Autonomous Systems Lab

Prof. Roland Siegwart

Bachelor-Thesis

Parachute Recovery
Systems for Unmanned
Aerial Vehicles

Autumn Term 2009

Supervised by: Author:

Stefan Leutenegger Gerig, Nicolas
Dario Schafroth
Contents

Abstract v

Symbols vii

1 Introduction 1
1.1 Definition of a Parachute Recovery System . . . . . . . . . . . . . . 2
1.2 Background, Goals, Approach and Outline . . . . . . . . . . . . . . . 2

2 Parachute Recovery Systems 3

2.1 PRS Diversity Factors . . . . . . . . . . . . . . . . . . . . . . . . . . 3
2.2 General PRS Design Criteria for light UAV . . . . . . . . . . . . . . 4
2.3 Description of Function and PRS Elements . . . . . . . . . . . . . . 5
2.3.1 Preparatory Mechanisms . . . . . . . . . . . . . . . . . . . . 5
2.3.2 Deployment Mechanisms . . . . . . . . . . . . . . . . . . . . . 6
2.3.3 Parachute Assembly . . . . . . . . . . . . . . . . . . . . . . . 9
2.3.4 Impact Attenuation Assembly . . . . . . . . . . . . . . . . . . 13
2.3.5 Continuative Tasks . . . . . . . . . . . . . . . . . . . . . . . . 13
2.4 Case Studies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
2.4.1 Manned Aircraft, Ballistic Recovery System . . . . . . . . . . 15
2.4.2 Phoenix UAV Recovery . . . . . . . . . . . . . . . . . . . . . 15
2.4.3 Maneuverable UAV Recovery, Experimental Program with
Banshee UAV . . . . . . . . . . . . . . . . . . . . . . . . . . . 16

3 PRS for ASL Solar Airplane 17

3.1 Requirements and Constraints . . . . . . . . . . . . . . . . . . . . . . 17
3.2 Evaluation Criteria . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
3.3 Concepts Compliant with Requirements . . . . . . . . . . . . . . . . 18
3.4 Evaluation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18
3.4.1 Maneuverable Concepts . . . . . . . . . . . . . . . . . . . . . 18
3.4.2 Main Canopy . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
3.4.3 Deployment Mechanism . . . . . . . . . . . . . . . . . . . . . 20
3.5 Preliminary Suggestion and Closer Specifications . . . . . . . . . . . 22

4 Summary and Discussion 23

5 Conclusion 25
5.1 Future Work . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25

A Knacke’s Tables of Common Parachute Decelerators 27

B Knacke’s Table of Common Canopy Material Characteristics 30

C Main Canopy Spread Sheet 33

i
D Timetable 34

Bibliography 37

ii
List of Figures

1.1 Historic Parachute Drawings [1] . . . . . . . . . . . . . . . . . . . . . 1

2.1 Controlled Parachute Deployment by Knacke[2] . . . . . . . . . . . . 6

2.2 Cross-Wind Ejection by Knacke[2] . . . . . . . . . . . . . . . . . . . 8
2.3 Downward Ejecting System by Knacke[2] . . . . . . . . . . . . . . . 9
2.4 Influence of Suspension Line to the Drag Coefficient by Knacke[2] . . 12
2.5 Ballistic Recovery System of Cirrus Aircraft, Wikipedia[1] . . . . . . 15
2.6 Phoenix UAV, 1999-2008[10] . . . . . . . . . . . . . . . . . . . . . . 15
2.7 Banshee UAV Under Parafoil Canopy, Wyllie[3] . . . . . . . . . . . . 16

3.1 Unscaled Deplyment System by Tristan Foon[6] . . . . . . . . . . . . 22

iii
iv
Abstract

Parachute Recovery Systems (PRS) are used today in numerous different tasks and
as components of aerial systems. This literature study categorizes and compares
common used airplane and Unmanned Aerial Vehicle (UAV) recovery system con-
cepts. Advantages and drawbacks are shown in particular for light UAV. Different
deployment methods with pilot chutes and deployment bags as well as spring ejec-
tion are described. Different common canopy shapes are compared and discussed.
Requirements and constraints for the 3kg ASL UAV are defined and a closer look
at applicable concepts is made. Preliminary approaches and calculations led to an
non-maneuverable polyconical parachute of about 1.75m nominal diameter which is
deployed with a spring ejected pilot chute. Without having performed substantial
material research, the preliminary calculations show that it is possible to design a
parachute recovery system lighter than 300g and options for weight optimization
are mentioned.

v
vi
Symbols

Symbols
m mass [kg]
g gravity acceleration [m/s2 ]
ρ density [kg/m3 ]
V velocity [m/s]
S surface/area [m2 ]
D0 nominal diameter [m]
DC constructed diameter [m]
SC ”constructed” suface (material need) [m2 ]
sf filling distance [m]
CD drag coefficient [-]
CX opening force coefficient [-]
φ mean angle of oscillation [◦ ]
Le suspension line length [m]

Indices
x x axis
y y axis

Acronyms and Abbreviations

ETH Eidgenössische Technische Hochschule
ASL Autonomous Systems Lab
UAV Unmanned Aerial Vehicle
PRS Parachute Recovery System
GPS Global Positioning System
SI (units) System International units

vii
viii
Chapter 1

Introduction

The Idea of slowing down falling objects or people to avoid damage is very old.
According to Wikipedia[1] the first known drawing of a parachute like system has
been created during Renaissance in 1470s by an unknown author.
Today the invention of modern parachutes is dated in 1783 when Louis-Sébastien
Lenormand jumped from the tower of the Montpellier observatory. As a first
Parachute Recovery System (PRS) of an aerial vehicle, André Garnerin’s concept of
a safe hot-air balloon car landing is assumed. He has tested his concept on October
22nd in 1797 and made history as the first user of a frameless parachute.

(a) (b) (c)

Figure 1.1: Historic Parachute Drawings [1]: (a) First Known Parachute Drawing,
in 1470s; (b) First Recorded Parachute Jump, 1783; (c) First PRS for an Aerial
Vehicle, 1797;

A lot of parachute development has been done by military and great efforts have
been made in the early 20th century. Due to the fact that parachute development is
strongly bound to knowledge in material science, fluid dynamics and manufacturing
technology, the 20th century brought a huge amount of progress in efficiency and
reliability. Therefore it is nowadays an applicable concept to save complete aerial
systems with parachutes.

