of rotor blade is derived
using Hamilton’s principal. The change of root boundary condition (blade attachment to the hub ) in
aeroelastic code is also analyze. With the change in root boundary condition the loads on the blade
correlates very well with industry results.

1. Introduction

Helicopter operates in a very complex dynamic and aerodynamic environment. The complex loading environment
of the helicopter is due to the continuous interaction of aerodynamic, structural, centrifugal and inertia forces
acting on it. Helicopters with articulated and soft in plane hingeless rotor are known to be susceptible to
aeromechanical instabilities such as ground and air resonance [2]. These instabilities arise due to the coupling of
the poorly damped rotor cyclic modes with fuselage modes. The ground and air resonance instabilities cause
uncomfortable and dangerous conditions for the helicopter and its occupants. In the case of ground resonance,
the instability can result in the complete destruction of the aircraft [3]. To eliminate the danger of these
instabilities, root end lead lag dampers have been used.

2. Formulation

The coupled equation of motion for rotor blade is derived by using Hamilton‟s variational principle. Damper
kinematic equation is derived using the variation of strain energy stored in the damper.

The generalized Hamilton principal is expressed as equations


t2

   ( U   T   We )dt  0
t1
Where δU is the virtual variation of strain energy, δT is the virtual variation of kinetic energy and δWe is the virtual
work done by the external forces[1] .

31
Aerospace Engineering Doctoral Students’ Symposium (AEDSS)
IIT Kanpur, May 12-13, 2014
AEDSS-2014-301

3. Result
The performance characteristic of a rotor is validated with whirl tower test data from industry. Figure 1
shows the thrust and power curves. Figure 2 shows the variation of root bending moment in lag and flap
direction with respect to collective pitch angle of the rotor blade. It can be seen that analysis result
match with whirl tower test data fairly well.
100000 2000 analysis
analysis
80000 whirl tower whirl tower
1500

Power
60000
1000
Thrust

40000
500
20000
0 0
-2 0 2 4 6 8 10 12 14 16 18
-20000 -2 0 2 4 6 8 10 12 14 16 18
Collective@0.7R(Deg.) Collective@0.7R(Deg.)

Figure-1: Thrust and power curves for whirl tower case

2000 analysis 6000 analysis


Lag moment@0.8m(Nm)

Flap moment@0.435m(Nm)

1000 whirl tower 4000 whirl tower


0 2000
-1000 -2 0 2 4 6 8 10 12 14 0
-2000 -2000 -2 0 2 4 6 8 10 12 14
-3000 -4000
-4000 -6000
-5000 -8000
Collective@0.7R(Deg.) Collective@0.7R(Deg.)

Figure 2: Bending moments variations with collective angles

4. Conclusion

The loads on the blade from analysis were found to correlate well with the whirl tower test data.

5. References

[1] Rohin Kumar M , “ Comprehensive Aeroelastic and flight Dynamic Formulation for the Prediction of Loads and
Control Response of a Helicopter in General Maneuvering Flight ,” Ph.D Thesis, Dept. of Aerospace Engineering.
I.I.T Kanpur, Feburary. 2014
[2] Conor Marr, “Conceptualization, Modeling and Characterization of a CF Driven Multi State Lead Lag Bypass
Damper,” Ph.D. Thesis, Dept. of Aerospace Engineering, Pennsylvania State University, May 2012.
[3] Byers, L.K. , “ Helicopter Rotor Lag Damping Augmentation based on radial absorber & Coriolis Coupling,”
Ph.D. Thesis, Dept. of Aerospace Engineering, Pennsylvania State University, August 2006.

32
Aerospace Engineering Doctoral Students’ Symposium (AEDSS)
IIT Kanpur, May 12-13, 2014
AEDSS-2014-302

Development and Structural Dynamic Analysis of Bio-inspired MAV Flapping Wings


David Kumar, Vemuri Shyam, Tigmanshu Goyal, P M Mohite, Sudhir Kamle

Indian Institute of Technology Kanpur, Uttar Pradesh, 208016.


Corresponding author‟s email: david@iitk.ac.in
Corresponding author‟s Ph. No.: +91-8953342358

Extended Abstract

The aim of this work is to develop and apply polypropylene (PP), carbon nanotubes (CNTs) composites, for
making micro air vehicle (MAV) flapping wings. MAVs (small unmanned aircrafts) can be used for
surveillance, reconnaissance, biochemical sensing, targeting, tracking, etc. [1]. While they can be fixed wing,
or rotary wing, flapping wing MAVs have the advantage of being able to fly at low speeds and have high
maneuverability. They generate both lift and thrust using their wings only. The development of a flapping
wing MAV presents many technical challenges. Some challenges include materials and structural aspects,
low Reynolds number aerodynamics, weight and volume constraints, high propulsive power to weight ratio
requirements and ability to adapt to all environments. MAVs require to do hovering for most of the missions.

