Richard Nakka’s Experimental Rocketry Web Site
________________________________________________
Introduction
to Rocket Design
9. Recovery
System
Introduction
My very first amateur rocket, the A-Rocket as it was deemed, was not equipped with any sort
of recovery system. Being a small rocket expected to achieve a modest altitude,
it was simply launched at an angle such that its ballistic flight would
terminate harmlessly at a (more-or-less) predictable point well away from anyone.
The rocket itself was sufficiently robust that such a hard landing caused
minimal damage (e.g. broken nosecone) allowing the rocket to be readily
refurbished for re-use. This rocket was flown a grand total of four times. It
admirably served its intended purpose, demonstrating that the small rocket
motor that I had developed could reliably loft a rocket into the blue. A short
while later, I had developed a substantially larger and more powerful rocket
motor that was capable of boosting my new and larger B-Rocket to a much greater altitude. A recovery system
was needed, both for safety and for sake of the rocket itself, as a ballistic
return would mean near-total destruction of the rocket. Unfortunately, my early
recovery systems proved to be unreliable. The outcome of many of those early
flights, unhappily, was a ballistic return, with the aforementioned consequence.
Why were there many lawn-dart flights,
as rocketeers refer to rocket flights that end up crashing to the ground? The
answer to that question first requires that we delve into the basics of rocket
recovery system.
Recovery System basics
There are three key systems that comprise a hobby or EX rocket recovery
system. I refer to these as:
1) Recovery Control
system
2) Recovery Deployment
system
3) Recovery Descent
system
The Recovery Control system activates
at liftoff, or some related event such as motor burn-out, and typically senses
the status of the rocket flight and at a predetermined point in the rocket’s
flight, the system activates the Recovery Deployment system. The Recovery Deployment system,
once it receives its signal from the Recovery Control system, generates an action that serves to initiate the Recovery
Descent system. The latter serves to retard the
rocket’s flight by reducing its velocity, in some manner, to ensure
that landing occurs at a sufficiently slow speed such that the rocket survives
with minimal or no damage, allowing for reflight with nominal refurbishment.
In the simplest case, such as that used for commercial model rockets, the Recovery Control
system is a length of pyrotechnic delay composition (or delay grain, an
integral part of the rocket
motor) that begins to burn once
the propellant has been fully consumed. The burning occurs at a certain rate
such that the length of delay grain is fully consumed by the time the rocket is
expected to complete its coast to apogee. It then ‘signals’ the Recovery Deployment system, which in essence, ignites an ejection charge. The ejection charge is a very fast burning
pyrotechnic material that generates a burst of pressure within the rocket body,
the result of which is an impulsive force that activates the Recovery Descent system, which typically is comprised of a
small parachute that is ejected out of the rocket body, usually by way of the
nosecone. The parachute, tethered to the rocket, retards the descent and allows
for a safe landing. Figure 1 illustrates the components a model rocket motor.
Figure 1: Model
rocket motor cut open (courtesy
Apogee Components)
Recovery Control System
As mentioned, the simplest Recovery Control system is a delay-element that is usually integral with the rocket motor.
As well as being the system employed for model rockets, this same method is also used for some mid-power
commercial rockets. With re-loadable versions of commercial motors, the delay grain
(or delay-element) is a separate, replaceable component that is housed within
the forward closure (or bulkhead). Depending on the effective length of the
delay grain, the time delay can be chosen to suit the expected coast to apogee.
Hi-Power commercial rocket motors as large as L-Class are available with a delay
element, with a delay period as long as 18 seconds, firing an ejection charge
at the end of this delay period.
Interestingly, a delay-element system (with ejection charge) was
utilized for certain commercial sounding rockets that featured boosted darts.
The Judi-Dart and the Super Loki-Dart both featured a similar system. The dart portion
of the rocket featured a delay-element which was electrically ignited at
launch, necessary as the dart was inert (being boosted, the dart had no
inherent propulsion system). The delay period for the Judi-Dart
was 100 seconds, which was the coast time to apogee, typically 240,000 feet (74
km).
Judi-Loki Dart cross-section
Super Loki-Dart cross-section
A delay-element system can be adapted to EX rocketry. In fact,
some years ago I developed a delay device, incorporating an ejection charge,
that I deemed Pyro-DED. This is a simple, motor-mounted,
pyrotechnic-based device that was designed to be adapted to any EX rocket
motor. The required delay time-period is achieved by adjusting the touch-hole
depth. The body can be fabricated from aluminum, for single-use, or of steel,
for re-usability. Figure 2 illustrate
the Pyro-DED.
Figure 2: Pyro-DED
The pyrogen charge serves two purposes. One, it provides auxiliary
motor ignition and two, it helps ensure that the delay grain ignites reliably.
My SkyDart rocket utilized a recovery system that utilized
a Pyro-DED fitted to my A-100M rocket motor.