1
Chapter 1. Introduction 2

1.1 Definition of a Parachute Recovery System

Parachute recovery systems use a parachute canopy to slow down flying or falling
objects or people to a safe landing. This is done by opening the canopy on demand
in order to use aerodynamic drag. Hence the rate of decent, also called steady state
velocity, is lowered to a designated impact velocity.
Parachute recovery does not implicitely denote an emergency procedure. It may
also be a standard operating procedure, a UAV autonomous landing for example.
There are other systems protecting an object from taking damage on impact such as
airbags or crash pads for example. Stand-alone alternatives to parachute recovery
systems will not be discussed in this report because they misfit our needs in dimen-
sions and weight. Airbags, crash pads and other concepts to support a parachute
in its function will be treated as parts of a parachute recovery system.
In this report the notion parachute recovery system is often used denoting only
parachute recovery systems for aerial vehicles, UAV and objects.

1.2 Background, Goals, Approach and Outline

Autonomous Systems Lab is developing a new autonomous motor glider that will
be equipped with expensive sensors and technology. Therefore there is a need for
a light-weight parachute recovery system. This PRS should allow safe landing in
emergency situations and may also become a standard landing procedure in case of
limited landing space.
Goal of this work is to give an overview of existing Parachute Recovery Systems for
aerial vehicles, especially for light UAV. Common concepts are discussed in section
2.3 and examples are given in section 2.4.
Section 3.1 characterizes the requirements and constraints for the new ASL solar
airplane. In section 3.3 evaluation criteria of different concepts are shown, followed
by an evaluation in section 3.4. This yields to a preliminary suggestion (section 3.5)
for which more detailed specifications are given.
Chapter 2

Parachute Recovery Systems

This chapter contains information about general preliminary considerations when

designing a parachute recovery system, as well as short characterizations of different
concepts with some examples.

2.1 PRS Diversity Factors

In general, there is a large amount of different parachute recovery systems. They
can be categorized in many different ways. Common diversity factors are:

manned vs unmanned aircraft This basic categorization yields to different re-

quirements in opening shocks, impact shocks and tolerances to aerial vehicle
movements (like bottom up landing). The steady state drop rate for manned
aircraft recovery usually is smaller than for UAV recovery and the canopy
opening more complicated.
emergency vs standard operating procedures While emergency procedures are
usually not planned exceptions in aircraft operation, parachute recovery may
also be a standard operating procedure for landing an aerial vehicle. In this
case reusability and non destructive methods gain importance.
destructive vs non-destructive recovery As an emergency operation there are
various concepts that actually damage the aircraft or abandon parts of it
during operation, those are called destructive recoveries. They provide more
liberty in designing, but also provide higher costs and efforts to treat the
aircraft for the next mission.
completely reusable vs non reusable A non destructive PRS is denoted as com-
pletely reusable if no parts of the UAV or of the PRS are being consumed
during PRS operation, for example absolutely no material like pyrotechnics,
rockets or airbag gas has to be refilled after usage.
maneuverable vs non maneuverable A maneuverable PRS has the ability to
control the parachute flight, usually realized by servos that pull suspension
lines, what ends in a change of parachute form like para gliders do. Non
maneuverable systems just lower the drop velocity, horizontal movement is
not controllable.

3
Chapter 2. Parachute Recovery Systems 4

2.2 General PRS Design Criteria for light UAV

It is important to get the requirements and constraints straight before designing a
PRS. Typical Design Criteria for light UAV composed from Knacke’s PRS Design
Manual[2] and Tim Wyllies contributed paper about PRS for UAV[3] are listed
below:
• Safety / Reliability of operation • Maximum System Loading and
Growth Potential (UAV usually
• Stability / Indifference to Dam- gain additional weight during
age development)
• Low Opening Forces • Simplicity of Design and Manu-
• Low System Weight and Vol- facturing
ume • Simplicity of Maintenance and
• Repeatability Service

• Accuracy on pre-determined • Low Acquisition Cost

landing spot • Low Life Cycle Cost
• Automation
5 2.3. Description of Function and PRS Elements

2.3 Description of Function and PRS Elements

To clarify the PRS function, it is split into different smaller tasks. Afterwards the
PRS subsystems are focussed, and described with information found in Knacke’s
Design Manual [2] and Tim Wyllie’s paper [3].
Determinate Time of Activation. Devices such as antennae or control systems
that receive, cause or delay the time of activation are called preparatory
mechanisms.
Parachute Discard and Inflation Starting. All parts like opening flaps, dis-
carding springs, explosives and pilot chutes that care for correct parachute
opening with compliant opening forces are called deployment system.
Lowering Drop Rate or Velocity. This is done by the parachute assembly.
Different parachute canopies or balloon type bodies cause a slowdown due to
their aerodynamical drag.
Impact Attenuation An optional sub function to prepare and protect the aerial
vehicle from ground impact that is necessary if the steady state drop rate is
higher than tolerance on impact speed of the aerial vehicle. There are different
impact attenuation assemblies like air bags or crash pads for example.
Continuative Tasks There are different tasks that may occur after landing. If
the object will not be collected immediately, it is often necessary to cut off
the parachute to avoid winds lifting and dragging around the UAV on ground.
Other tasks could be sending positioning information, sounds, light or smoke
effects to simplify picking up the UAV.

2.3.1 Preparatory Mechanisms

As the PRS is a subsystem of an aerial vehicle, it usually does not need sensors and
electronical control devices by itself, as it may use those that are already built in
the aerial vehicle. However this circumstance has to be reconsidered when defining
the emergency cases.
The simplest event for triggering a PRS might be a remote control command. An-
other factor for starting could be done by an autonomous control system that com-
pares set points with sensor data. Before parachute ejection, there may be time
delays caused by control systems. These control systems could for example try to
control a predefined or dynamically chosen landing spot (using air speed sensors
combined with GPS for it’s calculation). Another control system could check flight
speed (or drop rate) and the air vehicles orientation to ensure parachute inflation.
When designing the preparatory mechanisms, operational conditions have to be
clearly defined.
There has to be a set of emergency cases and or standard operating procedures
defined as part of the requirements of the PRS.
Some Example Events for triggering the PRS could be
- the aerial vehicle leaving a predefined operation area
- engine or rotor damage
- loss of electronic control
- loss of energy
- damage to necessery sensors
and many more.
Chapter 2. Parachute Recovery Systems 6

2.3.2 Deployment Mechanisms

Deployment Mechanisms are all parts of the PRS operating when first action on
parachute ejection took place, until main parachute inflation starts. At first, basic
concepts for deployment mechanisms are shown, afterwards different possibilities to
place an PRS in airplane shaped UAV are given.
Beside providing the preparation for main chute inflation, it is also a task of the
deployment mechanisms to reduce the opening forces when the parachute starts
inflating. According to Dan Poynter’s parachute manual [4] the opening forces
mainly consist of two distinct forces, snatch force and the opening shock.
The snatch force occurs to the load as the mass shock resulting from the instant
acceleration of the combined parachute mass relative to the UAV velocity when
the suspension lines and risers are fully stretched. Combined parachute mass is
noted as the mass of the parachute assembly and the mass of the air filled in those
components. Thus to reduce snatch force, trying to minimize the mass of air already
filled in the canopy is a common strategy.
While inflating the canopy, air is filled in and compressed until the canopy reaches
it’s fully inflated shape. The opening shock is occurring due to the compression
of air, when the stage of maximum inflated diameter is reached. It is heavily
dependend on the rate the inflation is taking place, therefore an optimal balance
between deployment time and opening shock has to be found.
When considering information provided by Knacke [2], it is important to know that
he defines the opening forces as the differences between the instantanious drag force
and the steady state drag force.