Hummingbirds (Giant Hummingbirds), having excellent flight (mainly hovering and backward motion) and
structural characteristics (make reverse camber to generate lift in both strokes) [2], were chosen as the bio-
inspiration for wing development. The wings are required to be light, strong and fatigue resistant, to perform
flapping motion during flight. Therefore, wing-material becomes a crucial component. An optimization
analysis, on the basis of density and fundamental frequency obtained through Ansys, was done for selecting
wing material. Polypropylene, a thermoplastic polymer, was observed to have desired properties such as
light weight, flexibility, good strength, fatigue resistance, good heat and chemical resistance etc. Mixing CNT
with PP further increased the strength significantly, making it more suitable for MAV application. Xylene was
used as the solvent for making PP-CNTs composite. The films were developed using two flat plates and
putting them in UTM thermal chamber, at 230-240 ºC, under compression mode. The fabricated films (0.1-0.2
mm thin) were characterized mechanically (using UTM) and thermally (using TGA and DSC). 20% increment
in strength of PP was observed at 0.1 wt% of CNT. At CNT concentration higher than 0.1 %, the strength of
PP decreased to significant level. There was no improvement or degradation observed in thermal properties.

The results repeatability is one of the key factors of manufacturing techniques. One must be able to
manufacture the same object with same properties repeatedly. Here, we need at least two wings for MAV
application and with same characteristics to avoid uncontrolled asymmetric flapping. A mold was designed
to cast the desired wings. The mold for the polypropylene wing was made out of brass from the 3-Axis CNC
machine. For fabrication of wings the mold, with the material within it, was put in the thermal chamber of
UTM in compression mode. The temperature of chamber was set up to 230-240 ºC. The molds are
compressed till they completely fit as designed. The wings were characterized by their structural dynamic
characteristics. Wing modal analysis was done to obtain natural frequency and mode shapes. They were
observed through stroboscope and recorded using high speed camera. The analysis was aimed to get the
fundamental mode in the flapping range (8-15 Hz) of hummingbirds [2]. The testing was also done inside
vacuum chamber to observe the effect of air on the natural frequency and modes. The Ansys results were
compared with the experiments in vacuum for validation of results.

Keywords: Hummingbird wings, MAVs, PP-CNT composites, Structural dynamic analysis

References:
[1] Svanberg CE., 2008, “Biometric Micro Air Vehicle Testing Development and Small Scale Flapping Wing Analysis”,
MS Thesis, AFIT/GAE/ENY/08-M27, AFIT (AU), Wright-Patterson AFB OH.
[2] Raney DL and Slominski EC, 2003, “Mechanization and Control Concepts for Biologically Inspired Micro Aerial
Vehicles”, AIAA 2003-5345.

33
Aerospace Engineering Doctoral Students’ Symposium (AEDSS)
IIT Kanpur, May 12-13, 2014
AEDSS-2014-303

A-Posteriori Error Estimation for Non-Linear Problems


P. M. Mohite, C. S. Upadhyay and Y. D. Jaiswal

*Indian Institute of Technology, Kanpur, U.P., 208016


Corresponding author‟s email: yjaiswal@iitk.ac.in
Corresponding author‟s Ph. No.: +91-9919968699

Extended Abstract

A-posteriori error estimate are so called because they are estimated after the computation of FEM solution.
These estimates are important as it gives an estimate of the error associated with the FEM solution and thus
shows us the reliability of FEM solution. Although there are numerous literatures on FEM but still the studies
on error estimation are immature and are under development. Many commercial codes which are based on
FEM are usually complicated and unreliable, further the algorithm employed in solving these problems are
often unknown to the user due to proprietary reasons. Hence, it is important to device some mathematical
model to estimate the error incurred a-posteriori in the obtained FEM solution and thus, determine the
reliability of FEM and improve the accuracy of obtained solution.

1. Introduction
We have constructed and validated, as given in Babuska et al [1], several versions of residual-based error estimator
for a cantilever bar with a spring (having stiffness coefficients k0 (N/m2) and k1 (N/m4)). The error problem was
constructed using the governing differential equation and solved using the known equilibrated residual method (ERM)
and various versions of partition of unity (POU) by taking appropriate shape functions. This problem was solved for
extreme values of (k/EA) and also with higher order of approximation. This helps us to understand the nature of the
error estimates because the same method can be extended to beams and shells and there we expect similar behavior
of error estimator. In order to validate the results, we used the ratio of strain energies of the estimated error with the
exact error.

2. Result
Result obtained using POU, when error problem was solved for the problem corresponding to the exact solution as x 2,
is shown in Figure 1. The ratio of strain energy between exact and estimated error for few extreme cases with
constant 'EA' are shown in Table 1, values close to unity signify good accuracy of estimated error.
-3
x 10
1.5
Ratio of Strain
k0/EA k1/EA Energy as obtained
Exact
Error in Axial displacement (m)

1 (at tip) (at tip) from


Sol.
(m-2) (m-4)
ERM POU
0.5
1 1 0.999782 1.022623
0
1 1000 0.860885 1.147323
x2
-0.5
Exact Error
1000 1 0.385023 1.245680
Estimated Error
-1 1000 1000 0.770046 1.234956
0 0.2 0.4 0.6 0.8 1
Distance from the fixed end of the bar (m)

Figure 1. Plot between estimated error and exact Table 1. Strain energy ratio between exact and
error with k0/EA=1000 m-2 & k1/EA=1000 m-4 (at tip) estimated error (element length = 0.0625 m)
3. References
[1] Babuska I, Strouboulis T, Upadhyay CS, Gangaraj SK and Copps K, Validation of a posteriori error estimators by
numerical approach, International Journal for Numerical Methods in Engineering, 37, (1994) 1073-1123.
34
Aerospace Engineering Doctoral Students’ Symposium (AEDSS)
IIT Kanpur, May 12-13, 2014
AEDSS-2014-401