There are two other Recovery Control systems that were utilized in
my earlier rockets that I will discuss briefly. An Air-speed Switch system and an Electronic Timer. The Air-speed Switch system served to sense
the velocity of the rocket via an external vane or flap. As the rocket slows
when apogee is approached, a spring-loaded flap becomes less deflected in the
airstream. At some point, spring force overcomes drag from the airflow, and
closes a micro-switch. This initiates the Recovery Deployment system
which is a standard ejection charge (a.k.a. ejection pyro).
The Air-Speed Switch system was used with reasonably good success on many of my
C-Series rockets (circa 1973-1984).
The Electronic Timer system consisted of a self-made electronic
circuit that featured a time delay that was initiated at motor burnout by a mercury
switch. The timer system is
essentially an electronic analog to the pyrotechnic delay grain system, and
requires knowledge of how long a rocket will take to coast to apogee. I first
used this system with my Boreas rocket as part of a dual-event
recovery philosophy (to be discussed later). This was an innovative hybrid recovery
system that utilized an Air-Speed Switch for
initiating a drogue chute deployment and an Electronic Timer for
initiating main chute deployment. The same system was used for the first two
flights of my Zephyr rocket. Overall, the reliability of this hybrid
Recovery Control system was near perfect, despite challenges of launching in
cold winter weather.
A contemporary approach to Recovery Control, one that is now used
for most Hi-Power and EX rockets, utilizes a microprocessor-based
electronic system. Such units are commonly referred to as an altimeter or flight computer.
This is a smart system that senses physical parameters that change over the
course of the flight (such as barometric pressure, rocket acceleration or tilt
angle). When a predetermined value of the pertinent parameter is sensed (or a
combination of such), the Recovery Control system sends an electronic
signal to the Recovery Deployment system. There are many affordable commercial
(COTS) flight computers now available.
My first rocket to utilize a COTS electronic flight computer for
Recovery Control was Zephyr which was launched in late 2003. This rocket
featured the R-DAS unit, which was quite sophisticated for its
time (and rather expensive). My Cirrus rocket, launched a few years prior, also
featured an electronic altimeter for Recovery Control. This was not a commercial
unit. Rather it was designed and built by fellow EX rocketeer Paul Kelly and
was state-of-the-art at the time. This altimeter was designed for dual-event
deployment. Most, if not all, modern flight computers have the capability to
support dual-event recovery.
In addition to serving as a Recovery Control system, flight
computers can sense and store flight parameters several times per second,
giving the user a complete history of a
rocket’s altitude, and other parameters such as acceleration, throughout the
flight and allows for detailed post-flight analysis. An invaluable bonus. The Blue
Raven (utilized in my later Xi rocket flights) senses and records a vast array of flight
parameters including axial and lateral acceleration, gyroscopic angular rates,
tilt and roll angle, vertical and lateral velocity, temperature and more).
Considering the availability, reliability and
relatively low cost for modern COTS flight computers , it does not make sense
to try to engineer one’s own Recovery Control system, especially when one
considers the critical function for safe recovery of a rocket.
For example, the Eggtimer Quark (also utilized, for redundancy, in my later Xi rocket flights) features dual-event Recovery Control and
costs a mere $20 USD. It comes as a kit that requires soldering SMD parts to a tiny board which was an interesting
challenge in itself. But if I can manage to succeed at that with my innate ham-fistiness,
any EX rocketeer can certainly manage this.
If an EX rocketeer does wish to
design and construct his/her own flight computer for Recovery Control, for
example as a learning exercise, a COTS flight computer should certainly be flown as a backup Recovery Control system.
Recovery Deployment System
With model rockets, the Recovery Deployment system, as mentioned
earlier, consists of an Ejection
Charge, wadding
( or other means) to isolate the resulting hot gases that are generated by combustion
of the charge, and a section of the body tube
. The ejection charge that is integral to a commercial model rocket motor is a
small quantity ( ½ to ¾ gram) of granular
Black
Powder. The section of body
tube serves as a pressure vessel, or means of
containing the hot, pressurizing gases that are generated (Figure 3). This
burst of pressure, acting over the cross-sectional area of the body tube,
generates an axial force that serves to activate the Recovery Descent system.
Figure 3: Recovery Deployment System
for a basic rocket
It is important to recognize that the force generated by an
ejection charge acts on all surfaces of
the pressure vessel. The magnitude of the pressure developed by combustion of a
given mass of ejection charge is dependent upon the degree of
containment. Containment can be considered to be the walls and closures of the pressure vessel. For example, if the
wadding, parachute and nosecone were not present, the forward end of the
pressure vessel would be open to the air and there would be no containment. As
such, minimal pressure would be developed when the ejection charge fires (some
containment would exist due to the mass inertia of the air that is present in
the body tube). If the nosecone was to be securely attached into the body tube,
such that it would not come off, maximum pressure would develop. If sufficient
mass of ejection charge were to be ignited under this condition, the body tube would
rupture or the retainer clip would fail.
Figure 4 illustrates the
concept of the rocket acting as a pressure vessel, for a simplified rocket. The
pressure forces shown are those acting on the walls and closures at the instant
the ejection charge fires. The closures consist of the rocket motor-nozzle
assembly (now depleted of propellant) and the wadding. Some items of interest worth
noting:
1.