Figure 2.1: Controlled Parachute Deployment by Knacke[2]

Controlled Deployment as seen in figure 2.1, is the most dependable concept.

At first a forced ejection of a pilot chute or a opening flap pulling the pilot
chute is taking place. This first pilot chute inflates and pulls out the main
parachute which is stored in a deployment bag. With the deployment bag
partial main canopy inflation is avoided and therefore the snatch force is
reduced. A designing problem for controlled deployment is the size of the pilot
chute. It has to provide enough power at low speeds to pull out the parachute
assembly but at high speeds it could exceed the opening force tolerances itself.
Therefore deployment velocity range is restricted.
In figure 2.1 a line-first deployment is shown, where there air vehicle is bound
to the riser, followed by the suspension lines and then the canopy. There
are other concepts like canopy-first deployments, where the main parachute
deployment bag is placed permanently on the riser between canopy and air
vehicle. Canopy-first deployment has a lot of disadvantages in comparison
to line-first deployment like higher snatch forces and a higher uncertainty of
operation.
Semi controlled Deployment denotes deployment with a pilot chute but with-
out any deployment bags. This concept still provides an enhanced deployment
7 2.3. Description of Function and PRS Elements

reliability but has no directed impact on the snatch force.

Uncontrolled Deployment is the easiest way to deploy a parachute, by just re-

leasing or ejecting the main parachute all alone. It is a very cheap concept
only applicable for small parachutes. Because there is no mechanism that
controls suspension lines being stretched before parachute inflation, very high
snatch forces may occur. There is also a higher risk for suspension lines to
invert or become entangled with parts of the canopy or other lines.

But despite all the advantages of fully controlled deployment, it is much more
complex and therefore in general has a higher system weight and a larger volume
than uncontrolled or semi controlled deployment. This is the main reason for the
other two concepts still being applicable for smaller UAV.

There are some general approaches for the distance a parachute or pilot chute has
to be thrown away from the vehicle to be able to inflate properly. This distance
is necessary because of aerodynamic turbulences caused by the forebody. Knacke
[2] states that Tests conducted by the NASA have determined that for vertical
descending bodies:

”the parachute should be ejected to a distance equivalent to more than

four times - and preferably six times - the forebody diameter, into good
airflow behind the forebody.”

Where the forebody diameter is the diameter of a cylindrical shaped model load
that would cover the same surface of the fully inflated parachute canopy as the true
forebody does.
This is a general approximation, exact distances have to be calculated or measured
for each PRS for specific velocities.
For deployment with a pilot chute, Knacke[2] states empirical tests of pilot chute
failure led to the thumb rule, that the pilot chute extraction force should be equal
or larger than four times the weight of the parachute assembly.

For a controlled ejection there are several methods. On one the hand, we have
destructive methods that work with pyrotechnics, rockets or mortars, and on the
other hand, there are re-usable concepts with mechanical springs or pneumatics.
Usage of drogue guns, mortars or rockets is often denoted generally as Drogue Gun
Ejection. A simple drogue gun shoots out a drogue slug to which the pilot chute
is bound. Bigger systems that fire out complete parachute assemblies are often
denoted as mortar deployment systems. These mortars are very useful for fast
velocities and rotating loads where the forebodies orientation is unknown. They
have some disadvantages in size, weight and additional shot reaction forces but
are very dependable. For example the Apollo space craft command module used a
system of parachute mortars to deploy their first stage parachutes.
There are also rocket deployments that pull out parachute assemblies or pilot chutes:
their advantage is that no reaction forces occur on the system, but they are complex
in designing.
For UAVs that have a defined orientation during drop phase, a common concept is to
store and deploy the PRS sideways to the main air stream. Due to aerodynamics, the
stall area sideways is thinner than the area behind the object, so that the distance
the pilot chute has to be ejected through is way shorter. This is called Cross-Wind
Ejection. For lightweight PRS this concept is most often the only applicable
concept. In Figure 2.2 three different cross-wind deployments are shown. The
first uses actuators like springs, pyrotechnics or pneumatics to eject the parachute
assembly. The second uses a lift web that gets stretched by the pilot chute and
Chapter 2. Parachute Recovery Systems 8

then pushes out the parachute assembly. The third is called 180◦ Catapult-Method,
where the parachute assembly is slightly pulled out by the pilot chute and then gets
gushed out by the airstream.

Figure 2.2: Cross-Wind Ejection by Knacke[2]

A very important factor of parachute deployment is also the installation location

of the PRS in the UAV. As seen above, there are some limitations to different
concepts. There is no general design guide where to locate the PRS. It is important
to consider the whole deployment process and final parachute forebody orientation
to ensure function. It is important that no parts of the UAV restrain the PRS
and no sharp edges should be inside and around the PRS storage. If there are
strict requirements in landing position of the UAV and rotating speeds, it may be
necessary to have several attachment points with risers connected to the UAV. This
may result in a very complex designing. Most often a single riser concept is chosen.
Vehicle orientation is controlled by choosing the attachment point relative to the
center of mass and rotation is affected by the chosen vehicle orientation. In order
to avoid chafing off the risers at the edges of the storage compartment, the usage of
webbing rings made of tear-resistent fabrics is a common solution. A webbing ring
is placed between the attachment point within the compartment and the riser, in
order to replace the part of the riser that would overlie the edges.
A lot of airplane shaped UAVs combine PRSs built in the rear with some impact
attenuation on the nose cone. A benefit from this concept is that the airplane is
orientated in a practical way so that disturbances to the air flow are small, what
makes the PRS gain stability against rotation and oscillation.
The contrary concept, to store the PRS in the nose cone, has also been built. It
is obvious that this concept has disadvantages during deployment and is not really
applicable if there is a main rotor at the front of the airplane.
For Cross-Wind deployment, there are two major concepts. One is to eject the
PRS upwards and the other to eject it downwards. For manned aircraft, only the
upward ejecting system is used as it is a requirement to optimize comfort for the
crew members. For drones and rockets there is also an option to to eject the PRS
down or sidewards as seen on Figure 2.3.
There are even some aerial vehicle where the PRS is not located inside the vehicle
but attached to the top of the vehicle. This might be more adjustable for future
developments, but in most of the cases it is avoided because of negative aerodynamic
consequences for the aerial vehicle during standard operations.

