Content uploaded by Himanshu Novis
Author content
All content in this area was uploaded by Himanshu Novis on Dec 12, 2021
Content may be subject to copyright.
Research Paper
IAD design for Spent Stage Recovery
Alok Ranjan1, Himanshu N1, Jay Rathod1, Subrahmanya V Bhide1, Harshit Vats1,
and Deshmukh Vinay Vivek1
1Aerospace Department, Indian Institute of Space Science and Technology,
Thiruvananthapuram
December 12, 2021
Abstract
This paper is a detailed summary of IAD based recovery for spent stage of a rocket. Initially
three IAD designs were selected for study, based on literature, and one shape was selected by
carrying out comparative CFD simulations. For the recovery the possibilities of usage of Grid
fins, different materials for the IAD, Pressurization systems, parachute system and landing
legs are studied. Finally CFD simulations are carried out for shape optimization and based on
this trajectory is worked out keeping in mind the specifications for the smooth and comfortable
landing.
Introduction
The history of Reusable Launch Vehicles dates
back to the 1970s when NASA developed the
partially reusable Space Shuttle which had a
lifespan of nearly 30 years. The motivation
came from the aircraft industry and also the
high cost of access to space. Since then multi-
stage expendable launch vehicles have been the
primary approach to place payloads in orbit. Al-
though being a reliable approach, the high cost
of launch is a challenge yet to be addressed. Sci-
entists attempt to design a fully reusable launch
vehicle which reduces the cost of launch and to
increase the launch rates.
Without compromising on the quality of ma-
terial and resources, ISRO has been churning out
successful missions despite limited budgets. Of
course, ISRO has an advantage over NASA and
other international space agencies in terms of
low manpower costs and infrastructure running
costs. “The cost of manpower that includes the
salaries of the scientists and the cost of mak-
ing a satellite and a spacecraft is much lower
in India when compared to western countries.
Most of the requirements are also outsourced to
local manufacturers who provide finished ma-
terial to ISRO to build satellites at less than
half the price when compared to their foreign
counterparts”(3). A system can be designed
to recover a complete launch vehicle, separate
stages or other components while making access
to space economical. In order to decide which
components to recover for reuse, it is impor-
tant to understand the relative value of different
launch vehicle components that are candidates
for recovery and reuse, while minimizing the
cost of recovery.
Recovery requires atmospheric reentry, de-
celeration, and landing. Reentry of the first
stage of PSLV can be accomplished either
via retro-propulsion, mid-air recovery, aeroshell,
IAD or landing impact attenuation. Realising
the future need for a system which ensures re-
covery, without compromising much of scientific
payload, is realisable and scalable for heavier
payload recovery, IAD is chosen over other com-
petitors as subject for further study. An IAD in
simple terms is an inflatable device which when
inflated or deployed provides sufficient drag to
slow down the body attached to it and ensure
a safe landing. It is better than a parachute as
it avoids the excessive oscillations at supersonic
speeds. There is no standard shape for an IAD,
however axisymmetric shapes are prevalent in
available literature.
Mission Objectives
Our problem statement is as follows ’To study
analyse and design an IAD based recovery sys-
1
tem for PSLV Stage 1’. The following are the
parameters related to the PS1 stage:
1. Stage weight at burnout - 30 Tonnes
2. Separation Altitude- 68.5km
3. Velocity - 1.8443 km/s
4. Dimensions - 2.8m Diameter and 20.3 m
Length
5. Time of separation - 113 s
The mission flow for the recovery system
is that, after the separation, we maneuver the
spent stage suitably to facilitate the IAD infla-
tion.
After this the IAD is inflated and we are target-
ing to achieve a velocity of about 5-10 m/s before
touchdown which will happen using Landing
Legs on a barge. Additionally if the velocity
decrement is not enough since the PS1 stage
is considerably heavy we also keep the option
for addition of a parachute open and study the
possibilities. We also intend to use Grid Fins
during the descent to stabilize the system which
is also studied in the course of this project.
Some literature based survey and comparison
for the material that can be used for an IAD are
also studied. Also the pressurisation systems
are studied and control loops for pressurization
is designed.
Figure 1: First stage of PSLV that we intend to
recover
IAD Shape Selection
Generally, IAD designs can be divided into two
categories.
Trailing IADs : In trailing IADs the decelera-
tor is trailing the payload. The payload and IAD
are usually attached by a towline or a group of
towlines. Trailing IADs include sphere-shaped,
teardrop-shaped, isotensoid shaped, torus and
tension cone, etc.
Attached IADs : In attached IADs the decel-
erator is directly attached to the payload. This
includes attached isotensoid, attached tension
cone, stacked toroid blunted cone, etc. These
types are shown in figure 2.
Figure 2: Different types of IADs (11)
From these shapes the following 3 commonly
used shapes were chosen for further analysis.
•Trailing Ellipsoid
A trailing ellipsoid is a trailing type IAD.
The IAD is connected to the payload using
only one towline. The shape of the decel-
erator is ellipsoid i.e. any cross-section of
this shape along the central axis will be an
ellipse.
•Attached Ellipsoid
The attached ellipsoid is similar to the
trailing ellipsoid. But in this case, the IAD
is directly attached to the payload.
•Attached Tension Cone
The tension cone consists of two primary
fabric components: a flexible shell that re-
sists shape deformation by remaining un-
der tension and an inflated torus. This
torus sustains its internal pressure via a
gas generator or pressure tank.
CFD analysis was done on these three shapes for
different Mach numbers. After the analysis fol-
lowing results were obtained for Mach number
2.2
•Trailing ellipsoid : Cd=0.6939278
•Attached ellipsoid : Cd=0.6462677
•tension cone: Cd=0.631842
This analysis was done keeping the projected
area (113.0973 m2) and thickness(6.0012 m) of
the IADs constant.
From this analysis, it was found that the trail-
ing ellipsoid produced maximum drag values
followed by attached ellipsoid and then the ten-
sion cone. But in determining the shape, drag
is not the only factor. Other requirements are
weight of IAD and complexity of vehicle inte-
gration mechanisms. A considerable part of this
2
weight will be due to the IAD integration mech-
anism. This includes an attachment, storage,
deployment and inflation mechanisms.
Table 4compares weight vectors of different
mechanisms for different types of IADs. If this
weight vector is high then it means that the
weight will be lower.
Figure 3: Vehicle integration mechanism weight
vectors(12)
The ellipsoid decelerator, has a less complex
inflation mechanism since an on board inflation
system is not required. The attachment vec-
tor favoured the trailing ellipsoid mechanism.