Angle of Attack, Pitch Angle and Glide Angle Modeling at Various Thrust Inputs for a
Powered Parachute Aerial Vehicle

Vindhya Devalla, Om Prakash† and Amit Kumar Mondal


*Doctoral Research Fellow, Dept. Aerospace, UPES, Dehradun, Uttarakhand, 248007.
† Professor, Head, Dept. Aerospace, UPES Dehradun, Uttarakhand, 248007.
 Doctoral Research Fellow, Dept. Electronics, Instrumentation and Control, UPES, Dehradun, Uttarakhand, 248007.
Corresponding author‟s email: vindsdevalla@yahoo.com, vindsdevalla@gmail.com
Corresponding author‟s Ph. No.: +91-7895663390

Extended Abstract

The biggest problem oil and gas industry facing today is the safety and security of the pipelines which contributes to
the increase in the maintenance cost of pipeline directly affecting the cost of fuel. An unmanned powered parachute
aerial vehicle was developed for monitoring the oil and gas pipeline. It is envisioned that the vehicle would follow a
planned trajectory to the target, and FLIR based system would be employed for extracting details about pipeline
leakages, thefts,internal corrosion, internal waxing, etc. In this paper, a 9DOF parafoil simulation was created in the
Matlab/Simulink environment to study the parafoil dynamics and assess the feasibility of the system. Thrust , pitch
angle, glide angle and angle of attack was studied and was studied for the developed model and was compared with
lingards’s model.

1. Introduction

A 9 DOF dynamical model of the parafoil was created to assist in the development of the precision delivery project. It is a
critical component of the parafoil simulation and is essential to the creation of the control and guidance systems. The Powered
Parachute (PPC) is an aircraft which derives lift from a ram-air inflated canopy, under which the fuselage is suspended. The
parachute is inflated by the dynamic pressure of the air flowing into the canopy which has a cross section in the shape of an
airfoil. This process helps the vehicle to create lift. This feature differentiates these parafoils from conventional parachutes
which are used to simply create drag. Powered parachutes have been utilized mostly for recreation activities, but some of the
special properties make them a suitable platform for unmanned aerial vehicle (UAV) and remote sensing applications.
Powered parachutes have existed since 1981 [1]. The concept was introduced at the Sun & Fun aviation event by the
ParaPlane Corporation. They represent aircraft that are somewhere between balloons and fixed wing aircraft when control is
considered as shown in Fig 1. The direction of a powered parachute is controlled by the pilot pushing on either a left or right
steering bar that pulls down on a line attached to the trailing edge of the canopy. The increased drag causes the aircraft to
turn.

2. Formulation

The 9 DOF model is given as in equation 1. The parafoil payload model is modeled as two body system consisting of canopy
mass and payload mass suspended below the canopy using the suspension lines. The steering configuration used is known
as a “fly-bar.” In this design, the Parafoil is connected to the ends of the flybar. This bar can be pulled either side of the
aircraft, changing the direction of the lift and making the aircraft turn. This type of parafoil, payload model uses 9 DOF model.
The separation between the Aerodynamic center of the canopy and the payload center of gravity produces a swinging motion.
These two centers are joined by massless links Rp and Rb, At joint C, resulting in 3DOF rotational and 3 DOF translational
motions [3].

35
Aerospace Engineering Doctoral Students’ Symposium (AEDSS)
IIT Kanpur, May 12-13, 2014
AEDSS-2014-401

Tb  b   B1 

 M b Rcb 0  M bTb
 0  
  ( M p  M F ) Rcp  ( M p  M F )T p  Tb     B2 
p   
 Ib 0 0  RcbTb   V   B3 
  c   
 0 I p  IM 0 Rcp T p   F   B4 
 c
B1  FbA  FbG  FbT   b  M b b  Rcb
B2  F pA  F pG    ( M p  M F ) p  Rcp  M F  p  T pVc   p  M F T pVc
B3   b  I b b
B4  M pA   p  ( I p  I M ) p

3. Result

The developed PAV model has been validated using the lingard model. Angle of attack (alpha), pitch angle (theta) and glide
angle (gamma) values obtained are compared with the Lingard values under no thrust conditions

Figure1. Parachute Aerial Vehicle under no Thrust condition

The model is validated using the lingard results by checking the values of angle of attack (alpha), pitch angle (theta)
and glide angle (gamma) of the parafoil with respect to time.

Figure2. Validation of Parachute Aerial Vehicle with Lingard model

The developed PAV model has been simulated at various thrust conditions, keeping the thrust as 4.5N, 7.75N and 10N.
It is observed that the gamma value is zero at the thrust input of 7.75N.

36
Aerospace Engineering Doctoral Students’ Symposium (AEDSS)
IIT Kanpur, May 12-13, 2014
AEDSS-2014-401

Figure3 . gamma values for various thrust conditions

Keeping the thrust value 7.75N the alpha, theta and gamma values are seen in figure 4

Figure.4. Parachute Aerial Vehicle with 7.75N thrust

Figure.5. PAV landing

From the figure 5, it can be observed that The theta value, while landing the there is glide angle of approximately 8 degrees.

