LAUNCH VEHICLE SYSTEM ENGINEERING & INTERFACES
- LESSONS LEARNED -
(N.Narayanamoorthy)
Associate Director (Projects) Retd.
Senior Advisor (Launch Vehicles)
Vikram Sarabhai Space Centre
Thiruvananthapuram- 695022
“Space is unforgiving, thousands of good decisions can be undone by a single engineering flaw
or workmanship error, and these flaws and errors can result in catastrophe”
INTRODUCTION
The Indian Space Program has carved a niche for itself with unique application driven programs
oriented towards National development over the last four decades. The success of these space
missions is based on established practices and procedures augmented by the lessons learned
during the entire life cycle of the launch vehicle project encompassing design, development,
realisation, assembly, testing, integration and launch.
As failure can propagate from virtually every critical activity in a space system life cycle, the
lessons learned perspective reminds one, of the need for thorough and complete understanding of
the space systems involved, proper interface management and assurance process implementation
over the life of a project or mission.
SYSTEMS ENGINEERING
Systems engineering is an interdisciplinary field of engineering that focuses on how complex
projects should be designed and managed over the life cycle of the project. This interdisciplinary
approach is a means to enable the realization of successful systems. It focuses on defining
interfaces and required functionality early in the development cycle, documenting requirements,
then proceeding with design synthesis and system validation.
VEHICLE SYSTEMS ENGINEERING & INTERFACES
Understanding of the Launch vehicle systems and related interfaces is important to achieve a
reliable launch vehicle. Defining the vehicle architecture, understanding the interdependencies of
each system and interface control are the essential constituents of vehicle system engineering.
A launch vehicle is constituted by various technological elements such as the Mission design,
which addresses the aerodynamics trajectory design, guidance, control system design and flight
software. The structural system is the skeleton of the vehicle with inter-stage structures, motor
cases and propellant tanks constituting these systems. The control actuators and associated
1
electronics, control and steer the flight Stage separation, ignition, destruction systems are the
auxiliary systems deploying mechanisms and using pyro elements in their functioning.
The mission computers, inertial navigation, guidance and control systems, telemetry, tracking,
tele-command and power systems are the major elements of vehicle avionics. Integration and
checkout of the vehicle addresses the stage by stage assembly and testing of the vehicle systems
to ensure satisfactory performance before launch.
Ground support systems include the down range ground stations for tracking the vehicle and
receiving the telemetry data. In the launch complex facilities, assembly of stage sub-systems,
vehicle integration, propellant servicing, launch processing and launch control are established.
The System engineering tasks interface each of these vital areas with a view to proper definition
and execution during the vehicle development, realisation and operation.
Systems engineering challenges
A launch vehicle is a complex safety critical system whose safety cannot be demonstrated solely
by test, its system interdependencies are difficult to comprehend without analytical tools and
might directly or indirectly put human lives at risk.
The stringent requirements as well as the development and testing of these critical systems adds
complexity to standard engineering activities as it requires consideration of all diverse conditions
of failure such that the system can function in the event of identified failures or operational
errors. This requires adoption of formal methods of safety and risk assessment in the
development process and an integration of risk assessment and design in the system engineering
process as well as designing strategies to manage hazard and risk such as
Designing for minimum risk
Incorporation of safety features and devices
Developing procedures and training
Learning from past experience
Thus the additional challenge in vehicle system engineering is integration of safety and risk
assessment as well as developing and testing associated strategies.
FAILURES AND LESSONS LEARNED
Launch failures have been a fact of life for most space-faring nations since the space age began
in 1957. A space mission involving a launch vehicle and a sophisticated satellite is a costly
process, the cost of experience being extremely high. Hence the emphasis on “Right the first
time, every time”
2
A lot of the most valuable information for space engineers is derived from experience. The
basics are easily found in data sheets, applications notes and text books. The hard things are
learned by failures. While unfortunate, they will happen. What is critical, is that they do not
happen again.
Any lessons learned from past experiences should serve to mitigate the deviations and failures in
the future with a view to improving the success rate and ensuring uninterrupted dissemination of
space based services.