The attached type IADs require multiple attach-
ment points, which increases overall complexity
of that system. The deployment mechanisms
favour the heritage hardware of the trailing el-
lipsoid. Trailing ellipsoid IAD has the highest
overall weight vector. This means that the trail-
ing ellipsoid shape will require less weight for ve-
hicle integration. Hence, the trailing ellipsoid
was chosen as the shape for IAD. Since, it
produces more drag, requires less weight and has
relatively simple mechanism for vehicle integra-
tion.
IAD’s working
Inside Pressure
As soon as stage will get seperated from our
launch vehicle and before it reaches very high
speed heading toward earth, IAD will get in-
flated and a particular/required amount of
gas will be injected for doing this so that
Pinside > Poutside but due to heat exchange
with outer atmosphere Pinside may change with
respect to height and we know Poutside must
increase rapidly as height decreases. So, by con-
sidering these factors before Pinside becomes <
Poutside we will again inject some moles of air
inside our IAD to make Pinside > Poutside, and
same thing can be repeated in more stages to
make this process more effective and efficient.
Temperature variations
In the troposphere, the temperature generally
decreases with altitude. The reason is that the
troposphere’s gases absorb very little of the in-
coming solar radiation. Instead, the ground
absorbs this radiation and then heats the tro-
pospheric air by conduction and convection.
Since this heating is most effective near the
ground, the temperature in the troposphere
gradually decreases with increasing altitude un-
til the tropopause is reached. This is the be-
ginning of the stratosphere. In the stratosphere,
the temperature remains isothermal until about
20 km. After that temperature actually begins
to increase with altitude. From temperature of
about -56.5C at 20 km, it increases to -2.5C at
50 km. The reason for this temperature fluctua-
tion is that ozone absorbs the UVB (280-314 nm)
radiation in the lower atmosphere. Higher in
the atmosphere, however, normal diatomic oxy-
gen absorbs the UVC (100-279 nm) radiation.
Once it is absorbed, it is re-radiated at different
wavelengths, thereby warming the stratosphere.
At the top of the stratosphere (about 50 km,
the stratopause), the temperature begins to de-
crease again as the altitude increases. Above the
stratopause, in the mesosphere, thermosphere,
and exosphere harmful gamma rays and X-rays
are absorbed so temperature again increases.
Car airbag system
Airbags work by inflating as soon as the vehicle
starts to slow down as the result of an accident.
They then begin to deflate as soon as the driver
or passenger’s head makes contact with them.
If the airbag was not designed to deflate, then
it would not solve the problem of the sudden
3
backward movement of the head and neck, as
your head would simply bounce off it. All of
this is possible because of a range of sensors and
a small explosion. The airbag includes an ac-
celerometer that detects changes in speed. If it
detects deceleration above a preset speed, which
is greater than normal braking speeds, it triggers
the airbag circuit. The circuit passes an electri-
cal current through a heating element, which in
turn ignites a chemical explosive. This generates
a large amount of harmless gas that rushes into
a nylon bag. The bag, which is packed into a
space behind the steering wheel or on the pas-
senger side dash, inflates. When the driver or
passenger’s head makes contact with the bag, it
begins to deflate with the gas escaping through
small holes around the edges of the bag. By the
time the vehicle has come to a full stop, the bag
should have completely deflated. Similar process
is being used in our IAD for inflation purpose
i.e., injecting gas inside IAD to increase inside
pressure. But instead of single stage our process
will be in multiple stages.
Chemical Reactions Used to Generate the
Gas
Pressure controller
Figure. (4), shows a pressure controller which
will help us to keep IAD’s inside pressure more
than outside by injecting extra required moles
of gas inside IAD. After discussion and consid-
ering aerodynamics and structural factors we
took a logical assumption i.e., ∆P=Pinside −
Poutside = 0.25. According to this design we pro-
posed an active closed loop pressure controller
which will take ∆Pinput from the pressure sen-
sor and will run based on the conditions shown
in figure. (4). By doing so we will be able to
keep our IAD inflatted throughout its journey.
In figure. (4), injector is a subsystem which will
inject or remove gas inside IAD as instructed by
the controller.
Figure 4: Closed loop pressure controller
Grid Fins
Grid fins are a type of flight control surface used
in rockets and missiles. Grid fins have often been
used on conventional missiles and bombs such as
the Vympel R-77 air-to-air missile; the 3M-54
Klub family of cruise missiles. In 2014, SpaceX
tested grid fins on a first-stage demonstration
test vehicle of its reusable Falcon 9 rocket.
Guidance with grid fin is achieved by deflect-
ing the fins from their original positions. Doing
this, relative air flow changes its direction and
aerodynamic lift force is generated. A grid fin
rotates about its rotation axis, which is referred
to as "hinge line", with the aid of an actua-
tor motor aligned with the fin centerline. As
a result, guidance commands coming into fin
actuators create a rotational motion about the
hinge line of each grid fin.
The flow field around lattice surfaces of a grid
fin is classified by flight regimes which are distin-
guished with well-defined Mach number thresh-
olds. Transonic flight with grid fins primarily
suffers from high drag force. The reason for
elevated drag is the internal flow field charac-
teristics within the cube-like cells of grid fins.
In the transonic regime, the flow inside the cells
is choked and flow in upstream of the fins de-
celerates. This means that grid fins act as an
obstacle to the flow so that drag force increases.
4
Figure 5: Flow patterns in different flight con-
ditions
Primary advantage of using grid fins comes
with significantly low magnitudes of hinge mo-
ment and minimal variation of center of pres-
sure with changing flight conditions. On the
other hand, grid fin is likely to be subject to high
axial force for large number of lattice surfaces;
however, it is possible to deal with this by care-
ful web and frame design. Efforts on reducing
the transonic drag penalty are becoming popular
and this issue is currently being the main con-
cern related to grid fin applications. Drag force
on a grid fin at transonic speeds increases due
to choking for Mach numbers below unity and
shock reflections for those larger than unity. Lo-
cally swept grid fins, where sweeping is done for
each individual cell, were validated as an effec-
tive tool for supersonic drag reduction, employ-
ing the sharp corners that diminish the strength
of shock waves.
Geometric parameters
Figure 6: Grid fin geometric parameters
The above figure shows all the geometric param-
eters that results in a well-defined grid fin that
has flat planform and blunt leading and trailing
edges. Those parameters related to overall di-
mensions of grid fins are called as span (s) total
width (b) and chord (c). The space occupied by
the grid fin is driven by these three parameters.
Width (w) and thickness (t) are those describ-
ing the dimensions of the individual cells. In
addition to these, frame thickness (tf), defines
the dimension of the outer frame, which is can
be utilized as the structural elements to keep
the grid fins cell together.