Another important factor for deployment is parachute packing. It does not only
affect the volume needed to store the chutes, but it also may have an influence on
the time needed to inflate the parachute.
In packing a compromise between maximum efficiency and effort for packing has to
be made. An easy way of packing is the ”Fruity Chutes” packing method. ”Fruity
Chutes” [8] is a manufacturer for parachutes that has published a simple way for
9 2.3. Description of Function and PRS Elements

Figure 2.3: Downward Ejecting System by Knacke[2]

hand-packing parachutes. More complex and efficient method is the ”air channel”
packing, where a special and complicated packing technique allows the parachute
to be inflated faster. An additional procedure to save volume is pressure packing,
where presses combined with air pumps vacuum achieve the parachute storage form.

2.3.3 Parachute Assembly

The parachute assembly consists of all parachute canopies, suspension lines and
risers that are ejected from the air vehicle. In this section an overview of different
canopy geometries, suspension and material choice is given.
The parachute assemblies aim is to reach a defined steady state drop rate V by
increasing the aerodynamical drag. Steady State is reached when the aerodynamical
drag force is equal to the gravity force. According to Tim Wyllie[3] the conventional
definition is
1
mg = ρV 2 SCD (2.1)
2
Where S is the parachute drag area and CD the drag coefficient described below.
According to Knacke[2] the most important characteristics of the parachute assem-
bly are

CD the drag coefficient is a dimensionless form factor indicating the drag char-
acteristic of a specific shape.
CX the opening force coefficient.
φ the mean angle of oscillation. A very imporant parameter for stability of
operation.

The parachute strongly affects weight and volume of the complete PRS. It is thus
important to find good compromises between the different design criteria.

Canopy Geometries
There are various different canopy forms used for different applications. Knacke[2]
has divided the different canopies into four groups.

- Solid Textil Parachutes

- Slotted Parachutes
Chapter 2. Parachute Recovery Systems 10

- Rotating Parachutes

- Maneuverable Parachutes

- Balloon Type Decelerators

Knacke’s tables of common parachute decelerators are included in the appendix A

to give a more complete overview with intuitive symbols for different concepts.
When having a closer look on solid textile parachutes, best performance is given by
polyconical or cruciform parachutes.
Polyconical parachutes were developed to increase drag characteristics of hemispher-
ical shaped parachutes and reduce the difficulty for manufacturing. In general they
have highest drag coefficients and thus reach the best drag/weight efficiency.
Cruciform parachutes in comparison have lower opening forces and a strongly re-
duced average angle of rotation, what yields in an increased stability of operation
and decreases the risk of impact damage as the vehicle lands more precisely in the
planned orientation.

Table 2.1: Best Performing Solid Textile Parachutes

Canopy Type Plan CD 0 Cx φ (Degrees)

Polyconical 0.8 to 0.96 1.8 10 to 20

Cruciform 0.6 to 0.8 1.1 to 1.2 0 to 3

Another important advantage for cruciform parachutes is the simplicity in manu-

facturing, they may be cut out from a single plane of textile without any glued or
sewed edgings. Tim Wyllie[3] describes that cruciform canopies have a tendency
to rotate around the axis of the main velocity more than polyconicals do. Never-
theless he states that this rotation usually does not cause problems and cruciform
parachutes are the most popular canopies for UAV recovery.

Before looking at slotted parachutes it is important to understand what porosity

means to parachute behavior. In parachute design porosity is devided in geometrical
porosity and fabric porosity. The first comes from a slotted design or holes in the
canopy and the second is an attribute of the used material. Porosity has a direct
influence on how much airflow in main velocity direction flows through the canopy
and therefore directly decreases drag, opening forces and oscillation φ. However, in
cases with high velocities and or big forebodies it may be a superior solution to use
bigger but slotted parachute canopies that use a geometric porosity (ratio between
open areas and total canopy area) of 10 to 35 percent. They are commonly used
for supersonic deceleration like first stage decelerators of a space craft command
module. But they lack in performance for lightweight UAV recovery.

Rotating parachute canopies have a special geometry that increases the rotation
around the axis of main velocity. As this is in general not really wanted in UAV
recovery, they are not used in common concepts.
Balloon type decelerators are also not specified here as they lack drag and weight
efficiency.
11 2.3. Description of Function and PRS Elements

Maneuverable parachutes could be an interesting option for UAV recovery. With ge-
ometric variation of canopy form by pulling break cords they enable maneuvering the
UAV during drop phase. Commonly most used maneuverable or gliding parachutes
are ram formed parafoils. Designing a maneuverable PRS is very complicated, as
there have to be additional servos to control the break cords. Deployment of the
parachute assembly will also result in a more complex operation as the servos have
to be protected from opening forces.

Parachute systems may have more than one main canopy. Those systems are usually
non-maneuverable and almost only used for high loads, where a single parachute
would be too big to ensure proper inflation or for very fast and rotating bodies like
spacecraft to decelerate supersonic velocities.

Pilot Chute
If a controlled or semi-controlled deployment is chosen, a pilot chute has to be
considered in design too. There are two different concepts for pilot chute handling
after deployment, one option is to cut off the pilot chute, another is to leave it
permanently attached to the main parachute. For permanently attached pilot chutes
it is important to test the canopy suspension and the pilot to main parachute
suspension. As the pilot chute may deform the main parachute after inflation it is
important to consider the ratio between pilot chute drag area and main canopy drag
area. According to Knacke[2] empirical tests showed that this ratio is constrained
to deployment velocities and should not exceed 0.03 for deployment velocities lower
than 277,8 km/h EAS (where EAS denotes Equivalent AirSpeed defined as true
airspeed multiplied by the square root of the ratio between the actual air density
and the standard air density at sea level).
As a lightweight UAV probably has a small main canopy, this could be a critical
factor for permanently attached pilot chute concepts.

The most important responsability of pilot chutes is to open quickly and reliably:
they must have sufficient drag to pull out the parachute assembly and they are not
allowed to exceed a maximum supported snatch force. Pilot chutes are often de-
signed with stabilizing vanes or with internal springs that cause a very fast inflation.
If a permanently attached pilot chute is used, main canopies form and suspension
have to be matched. Usually the main canopy crown has a round hole through
which the pilot chute riser is connected to the main canopy suspension lines.