PERSPECTIVE FROM HERITAGE ISRO LAUNCHERS
The development of SLV-3 and ASLV and the subsequent lessons learned played a significant
role in mastering the technologies for the successive PSLV, GSLV and GSLV MK-III
programmes.
Some of the lessons learned from the development of SLV-3 and ASLV are worth mentioning.
The SLV-3 development established a good foundation in the solid propulsion technology,
development of control systems based on fin-tip deflection control, SITVC, on-board computer
systems, Inertial navigation systems, orbital design and mission management. The first flight of
SLV-3 could not achieve the mission goals, the reason being identified to the second stage
control system. This established the need of clean rooms for precision assemblies.
The experience of the failure and subsequent analysis established a rigorous design audit and
review, as well as strengthened many weak areas. The ASLV introduced many new concepts
including strap-on technology required for PSLV. A closed loop guidance system was
introduced which improved the injection accuracy of the satellite. Real Time Decision making
(RTD) using on-board measurements was introduced. The rocket was vertically integrated and
new concepts of assembly and testing were introduced.
The inadequacy of the control transfer algorithm between the first and second stage in ASLV-D2
led to vehicle break up. The corrective action leading to improvisation in mission design was one
of the most significant lessons learned from the ASLV which ensured a robust design of the
PSLV vehicle.
LESSONS LEARNED - CASE STUDIES
A few salient case studies drawn from heritage missions and current launchers are
addressed in this article, especially on interface design as well as system engineering aspects.
3
EXPERIENCE DURING SLV3 PROGRAMME
The SLV3 Realisation Phase was a real learning ground for all of us as most of the Team
Members associated with the programme was fresh from the college. As part of SLV3
Realisation, sounding rocket flights along with certain SLV3 Avionics Systems were planned
initially. In fact this activity enabled us to understand many major issues which are relevant even
today.
1. LOOKING FROM TOP OR BOTTOM
A Monopropellant Reaction Control System was planned in one of the Centaur missions
to reduce the spin rate by firing the thrusters. After the assembly was completed, during an
explanation of the system, Shri Aravamudan, the then Director, TERLS remarked, “If this is
how you have planned, then the spin rate will further go up”. All were stunned to hear the
remark and the folly was understood. In defining the roll axes, R+ and R-, the error had
happened. For the system, the roll direction was defined looking from Top, where as the rocket
Roll was defined as “Looking from Bottom”
Lesson:
Roll Directions to be specifically addressed uniformly in all documents.
2. ROLE OF SOURCE IMPEDANCE IN TELEMETRY SYSTEM
In the earlier days of ISRO, we were using Analog FM/FM Telemetry along with PAM
commutator. As part of Technology Development, it was decided to fly a 16 channel PCM
Telemetry System in Centaur Rocket. As part of interfacing the signal with the Telemetry
system based on the text book knowledge, we implemented a pre-sampling filter for every
channel. This was done as part of Integration. During testing, to our surprise we found what
ever the input voltage, the output showed full scale. Whenever the input to the Telemetry
System was fed from power supply every thing is normal. The issue was puzzling us for all of
us for almost one week. Then based on the Literature Survey of similar circuits, we could
understand the problem. Solution was simple – The source impedance has to be limited for
High speed multiplexers.
Lesson Learned:
Before sampling a channel, the previous channel input has to be totally discharged
through the source. If the source impedance is high and the next input is sampled, as the
discharge is incomplete, the data will be erroneous. This was a lesson for all of us and even
4
today the source impedance (1kΩ & 600 pf) spec is followed for Telemetry System
measurements.
3. RESPONSE OF SOLENOID VALVES
Everybody is aware that when power is interrupted to a solenoid valve, a high back emf
is generated (-L di/dt). The voltage can damage the Electronic Driver circuits and will generate
interference. As per the usual practice a surge suppression diode was connected across the
solenoid valve. The RCS Design team came out complaining that the response time (off time)
was not acceptable. Then the efforts to reduce the off time resulted in the introduction of Zener
diode in series with the diode. This configuration is being still followed where off time response
is to be controlled. Also the procedure for measurement of ‘ON’ time and off time for the valve
was evolved in this process.