The approximate relation between diameter (D)
of the launch vehicle and the parameters defin-
ing grid fins geometry are :
s= 0.416D
b= 0.333D
c= 0.118D
Accordingly , for our case, span of the fins can
be taken as 1.1667 m, total width as 0.9324
mand chord as 0.3304 m
Behaviour of grid fins in different
flight regimes
Use of grid fin in different flight regimes results
in distinctive flow characteristics which usually
determine the constraints of the design efforts.
In general, Subsonic(M <0.7)and high super-
sonic (M >2.0) regimes are admitted as the best
performance conditions for grid fins. Literature
indicates that subsonic operation of grid fins is
never seen as problematic because of absence of
complex flow particular to this regime. Aero-
dynamic behavior of the grid fins is predictable
with no special attention to flow field character-
istics.
For the transonic flight, depending on the area
ratio and free stream Mach number, it is possi-
ble to observe choking, when local Mach number
at any region within the grid fin cell reaches to
unity. For a specific Mach number, choking is
observed below a certain ‘critical area ratio’.
This threshold can be calculated by the equa-
tion given below (air as the fluid) for any free
stream Mach number.
A
A∗=1
M∞"5
6(1 + 0.2M2
∞)#3
(1)
A grid fin in supersonic flight is exposed to
shock waves in different patterns, depending on
the Mach number of the flow. For the Mach
numbers that are slightly larger than 1.0, a bow
shock occurs in front of the leading edges of
the grid fin. At a certain Mach number value
shock waves become attached to leading edge.
After this threshold, oblique shock and rarefac-
tion waves are observed in case of any change in
5
direction of the air flow.
The grid fin flow regimes are separated by def-
inite Mach number thresholds, called ’critical
Mach numbers’ (Mcr 1,Mcr2,Mcr3).
Flow regime Lower bound (M∞) Upper bound (M∞)
Subsonic 0 Mcr1
Choked flow Mcr11
Bow shock 1 Mcr2
Reflecting wave pattern Mcr2Mcr3
Non-reflecting wave pattern Mcr3∞
From literature, the approximate values of
Mcr1,Mcr2and Mcr3are 0.8, 1.4 and 1.9 re-
spectively. So, after all these studies it is clear
that the grid fins should be used in the subsonic
regions because it does not require any kind of
shock formation, and shock formation can very
much interfere with the whole body.
Fundamental consideration in de-
sign of a grid fin
Design procedure aims to arrange an effective
configuration of grid fins in terms of aerody-
namic stability and control. This is ensured
by creating sufficient amount of moment about
center of gravity of a missile. In this design pro-
cess, longitudinal static stability and pitching
moment control are of interest.
In general, the stability parameter, which is
change in pitching moment coefficient with re-
spect to angle of attack, Cmα, has a negative
value for stable missiles. The higher magnitude
of this parameter, the stronger static stability.
Logitudinal control of the stage is performed
by means of pitching moment created by de-
flection of grid fin control surfaces. The change
in pitching moment coefficient with respect to
elevator angle, Cmδe, is linear in nature within a
few degrees of rotation. By definition, positive
deflection angle yields a positive moment.
Control characteristics is generally influenced
by Cmαand Cmδeparameters. A new parame-
ter, pitch control effectiveness, in introduced as
α
δe=Cmδe
Cmα, which expresses, roughly, how much
deflection angle is required to obtain a certain
angle of attack value. According to literature,
this value should be greater than 1.
Overall dimensions of the design should be taken
as one of the considerations in evaluation of the
design alternatives. Design efforts should also
aim at a manufacturable, utilizable and main-
tainable solution.
Neglecting heat transfer and assuming constant
geometry for the grid fins, we can determine few
simple dimensionless parameters that influence
the performance characteristics.
CN=FN
0.5ρv2
∞L2
∞
=f1(Re, M∞, α, δ)
CA=FA
0.5ρv2
∞L2
∞
=f2(Re, M∞, α, δ)
Since, grid fins is also used at the time of re-
entry of the first stage, therefore a material with
very high heat resistance should be used in its
manufacturing. Generally, titanium is used
for this purpose. The mass of grid fins used in
launch vehicles is approximately 30-40 Kg.
Landing Subsystem
Study on Landing Leg Configura-
tions
The major components involved in the landing
subsystem are Legs and their energy absorbing
capabilities. In general, the configuration of legs
for landing, are classified into two :
1. Fixed Leg Configuration
2. Deployable Leg Configuration
Fixed Leg Configuration
As the name suggests, the fixed legs are at-
tached to the body before liftoff, and remain
attached throughout the flight duration. Usu-
ally, the number of legs attached are four and
they are symmetrical. In figure 8, the overall
scheme of fixed legs and schematic representa-
tion of damper is shown. The overall scheme
consists of the retraction system and attenua-
tion system. In which, piston and cylinder are
involved in retracting, and hydraulic or honey-
comb damper as well as foot pad is involved in
attenuating. Typical bearings are used with pis-
tons to complete the overall basic structure.
6
Fixed Leg Configuration
(a)
(b)
Figure 8: Fixed Leg Configuration
Deployable Leg Configuration
In contrast with the previous one, this configu-
ration allows the legs to deploy out during the
landing of the recovering stage. Mechanism of
retracting and deploying the legs can be con-
trolled. In figure 9, retracted and deployed state
of overall scheme, as well as retraction pro-
file and deployment profile of landing legs are
shown. Usually, the auxiliary legs have constant
length and the main landing legs have three
stages of struts and lockers involved, as can be
seen in figure 10.
(a)
(b)
Figure 9: Deployable Leg Configuration
Selection of Feasible Leg Configuration
Figure 10: Chosen Deployable Leg Configura-
tion
After the literature survey, the performance of
fixed and deployable leg configuration, are found
to be approximately equivalent for particular
kind of recovering body. Hence, to choose a spe-
cific configuration, the following aspect can be
emphasised. For the fixed leg configuration, the
legs are attached to the body throughout the
flight interval, due to which, during ascent, the
legs can act as an extra drag creator and even
can lead to shock formation, which catastrophes
the whole objective of landing. Whereas, the use
of deployable leg configuration do not affect the
7
flight in terms of extra drag creation, since the
legs are well attached to the surface of the body
and even are aerodynamically designed.
Therefore, the Deployable leg configuration
is chosen for the landing purpose in this project,
for which, the overall architecture is shown in
figure 10, and corresponding analysis is per-
formed in further sections.