The retainer clip (or motor retainer) must be sufficiently robust
to withstand the force acting on the motor assembly.
2.
The wadding itself provides minimal containment. Containment due
to wadding is primarily due to friction of the wadding against the body tube
(hence it should not be overly tight fitting).
3.
The parachute provides no significant containment.
4.
The nosecone is therefore expected to provide containment. This is
achieved by attachment means (such as a snug fit in the body tube for our
example rocket).
5.
Force acting on the wadding causes it to move forward (like a
piston), compressing air in the parachute compartment until enough pressure is
developed to blow off the nosecone. Residual pressure pushes out the parachute
and wadding.
6.
As the wadding and air within the parachute compartment become
compressed, the volume of the pressure vessel grows. Boyle’s
Law tells us that as the
volume increases, pressure drops
proportionally. If the volume of the pressure vessel increases too much, the force acting to blow off the nosecone
diminishes and may not be sufficient to do such.
7.
The motor-nozzle assembly has leak paths, such as the nozzle
throat hole and the motor casing/body tube interface. Although there is nothing
we can do about the throat hole, other than to consider this pressure loss when
sizing the ejection charge, the motor/body tube interface should be well-sealed
to eliminate leakage.
Figure 4: Pressure vessel
interpretation of ejection charge action
There are variations to this concept. Instead of using wadding,
which is essentially flame-resistant ‘toilet paper’, to isolate the parachute,
there are various baffle techniques that have been
used with hobby rockets. Resource R-1 describes
several of these. Baffles work by trapping combustion residue and cooling the
gases that flow through. As there is no wadding to serve as a piston, ground
testing should be conducted to ensure reliability.
With re-loadable
versions of commercial motors, the ejection charge is loaded by the user into a
well that is integral to the bulkhead of
the motor. The ejection charge material is typically 3F
or 4F Black Powder that is supplied by the user. My early rockets used granulated
KNO3-sucrose as the ejection charge material. The KNO3-sucrose
was first melted, then ground up into granular form using a mortar
and pestle. This material worked
reasonably well, but was hard to grind and had a tendency to absorb moisture
from the air. All my later rockets, starting with Zephyr in 2003, employed Crimson Powder as the ejection charge material. I use this
material exclusively. Crimson Powder has significant advantages over Black
Powder:
1.
Higher impetus (i.e. more potent)
2.
Cooler burning (results in less damage to components exposed to
the combustion heat)
3.
Combustion residue is odourless and cleans up with warm water (no
sulphur stink)
4.
Safer to make and handle (personal anecdote)
Crimson Powder is somewhat hygroscopic and to maintain potency, it
must be stored in a sealed container with a desiccant (such as calcium chloride
or silica-gel).
Appendix
E provides details as to
sizing an ejection charge using Crimson Powder (or Black Powder) as the
pyrolant.
Energy needed to power a Recovery Deployment System does not
necessarily have to be from combustion of a pyrotechnic material. There are
COTS system that utilize liquified gas, specifically CO2. The CO2
is stored in a single-use cartridge (a.k.a Powerlet). The cartridge, which is typically one-third
filled with liquid CO2 has the remaining volume as gaseous CO2
in equilibrium with the liquid at its vapour pressure. A typical system utilizes
a small charge of Black Powder that, when fired, generates an action that
drives a pin into the end of the cartridge, puncturing it, thereby releasing
the CO2 which instantly flashes into a pressurized gas.
There are also cold-gas Recovery Deployment systems that are
completely mechanical and use an electric servo system to puncture the CO2
cartridge. Cold-gas systems such as this have a couple of potential
advantages over a pyrotechnic-based (hot-gas) system. One, as the pressurizing
gas is cold when vented, there is no risk of flame or hot particles causing
damage to the parachute or other components exposed to the pressurized gas. A
second advantage is that a CO2 system may operate effectively at high
altitude, even in low ambient pressure conditions such as that present at very
high altitudes. Pyrotechnic materials burn less reliably under conditions of
low ambient pressure, such as that which exists at high altitude such as >
50k feet (15k metres). Proper containment of a pyrotechnic charge is necessary to
ensure expected performance. There are also disadvantages. A cold-gas based Recovery
Deployment system is more expensive, heavier and more complex (and therefore potentially
less reliable) than that based on a simple pyrotechnic squib. The impetus of pressurized CO2 is
about 1/5 that of Black Powder, requiring five times the amount (mass) to
generate the same useful pressure. Another more significant disadvantage, as least
for those who launch in cold weather, is that vapour pressure of a CO2
cartridge decreases significantly with ambient temperature. As shown in this graph, the vapour pressure at -20°C. is one-third the vapour pressure at +20°C.
Resource R2 describes an amateur designed and built CO2 deployment
system.
It may also be feasible to utilize a mechanical
Recovery Deployment system, although I have never attempted this approach. An
example of such a deployment system is one that would utilize a compressed
spring to store energy. The spring would be placed at the bottom of a parachute
canister. A spring-release mechanism (either mechanical or pyrotechnic) would
release the spring when commanded by the Recovery Control system. The potential
energy of the spring would then serve to accelerate the parachute out of the
rocket. There are a number of advantages to a purely mechanical system. Being a
“cold” system, there is no chance of damage to the parachute or other
components. Ground testing is simplified compared to pyrotechnic or CO2
systems, as nothing is expended which makes reloading simple and cost-free.