4. GROUND TEST OF SECOND STAGE CONTROL SYSTEM – GIVE RESPECT
TO SAFETY NORMS
The SLV3 second stage Bi-Propellant system was being made ready for static firing in
TERLS. The fuelling of RFNA was in progress. In those days, due to over enthusiasm, more
than 50 people were crowding around the Test article in spite of repeated warning from Safety
Officer Mr.Gurjar. Suddenly we heard a big blast and splashing of RFNA all around. Many
persons were severely injured. This incident really brought to focus the importance of adherence
to safety practices and also design improvements to be carried out in fibre glass pressurant tanks.
5. NEVER OPERATE CONSOLES WITHOUT AUTHORISATION / TRAINING
During second stage control system pressurization test at SHAR, the Gas Bottle was
pressurized to 210 bar. The Hardware was located in Integration bay and the control of the
solenoid valve and pressure monitoring was being done from Block house. The snoop test of the
system was progressing to ensure leak tightness. The pressure was being regularly
communicated through Intercom from Block house to Test bay by the console operator. The
5
console operator moved out of the console for attending a telephone call and hence did not
attend to queries from Test bay. A fresh Engineer who was not connected with this activity
thought it prudent for him to answer the call. But instead of simply reading pressure from the
display of communicator he did something more.
He operated the console switch and then started announcing the pressure. There was
panic in the Test bay because of a big sound and gas leakage. People started running for their
lives as the switch that was operated happened to be vent valve which released the gas with a
bang!
Lesson : Never operate an equipment or console without proper authorization or training.
6. LESSONS FROM SLV3 E1 FAILURE
The first failure of SLV3 was a lesson to all of us on many fronts. During the Launch
countdown and final operations, an operation called “Warm up Pulsing” command is issued
from Checkout to test the control system thrusters of both second stage and third stage.
Immediately after warm up pulsing sequence was completed, the gas pressure of S2 control
system came down suddenly from 220 bar to 180 bar. There was a commotion and debate
started whether to stop the launch. But based on opinion sought from the experts that 180 bar
was adequate to meet the launch requirements, the mission was cleared. However, the mission
failed during second stage regime due to loss of control. On inspection, the launch pad had
traces of RFNA splashing, clearly indicating that RFNA had leaked. The problem was traced
due to permanent opening of the solenoid valves in the RFNA circuit. Several lessons were
learned from this experience. Let us look into them.
(1) The control system was a blow down system using a bladder. The gas is filled between
bladder and the casing where as the propellant is filled inside the bladder. So what happens? If
the liquid totally leaks, the gas will occupy the space thereby drastically reducing the pressure.
This failure mode was overlooked during clearance for the flight.
Hence, any pressure monitoring anomaly shall be well understood before proceeding
with flight.
(2) The post flight analysis of the Video Recording, gave clear indication of leakage of
RFNA through the valve. Then, how did we miss it? Only one monitor was used for
monitoring S2 &S3 control system firing during launch countdown. Hence immediately after
S2 warm up firing operation, the monitor was zoomed to S3 location. Subsequently for all
flights the surveillance and observation system was strengthened.
6
(3) Why did the solenoid valve fail?
The solenoid valve opening is of the order of microns and any contamination inside can
jam the piston. Subsequent to the event, clean rooms of class 10,000 and class 1,00,000/- were
set up for assembly for control components to ensure contamination free assembly.
7. CRIMP TOOL INSPECTION AND PULL TEST OF CRIMP SAMPLES
During the post flight analysis of SLV3 C1 flight, it was noticed that one half of the
solenoid valve for control system was not operating due to power supply not reaching the
valves. In spite of our best efforts we could not identify a cause for this failure. But during the
Harness preparation of the next flight, the probable cause could be correlated. It was noticed
that the crimp joint of a circular connector was undercrimped and the cable was coming out of
the crimp barrel easily with a slight pull on the cable. Reason was established as due to the
defective crimp tool. The previous flight harness was also carried out using the same crimping
tool and some cables could have come out of the crimp barrel during flight, resulting in the
failure mode.
Lessons Learned:
Crimping Tool Inspection was tightened and subsequent to the observation, sample level
pull test in the crimp joints (3 for every 100) was introduced. Also inspection procedures and
training the personnel for inspection of the joints during crimping helped to eliminate such
problems.