Geometric Topological Model of
Landing Legs
The overall scheme includes four landing legs
symmetrically arranged. Each leg contains a
retractable pillar with the absorber, an auxil-
iary shell, and a footpad. Among them, the
retractable pillar is composed of a three-stage
sleeve. During the launching, the sleeve is in a
retracted state. The landing legs are wrapped
and protected by an auxiliary shell closely at-
tached to the surface of the vehicle body by a
locking device. While ready for landing, the
retractable pillars are elongated. The entire
mechanism is unfolded around the auxiliary pil-
lar hinge point, and the sleeves are locked by
built-in locks when in place. The landing en-
ergy is absorbed by the dampers.
2D Topology with Coordinate System
A coordinate system called O-XYZ is shown in
figure ??. O at the center of the body’s bottom
surface. The positive direction of the OY axis
is upward along the axial direction of the rocket
body. The OX axis is located on the plane of
symmetry of the vehicle whose positive direc-
tion points to the left. The landing legs are
symmetrically distributed in the three dimen-
sional space. Projecting the landing legs in the
XY plane. The deploying and retracting two
dimensional topology map of the left landing leg
is shown in ??, which shows three topological
states of the legs in the release process before
landing: Triangle ABE (State 1: the state of
landing leg locked on the surface of the body),
Line ABD (State 2: the state of the retractable
pillar fully retracting) and Triangle ABC (State
3: the state of landing leg fully expanding).
Lmax,Lmin represent the maximum (State
3 (BC)), minimum length (State 2 (BD)) of a
retractable pillar, Lris the length of the re-
tractable pillar when the landing leg is fully re-
tracted (State 1 (BE)). Within the coordinate
system O-XYZ, the coordinates of point A, point
B, and point C can be given as (x1,y1,z1), (x2,
y2,0), and (x3,y3,0). The relationship of y1 and
z1 can be expressed as:
y2
1+z2
1=R2
where R is the radius of the vehicle body,
which is 1400 mm.
According to the geometrical relation in the
figure ??, the length of the pillars in each state
can be obtained as follows (14) :
Lmax =p(x3−x2)2+ (y2−y3)2
La=p(x3−x1)2+ (y1−y3)2
Lmin =La−p(x2−x1)2+ (y2−y1)2
Lr=qp(x1−x2)2+ (y1−y2)2−L2
a+ 2Lap(x1−x2)2+ (y1−y2)2cos (b)
where,
b= arcsin ((x2−x1)La)
α= arcsin Lasin (b)p(x2−x1)2+ (y2−y1)2
Simplification of Deployable Strut Model
To facilitate the analysis of the length of the re-
tractable pillar, the pillar is simplified as follow:
The length of each sleeve is equal and ’L’. Each
lock is of the equal length L1. The upper of
the first stage sleeve is provided with a cap, and
its length is L0. The simplified model of the
retractable pillar is shown in ??, where Ldis
the length of the damper.
The minimum and maximum length of the
structure can be written as:
Lkmin =Ld+L+ 2L1+L0
Lkmax =Ld+ 3L+L0
To meet the requirements for deployment,
8
the minimum length of the structure should be
less than or equal to the overall minimum re-
quired length, which can be denoted as:
Lkmin ≤Lmin
The design margin of the retractable pillar
length is defined as follows:
e=Lmin −Lkmin
Mechanisms Involved
When the rocket is launched, the landing gear
struts are in a retracted position, close to the
rocket body, and do not affect the rocket launch.
When the rocket lands, the strut is extended
and locked to 135 ◦. The landing gear strut is
composed of a rocket body, a new type three
stages of struts of pneumatic extension mecha-
nism with locking function, a shock absorber, a
landing support with a rocker arm and fairing
shell. The hydraulic shock absorber device is
designed fixed with the third stage strut. The
design principle is based on the compressibility
of the fluids. Its basic components are an outer
cylinder filled with liquid, a piston rod, and a
valve or a special oil limit head. When the pis-
ton begins to move from zero load, the liquid is
compressed by the piston rod which gradually
occupies more volume. A valve is mounted on
the piston rod to open during the compression
of the fluid and to close to limit fluid flow when
it is bounced. The seal of the outer cylinder is
realized by the sealing of the inner wall of the
outer cylinder and the piston rod. Mechanical
Procedure of absorber can be given as follows:
the hydraulic absorber device is composed of
a piston rod, a piston head, outer cylinders of
damper,a sealing device, and a screw fasten-
ing device. As the landing struts contact the
ground, the piston cylinder receives the counter
force of the ground and moves upward. Then,
the oil in the oil tank is compressed and moves
to the oil chamber through the damping oil hole,
and the damping force of the oil is generated.
When the piston rod moves to the limit position
in the outer cylinder, all the energy of the car-
rier is absorbed by the buffer. The energy stored
behind the buffer forces the piston rod to move
back and forth, and finally the carrier’s attitude
is restored.
Dynamic Equations for Hydraulic
Absorber
The axial force of hydraulic absorber F consists
of damping force of oil (fh), oil spring force (fa),
internal friction force (ff)and structural restric-
tion (fs).
F=fh+fa+ff+fs
where,
fh=
ρA3
h˙
S2
a
2C2
dA2
+
˙
Sa≥0
−ρA3
h˙
S2
a
2C2
dA2
−
˙
Sa<0
fa=AaP0V0
V0−AaSaγ
−Patm
ff=µmfa
˙
S
|˙
S|+µb(|Nu|+|Nl|)˙
S
|˙
S|
fs=
Ks.S S < S0
0S0≤S < Smax
Ks.(S−Smax)S≥Smax
The interpretation of involved parameters are as
follows : ρis the density of the oil, Ahis the
compressed oil area, A+is the orifice of the com-
pression stroke, A−is the orifice of the extension
stroke, Cdis the coefficient of contraction, Aais
the area of compressed air, P0is the initial air
pressure, V0is the initial air chamber volume,
Patm is the atmospheric pressure, and γis the
gas change index, Sais the stroke of the piston.
µbis the coulomb friction coefficient, Nuis the
pressure generated by the upper support points
of the piston rod and buffering strut, Nlis the
pressure generated by the lower support points
of the piston rod and buffering strut. S0is the
initial stroke of the piston, Smax is the maxi-
mum stroke of the piston, and Ksis the contact
stiffness between the piston and the envelope.
The damper is considered to have a single
air cavity. When the external force pushes the
damper, the oil in the chamber is squeezed into
the gas cavity. The oil flows through one or more
small holes, which dissipates energy in the form
of heat energy. Meanwhile, part of the energy is
stored by the compressed air.