There are, of course, disadvantages. A sufficiently strong spring is heavy and
takes up valuable volume. When cocked, a spring stores considerable energy and
careful design is required to ensure the spring is not inadvertently released,
either during or after cocking, to protect the operator from possible injury.
Designing a reliable release mechanism can be challenging, considering the
force that is being held back by a spring that is powerful enough to reliably
eject a parachute (or to separate the rocket body sections, for the case of a
dual-event system).
As mentioned earlier, the Judi-Dart
sounding rocket featured a pyrotechnic delay-element as the Recovery Control system.
The Recovery Deployment system for this rocket consisted of a pyrotechnic ejection charge and piston-staves system that deployed a parachute at (or near) apogee. When
triggered by the last of the burning delay grain, the ejection charge fires,
pressurizing the body section (which had been filled with delay grain). The resulting
pressure applies a force against the face of the piston. The piston presses
against the rigid staves, which butt against the nosecone. The force
transferred by the staves to the nosecone causes the nosecone to separate. This
allows the residual pressure from the burnt ejection charge to propel the
piston, staves, parachute and instrument package out of the dart. For this
dart, solely the instrument package descended to earth by parachute, the rest
of the dart (and booster rocket) free-falls to the ground.
Based on the preceding
descriptions of Recovery Deployment system concepts, it is clear that there are
two aspects of the system.
1.
Energy source required to power the system.
2.
Physical means to harness the energy in order to deploy and/or activate
the Recovery Descent system.
With regard to the energy source for
a Recovery Deployment system, for the pyrotechnic system, the energy is in the
form of chemical energy. Only a small fraction of the total chemical energy can be
harnessed to do useful work.
For the mechanical system, the energy source is elastic potential
energy stored in the compressed spring as a result of
the work performed in compressing it:
P.E. = ½ k DL2
where k = spring
constant lbf/inch or N/cm
DL = change in length of spring due to
compression inch or cm
P.E. = Potential energy lbf-inch or N-cm
Essentially all the potential energy of a spring is available to
do useful work.
Appendix
F provides an example
means of designing a spring-based parachute ejection system.
Curiously, for the CO2 system, only a small portion of
the energy is potential energy of the compressed gas in the cartridge. Most the
energy of the system is heat energy drawn from the surrounding environment -
the heat
of vaporization needed to flash the
liquid CO2 to gaseous form. As such, the emptied cartridge and
expelled gas become very cold as heat energy is extracted from whatever the
liquid CO2 comes in contact with.
An example of the physical system that
harnesses the energy to deploy its Recovery Descent system was given for model
rockets and for the Judi-Dart
sounding rocket. What about and Hi-Power and EX rockets? Truth is, I have never
launched a commercial Hi-Power rocket. As such, my knowledge of the Hi-Power
rocket Recovery Deployment systems consists solely of what I have picked up
from others over the years and from what I have researched on the web. My
understanding is that, in most cases, the Deployment System is basically the
same as that for a model rocket, scaled up and having more features
necessitated by the larger size of many Hi-Power rockets and more sizeable
ejection charges employed. These additional features are mainly to provide
protection of the parachute from the heat of the ejection charge and include
such items as parachute protector, parachute liner, nomex blanket and deployment
bag. The deployment bag is used for larger rockets and serves an additional
purpose of deploying the parachute in a more controlled manner that helps it to
inflate reliably and to reduce chance of shroud line tangling. Pistons are also
used in some systems, along the lines of the Judi-Dart
sounding rocket.
Figure 5 illustrates, diagrammatically, the basic principle of a
typical Hi-Power rocket Recovery Deployment system. This particular example is one
that employs a dual-event
recovery method (covered in more detail later). The rocket consists of three
sections. The aft section which houses the motor and main chute, the forward
section which houses the payload and drogue chute, and the Avionics Bay, which
houses the flight computer and serves a second important purpose as a coupler
that joins the forward and aft sections. The section joints are fastened with nylon
shear screws. These screws provide for a critical degree of containment. When
the ejection pyro fires, pressure rapidly builds up in the chute compartment,
sealed by a fixed bulkhead at one end, and by the Avionics Bay at the other
end. When the pressure reaches a certain critical level, the nylon screws fail
in shear, causing the sections to forcibly separate, thereby pulling out the drogue
or main parachute. Thermal protection is required for the parachutes as they
are directly exposed to the ejection pyro combustion products, which consist of
both hot gases and hot particles.
The tethers (also called shock cords)
which connect the chute to the rocket are also exposed to this heat and are
usually fabricated of a heat-resistant fabric such as Kevlar.
Note:
For clarity, tethers not shown in Figure 5.