8. EXPERIENCE OF ASLV-D1 – IGNITION CIRCUIT INTERFACE
The ASLV configuration consists of two S9 strapon motors and one S9 core stage as part
of the first stage. As per the flight safety norms, in case destruct command is to be issued during
strapon firing regime, it should be preceded by S9 core stage ignition. This means S9 core stage
ignition squibs are to be commanded by either tele-command decoder or on-board sequencer. A
flaw in interfacing the two systems for squib firing was noticed during ASLV-D1 post flight
analysis.
Diode not
present in
implementation
Bat Rtn
Telecommand
Squib
PYRO BAT +ve 1 Ohm
Sequencer
Bat Rtn
Diode not
present in
implementation
7
As a design practice in tele-command decoder, the squibs are kept shorted in the safe
condition. This has resulted in current sharing during commanding of the squibs through
sequencer.
Analysis:
While interfacing two systems, isolation diode should have been used and this was over
looked. Also the configuration of relay circuit within the tele-command decoder was not
considered. The ground testing of the system could not bring out the defect as the squib was
firing even in this condition as no squib current monitoring existed those days.
Lesson Learned:
Sneak path analysis of all interface circuits introduced as part of electrical integration
circuit review.
Squib current monitoring during ground test.
PSLV EXPERIENCE
9. WIRE LOCKING – BE CAREFUL
Many joints are wire locked to avoid loosening of the joints. Though it is a good
practice for mechanical integration, one has to be extremely careful when wire locking is done in
the vicinity of cable bunches. One such incident occurred during PSLV C5 Assembly operation.
The satellite Electrical Umbilical Cable is interfaced to satellite separation connector
through a cable assembly which is passing through equipment bay. This cable is normally
installed during satellite assembly operation. After the Heat shield was closed, when the
satellite was put on, one of the supply lines was shorting. All efforts for locating the fault could
not succeed for almost 12 hours as there was no access to satellite and equipment bay. On
further inspection through a cut out in the heat shield, we observed that the cable bunch along
with the Thermal Protection tape was wire locked for anchoring. On close inspection, it was
noticed at one of the wire locking has pierced the insulation and this has resulted in the short.
Lessons Learned:
Do not use wire locking for anchoring cables.
The edges after wire locking shall be bent to avoid sharp edges.
8
10. PRESSURE MEASUREMENT RESPONSE DURING PS2 SHUT-OFF
The PS2 stage chamber pressure was generally monitored using two types of
Transducers. (i.e. one with LPSC make 21 NA and the other by BHL- USA). In variably the
PS2 shut off transient measured by BHL was higher than that measured by 21 NA. The
response measured at satellite was corresponding to the transient recorded by BHL. Question
was frequently asked why 21 NA transient should record lower valves. Finally the question was
answered. On inspecting the pressure Transducer Mounting Scheme, it was noticed that the
BHL Transducer was mounted to the chamber with a short plumbing where as 21 NA was
mounted with another plumbing which has two more branches to mount Differential Pressure
Transducers which had a big Dead Volume. The increased Dead Volume in the 21 NA scheme
has damped out all oscillations resulting in Lower Transient Recording.
Lessons learned:
The following points need to be addressed for Response characteristics of the pressure
measurements.
Canalization Tube Volume
Dead Volume of the transducer.
Effect of oil or grease which is used to reduce the transducer thermal
environment.
11. THE EXPERIENCE OF PYRO ARMING FOR PSLV-D1
As part of the final countdown operations, pyro arming operations are done to connect
the command output to the respective explosive circuits. To check the integrity of connector
mating, connector mating status is introduced in each ‘Arm’ connector. To our surprise after
Arming operations were completed, it was noticed that Arming connector status was not
available in Telemetry for PS3 stage functions. Also loading of +5V supply was noticed.
PSLV D1 vehicle was fully ready for the final operation and one can visualize the
commotion at the instance. Immediately the Integration Harness details including the PS4/PS3
interface circuits were verified. It was clearly understood that the mandatory current limiting
resistor for monitoring was not used either in PS4 or PS3 stage. This simple error caused major
event of postponing the Launch by a day as well as re-verification of all Arming Connectors.