4. Conclusion

The following cases are considered in the above mentioned model validating along with the lingard report.

Lingard Parachute Arial vehicle


(Thrust = 0N)
Alpha_p (deg) 3 5
Theta_p (deg) -9 -10
Gamma(deg) -18 -18
37
Aerospace Engineering Doctoral Students’ Symposium (AEDSS)
IIT Kanpur, May 12-13, 2014
AEDSS-2014-401
Gamma value with different thrust conditions

S.No. Thrust (N) Gamma(deg)


1 4.5 -8
2 7.75 0
3 9 7

It is observed that at thrust 7.75N the gamma value is 0, i.e. at level flight. Considering thrust value to be 7.75 N, angle of
attack, pitch angle and glide angles are plotted and the following table is obtained.

Parachute Arial vehicle Parachute Arial vehicle Parachute Arial vehicle


(Thrust = 4.5N) (Thrust = 7.75N) (Thrust = 9N)
Alpha_p (deg) 10 8 7
Theta_p (deg) 0 8 10
Gamma(deg) -8 0 7
It is observed that as the thrust increases the angle of attack decreases and the pitching angle increases. The gliding angle
indicates that the descending or on steady level flight.

5. References

[1] Damian Toohey, “Development of small parafoil Vehicle for Precision Delivery”. Massachusetts Institute of Technology,
2005
[2] O. Prakash and N. Ananthkrishnan, “Modeling and Simulation of 9-DOF Parafoil-Payload System Flight Dynamics,” AIAA
Atmospheric Flight Mechanics Conference and Exhibit, Keystone, Colorado, Aug., 2006
[3] Lingard. J. S, “Ram Air parachute Design” 13th AIAA Aerodynamic decelerator system technology conference,
Clearwater Beach, May 1995.

38
Aerospace Engineering Doctoral Students’ Symposium (AEDSS)
IIT Kanpur, May 12-13, 2014
AEDSS-2014-402

Flight Dynamic Modeling for Mini Helicopter with Stabilizer bar for Trim and Stability
Analysis
T.Sakthivel* and Dr.C.Venkatesan †

*Graduate Student, IIT Kanpur,UP-208016.


† Professor,Department of Aerospace Engineering IIT Kanpur,UP-208016.
Corresponding author‟s email: sakvel@iitk.ac.in
Corresponding author‟s Ph. No.: +91-7897768700

Extended Abstract

This study describes the development and validation of a flight dynamic modeling for analysis of trim and
stability of a mini helicopter including the effect of stabilizer bar. Designing a helicopter meeting the
required handling quality aspects ensures low pilot workload, increased mission effectiveness and
improved safety of operations in all weather and visibility conditions. Lack of an indigenous tool for
helicopter design industry in India is the motivation behind this study. The model consists of development
and integration of sub-models for main rotor, tail rotor, stabilizer bar fuselage, horizontal stabilizer and
vertical stabilizer. The simulation model has the capability to analyze any flight condition including level,
sideslip, climb, descent or turns. The simulation model can be improved by including dynamic stall and
dynamic inflow effects.

1. Introduction

Helicopter design is purely an interdisciplinary activity involving dynamics, aerodynamics, flight control and
propulsion. Modeling of helicopter dynamics involves main rotor blade dynamics, its coupling with fuselage and the
forces associated with the tail rotor and other lifting surfaces. Compared to large helicopters, mini helicopters have
certain key features like variable main and trail rotor rpm, rigid rotor blade, stabilizer bar. The stabilizer bar
provides the passive control pitch input to main rotor blade. This feedback control pitch input depends on the
flapping response of the stabilizer bar, which is influenced by the perturbational motions of the fuselage and also
due to cyclic pitch input from the swash plate.

2. Formulation

In this work formulation part includes the force and moment calculation from the various lift generating parts in the
helicopter like Main rotor, Tail rotor, Fuselage, Stabilizer bar, Horizontal stabilizer and Vertical stabilizer. The main
rotor forces contributed by the inertial and aerodynamic loads. Inertial loads are formulated from Newton law of
motion. Aerodynamic forces are formulated from Blade element theory. Sectional loads moments are calculated
and integrated over the length of the blade to get net force and moment. Most simplified models are taken into
account for force and moment calculation. For the main rotor no lead lag and twisting of the blade taken into
account. For tail rotor only thrust is calculated and other forces are neglected. Stabilizer bar flap dynamics is
considered and the forces and moment are not taken into account. Fuselage drag is calculated from the equivalent
flat plate area. Empennage surface are generating only forces normal to their surfaces. For the trim analysis 6
equilibrium equations are solved for the control variables. Using small perturbation theory, the stability of the mini
helicopter is studied.

3. Result
The mathematical developed in the previous section is simulated for the mini helicopter of 10 kg weight
and 0.91m of blade radius with the stabilizer bar.The control angles to trim the helicopter for various
forward speeds are plotted.