Figure 5: Hi-Power rocket Recovery
Deployment system (typical)
Figure 6 illustrates deployment of the Drogue Chute, which
typically occurs at apogee. Figure 7 illustrates deployment of the Main Chute,
which occurs at an altitude nearer the ground. Typically, all three sections of
the rocket descend together, connected by tethers. However, some designs have
the forward section, with Drogue Chute, descent separately
from the rest of the rocket.
Figure 6: Deployment of Drogue Chute
Figure 7: Deployment of Main Chute
With regard to EX rocketry, my first rockets were based on the
model rocket deployment system from which I had derived much experience. This approach
seemed to be the natural way to go, as my early rockets were indeed scaled up
model rockets in many respects. Over the years, my Recovery Deployment system
evolved as I recognized faults and came up with incremental improvements. One
key feature was the development of a Non-fixed Bulkhead to
isolate the parachute from the ejection charge. I first used this isolation
method with my Boreas rocket. A Non-fixed Bulkhead (NFB) is similar in
some respects to a piston (in fact, for convenience I often refer to it as a
piston). It looks like a piston and moves like a piston does, however, it
serves as a bulkhead that separates the ejection
charge pressure vessel section of the rocket from the parachute compartment.
Figure 8 shows a PLA 3D printed NFB used in my Xi
rocket.
Figure 8: Non-fixed Bulkhead for Xi rocket deployment system
Figure 9 illustrates, diagrammatically, the principle of the
Recovery Deployment system for my newest generation of EX rockets. All my EX
rockets now feature dual-event recovery. There are similarities to that of the
Hi-Power rocket recovery system discussed earlier. The rocket consists of three
sections: the aft section which houses the motor (and optionally an auxiliary
payload bay), the forward section which houses the payload and parachute, and
the Avionics Bay, which houses the flight computer and serves as a coupler that
joins the forward and aft sections. The Fixed Ring is a thin-walled ring that is rigidly fastened
to the body tube. The purpose of the Fixed Ring is to provide a stop for the
NFB to react against when the compartment housing the pyro is pressurized at
the moment the pyro fires.
Figure 9: My EX rocket Recovery
Deployment system
Figure 10 illustrates the apogee separation event. No drogue
parachute is typically deployed, rather, the rocket, now separated into two sections
connected by a long tether, simply free-falls (Fig.13). On a couple of flights,
I have deployed a drogue parachute. However, the descent rate was found to be
essentially the same (of course, this depends on the size of the drogue chute).
Figure 10: Apogee separation event
Figures 11 and 12 illustrate deployment of the Main Chute, which
occurs at an altitude nearer the ground. All three sections of the rocket
descend together, connected by tethers.
Figure 11: Apogee separation event
As seen
in Figure 11, when the parachute pyro fires, the nylon shear screws fail when
sufficient pressure is developed, causing the Avionics Bay to forcibly separate
from the forward section of the rocket.
Figure 12: Apogee separation event
Figure 12
illustrates what subsequently happens. The momentum of the Avionics Bay
extracts the Non-fixed Bulkhead, and in turn, pulls out the parachute.
The NFB system
has been used with my DS-series of rockets and my Xi-series of rockets,
totaling nearly 50 flights to date. Following improvements made after
experiencing initial teething problems, this Recovery Deployment system has worked with
100% reliability.
What are
the advantages of the NFB-based EX Recovery Deployment system over that
described earlier for typical Hi-Power rockets such as that shown in Figure 5?
I see the following advantages:
1.
No drogue
chute is required (admittedly the drogue chute can be omitted from the Hi-Power
system).
2.
Parachute
is completely isolated from the hot gases and particles of the pyro charges. No
deployment bag or other protective features are needed, features that add
complexity and weight. Parachutes are expensive items, and eliminating risk of
damage due to, say, not properly packing the thermal protection item, is a nice
form of insurance.
3.
The
volume of the pressurized sections of the rocket (outlined in orange dashed boxes
in Figures 5 and 8) is much smaller, requiring a significantly smaller ejection
pyro. As a consequence of using a smaller amount of pyrotechnic material, much
less energy is imparted to the separating sections of the rocket. The result is
less work needed to be performed by the tethers to safely absorb the energy (by
stretching) without breaking or generating rebound which can lead to colliding
and damaged airframe sections.
An important consideration for any Recovery Deployment system that
utilizes nylon shear screws is selecting an
appropriate size (diameter) and number of
screws. In other words, screws that provide appropriate shear
strength. For the apogee event, the screws should be chosen to have
a minimum strength, being strong enough to hold the rocket body sections
together during handling and to withstand the pressure differential between
ground level and apogee. The screws joining the parachute section must be
carefully sized to prevent inadvertent failure of the screws due to the momentum
of the separating sections that occurs when the apogee pyro fires. If the
screws are too weak, this joint could fail, resulting in deployment of the
parachute at apogee. Obviously, this is undesirable, as the rocket could drift
for kilometres before landing. This unfortunate circumstance occurred on two of
my early flights that featured nylon shear screws.