9
Lesson:
The interface design detailing between different stages were thoroughly reviewed
including preparations of chain diagram.
The connector status monitoring scheme was introduced as part of the initial
level vehicle checks.
12. CONTROL SYSTEM
During PSLV D1 flight, there was a constant engine deflection of 0.7 degrees in the
second stage. Yaw error build up was seen which was not accounted for. This was analysed and
found to be due to the deflection of the thrust frame during the thrusting of the liquid engine
which is to be accounted in the Control system design.
Lesson learned:
Deflection of thrust frame has to be accounted for in the engine control system during the
thrusting phase.
13. BE SPECIFIC IN YOUR COMMUNICATION
The PSLV C5 countdown operations were in progress. During the first stage stage
control system Gas Bottle pressurization, suddenly the control system consoles started showing
anomalous readings. On rushing to the Launch Pad, we observed that all the Base shroud
Electrical Umbilicals were in the demated condition. How did it happen? The incident
happened due to a communication error as well as lack of understanding of the system by the
Technician. Normally the Base shroud Housings are covered by an antistatic cloth which is
black in color.
As part of the final preparation the concerned Engineer had instructed the Technician to
remove “Black Items” from the Base shroud assembly. Unfortunately the Technician removed
the umbilical connector also which were Black in color!
Lesson: Be specific in communication
10
14. GRAPHITE – PSOM NOZZLE
In PSLV D2 flight , one of the PS0M motors showed sudden increase in pressure of the order of
2 KSC at burn out before start of tail off and sharp decay thereafter. This has resulted in vehicle
roll disturbance calling for PS0M SITVC into operation.
This was attributed to failure and ejection of a portion of graphite throat insert, recommended
use of higher-grade graphite.
During PSLV C2 flight, few pieces of graphite (0.7 kg) were recovered from the vicinity of
launch pad after launch. Pressure spike was also seen in one of the PS0M at 7.9 sec
Graphite failure at the stepped portion has caused this observation. Redesigned the graphite
throat to convergent liner joint and introduced UT inspection for any flaw in the graphite.
Lesson learned :
• Interface design shall consider the thermal behavior of materials involved
• Mechanical properties of the materials at operation temperature shall be assessed in
addition to room temperature
• All available NDT techniques shall be employed for assessing the integrity of materials
even at product level
15. UNDERSTANDING DEVICE BEHAVIOUR DURING POWER-OFF
The output state of digital circuits change state during power off and at low voltages. During
PSLV vehicle level checks, the SCOUT package was used for commanding and data acquisition.
This used 5V supply for digital devices and 28V for the commanding relays.
Observation was that during the switching off sequence, commands were getting inadvertently
exercised.
The reason was during the switching off of the 5V supply, the buffer circuits changed their state
at a lower voltage of 2.5V thereby turning on the relay drivers.
Lesson
Device characteristics at power off are to be understood.To protect against a failure of supply
voltage, the output circuits are disabled from generating erroneous commands.
11
LESSONS LEARNED IN THE DEVELOPMENT OF GSLV MK-III
During the evolution and testing of GSLV Mk-III systems and structures, lessons are being
learned and understood. Some of the important lessons are:
16. Vehicle configuration definition & payload growth
The configuration of GSLV MK-III was evolved during 2000-2002 period where the mission
studies and preliminary designs of propulsion systems and structures were carried out by LVDE.
Some of the configuration changes made at the implementation phase give us an opportunity to
understand the need of detailed design while finalizing the configuration. The payload capability
of the vehicle was 4500 kg to GTO with sufficient growth options. This was with C25 stage with
25 ton propellant loading and two burn sequence of C25. The present payload capability in only
3600 kg with no major payload improvement options but for significant changes in propulsion
systems. The lower stages of S200 and L100 together were supposed to impart a velocity of
about 5.5 km/s and also meet the impact constraints. While understanding the CUS, it was
assessed that two burn sequence of C25 as risky and should not be attempted in the first phase.
This brought down the payload to 4000 kg. To bring back the payload to 4000 kg considering
further drop due the increase in the mass of C25 as well as the lower stages, 25 ton propellant
loading was increased to 27 tons.