Canopy Suspension
It is important to notice that the length of risers and suspension lines have to be
long enough for the parachute to have a proper air stream behind the forebody.
Knacke[2] provides a very good illustration of suspension line length influence to
drag characteristics for different canopy shapes as a collection of empirical data.
This illustration is shown in figure 2.4, where the length of a suspension line (from
it’s anopy attachement point to the riser’s node) is denoted with Le . D0 denotes the
nominal canopy diameter. As an approach for polyconical parachutes, the behavior
of hemispherical parachutes may be considered.
When designing the parachute suspension one should first get some references for
the specific canopy shape how much risers are needed and or optimal for correct
canopy inflation. The second part should be done directly with material choice.
As the Suspension lines must endure opening forces together it is sometimes a
more optimal way to choose more suspension lines than absolutely necessery but
thinner and lighter ones. As canopy inflation may take place irregularly, canopy
suspension most often has to be tested experimentally. For common parachute
Chapter 2. Parachute Recovery Systems 12

Figure 2.4: Influence of Suspension Line to the Drag Coefficient by Knacke[2]

concepts suspension references can be found in Knacke[2] and in the Parachute

Manual[4].

If the canopy attachment points do not support the opening forces there are different
options like ”suspension line tear out”, where a suspension line is attached to a sort
of web that distributes the inflation loads better on the canopy. This may end in a
very complex manufacturing but could give the possibility to save weight in canopy
material.

Material Selection
The most important canopy material selection criteria are porosity, tensile strength,
resistance to environmental influences and specific weight (weight per surface).
Knacke’s table of common canopy material characteristics is included in the Ap-
pendix B.
Today most parachute canopies are made of rip stop nylon because of it’s low weight
compared to very good characteristics in tensile strength. As it is a synthetic fiber
there are lots of variations in strength and thickness. Rip stop denotes a special
way of weaving, where filaments of different diameters are used to increase tear
resistance with low weight gain. In references, material is often shortly notated as
13 2.3. Description of Function and PRS Elements

”rip stop nylon 1.1oz” for example, where 1.1oz denotes specific weight of 1.1oz/yd2
being equivalent to 0.0373kg/m2 .
As parachute canopy stress analysis is a very complicated mechanical problem, it
will not be further developed here.

Suspension lines and risers are commonly made of Paracord. Paracord is a nylon
kernmantle rope originally developed by the US military for parachutes. According
to Wikipedia[1], the Paracord Type I, as the lightest of these, supports a breaking
load of 43kg and has a specific weight of 1.567g/m.
As the different standard types of Paracord are probably too strong and thus too
heavy for a lightweight UAV, additional research on materials has to be done.

2.3.4 Impact Attenuation Assembly

Vertical and horizontal landing energy has to be absorbed at ground impact. In
the ideal case horizontal landing energy is absorbed by elastic vehicle deformation
and ground squeezing, while horizontal energy is absorbed by sliding on the ground.
But as the ground conditions are not predefined, there could be rocks or asphalt
for example, it is a common concept to add impact attenuators instead of lowering
steady state drop speed.
One concept is to use crushable impact attenuators like honeycomb structures or
different types of foam. Those impact attenuators lower the deceleration force
by increasing the deceleration stroke distance and absorb energy due to plastic
deformations. There has been a lot of research done on common material like
different paper or aluminum honeycomb structures and foam plastics.
Another common concept are airbag systems. An ideal airbag system consists of
an textile bag with zero porosity that is inflated before impact. On impact the
pressure within the airbag rises and pressure valves deflate the bag to flatten the
impact pressure to an allowable limit.
A third concept are retrorockets that actually decelerate the vehicle before ground
impact. Retrorockets are a very advisable concepts for heavy air vehicles as their
energy-weight ratio then outperforms airbags and crushable impact attenuators.

A big disadvantage of crushable impact attenuators is their required volume what

has a negative influence on the aerodynamics of the air vehicle.
Airbags that do not require that much volume before need additional compartments
and a pressure vessel that has to be refilled after each usage. The additional com-
partments and the final weight therefore are a critical factor in lightweight UAV
recovery.
A more lightweight UAV recovery applicable concept is to have a sort of semi
destructive recovery, where parts of the UAV were designed with predetermined
breaking points that absorb energy and protect the expensive parts.

2.3.5 Continuative Tasks

After landing, there are some small tasks that may have to be done by the UAV to
avoid damage until it is picked up or to enable easier UAV localization.
The most common task, that is mentioned in references, is to release the parachute
after impact to avoid the UAV being damaged by being pulled around by winds
after landing. For this purpose, there are standard solutions such as e.g. the ”three
ring parachute release mechanism”. Release also could be done with sensors that
identify standstill or impact and activate a cut off or riser release mechanism. A
major problem of those systems is that sensor failure may cause the subsystem
to release the parachute assembly before landing which results in complete PRS
Chapter 2. Parachute Recovery Systems 14

failure. Another negative point that has to be mentioned is: the parachute assembly
may be lost even in cases the parachute release would not have been necessary. It
is important to consider typical operation terrains, weather and the time that is
needed to pick up the UAV, when deciding about parachute release systems.
Other continuative tasks might be sending location information, optical or visual
signals for easier UAV localization. For example GPS data could be transmitted as
well as smoke or peep signals could be emitted.
15 2.4. Case Studies

2.4 Case Studies

2.4.1 Manned Aircraft, Ballistic Recovery System
BRS Aerospace and Cirrus Aircraft[9] produce non maneuverable recovery systems
for manned aircraft with the goal to save the crews lives. They use a single main
canopy. It is a hemispherical high drag canopy that is deployed with rocket extrac-
tion. As the rocket and the landing usually damage the aircraft, it is a destructive
recovery system. Because of the very big canopy, reducing the high opening forces
to a tolerable limit for the crew was the biggest problem to solve. The engineers
solved this problem with a complex reefing system called CAPST M . A so called
”slider” causes the canopy to inflate partially in different stages. The slider can be
seen as a white ring in the suspension lines on part 4 of figure 2.7.

Figure 2.5: Ballistic Recovery System of Cirrus Aircraft, Wikipedia[1]

2.4.2 Phoenix UAV Recovery

The Phoenix UAV System used by the British army has a recovery system con-
sisting of a cruciform parachute and fan-inflated airbags. The parachute is stored
in the tail-cone and deployed underneath the horizontal tail plane. The airplane
is orientated nose done-down during drop phase and an airbag is inflated at the
nosecone before impact and causes a landing on the UAV backside to protect the
sensors that are placed at the body.

Figure 2.6: Phoenix UAV, 1999-2008[10]

Chapter 2. Parachute Recovery Systems 16

2.4.3 Maneuverable UAV Recovery, Experimental Program

with Banshee UAV
According to Wyllie[3] there are only a few UAV systems using maneuverable
parafoils. In 1996/1997 he was working on a 15 month lasting experiment to design
an autonomous maneuverable UAV Recovery system. They have chosen an 80kg
military targeting UAV called Banshee. A major problem was that the attachment
points needed to be spread as widely as possible for a stable controllable flight, but
the airframe only had a few possible points where the deployment loads would be
supported. Of course this problem occurred because they have taken an existing
airplane, but this may show what considerations in designing the UAV have to be
made if the use of a maneuverable PRS is planned. They found a solution, but the
deployment process had uncertainties and the risers tended to twist. They were
experimenting with staged deployment processes but did not find a very satisfying
solution. Even if deployment was succesfull without twisting it needed at least 15.5
seconds to deploy to a controlled flight, what represents about 80 meters of height
loss in their tests. Their conclusion to this experiment was that maneuverable PRS
are not suitable for small UAV because of the additional weight from lines and
actuators as well as the complexity of deployment.