39
Aerospace Engineering Doctoral Students’ Symposium (AEDSS)
IIT Kanpur, May 12-13, 2014
AEDSS-2014-402

16 TR Collective (deg)
18 MR Collective (deg)

12
12
8
6
4

0 0
0.00 0.08  0.16 0.24 0.00 0.08 0.16 0.24
0 MR Longitudinal Cyclic (deg) 8 MR Lateral Cyclic (deg)

-2 0

-4

-4 -8
0.00 0.08 0.16 0.24 0.00 0.08 0.16 0.24
0 0 H/c Roll Attitude (deg)
H/c Pitch Attitude (deg)

-5
-4
-10
-8
-15

-20 -12
0.00 0.08 0.16 0.24 0.00 0.08  0.16 0.24

Figure 1. Trim Control angles for various forward speeds

4. Conclusion
Trim and stability analysis are performed for the mini helicopter with stabilizer bar and without stabilizer
bar.By comparing both the results we can conclude that stabilizer bar will increase the stability of the
helicopter. Without stabilizer bar helicopter is unstable at hover and small forward speeds. After
including the stabilizer bar that unstable modes are brought to neutrally stable.

5. References

[1] Venketesan C, 1997, Short course on Helicopter Technology Department of aerospace Engineering, IIT
Kanpur
[2] Gagan Deep Singh, 2012,Helicopter flight dynamics simulation for analysis of trim, stability and control
response, IIT Kanpur.
40
Aerospace Engineering Doctoral Students’ Symposium (AEDSS)
IIT Kanpur, May 12-13, 2014
AEDSS-2014-403

Influence of Main Rotor downwash on the Horizontal Tail and its effect on Trim of the
Helicopter
Aravindhan D and Dr.C. Venkatesan†

*Graduate student, IIT Kanpur, UP, 208016.


† Professor, Dept. of Aerospace Engg. IIT Kanpur, 208016.
Corresponding author‟s email: aravindh@iitk.ac.in
Corresponding author‟s Ph. No.: +91-9886634377

Extended Abstract

1. Introduction

The helicopter loads and vibration prediction analysis is simulated using the Comprehensive Aeroelastic
Response (CAR) code. For the purpose of calculating downwash effect on the horizontal tail, Peter-He Dynamic
wake model is used. Addition of main rotor induced velocities to the horizontal tail with its downwash effective
factor and its percentage impingement is considered for the calculation of loads arising out of the empennage for
the improvement of helicopter trim conditions.

2. Formulation

The load generation by the horizontal tail is calculated by the following equations. The wake skew angle ( ) is
calculated to find the percentage impingement of the rotor wake on the tail plane. The horizontal tail setting angle
( ) is considered for calculating the effective angle of attack ( ) due to the induced velocity component
( .

3. Result
The effect of downwash on the empennage of the helicopter was studied for various speeds of the
aircraft. The downwash effective factor ( ) is varied from 1 to 2. The increase in downwash effective
factor reduces the collective angle (θ0) and longitudinal cyclic control angle (θ1s) whereas it increases
the Roll attitude (φ), pitch attitude (θ) and Lateral cyclic angles (θ1c). The tail rotor collective angle (θTR)
does not vary much with the downwash effective factor. Figure 1 depicts the variation of control angles
with downwash effective factor for advance ratio of 0.2. Table 1 shows the variation of Trim angles with
respect to downwash effective factor for advance ratio of 0.2 and 0.25.

41
Aerospace Engineering Doctoral Students’ Symposium (AEDSS)
IIT Kanpur, May 12-13, 2014
AEDSS-2014-403

θ φ θ0 θ1c θ1s θTR


control angles 10

-5
1 1.2 1.4
downwash effective1.6
factor 1.8 2

Figure 1. Downwash effective factor on control angle variation

θ φ θ0
μ downwash downwash downwash
Flt Flt Flt
with With out with With out with With out
0.20 1.96 2.058 2.653 -2.24 -0.992 -0.759 7.85 8.187 9.287
0.25 -0.60 0.809 1.288 -2.04 -1.574 -1.374 8.38 8.755 10.552

θ1c θ1s θTR


μ downwash downwash downwash
Flt Flt Flt
with With out with With out with With out
0.20 2 1.748 2.148 -4.04 -1.662 -2.750 3.9 4.075 3.973
0.25 1.4 1.583 1.863 -5.07 -2.902 -3.757 3.9 4.477 4.355

Table 1.Trim control angles comparison for advance ratio (μ) of 0.2 and 0.25

4. Conclusion
The control angles for the trim of helicopter are improved by considering the main rotor downwash on the
horizontal tail as is evident from the above results.

5. Acknowledgement

I would like to express my gratitude and regards to my guide Prof. Dr. C. Venkatesan for his insightful guidance
throughout this course of work. I would also like to thank Mr. DSD Prasada Rao, DGM, RWR&DC for his support
and guidance. This program is made possible by generous support from Hindustan Aeronautics Limited.

6. References

[1] David A. Peters, David Doug Boyd, Cheng Jian He, 1987, Finite-State Induced-Flow Model for Rotors in
Hover and Forward Flight, Proceedings of 43rd Annual Forum of AHS, St. Louis, Mo., May 18 – May 20.
[2] Rohin Kumar M, 2014, Comprehensive Aeroelastic and Flight dynamic formulation for the prediction of loads
and control response of a helicopter in general Maneuvering flight, Doctoral Thesis, Indian Institute of
Technology Kanpur, February 2014.
[3] Robert T.N. Chen, A Survey of Nonuniform Inflow Models for Rotorcraft Flight Dynamics and Control
Applications, NASA Technical Memorandums, Ames Research Center, Moffett Field, California, 64(1) – 64(62),
November 1989.
[4] Gareth D. Padfield, Helicopter Flight Dynamics, 2007, AIAA Education Series, 2nd Edition, Blackwell Publishing,
North America.
42
Aerospace Engineering Doctoral Students’ Symposium (AEDSS)
IIT Kanpur, May 12-13, 2014
AEDSS-2014-404

Estimation of Nonlinear Parameters from Simulated Data of an Aircraft

Dhayalan.R, A.K.Ghosh†

*Doctoral Student, Department of Aerospace Engineering, IIT Kanpur, India.