Figure 13: Free-falling phase of Xi rocket descent
The Recovery Deployment System is arguably the most prone to
failure of the three rocket recovery systems. There are many opportunities for
things to go awry, either in design, construction or implementation. Factors,
such as launch temperature, wear-and-tear and simple mistakes in preparation can
play a role in how reliably a Recovery Deployment system works. Considering the
critical nature of the Recovery Deployment system of a rocket, exceptional care
must be followed in the design, followed by ground testing, and meticulous implementation
in preparation for each flight.
Appendix G 
details sizing of nylon screws for the Recovery
Deployment system.
Appendix H provides details on the design of a Recovery
Deployment system utilizing a NFB.
Appendix I provides details on the design of a Recovery Deployment
system utilizing a delay-grain.
Appendix J provides details on the design of an Ejection
Charge.
Recovery Descent System
The Recovery Descent system is the last of the three systems that
constitute the complete recovery system of a rocket. Recovery Descent system is
that part of the system whose function is to implement safe return of a rocket
to the ground following activation of the Recovery Deployment system(s). In the
simplest case, such as with a model rocket, the Recovery Descent system is a
device such as parachute or streamer that is deployed near apogee. The
parachute or streamer serves this function by generating enough aerodynamic
drag force that the rate of descent is sufficiently retarded such that the
rocket lands intact.
With model rockets, the Recovery Descent system can encompass a
wide array of different techniques. Parachute and streamer are the most common,
however, there are tumble, glide, lifting-body, helicopter, drag, para-wing,
aerobrake, magnus-rotor and other methods that seem to be limited only by one’s
creativity. The fact that a model rocket is very lightweight (in comparison to Hi-Power
and EX rockets) allows for these different methods to be implemented in a
relatively simple, safe and reliable manner. Tumble recovery is the most simple
method and is suitable for smaller model rockets. The Recovery Descent system
is the rocket itself, made unstable by the Recovery Deployment action which
moves the motor rearward (upon firing of the “ejection charge”), by a certain
amount limited by a retainer clip. This serves to move the rocket’s C.G. aft of
the C.P. thereby making the rocket unstable. The very low weight of the rocket relative
to the resulting large drag force present during descent ensures a soft
landing. Due to its simplicity, tumble recovery is clearly a method with high
reliability. The other mentioned techniques such as glide and helicopter have
no advantages over parachute, streamer or tumble. Complexity of such methods tend
to reduce reliability and may compromise safe recovery.
Deployment of a sole Recovery Descent device at apogee is referred
to as single-event recovery. Besides being the
method used for model rockets, this method is also suitable for many Hi-Power
or EX rockets. Dual-event recovery involves
deployment of a Recovery Descent system at (or near) apogee and a second
Recovery Descent system nearer the ground. Figure 5 and Figure 8 illustrated
rockets with dual-event recovery systems. In Figure 5, deployment of a drogue
chute at apogee constitutes the first of the two events. Deployment of a
parachute later in the flight constitutes the second event. In Figure 8,
separation of the rocket into two sections at apogee to allow for a controlled
free-fall descent constitutes the first event. Deployment of a parachute later
in the flight constitutes the second event. The compelling reason for choosing
a dual-deploy scenario over a single-deploy recovery is to reduce down-range
drift of the rocket. Deployment of a parachute at apogee, large enough to allow
for gentle touch-down, may result in a long descent duration (depending on
apogee altitude). If there is a wind, the rocket will drift during descent at
the same lateral velocity as the wind. Despite this, single-deploy may be advantageous,
due to its inherent simplicity relative to dual-deploy, if the following apply:
1.
Apogee altitude is low enough that drift, even with a moderate
wind. is not a concern with regard to landing site.
2.
Launch range is sufficiently large and obstacle-free such that
location of landing is not a concern, and
a.
rocket remains in visual contact throughout the flight.
b.
or a reliable GPS tracker
system is implemented as part of the payload
3.
Launching in an area where wind
is always light (i.e. rocketeer’s
paradise) such as the Amazon basin or perhaps in the Doldrums. The winds at the South Pole are also perpetually
very light, but this location has definite drawbacks as a launch site.
When deploying a parachute from a rocket, it is not enough to
simply release it from the rocket. It must be forcibly
ejected such that tension develops in the tether connecting the chute to the
rocket in order to reliably inflate the
chute. Otherwise, it may not ‘catch enough air’ to inflate. The result of such
can lead to the parachute remaining in a closed bundle
all the way to the ground with a resulting harder than expected landing. When
ejected with adequate force, the momentum of the chute’s canopy will tend to
unfurl it when the tether becomes taut and the parachute experiences a jerk. This anomalous scenario happened on a number
of my earlier Xi rocket flights, whereby the parachute was properly released,
but did not blossom, or did so only partially. This was puzzling and it took a
while (and a few dead-end solutions) to figure out the actual cause. Figure 14
illustrates diagrammatically the incorrect means to tether the parachute. In
this scenario, when the Avionics Bay and the Parachute Bay fly apart, the
tethers connecting the two (via the NFB) become taut. The parachute is pulled
out of the Parachute Bay and “drops” free. In this scenario, the parachute may not unfurl.