Figure 2.7: Banshee UAV Under Parafoil Canopy, Wyllie[3]

Chapter 3

PRS for ASL Solar Airplane

As the ASL solar airplane still is in development phase the following specifications
are a preliminary suggestion only as they are based on estimations and momentary
prospects.

3.1 Requirements and Constraints

The ASL UAV will be an autonomous lightweight solar airplane with two propellers.
The UAV will have a wingspan of 2,5m and a total system weight of about 3 kg.
The wing surfaces are equipped with solar panels and a set of expensive sensors will
be placed at the airplane body.
For emergency situations such as loss of energy, actuator failure, electronic failure
or structural damages the ASL UAV needs a parachute recovery system that at first
ensures that no people on ground become injured and at second minimizes damage
to the airframe and to the sensors on ground impact.
In cases of emergency, an angle of attack between -10 and +20 degree and a sideslip
angle from -10 to +10 degrees are considered.
As the ASL UAV should be a very energy efficient airplane, negative effects to the
aerodynamics due to the parachute recovery system should be avoided. In addition
to that, the PRS weight should be as small as possible and is not allowed to exceed
an absolute weight maximum of 300g.
For economic reasons, the PRS should be reusable and reallocating the PRS after
usage should be doable for normal operators. For example parachute packing should
be easy to do with a guideline and no specialists or advanced equipment should be
necessary.
As a primary estimation for an allowed impact deceleration tolerance 4 to 6 times
g (the gravity acceleration) is assumed.
A further specified requirement is that the PRS should be able to act correctly at
a minimum deploy height of 50m over ground.
Because the sensors are placed in the front part of the body, a preliminary idea was
to land rear first.

3.2 Evaluation Criteria

Essential Criteria

• no people seriously injured when hit by dropping UAV

• minimize damage to airframe and sensors on ground impact

17
Chapter 3. PRS for ASL Solar Airplane 18

• capable of carrying a 3kg load

• no specialists needed for reallocating PRS

• minimum deployment height of 50m over ground

• weight maximum of 300g

Desirable Criteria

• deploy reliably

• light weight

• reusability

• possibility to become a standard landing procedure

• simplicity of design and manufacturing

• minimize negative influences on UAV aerodynamics during normal flight

• independability of operation to other components that took damage

• minimize additional electronic components

3.3 Concepts Compliant with Requirements

At this point uncontrolled, semi-controlled and controlled deployment have to be
considered as they all could match our requirements. For ejection we may focus
on the different concepts of Cross-Wind Ejection because we may not support the
weight and volume drogue guns or mortar deployment systems would need.
For the installation location, it seems that we have two possibilities.
The first one is found within the exchangeable module at the fuselage area below
the wings. This provides an attachment point near the center of weight and would
end in a downwards ejecting system that turns the airplane around.
The second possibility is to place the PRS at the rear of the airplane within one of
the two tail beams.
For weight and volume restrictions, impact attenuators are not considered further.

3.4 Evaluation
3.4.1 Maneuverable Concepts
In cases of motor or propeller damage the UAV can still glide in a controlled manner
itself. Only in the case of electronic failure, blockage of two or more servos or
structural damage to flaps or other parts of the airplane, the PRS has to take
action. Because maneuverable concepts need additional electronic servos that have
to be controlled, in most of the emergency cases there would be no advantage to
unmaneuverable concepts. As unmaneuverable concepts are superior in simplicity,
weight and volume requirements, maneuverable concepts are not seen as an option
any further.
19 3.4. Evaluation

3.4.2 Main Canopy

There are still two different main canopy shapes that have to be considered. Cruci-
form canopies that provide more stability and are easier to manufacture and poly-
conical or hemispherical canopies that provide the best drag/weight ratio. Stewart
Cartwright[7] states that cruciform canopies have smaller opening forces because
they inflate more gently, which directly means that the UAV loses more hight un-
til full inflation is reached. He also states that the tendency of hemisphericals (or
polyconicals that behave very similar) to oscillate up to 30 degrees only occurs at
descent velocities over 9m/s.
As a common UAV deceleration velocities are about 5m/s [3] the only advantage of
cruciform canopies would be the easier manufacturing.
The better inflation behavior and the superior weight efficiency then favors the use
of a polyconical parachute canopy.
Before further analyzing deployment systems, it is important to get an overview
about the canopy scales.
With equation 2.1
1
mg = ρV 2 SCD (3.1)
2
introduced in section 2.3.3 we can now calculate the needed surface area.
2mg
S= (3.2)
ρV 2 CD
With m = 3kg, g = 9.81m/s2 ,ρ = 1.225kg/m3 (assumed T=15◦ C, p = 1atm) ,
V = 5m/s, CD = 0.8 we get S = 2.4m2 as a first approach.
The Nominal Diameter D0 can now be found as
r
4S
D0 = (3.3)
π
Therefore the Nominal Diameter results as D0 = 1.75m.
With Knacke’s Tables of Common Parachute Decelerators A we can infer the con-
structed diameter of
DC = 0.95D0 (3.4)
DC = 1.66m, and with an approximation that polyconical is almost hemispherical
we get a material surface of
1 2
SC = πDC (3.5)
2
Therefore the canopy material needed is about SC = 4.34m2 , which would end in a
main canopy weight of mC = 0.161.7kg if 1.1oz ripstop nylon was used.
Knacke [2] states that the canopy filling distance sf remains constant for a specific
canopy and is depending on the inflated diameter Dp .

sf = nDp = 0.7nD0 (3.6)