Analysis
Geometric Parameters
The standard values of following parameters are
considered from the literature : L0= 150 mm,
L1= 180 mm, Ld= 550 mm. For simplification,
values of some parameters are assumed such as
e = 0, Lkmax =Lmax , each of the three strut
length to be L = 800 mm, distance between A
and B to be 1000, distance between horizontal
components of A and B to be 100 mm. These
9
values are considered by physical appearance of
the overall structure. By Artificial Iterations
of all the expressions described in section geo-
metric topological model, the suitable geometric
parameters are obtained as follows:
Lα= 2860 mm, Lmin = 1860 mm, Lmax =
3100 mm and α= 54.88 ◦. Hence, Auxiliary
legs have constant length of 2860mm and main
landing legs (which consists of three retractable
struts) have length varying from 1860 mm to
3100 mm which takes the minimum value at α
= 54.88 ◦.
Mass Evaluation : The dimensions of the
main and auxiliary legs can be considered as
3100mm and 2860mm respectively, with mean
diameter of around 0.3m and thickness of 0.02m,
combination of all 4 legs results into volume of
0.2111 m3and with density of 2250 kg/m3, leads
to mass addition of around 475 kg to the overall
system.
Feasibility Analysis
The following assumptions are considered : All
the 4 legs touch simultaneously down. For a
standard hydraulic absorber, the required pa-
rameters are taken from literature, such as, den-
sity of oil ρ= 960 kg/m3, contraction coeffi-
cient cd= 0.8, initial pressure P0= 1 MPa, etc.
Energy conservation is carried out for transfer-
ring the overall kinetic energy of the system to
compression capability of absorbers. Since, all
legs are assumed to touchdown simultaneously,
summation of axial forces experienced by the ab-
sorbers must be less than the impact force gen-
erated by landing. Let the horizontal velocity of
the system at landing is 0.5 m/s with neglected
pitch angles. Hence, for maximum allowed com-
pression of 0.3 m for hydraulic absorber, the
maximum allowed vertical touchdown velocity
obtained is around 2 m/s.
Parachute and Material stud-
ies
A parachute is a flexible device who’s primary
purpose is to produce DRAG. The thin canopy
material and slender suspension lines allow a
big device to be packed in small volume. the
purpose of the parachute is to create large drag
area which in turn creates a large drag.
Types of Parachutes
•Solid type canopy : solid type canopy
is fabricated mainly with cloth materials
and it is relatively easy to made. these
type of parachutes have no opening other
than vent.
•Slotted type canopy: These Type of
Parachutes with canopies fabricated from
either cloth materials or ribbons. These
types of parachutes have extensive open-
ings through the canopy in addition to the
vent. the slotted type canopy parachutes
can be expensive to manufacture, these
type of parachutes are most commonly
used in planetary exploration missions.
Table 1: Parachute types and Cd ranges
Type of Parachute Cd range
Disk gap band 0.52
Ring sail 0.75
Conical 0.75-0.9
Tri conical 0.8-0.89
Circular slotted 0.75-0.8
Aeroconical 0.635
Conical ribbon 0.5
Parachute system components
Recovery system involves more than one
parachutes. The first thing to deploy is the
drogue parachutes, the drogue parachute is de-
signed to be strong so that it can withstand the
high velocities in heavy loads of initial deploy-
ment. The drogue parachute provides some ini-
tial drag and stabilizes the system so that main
parachute could come out from the packed de-
vice cleanly. The drogue parachute is attached
to the main parachute bag with a long towline,
The main canopy is inside the bag and it come
out after 10 to 15 second after line is cut along
the drogue to pull the main parachute bag out
of the payload. The main canopy is designed
to provide full deceleration of the payload and
provides the soft landing of the system. The sus-
pended line attached to the payload with weight,
weight is transferred to the suspension line and
suspension line transfer the load to the canopy.
Mathematical model for drag force
Drag and Weight act on the parachute as it falls
through the atmosphere , when the parachute
deploys, it creates more drag area, and produces
more drag. According to 2nd law of motion addi-
tion of more drag results in a lower acceleration.
When a parachute of mass m is falling under
the action of gravity, it is subjected to two dis-
tinct forces: weight mg, and the drag force, FD.
10
Applying Newton’s second law of motion in the
vertical direction yields the following differential
equation for the motion of the parachute:
mdV
dt +Fd−mg = 0 (2)
Fd=1
2ρV 2CDAref (3)
where, V is the vertical speed, t is the time, and
g is the acceleration of gravity.
When the body reach at equilibrium position,
acceleration becomes zero.
Drag F orce =weight 1
2ρV 2CDAref =mg
(4)
Selection of parachute and material
circular slotted parachute which we have chosen
to use for our mission . From the paper(16) we
found that the circular slotted parachute has less
opening shock and also the minimum size for the
same payload amongst the various parachutes.
performance and drag coefficient between the
circular slotted canopy and ringsail canopy are
nearly same but Circular slotted parachute is
easy to manufacture whereas ringsail requires
high skill to maintain the sail dimension, band-
gap, and porosity of the overall parachute. Ny-
lon /kevlar hybrid material can be use for main
parachute.
Parachute design
Main parachute design
In our project we are getting 400m/s at 5km
altitude using the IAD, dynamic pressure will be
around 58,864P a, which is not feasible for open
the parachute. If we can use any other method
to reduce the velocity upto around 160m/s, then
we can use the parachute system for recovery.
The calculation will be done for
160m/s velocity at 4km altitude
The size of the cluster is determined by cal-
culating the total effective drag area required.
qe=1
2ρ(z)V2, ρ(z) = ρ0(1 −22.57.10−6Z)4.256
(5)
Considering that in equilibrium descent the to-
tal drag of the system is very nearly equal to its
weight, the required effective drag area can be
obtained.
CDS0=W
qe
(6)
dynamic pressure at equilibrium for terminal ve-
locity 20m/s at altitude 230mis 238pa W =
30300kg so
(CDS0)1,2,3= 415.88m2
d0=r4.S0
π= 26.57m
to get the terminal velocity 20m/s , diameter of
the all three main parachute is 37.57m// per-
centage of drag loss is less than 5% for a cluster
of three parachutes.
Opening force
For reefed parachute
Df=ρ
2lfbdf bV2cfsf
Df= 4039.4N
¨x=∆v
∆t=(117 −31)
234 −220 =−6.1m/s2
(CDS0)R=2(m(g−¨x)−Df)
ρv2= 63.30m2
(S)0= 84.4m2
(D0)R= 10.36m
tf=nD0
v(CDS)R
CDS)00.5= 0.864second
F(X)reef ed = (CDS)R.qR.CX.(X1)R= 501579N
F(X)Disreef ed = (CDS)D.qD.CX.(X1)D= 163501N
Opening coefficient can be consider as CX=
1.05,(X1)R= 1,(X1)D= 0.4
Material Selection for IAD and
parachute
Materials that are used in the construction of an
IAD must be flexible, lightweight, and strong.