† Professor, Department of Aerospace Engineering, IIT Kanpur, India.

Corresponding author‟s email: rdhayal@iitk.ac.in

Corresponding author‟s Ph. No.: +91-7275467379

Extended Abstract

The current paper discusses an attempt for estimating Non-linear parameter by an improvement to well-
known Neural Gauss Newton (NGN) method, which makes the method capable of estimating nonlinear
parameters from flight data. The estimation is carried over for a set of simulated data with various
control surface combinations. Then the estimation is carried out for the simulated data with selected
control surface combination, for which noise is added, to test the handling capabilities of the Improved
Neural Gauss Newton (INGN) method.

7. Introduction

The NGN method [3] proposed by Peyada et. al, serves as an excellent tool in recent times for parameter
estimation from flight data. An improvement in NGN algorithm [1] has been tried out, which was proved to be
worth of trying the same base algorithm for estimating nonlinear parameters. The estimation process we used is
mainly based upon the angle of attack pitch rate ( ) and linear accelerations along X and Z axis ( and
). In NGN method, the input variables for training include the above mentioned. The force and moment
coefficients are provided along with the accelerations, angle of attack, pitch rate and pitch angle ( ) as outputs
for training the neural model. The data used for estimation is the trajectory model of the airplane under
consideration i.e. time history of the variables. Since the neural model is developed based on the input and
output vector, the idea of improving NGN is tried by including the first derivative of the important motion
variables, in the longitudinal case and for training. By these two inputs, we can characterize the force and
moment coefficients ( , and ) better than NGN method. The pattern following ability of the neural network
is exploited for this improvement. This improves the robustness of the neural model too, since we give the
derivatives of the most influential variables for training. This improvement has brought better estimation of linear
model with a much lower standard deviation. This is has been validated using the HANSA-3 flight data for
longitudinal case, along with NGN for comparison [1]. The improvement made has been proved very effective for
linear model. With this method, we can estimate nonlinear aerodynamic model.

8. Formulation

The simulation is carried out for a McDonnell Douglas F-4 aircraft, which has nonlinear aerodynamic derivatives
over 3 different angle of attack regions. The nonlinear derivatives include the some of the lateral variables too.
Hence, the data required for estimation process requires the inclusion of trajectory variable angle of side slip. To
generate such data, we have to simulate the data by deflecting both elevator and rudder. By deflecting both
elevator and rudder, the combined effect of angle of attack and angle of sideslip is captured in the data, which is
43
Aerospace Engineering Doctoral Students’ Symposium (AEDSS)
IIT Kanpur, May 12-13, 2014
AEDSS-2014-404

also carried out for four different combinations of the control surfaces which maximizes the energy of the whole
input[2]. From this the best combination of the control surfaces is chosen and used for estimation of nonlinear
parameters after the introduction of noise. The simulation is carried out for two Angle of attack regimes, i.e.
and .

Result

The estimation results are given below. The plots are presented for the comparison of Linear and Nonlinear parameters over
two angle of attack regions. The Improved NGN method is able to predict the nonlinear parameters from simulated data
without noise with better accuracy. This method is carried out for noise introduction too, which is not presented here.

9. Conclusion

44
Aerospace Engineering Doctoral Students’ Symposium (AEDSS)
IIT Kanpur, May 12-13, 2014
AEDSS-2014-404

The nonlinear aerodynamic parameters are estimated using the Improved NGN method. The estimation is
carried out for noise-introduced-simulated flight data, which will be presented in the seminar.

10. References

[1] Dhayalan.R and A.K.Ghosh. Improved NGN algorithm for parameter estimation from flight data. In SAROD
2011, 2011.

[2] R. V. Jategaonkar. Flight Vehicle System Identification A Time Domain Methodology. AIAA, Inc., Reston, VA,,
2006.

[3] N. K. Peyada and A. K. Ghosh. Aerodynamic parameter estimation using new filtering technique based on
neural network and gauss-newton method. In ARMS 2008,. ARDE, 2008.

45
Aerospace Engineering Doctoral Students’ Symposium (AEDSS)
IIT Kanpur, May 12-13, 2014
AEDSS-2014-405
Design, Instrumentation and Data Acquisition of 5 Degree of
Freedom (5-DOF) Dynamic Test Rig
Subrahmanyam S1, a, Dhayalan R2, b, Rishi Raj Singh3, c, Ghosh A K4, d.
1,2
Ph.D scholar, Flight Dynamics Lab, Department of Aerospace Engineering, IITK, Kanpur, INDIA.
3
Senior Research Associate, Flight Dynamics Lab, Department of Aerospace Engineering, IITK, Kanpur,
INDIA.
4
Professor, Flight Dynamics Lab, Department of Aerospace Engineering, IITK, Kanpur, INDIA.
a b c d
saderlas@iitk.ac.in , rdhayal@iitk.ac.in , rishi.iitk@gmail.com , akg@iitk.ac.in

Keywords: Instrumentation, Data Acquisition, 5-DOF Dynamic Test Rig, 3-DOF gimbal, 2-DOF gimbal,
Data telemetry, compensator.