Figure 14: Incorrect tethering of
parachute
Figure 15 illustrates
diagrammatically the correct means
to tether the parachute. When the Avionics Bay and the Parachute Bay fly apart,
the tethers connecting the two (via the NFB) are initially slack as the chute
is released. As the two sections depart with adequate velocity, the slack
tethers become taut, jerking the parachute enough that the canopy opens,
catches air and immediately blossoms. Since implementation of this fix, the
parachute has fully blossomed every time on subsequent Xi flights.
Figure 15: Correct tethering of
parachute
Appendix K provides information on parachute sizing and
loads.
Recovery Deployment Reliability
Now getting back to the question of why many of my early rockets
experienced lawn-dart flight termination. The answer lay in the design and
implementation of both the Recovery Control system and Recovery Deployment
system on these early rockets. Primarily, the fault lay with the Recovery
Control system. After initial, and ill-fated, attempts using inertia-type
switches to sense apogee (such as a mercury switch) I set out to design and
build my own crude (circa 1980’s) ‘flight computer’ which usually consisted of
a timer, alone, or in conjunction with an air-speed
switch. Simply put, my
electronic designs were not very good. The timer circuits proved to be
unreliable. The air-speed switch, when operated alone as a simple switch, proved
to be basically reliable, but had its limitations, such as when the rocket
experienced severe weather-cocking and the rocket did not slow sufficiently at
apogee (having attained significant lateral velocity). The Recovery Deployment
system on those rockets proved to be reasonably reliable, however, suffered
problems as well. Over time, the design of the Recovery Deployment system
evolved to become the system I currently use on all my rockets, which has
proven to be exceptionally reliable. The fallibility of the Recovery Control
system was fully overcome by the introduction and usage of COTS flight
computers.
Importantly, redundancy of
critical components played a key role in reliable safe recovery of my rockets
and striving to eliminate single-point
of failure features. Redundant
components of the rocket recovery systems on my current rockets include:
-
Flight COTS computers (or altimeters). Three fully redundant units
are employed including separate and isolated power supplies. Two different
manufacturers (Xi rocket: two Ravens and one Quark).
-
Tethers. Dual tethers are used for all connections.
-
Pyro charges. Three independent pyros are used for apogee
separation event, one per flight computer, as this is a flight-critical event.
Two independent pyros are used for chute deployment.
-
Eyebolts (for connecting tethers). Specifically, nuts attaching
eyebolts are secondly retained using thread-lock compound.
-
Avionics Bay static ports. Barometric altimeters sense the
pressure within the Avionics Bay compartment, which is maintained at
equilibrium with the local atmospheric pressure during the flight. A minimum of
four holes helps ensure this flight critical feature. As the holes are tiny, one
or more can get unknowingly plugged with dirt or paint.
With regard to other Recovery Deployment methods, redundancy can
also be largely achieved with a CO2 Parachute Deployment system by
having two units mounted side-by-side. However, this can only be done with
a sufficiently large (diameter)
airframe. For example, with the Peregrine CO2 system, a 4” (100mm) diameter,
or larger, AvBay is required. With a mechanical (e.g. spring-based) Recovery
Deployment system, building in redundancy is certainly more difficult. Having a
pyro squib as a backup would certainly work, but rather defeats the purpose of developing
an all-mechanical system.
Ground Testing
The final point of discussion regarding rocket recovery system is ground testing. Although the need for ground testing may
seem obvious, it is important to ensure that the testing will achieve its goal.
The goal being to help ensure that the rocket recovery system is sufficiently robust to work every time a
rocket is launched. Consider the
dictionary definition of the word robust:
1. quality of being
strong and unlikely to break or fail
2. able to survive being
used a lot
3. effective in all or
most situations and conditions
All three of these definitions of robustness are applicable to the
design, construction and to the process
of ground testing of the system. Ground testing
should duplicate as closely as possible the operation of the system in flight.
Performing a ground test should confirm that the Recovery Deployment system
works as expected and components neither fail or experience damage. Or suffer
signs of wear or weakness such that after repeated usage, the system continues
to perform nominally. Definition (3.) is one aspect of ground testing that is
perhaps less obvious, and one that applies with particular emphasis to both the
Recovery Control system and the Recovery Deployment system. One must give
careful thought to what possible off-nominal situations
could arise that may affect the functionality or reliability of the recovery
system. As well, what are the conditions that
may be encountered that may likewise affect the functionality or reliability of
the recovery system?
An example of an off-nominal situation might be the amount of
ejection charge material used in the Recovery Deployment system for a flight (for
instance, by measurement error). Ground
testing should anticipate this by conducting the deployment test with a minimum quantity. It may also be prudent to conduct a
deployment test with a maximum
quantity to ensure that the extra amount does not result in physical or heat
damage to any components, or result in breakage of tethers. On the subject of
tethers, ground testing should be conducted with a single
tether of suitable strength. This will ensure true redundancy assumed through
the use of a pair of identical tethers in
flight.
With regard to conditions that
may be encountered during a launch that should be taken into consideration with
regard to a ground test plan include:
1.
Launch site ambient temperature range expected to be acceptable
for launch.
2.
Ambient humidity.
3.