In addition, the materials must withstand tight
folding, temperature extremes ultraviolet (UV)
degradation, vibration, and other harmful en-
vironmental elements. There are many light-
certified flexible materials are available. kevlar
29 is one of the good material that is used in
many space missions.
Kevlar is an excellent material for primary
load-carrying members such as suspension lines,
and it reduces the parachute’s weight and vol-
ume. The drawbacks of Kevlar are its loss in
strength due to moisture absorption or smog
storage environment; it is not as resistant to
abrasion, and it is slightly more difficult to man-
ufacture compared to nylon. Another drawback
of Kevlar is its higher cost, about a factor of 3
greater than nylon. So we are using the nylon
11
hybrid material for IAD , because it is cost effec-
tive and exhibits good strength to weight ratio.
Recovery using parachute is not feasible for our
mission because it cannot open at high velocity.
it can be operate when load velocity can be re-
duced by any other method like retro thruster.
when it opens at high dynamic pressure it can
break the IAD and suspension line, also if we
use multiple cluster parachute then the weight
of the system will increase.
From basic calculations the weight of
the main canopy (nylon) is obtained as,
Wcanopy = 0.0598(D0)2.046 = 49kg, W3canopy =
147kg approx
Shape Optimization
Aerodynamic analysis is a vital element in the
present study for the design of the IAD for
spent-stage recovery. The main use of IAD is to
decrease the peak dynamic pressure during de-
scent. This is done by decelerating the stage in
the upper atmosphere. For this the IAD should
augment the drag of the stage sufficiently, but
should not add significant mass to the system.
A fixed aft mounted IAD has been proposed by
ISRO for stage recovery studies. Another as-
pect is the study the stability characteristics of
the stage. The configuration is expected to be
stable because of the high drag producing IAD
at the rear end. Hence, in the present study, the
aim is to obtain the drag characteristics of the
stage with trailing IAD and study the flow-field
around the two bodies.
Extensive CFD simulations have been car-
ried out to estimate the required drag. 2-D ax-
isymmetric simulations are carried out for the
purpose of characterization since the IAD and
stage are axisymmetric. As the aerodynamic
data generation is very compute and time in-
tensive and is required for a very large number
of configurations during the design optimization
studies.
Mesh Generation
The very first step in CFD analysis is discretiza-
tion, which involves splitting the domain of in-
terest into smaller sub-domains and cells. This
process of discretization is called Mesh genera-
tion or meshing. The current project utilises the
Pointwise software for meshing. The mesh qual-
ity determines solution accuracy, rate of conver-
gence and also time required for computation.
A good mesh should ensure all the expected
flow features are captured. The domain chosen
for analysis is a square of size twenty times the
IAD maximum diameter to ensure negligible up-
stream influence of body. Structured meshing
is undertaken throughout the domain owing to
the relatively less complex geometry, higher ac-
curacy, better convergence and lower memory
requirement. To ensure that all the flow fea-
tures are captured appropriately, specially near
the boundaries of the body, fine meshing (initial
cell height of 1e-6m)is done within 0.05m of the
body boundary to capture the resulting bound-
ary layer. The mesh quality is maintained by
ensuring that the area ratio is within 3.0 and
equiangle skewness within 0.85. Figure (13),
shows an overview of domain where cis the the
distance of IAD from stage and aand bare the
IAD geometry parameters.
Figure 13: IAD geometric parameters
12
Computational results
We have tried varying our design parameter i.e.,
a,band cat fixed mach number 3.2 and with re-
spect to each other it was observed that increase
in awas reducing drag while increase in bwas
contributing to increase net drag. While varying
cit was observed that drag was first slowly in-
creasing and suddenly reached a peak value and
got saturated just near the peak value. Further,
for trajectory design as required we had taken
a reference height i.e., 25km and we have varied
all three parameters individually and simulated
with respect to different mach no.
Figure 14: Cd vs Mach no. (a=5.5, b=5.8 and
c=11.8)
Figure 15: Cd vs Mach no. (a=5.5, b=7 and
c=11.8)
Fig. (14) and (15), shows a set of results.
In this way we have taken some set of outputs
manually to feed into trajectory equations from
where we were getting a feedback of requirement
of drag and according to that we have varied de-
sign parameters keeping structural and stabil-
ity factors in mind. For example, increase in b
was giving us more drag which was a favorable
thing but our stability of structure was reducing
which was not in favour. Fig. (16) shows veloc-
ity profile for a simulation at 3.2 mach no. which
clearly shows primary and secondary shock for-
mations as well.
Figure 16: Velocity profile for a simulation at 3.2 mach no.
Trajectory Design
The main elements of the trajectory design work
done are:
1. Trajectory Estimation without IAD
2. Thruster selection/design for pitch maneu-
ver
3. Trajectory Estimation and Optimization
with IAD
Governing Equations
We use the 3 kinematic equations and 3 dynamic
equations to simulate the trajectory of the Spent
Stage.
˙r=vsinγ (7)
˙
ξ=vcosγcosζ
rcosϕ (8)
˙
ϕ=vcosγcosζ
r(9)
The equations 7,8and 9are the kinematic equa-
tions.
13
˙v=F cosαT
m−µE
r2sinγ −D
m+ω2
Ercosϕ(cosϕsinγ −sinϕcosγsinζ)(10)
˙γ=F sinαT
mv −µE
r2vsinγ +v
rsinγ +L
mv +cosϕ(2ωEcosζ +ω2
Er
v(cosϕcosγ +sinϕsinγsinζ )(11)
˙
ζ=−v
rtanϕcosγcosζ + 2ωEcosϕtanγsinζ −ω2
Er
vcosγ sinϕcosϕcosζ −2ωEsinϕ. (12)
The equations 10,11 and 12 are the 3 dy-
namic equations. Here r is the distance from the
center of the Earth, v is the velocity magnitude,
ϕis the lattitude, ξis the longitude, zeta is the
heading angle and γis the flight path angle. The
axis system for the above formulation is shwon
in the Fig. 17.
Figure 17: Planet-fixed and local horizon frames
for atmospheric flight.
Since the Spent stage doesn’t produce any
Lift and also there is no Thrust produced, the
two terms L and F can be made equal to zero.
The Drag ’D’ can be found by the expression 13.