Where n is a specific constant for a canopy shape that has to be evaluated while
testing and the inflated diameter is 0.7 time D0 according to Knacke’s Tables of
Commonly Parachute Decelerators A. A first assumtion of n = 8 this yields to a
filling distance of sf = 9.8m. The height loss until line stretch is usually even
smaller than this distance. A deployment height of 50m should therefore not be a
problem for this PRS.
For common hemispherical and polyconical canopies, the used amount of suspension
lines vary with their size. Fruitychutes [8] use 14 suspension lines for similar size
parachutes. In figure 2.4 the suspension lines influence for different types of canopies
Chapter 3. PRS for ASL Solar Airplane 20

are shown. Therefore the typical length for hemispherical or polyconical canopies
is between 1.1 and 1.2 time D0 . Fruity chutes [8] and Tristan Foon [6] use 1.15D0 .
As a preliminary approach we can assume 14 suspension lines with a length of
2.06m what yields in a total suspension line material length of 28.16m and a weight
of mS = 44.1g, if Paracord Type I is used.
Therefore the main canopy and suspension weight would have a complete weight of
m = mC + mS = 205.9g what still is within the limit of 300g.
By recalculating this process for an optimal polyconical parachute with drag coef-
ficiant 0.96, a total weight of m2 = mC2 + mS2 = 161.3g resulted. An excel sheet
with detailed data is included in the appendix C.
The main riser is still excluded in this calculation, because the needed length has to
be around 4 times the nominal forebody diameter. The nominal forebody diameter
is very dependend on what orientation the airplane has during steady state drop
phase. As an approach, a maximum riser length of 5m would result in a rather low
weight increase of 8g if Paracord Type I is used.
As an estimation of a packed parachute volume requirement, there are no rules of
thumb for a preliminary calculation. The packed size of a similar fruity chute[8]
model may provide a first approach. The nominal diameter D0 of our preliminary
canopy suggestion is between 1.6m and 1.75m, a similar sized fruity chute (CFC-72
with 1.88m diameter) fits in a cylindrical case of 10cm diameter and 16.5cm length
what provides an upper limit for our volume requirement.
As an approach for the maximum opening shock, Tristan Foon [6] multiplies 4g
with the system load and a safety factor of 1.5. With this approach a maximal
opening shock of 3kg·4·8.91m/s2 ·1.5= 176.6N results that has to be considered
when choosing lines and risers.

3.4.3 Deployment Mechanism

Knacke [2] states that uncontrolled deployment is acceptable for parachutes with
less than 1.5m diameter and our approxmations led to a nominal diameter of 1.6-
1.75m. Because a pilot chute provides a great gain in reliability and has a relatively
low weigth, semi-controlled or controlled deployment is suggested. The pilot chute
should be permanently attached, because reusability is an important criterion for
S
our PRS. With the given surface area of S = 2.4m2 and the condition of pilotmaxS =
2
0.03 we get a pilot chute surface area of Spilotmax = 0.072m . Using equation 3.3,
a maximum nominal pilot diameter of D0pilotmax = 0.3m can be calculated. With
same sizing procedure and materials as for the main canopy, this corresponds to a
maximum pilot canopy mass of mCpilotmax = 4.8g and with 8 suspension lines a
0.35m mSpilotmax = 4.4g. This yields in a very small weight increase of about 10g.
Knacke[2] states that the extraction force of a pilot chute should be at least four
times the weight force of the parachute assembly to be ejected, when designing the
pilot chute. For this consideration, a minimum deployment velocity needs to be
defined.
For pilot chute ejection, a spring is recommanded, as it is reusable and very reliable.
Pneumatics are usually too heavy and the 180◦-Catapult-Method in general needs
to abandon the compartment door.
If a deployment bag has to be used for the main canopy is dependent on the occuring
opening forces. But as there is only a very low increase, it is usually recommanded
to use a deployment bag. If wind tunnel tests show, that the opening forces are
very low anyway, it may be omitted.
If deployment should be upwards or downwards and the UAV turns over when
hanging on the chute can not be decided that easily, because extensions of the
opening time, turn moments and forces to the airplane have to be considered. Most
21 3.4. Evaluation

likely, the attachement point of the PRS should be set very near to the center of
gravity for avoiding high turn moments to the UAV.
The concept the ASL UAV is planned now, it will probably have an exchangeable
module at the fuselage area below the wings. As the PRS is a considerable increase
in weight for the UAV, it is probably wanted to be able to fly without PRS for
establishing long flight records. Therefore it would be an interesting concept of
storing the PRS in that removable module and with the implication to eject the
parachute downwards.
Another possibility would be to place the PRS at the rear of the airplane within
one of the two beams. Problems in deployment could be caused by dissymetry as
well as high turn moments could result in high stresses. The available space within
one frame has to be defined more definitively for ensuring feasability.
Chapter 3. PRS for ASL Solar Airplane 22

3.5 Preliminary Suggestion and Closer Specifica-

tions
Considering the evaluation the suggested design concept is a spring ejected, per-
manently attached pilot chute that pulls out the main parachute in it’s deployment
bag and produces a controlled deplyoment. This yields the same concept as Tristan
Foon[6] advises for a 25kg UAV that can be seen in figure 3.1. Differences will occur
in canopy shape, sizes and proabably in material choice.

Figure 3.1: Unscaled Deplyment System by Tristan Foon[6]

Main and pilot chute are polyconical or hemispherical canopies that offer the best
drag-to-weight ratio.
For an assumed steady state drop rate of 5m/s, drag coefficients for polyconical
canopies and a avareged air density of ρ = 1.225kg/m3 , the main parachutes nom-
inal diameter will be between 1.6m and 1.75m. The pilot chutes diameter has to
be determined experimentally, but it will have a maximum nominal diameter of
D0pilotmax = 0.3m.
With standard material, the parachute assembly will have a final weight of about
180g and 220g. As there has not been a detailed stress analysis and the assumed
opening shock of 176.6N was not further verified, the chosen materials (1.1oz ripstop
nylon and Paracord Type I) are too strong and heavy. There is a lot of weight that
can still be saved by optimising material choice. Even if this weight does not include
the deplyoment bag and the spring components, this first weight assumption shows
that it is possible to create a PRS lighter than 300g.
As a volume estimation, the main canopy probably needs a cylindrical diameter of
maximum 10cm diameter and 16.5cm length.
As there are mainly two possible storage options and one of those provides modular-
ity and better attachement point positioning, storing the PRS within the exchange-
able sensor module and eject the parachute assembly downwards is suggested.
When designing this PRS in detail, it should be noticed that every reference con-
cludes that testing the deployment process is a very important part of developement.
Chapter 4