Reduced ambient temperature and pressure at altitude.
4.
Peak acceleration during launch and its effect on recovery system
components.
With regard to the first point, it may not be practical to do all-up ground testing at the coldest or hottest temperature
extreme. However, this condition can be taken into account by testing
components that may be affected by temperature extremes (so-called detail testing). An example is plastic components. My own
ground testing revealed significant weakness in certain plastics such as PVC at
sub-zero temperatures. Another example is electronics, such as power supplies
and flight computers. Cold-temperature ground testing of candidate batteries
(to power the electronics) can be done by placing the battery in a refrigerator
or freezer for a suitable length of time. This is how I learned that alkaline
batteries have greatly diminished output at low (sub-zero C.) temperatures. The
same testing method confirmed that primary Lithium cells perform well at -25°C. Electronic devices may also be affected by
temperature extremes. Importantly, a flight computer may not function reliably
in the cold (or extreme heat). Testing of flight computers can be done in a
manner similar to that done for batteries. Before being used for cold-weather
launches, I tested the Raven 3 unit
that I had originally used for my Xi rocket. It was left for 12 hours at -20°C then tested for functionality at that
temperature. It was found to be unaffected by the cold. On the other hand, the BREO altimeter unit that I used as a backup experienced
reduced functionality at a temperature below 0°C.
Ambient humidity may affect components of a rocket recovery system
in an adverse way. Black Powder and Crimson powder are both somewhat
hygroscopic. It may be wise to conduct all-up testing using pyros that have
been conditioned at a certain humidity level (even though pyro devices should
be hermetically sealed as good practice). Composite and plastic materials
absorb moisture to some extent, changing their mechanical properties -- another
candidate for detail ground testing. When employing nylon shear screws, the
effect of both temperature and humidity on the shear strength (which is related to tensile
strength) is significant. This needs to be taken into consideration when
assessing the results of ground testing of the recovery deployment system, as
well as during the design phase.
For high-altitude flights, say greater than 10k feet (3 km), the
ambient air pressure
and temperature are markedly lower.
Pyro devices may be less reliable and may not combust fully (due to reduced
heat transfer). This should be taken into consideration when ground testing.
Delay grains may burn slower at reduced ambient pressure and may even
extinguish. Under consideration as a delay element, RNX has been tested at reduced pressure and found to burn much slower
and tended to self-extinguish at a critically reduced level. CO2
-based Recovery Deployment systems, if being considered for extreme altitude
flight, should be tested at reduced ambient pressure and temperature. When a
cartridge is activated, liquid CO2 must evaporate in order to
generate pressure, and thereby absorb heat from its surroundings. Ambient air
at extreme altitude has low density, therefore much lower heat capacity
(compared to ground level air), retarding its ability to give up heat,
especially when cold to begin with. This, plus reduced vapour pressure at cold
temperature (as mentioned earlier) may reduce the effectiveness of such a
deployment system.
Finally, peak acceleration during powered flight can affect the
recovery system of a rocket, sometimes in an unexpected manner. A parachute and
tether system has mass, therefore, high acceleration acting upon these
components can cause shifting (aftward) within the rocket. A parachute can
become compacted and consequently not pull out of the parachute bay as readily
as assumed. Ground testing can replicate this effect by intentionally packing a
parachute in a less-than-ideal form. Any other items of mass need to be
assessed in a similar way.
All-up ground testing can reveal unexpected flaws in a Recovery
Deployment system design. This happened when I began ground testing the Xi parachute
deployment system. Although, the pyro charge separated the AvBay from the
forward section of the rocket with abundant force, and the NFB pulled out as
expected, the parachute was barely extracted. Investigating this odd anomaly,
the cause was soon apparent. The parachute acted like a piston as it began to
be pulled out of the parachute bay. The result was a partial vacuum which
generated a force (due to ambient air pressure) opposing the motion of the
chute. This issue was readily solved by adding vent holes in the parachute bay
to allow air to enter to fill the volume vacated by the chute (see Figure 9).
Resources
Res. R1 Using Ejection Charge Baffles, Tim Van
Milligan, Peak of Flight Newsletter, Issue 129,
August 10, 2004
Res. R2 Parachute
recovery system design for large rocket vehicles, Rick Newlands,
Aspire Space Technical paper
Res. R3 Design and application of a parachute deployment
mechanism for sounding rockets based on commonly available and affordable
components, Artur
Kłosiński & Michał Jasztal, Journal of KONBiN 2024, Volume
54, Issue 1
Res. R4 Techniques for Selection and Analysis of Parachute Deployment
Systems, Earle K. Huckins III, NASA TN-D-5619, January 1970
Res. R5 Tinder Rocketry’s – Peregrine Exhaustless CO2
Ejection System (user guide)
Res. R6 Nylon Fasteners
, Test Data Sheet, Micro Plastics (Chart 1) (Chart 2)
Videos
1. Static firing of A-100M motor fitted with Pyro-DED for parachute deployment
2. Alpha rocket cold-weather
ground test of
parachute deployment
Next-- Weight Control
Last updated December
24, 2024
Originally posted October
4, 2024