D=CD
1
2ρv2Sref (13)
The term ωEis the angular speed of the Earth
which is equal to 7.2925 ×10−5rad/s. The term
µEis the standard gravitational parameter for
Earth which is equal to 3.986×1014 m3/s2. The
term r is the sum of the altitude and the radius
of earth taken to be 6378 km.
Trajectory Estimation without IAD
The Spent stage will follow a parabolic trajec-
tory. This trajectory needs to be estimated
initially so as to give us an initial idea regarding
the timescale of the ascent so that suitable pitch
maneuver can be designed.
This is done using the trajectory equations from
the previous sections. The spent stage can be
assumed to be a cylinder, the CDfor a cylinder
in cross flow is known but in our case we require
an initial estimate for the CDfor a cylinder
with the flow along the axis of the cylinder. As
an initial guess we we take the CDvalues esti-
mated for a fuselage which is almost similar to a
cylinder but except the fact that the nose of the
fuselage is aerodynamic but in our case it will
be flat. Thus we conservatively estimate the CD
to be equal to 0.02. Using this in the trajectory
equations we obtain the trajectory and velocity
plots as shown in Fig. 18 and 19. (2)
Figure 18: Trajectory estimate for the spent
stage without IAD
Figure 19: Velocity estimate for the spent stage
without IAD
From the plots we can make the following
observations,
1. The peak of the parabolic trajectory is at-
tained after around 120s after separation.
2. The minimum velocity attained during
this phase is around 1400 m/s.
14
Thus we have sufficient enough time for the pitch
maneuver to be executed after which the IAD
can be inflated at the top most point. This pitch
maneuver and maintaining it as such is essential
since the descent we require is vertical and only
1 IAD is used.
Thruster design for pitch maneuver
The pitch maneuver is essentially rotating the
spent stage such that it becomes perpendicular
to the local horizon. The pitch maneuver can be
executed by using 4 cold gas/mono-propellant
thursters. These thrusters provide thrust of the
order of 10-100 mN using which the pitch ma-
neuver can be executed in around 10-20 sec. The
timing of this maneuver can be either of the
three following options,
1. After the separation occurs (0-40 secs)
2. Somewhere in between (40 - 180 secs)
3. Just before the maximum height is reached
(80 -100secs)
If the maneuver is done just after separation or
somewhere in between, then for a considerable
amount of time the velocity direction and ori-
entation won’t be aligned and thus would cause
moments on the stage. Hence a better option
is to execute the maneuver just before the top
most point so that after the maneuver the IAD
can be inflated.
This maneuver would involve firing of thrusters
to make the spent stage vertical and thus suit-
able for inflating the IAD. Small Cold gas /
mono propellant thrusters can be considered for
this purpose.
In a general understanding Cold gas propellant
systems are much simpler and less complex than
mono-propellant systems. But along with this
Cold gas systems have lesser specific impulses
too. The variation of propulsion system mass
and Total impulse is given in Fig. 20. From
the figure it is evident that for lower total im-
pulses required cold gas propellant system have
lower mass than mono-propellant system. But
at higher total impulse Mono-propellant would
have lesser system mass
Figure 20: variation of propulsion system mass
and Total impulse
Cold gas propellants mostly have Specific
impulse lesser than 100 s, where as for mono-
propellant systems hydrazine can provide a spe-
cific impulse of about 230 s. A mono-propellant
system would require some additional compo-
nents if it is a pressurised system. If it is a blow-
down system then the mass and complexity of
the mono-propellant would almost be the same
as that of the cold gas system. Before finalizing
the propellant system type we would want to
have an idea of the range of thrust required for
the maneuver. For this we solve the two point
boundary optimization problem. We know the
initial pitch angle of the spent stage and assum-
ing a angular rate of zero, we need to reach the
final point of 90 degree with the horizontal and
zero angular speed. The thrust that we apply is
the input and this has to be minimized. Employ-
ing the fmincon of MATLAB to minimize thrust
given the boundary conditions we arrive at a lin-
ear thrust profile that would be required. The
initial angle is around 45◦considering some dis-
turbance we assume the initial angle to be 40◦.
If the maneuver has to be completed within 14
secs we would require a maximum thrust value
of 0.8N. This can be seen from the Fig. 21 and
22. For the worst case scenario if the IAD has
to be turned from −90 to 90 we would require a
thrust of 2.9N. This can be seen form the Figs.
23 and 24.
Figure 21: The variation of Angle with time for
the pitch maneuver
15
Figure 22: The variation of Thrust with time for
the pitch maneuver
Figure 23: The variation of Angle with time for
the pitch maneuver in the worst case scenario
Figure 24: The variation of Thrust with time for
the pitch maneuver for the worst case scenario
Therefore now we search for cold gas and
mono-propellant systems which can provide
thrust is similar range that is in the order of
1N to 5N and compare their properties. We will
compare the MOOG MONOARC-5 and MOOG
GN2058-118 cold gas propellants. A compar-
ison table made 3. From the table we can see
that in terms of mass, dimensions Cold gas pro-
pellant has an upper hand but it requires more
power to operate than the mono-propellant sys-
tem.
Table 3: Comparison of Cold Gas and Mono-propellant system
Property Cold Gas System Mono-propellant system
Nominal thrust (N) 3.6 4.5
Specific Impulse (s) 57 226
Power 30 18W
Mass 23g 490g
Dimensions 6.6 mm, 25.4 mm 41.8 cm, 2.5 cm
Considering the fact that PSLV is a working
LV and the recovery system must cause as less
payload penalty as possible it will be better to
chose for the cold gas system rather than the
heavier mono-propellant system. Also we would
require control systems to control the Thrust in
the fashion as shown in Fig. 22. (1) (9) (10).
Finally we can conclude this section by pro-
viding a mass approximate that is incurred due
to the propulsion system. For the cold gas
thrutser we have chosen, the thrusters have a
mass of 23g and we use 4 thrusters thus we would
have 92g.
Trajectory Estimation and Opti-
mization with IAD
We now employ the trajectory equations and in-
tegrate them using odeint of the Scipy module
which is a wrapper for the Runge-Kutta meth-
ods. The top most point is our start point and
for the integration and is terminated once we
reach the ground. For specific shape of the IAD
the variation of Cd with Mach number is ob-
tained from the IAD shape optimization studies
and later they are employed to obtain the tra-
jectory.