Abstract. The present paper discusses about the design of various components namely 3-DOF
gimbal, 2-DOF gimbal and compensator etc., of 5 degree of freedom dynamic test rig. The paper
also figures out the advantages of the 5-DOF rig in estimating the dynamic derivatives of the
flight vehicle of interest. Importance of the instrumentation of various components of 5-DOF rig
as well as flight vehicle has been explained. It also describes the onboard data acquisition system
as well as data telemetry to ground station which is obtained from the sensors during the tests.
Finally at the end of the paper a sample data is presented which obtained during the wind tunnel
testing of the flight vehicle mounted on 5-DOF dynamic test rig.

Introduction

Any flight vehicle is characterized based upon its aerodynamic derivatives, which
includes the stability derivatives both static and dynamic, and also the control derivatives. Study
of these aerodynamic derivatives of flight vehicle is a part of the design process. Many practical
applications like flight control system design, development of simulators, flying quality analysis
and autopilot design requires the estimation of accurate static, dynamic and control derivatives
[1]. By assuming the flight vehicle as a rigid body the six degrees of freedom (6-DOF) equations
of motion of an aircraft are derived and are expressed below as follows [2].

(1)
(2)

If we observe the above dynamic equations of an aircraft, the total external forces and
moments acting on the flight vehicle is the summation of gravity, propulsive and aerodynamic
forces and moments respectively. These aerodynamic forces and moment coefficients are
modeled as a function of static stability, dynamic stability and control derivatives respectively.
The following first order Taylor series expansion explains the force or moment coefficient as a
function of aerodynamic derivatives [3].

(3)
The above mentioned aerodynamic parameters of flight vehicle can be estimated using
different methods – namely, empirical methods, computational fluid dynamics (CFD) analysis,
wind tunnel testing and, finally from real flight data. In modern engineering, aircraft
46
Aerospace Engineering Doctoral Students’ Symposium (AEDSS)
IIT Kanpur, May 12-13, 2014
AEDSS-2014-405

development involves a combination of all these methods [4]. Empirical methods, which are
based on the experimental database, are used to estimate the static, dynamic and control
derivatives during the preliminary design of an aircraft [5]. Static parameters are required to
estimate the trim as well as to characterize the handling qualities of an aircraft, where as dynamic
stability derivative enables us to understand how the system responds for external disturbances
over a period of time.
The aerodynamic and stability parameters are usually estimated from the data obtained
through static wind tunnel tests. It is generally difficult to estimate dynamic derivatives through
conventional wind tunnel tests, in which the model is kept static and the data is obtained at each
and every position of interest. Dynamic parameters are in general estimated through flight tests,
which are expensive, time consuming and also fatal during dangerous maneuvers. Further, such
tests are performed on the frozen design, and the scope of modification and design improvements
are very difficult. To minimize the expenditure of real flight testing and to have useful
information at a mid-course of design stage, a 5-DOF dynamic test rig has been designed,
fabricated and tested in the National Wind Tunnel test Facility(NWTF), IITK. The proposed 5-
DOF dynamic test rig is capable for aerodynamic characterization of flight vehicle of any
configuration [6]. The model mounted on this 5-DOF rig can perform roll, pitch, yaw, heave and
sway and the aerodynamic parameters of the flight vehicle can be estimated through variety of
maneuvers. The dynamic data of the flight vehicle obtained from the 5-DOF dynamic test rig is
the time history of the variation of state variables. A 5-DOF mathematical model has been
developed and is used for the simulation of the dynamic motion of the flight vehicle along with
test rig [7]. Parameter estimation techniques, which utilize the generated flight data, are used to
estimate the dynamic parameters of the corresponding flight vehicle [7].

Design and Fabrication.


5-DOF dynamic Test Rig. The flight vehicle of interest, whose dynamic derivatives has to be
determined, will be mounted on the 5-DOF dynamic test rig. Unlike conventional wind tunnel
tests, the model mounted on the 5-DOF dynamic test rig will have the following degrees of
freedom namely, pitch, yaw, roll, heave and sway.

(a) (b)
Fig. 1 (a) Schematic diagram of 5-DOF dynamic test rig indicating various components [5].
(b) Photograph representing 5-DOF dynamic test rig.

The above mentioned DOF are enabled by means of the following components namely
3-DOF gimbal - provides the heave, sway and roll to the aircraft model
2-DOF gimbal - enables the flight vehicle to perform pitch and yaw
Rig Arm - connects the flight model mounted on 2-DOF gimbal to 3-DOF gimbal
47
Aerospace Engineering Doctoral Students’ Symposium (AEDSS)
IIT Kanpur, May 12-13, 2014
AEDSS-2014-405

Compensator - Excites the system such that the inertial effect of rig arm on model of
interest is negligible.

Instrumentation and Data Acquisition.