Summary and Discussion

The in-flight uncertainties by outside influences and the high component costs of
UAV have brought the need of emergency recovery for a lot of UAVs. As the landing
of a autonomous UAV is usually a difficult task some of them even use parachute
recovery systems as standard landing procedures.
There has been a lot of parachute research done and there are several parachute
recovery systems for manned and unmanned aircraft in use today. However, since
weight is a very critical criterion for lightweight UAV, PRS are often omitted. In
my research I have not found any documents about PRSs for a UAV below 25kg.
There are also maneuverable parachute concepts like parafoil gliders. References
have shown that the development for maneuverable concepts is very complicated
and the deployment of the parafoil often unreliable. The additional weight compared
to non maneuverable concepts finally led to the decision of not further considering
maneuverable concepts.
The most critical action of a PRS usually is the canopy deployment. The parachute
has to be ejected through the UAV wake effect area and it has to be inflated cor-
rectly under different circumstances in terms of UAV orientation, flight velocities
and turn moments. During deployment, it is also fundamental to reduce opening
forces to a minimum. Therefore, advanced concepts for deployment were described.
For reasons of reusability and simplicity in reallocating a controlled deployment is
suggested for the ASL UAV. A small pilot chute ejected by a spring is inflating and
causes the needed drag to pull out the main canopy in a deployment bag. The main
canopy stays within this deployment bag almost until line stretch to avoid partial
inflation and a high snatch force.
Drag performances of parachute canopies are strongly bound to different geometries
and fabrics. There is a lot of empirical research done on aerodynamics of parachutes.
Polyconical parachutes offer best drag to weight ratio, where cruciform canopies
have a better stability and are easier to manufacture. By assuming a steady state
decent rate of 5m/s the stability of polyconical parachutes is sufficient and the
better weight efficiency led to suggest this canopy shape.
Preliminary approaches for needed parachute dimensions showed that even if stan-
dard parachute materials are used, it is possible to design a parachute recovery
system lighter than 300g for the ASL UAV. A lot of weight may be saved by the
development of a more appropriate material choice. Therefore, operational cases
have to be defined and stress analysis while deployment should be done. The pre-
liminary calculations led to a main parachute diameter of maximum 1.75m and a
maximum parachute assembly weight of 220g.
At the moment it seems to be the best option to store the PRS in the exchangeable
sensor module and eject the parachute downwards. As the UAV is still in develop-
ment this concept may have to be discarded on changes in the UAV shape. Future

23
Chapter 4. Summary and Discussion 24

work has to be done when designing the parachute assembly in detail. It is impor-
tant to notice that those preliminary calculations are based on the assumption of
a 5m/s steady state drop rate, this velocity is found in literature to be common
for recovery of airplane shaped vehicles. Increasing this velocity will cause lower
parachute size and weight but increase the impact shock. An optimal trade-off
should be elaborated.
A very important task in development is testing. Reliability in deployment and
occurring opening forces should be tested in detail.
Chapter 5

Conclusion

As there are no straight forward guides on how to design a parachute recovery

system for small UAV, the designing is a very empirical process. In order to design
an efficient PRS a lot of testing of deployment processes in terms of reliability and
stresses has to be done to minimize weight.
Due to extensive documentation like Knacke’s ”Parachute Recovery Systems De-
sign Manual”[2] and ”The Parachute Manual” by Dan Pointer[4], a preliminary
suggestion for a PRS concept has been worked out.
As we have very strong limits in weight a maneuverable PRS is beyond the scope.
Therefore, the suggested resulting concept is very common. A sideways to the
airstream, spring ejected pilot chute should pull out the main parachute, which is
stored in a deployment bag. The main parachute is extracted from the deployment
bag when line stretch is reached. This avoids partial canopy inflation before line
stretch to reduce snatch forces. Polyconical parachute canopies should be used, as
those provide best weight efficiency. Polyconical canopies are close to hemispherical
shape, that additionally are easier to manufacture.
Research on material has not been done in detail yet. As the extensive references
were published around 1992, they do not provide information about the newest
fabrics. With standard materials that used by parachute manufacturers, a prelimi-
nary rough calculation has been done that shows that it is possible to design a PRS
within the absolute weight limit of 300g. The PRS packed volume may be increased,
but it should be possible to locate it inside the sensor pack without having a big
influence on the aerodynamics.
A suggested main parachute diameter of 1.75m has fulfilled the requirements for a
standard drop rate of 5m/s.

5.1 Future Work

Conditions for PRS activation and preliminary tasks have to be chosen. Detailed
operational cases for the PRS have to be defined, in terms of flight conditions, UAV
velocities and vehicle orientation.
The landing phase for the ASL UAV has to be analyzed in order to ensure that the
common drop rate of 5m/s produces tolerable impact shocks.
Then, the parachute assembly should be designed in detail and analyzed in wind
tunnel tests. With those experiments the parachute assembly may be optimized in
terms of weight with lighter materials.
After this very big milestone has been completed, the activation components and
deployment subparts (like the spring system or the compartment door) have to be
designed and tested in detail.

25
Chapter 5. Conclusion 26

If possible, the deployment process should also be tested with model aircrafts or
dummy weights.
Appendix A

Knacke’s Tables of Common

Parachute Decelerators

27
Appendix A. Knacke’s Tables of Common Parachute Decelerators 28
29
30
31

Appendix B

Knacke’s Table of Common

Canopy Material
Characteristics
Appendix B. Knacke’s Table of Common Canopy Material Characteristics 32
Appendix C

Main Canopy Spread Sheet

33
34
35

Appendix D

Timetable
Appendix D. Timetable 36
Bibliography

[1] Wikipedia: Fallschirm / Parachute / Paracord / Gesamtrettungssystem

http: // de. wikipedia. org/ wiki/ Fallschirm# Geschichte
http: // en. wikipedia. org/ wiki/ Parachute# History
http: // de. wikipedia. org/ wiki/ Paracord
http: // de. wikipedia. org/ wiki/ Gesamtrettungssystem

[2] Theo W. Knacke: Parachute Recovery Systems Design Manual. Published

by permission of the U.S. Navy (NWC TP 6575) at the Naval Weapons
Center, China Lake, CA 93555-6001, First edition 1992.
[Link]
html&identifier=ADA247666

[3] Tim Wyllie: Parachute recovery for UAV systems. Contributed paper
[Link]
contentType=Article&contentId=1500358

[4] Dan Poynter: The Parachute Manual: a technical treatise on aerodynamic

decelerators. Published by Para Publishing, Santa Barbara, Second Printing
1991
[Link]
frontcover&source=gbs_v2_summary_r&cad=0#v=onepage&q=&f=false

[5] Earle K. Huckins III: Techniques for Selection and Analysis of Parachute
Deployment Systems. Nasa Technical Note, Washington D.C., January 1970
[Link]
19700005898_1970005898.pdf

[6] Tristan Foon: RDTE of Parachute Recovery System for ADFA UAV.
University of New South Wales, Australian Defence Force Academy, Canberra,
ACT 2600, Australia
[Link]
viewFile/244/120

[7] Kirk Graham, Stewart Cartwright: Parachute Recovery System for

Small Research UAV’s. University of New South Wales, Australian Defence
Force Academy, Canberra, ACT 2600, Australia
[Link]
viewPDFInterstitial/81/51

[8] Fruity Chutes: Manufacturer for ”custom parachutes in fruit flavoured col-
ors”
[Link]

[9] BRS Aerospace and Cirrus Aircraft: Manufacturer for Parachute Re-
covery Systems for small manned aircraft

37
Bibliography 38

[Link]
[Link]
[10] Reme Museum of Technology: Phoenix UAV
[Link]