16
First Iteration
For the first iteration IAD shape and design the
variation of Altitude and Velocity is shown in
Fig. 25. This corresponds to the variation of Cd
as shown in Fig. 14
Figure 25: The variation of Altitude and Veloc-
ity
Second Iteration
For the second iteration IAD shape and design
the variation of Altitude and Velocity is shown
in Fig. 26. This corresponds to the variation of
Cd as shown in Fig. 15
Figure 26: the variation of Altitude and Velocity
In both these iterations the lowest velocity
achievable is 400m/s. This value must be re-
duced further so that smooth and proper land-
ing can take place. For this the IAD shape has
to be further optimized. If all the velocity can’t
be killed by the IAD, even for the parachutes
to be able to do it we need the velocity to be
around at least 160m/s. One other option can
be to use retro propellants used during sepa-
ration to reduce velocity but due to this more
mass would be added to the system leading to
payload penalty.
The representational image of the fully deployed
IAD and Grid Fins that we intend to design is
shown in Fig. 27
Figure 27: Representational Image of the system
The trajectory and sequence of events for
the system are shown in the Fig. 28 according
to our initial idea where IAD is sufficient to pro-
vide enough deceleration. If that is not possible
additional parachutes would be required.
Figure 28: Sequence of Events
Conclusions
Based on the studies that could be conducted
within the time frame of the project, the follow-
ing observations and conclusions can be made,
1. The velocity at touchdown for the initial
two simulations is around 400 m/s and this
is very high and at present the chances of
successful design of an IAD for recovery of
PS1 seems slim.
2. IAD shape optimizations can be conducted
further to improve the touchdown velocity.
The maximum Cd we have obtained is in
the second iteration with Cd about 1. This
happens at a mach number of about 1.75.
Optimizations must be done in the direc-
tion shifting the peak towards the Mach
numbers which occur during major parts
of the flight.
17
3. We can also explore the options of using a
retro thruster at around 20 km to reduce
the velocity further but this would increase
the mass of the system.
4. Due to structural stability we can not vary
geometry parameters beyond a particular
limit.
5. Even if we can achieve lower velocities say
in the range 100m/s which would necessi-
tate the use of parachutes, given the mass
of the PS1 stage at least 3 parachutes
would be required. The mass estimate of
the parachute is about 147 kg.
6. Also Addition of Parachutes along with
IAD can complicate the system and the
total integration can pose serious problems
as the IAD is itself trailing.
7. The mass estimate for the Landing legs it-
self is around 475 kg. The Average pay-
load capacity for PSLV is 1500kg. Thus
we will have a maximum payload penalty
component of about 475
40 kg from the Land-
ing legs.
8. Considering the above facts encountered
during the course of our project we are sure
that there will be payload penalties and
also complexities in the integration if both
IAD and parachutes are required. Not
only the mass penalty the highly optimized
PSLV performance will also get affected
due to changes in this. Hence another op-
tion can be, which is the case for current
launch vehicles that exhibit reusability , is
to design a new launch vehicle keeping in
mind the reusablility.
9. The main drawback for the PS1 stage is its
mass as IAD concept has proven success-
ful for sounding rockets and other lighter
structures.
Acknowledgements
We would like to thank Dr. Ayyappan G
(VSSC, ISRO) and Dr. Manoj T Nair (IIST,
ISRO) for their continuous guidance and sup-
port both technical and non-technical during
the project and also for providing this project
for the course. We would also like to thank
Dr.Pankaj Priyadarshi (VSSC, ISRO), Ankita
Hudedagaddi A&M (STS, VSSC) and alumni of
IIST for their guidance for the project.
18
References
[1] Development of a Cold Gas Propulsion System for the TALARIS Hopper, Sarah L. Nothnagel,
2011
[2] Moonan, William C., "Evaluation of the Aerodynamics of an Aircraft Fuselage Pod Using
Analytical, CFD, and Flight Testing Techniques. " Master’s Thesis, University of Tennessee,
2010. https://trace.tennessee.edu/utk_gradthes/823
[3] FirstPost NS, Union Budget 2021, DoS granted funds. https://www.firstpost.com/india/
union-budget-2021-mission-9264471.html
[4] http://arc.aiaa.org DOI: 10.2514/6.2013-1389
[5] PSLV C1 - IRS 1D Mission
[6] DESIGN OF A GRID FIN AERODYNAMIC CONTROL DEVICE FOR TRANSONIC
FLIGHT REGIME, ERDEM DIKBAS
[7] http://arc.aiaa.org DOI: 10.2514/6.2014-1092 Supersonic Inflatable Aerodynamic Decel-
erators for use on Sounding Rocket Payloads
[8] ISRO’S SOLID ROCKET MOTORSt R. NAGAPPA, M. R. KURUP and A. E. MUTHU-
NAYAGAM, Acta Astronautica Vol. 19, No. 8, pp. 681 97, 1989
[9] https://www.moog.com/content/dam/moog/literature/Space_Defense/
spaceliterature/propulsion/moog-coldgasthrusters-datasheet.pdf
[10] https://www.moog.com/content/dam/moog/literature/Space_Defense/
spaceliterature/propulsion/Moog-MonopropellantThrusters-Datasheet.pdf
[11] MORPHING HYPERSONIC INFLATABLE AERODYNAMIC DECELERATOR, https:
//arc.aiaa.org/doi/abs/10.2514/6.2013-1256
[12] Supersonic Inflatable Aerodynamic Decelerators for use on Sounding Rocket Payloads,
Matthew J. Miller, Bradley A. Steinfeldt, Robert D. Braun, (2014), AIAA Atmospheric
Flight Mechanics Conference
[13] Design and dynamic analysis of landing gear system in vertical takeoff and vertical landing
reusable launch vehicle, Ming ZhangDafu, XuShuai, YueShuai, YueHaifeng Tao, journal of
aerospace engineering.
[14] Optimization design containing dimension and buffer parameters of landing legs for reusable
landing vehicle, Bo LEI a, Ming ZHANG a,Hanyu LIN a, Hong NIE b, Chinese Society of
Aeronautics and Astronautics and Beihang University, Chinese Journal of Aeronautics.
[15] Liquid spring damper for vertical landing Reusable Launch Vehicle under impact conditions
Shuai Yue a,Branislav Titurus b, Hong Nie a,Ming Zhang a, Mechanical systems and signal
processing, elsevier.
[16] Design and Selection Criteria of Main Parachute for Re-entry Space Payload Mahendra
Pratap#,*, A.K. Agrawal@, and Swadesh Kumar#,Defence Science Journal, Vol. 69, No. 6,
November 2019, pp. 531-537, DOI : 10.14429/dsj.69.12681
[17] W. Mueller: “Parachutes for Aircraft”. National Advisory Committee for Aeronautics, NACA
TM-450, October 1927.
19