In order to acquire the reliable dynamic response of the flight vehicle using dynamic test
rig, proper instrumentation of flight vehicle as well as rig arm is required. Various electronic
instruments and their calibration setups, which are used for data acquisition and data telemetry,
have been explained.
Inertial Measurement Unit (IMU) - Two 9-DOF has been used to extract the linear
accelerations (ax, ay, az), angular rates (p, q, r) and Euler angles (ϕ, θ, ψ) of the flight model as
well as rig arm. This device can be easily mounted on rig arm and flight vehicle model, such that
axis system of device and rig arm/flight vehicle matches. This information is generally used as a
feedback in the control system design[7].
Actuators and Sensors - One of the aspects critical to precise measurement of control
surface deflection and actuation is calibration of both the potentiometers and servo motors, which
are used as sensors and actuators respectively. A calibration rig was designed for this purpose on
which the servo motors and potentiometers can be simultaneously or separately calibrated
2.5 3.5

3
2
2.5
Pulse width (in ms)

Voltage in volts
1.5
2

1.5
1

servo1 1
Potentiometer1
0.5 servo2
Potentiometer2
servo3 0.5
Potentiometer3
servo4 Potentiometer4
0 0
0 50 100 150 200 250 300 0 50 100 150 200 250
Angle (in deg) Angle in deg

(a) Calibration Rig for servo&Potentiometer (b) Calibration plot of Servo (c) Calibration plot of Potentiometer
Fig. 2 Calibration setup and plots for servo and potentiometer
Data Acquisition system - The following block diagram explains the data acquisition
module of the proposed 5-DOF dynamic test rig.

Aircraft model
IMU
Potentiometer
Servo

Model Mounted on
2DOF GIMBAL
Onboard Microprocessor Ground Station
Data
Data Data
Logging Telemetry Transmission Receiver GUI

Rig Compensator
IMU
Potentiometer
Servo

Fig. 3 Block diagram representing the data acquisition system of 5-DOF dynamic test rig.
Wind tunnel Testing.
48
Aerospace Engineering Doctoral Students’ Symposium (AEDSS)
IIT Kanpur, May 12-13, 2014
AEDSS-2014-405

The 5-DOF dynamic test rig was tested in National Wind-tunnel Testing Facility
(NWTF), at IIT Kanpur. NWTF is a low speed, closed circuit wind tunnel which is capable of
testing at a wind speed up to 80 m/s. It has the facility to interchange the test sections and the
cross section of each test section is 3 m x 2.25 m [8]. Tests were performed at two Reynolds no.
namely 1.69x105 and 1.13x105. The following figure. 4 represents a sample data acquired during
testing.
5
ax , m/s 2

-5
0 200 400 600 800 1000 1200 1400 1600 1800 2000
0
az , m/s 2

-5
-10
-15
0 200 400 600 800 1000 1200 1400 1600 1800 2000
20
, deg

0
-20
0 200 400 600 800 1000 1200 1400 1600 1800 2000
20
q, deg/s

0
-20
0 200 400 600 800 1000 1200 1400 1600 1800 2000
2000
ea, deg

Aircraft Data
0
Rig Data
-2000
0 200 400 600 800 1000 1200 1400 1600 1800 2000
c (pitch), deg

20
0
-20
0 200 400 600 800 1000 1200 1400 1600 1800 2000
Data Points

Fig. 4 Sample data of Euler angles and angular rates acquired during the wind tunnel tests of
5-DOF rig along with flight vehicle model.
References
[1] S.D. Carnduff, S.D. Erbsloeh, A.K. Cooke and M.V. Cook, “Characterizing stability and
Control of subscale aircraft from wind-tunnel dynamic motion,” Journal of Aircraft, Vol.46,
No. 1, pp. 137-147, Jan.-Feb. 2009.
[2] Robert. C. Nelson, Flight Stability and Automatic Control, 2nd edition, published by TATA
McGraw-Hill companies.
[3] Bandu N. Pamadi, Performance, Stability, Dynamics and Control of Airplanes, Published by
AIAA, © 2004, 2nd Edition.
[4] D. Greenwell and M. Goman, “Wind tunnel simulation of combat aircraft maneuvers,” Proc.
21st Congress of the International Council of Aeronautical Sciences, No. ICAS-98-3.9.2,
Melbourne, Australia, Sep. 1998.
[5] M.S. Rajamurthy, “Generation of comprehensive longitudinal aerodynamic data using
dynamic wind-tunnel simulation,” Journal of Aircraft, Vol. 34, No. 1, pp29-33,1997.

49
Aerospace Engineering Doctoral Students’ Symposium (AEDSS)
IIT Kanpur, May 12-13, 2014
AEDSS-2014-405

[6] A. Gatto and M.H. Lowenber, “Evaluation of a three-degree-of-freedom test rig for stability
derivative estimation,” Journal of Aircrafts, Vol. 43, No. 6,pp. 1747-1762, Nov.-Dec. 2006.
[7] Naba K. Payeda, Manoj K. Dhiman, A.K.Ghosh, “Aerodynamic Characterization of Scale
Model Aircraft using 5-DOF Dynamic Test Rig,” Proc. 2010 NSBE Aerospace System
Conference, Los Angeles, California, Feb. 2010.
[8]Saderla Subrahmanyam, Sunil Sharma and Ghosh A.K, “Analytical Modeling, Trajectory
Simulation and Control of Guided Projectiles,” Proc. Annual International Conference on
Control Automation and Robotics (CAR), Singapore, Feb. 2011.

50

You might also like