The assessment of inert mass of the engine and stage was on the lower side and increase in inert
mass of C25 resulted in further drop in the payload capability. The area ratio of the nozzle of
CE20 engine was reduced from 150 to 100, because of realisation problems anticipated. This
brought down the ISP from 450s to 435 s. Even after payload improvement options were
attempted such as composite PLF, ultimate strength based design of propellant tanks, increased
propellant loading in L100 to L110, were offset by the increase in inert mass. Presently the
capability of 3600 kg can be improved only by ensuring the mass budget of C25. Further
payload improvement in this configuration is feasible by two burn sequence and change in
launch azimuth.
The lesson learned in this process is that, while configuration studies are carried out, sufficient
margins are to be built in to handle inert mass increase resulting from detailed design of
structures and matured options in various sub-systems only shall be adopted even for a new
vehicle. (FLSC for Cryo temperature, Cryo interfacing structures FSA, ASA not conceived
earlier).
17. S200 GRAIN DESIGN
The propellant grain of S200 is designed with an identical concept as followed in S139, but
with improvements in nozzle end segment to accommodate the submerged nozzle and also to
12
have a long tail off and shape the tail off thrust so as to have the differential between two S200
to be minimised. The head end segment has a 10 point slotted grain to obtain the required initial
thrust to meet the lift off thrust requirements. The igniter chosen for S200 is similar to S139 but
with higher mass flow rate to match the increased surface area of S200. After the first static test
of S200, it was observed that the initial thrust is lower than expected and thrust after the first
peak thrust is higher. Mission studies with this thrust time profile showed that the design limits
of dynamic pressure are violated. The nominal dynamic pressure is 35 kPa and upper bound is
38.5 kPa. With this thrust time profile, the upper bound of dynamic pressure exceeded 40 kPa.
Hence a redesign of the head end segment grain with a 13 point star having higher surface area
and thus higher initial thrust and lower thrust during maximum dynamic pressure regime. This
change has resulted in increased payload, higher take off thrust and brought back the dynamic
pressure within limits. The new grain shape has improved structural margins compared to the
earlier grain. There are no other changes including igniter and nozzle to make this change.
However one more static test of S200 is necessary to qualify the new design.
The lesson learned here is that while the proven designs were adopted for the grain design, the
dispersions considered were not adequate to guard against the violation of mission
specifications.
18. DESIGN OF STRUCTURES OF L110 & FLANGED INTERFACES
The design of the structures of L110 was based on the proven methods of PSLV and GSLV
adopting the skin stringer concept with external flanged joints. The design of Core base shroud
which is to transfer the thrust of L110 was originally of skin stiffened shell. Seeing the
advantage of configuration, it was quickly changed to isogrid construction, which was proved
right in meeting the strength and stiffness requirements. But the design of IS1/2L, ITS and
IS1/2U were based on external stringer designs. This design is suitable only for a predominantly
compressive loaded structures. But the tensile loads were dominant. However successful optimal
designs evolved. During the structural testing a new problem surfaced. The failure of IS1/2L
brought to picture the sensitivity of fastening in creating unsafe situations in tensile loaded
structures. This skin stringer structure uses plenty of rivets and jo-bolts. The loads on each of
them were to be assessed to ensure good margins especially near cutouts and critical joints. In
PSLV and GSLV the design loads were seldom faced by the structures in flight due to lower
angles of attack and lower dynamic pressures. But in GSLV MK-III, the tensile loads are
generated by the inertial loads which are developed to the full level in every flight. Thus the
quality of each fasteners became critical.
Another issue was in the design of the flange joints. In PSLV and GSLV, the flange joints were
not as critical as in GSLV MK-III due to primarily being in compressive loaded state. Under
tensile loads, due to prying the fastener loads were very high. The number of interface bolts is to
be doubled from the original plan to meet the required margins. In some joints 1450 class
fasteners and higher size (M14 instead of M12 in IS1/2L and IS1/2M joints and M16 in IS1/2U
FWD joint) are to be adopted to get the required margins.
13
A safer and better design would be to adopt isogrid of waffle design with tongue and groove
joints at the interface. Fewer fasteners are required in this design and lower eccentricity in joints
will result in lower fastener loads. This design can be adopted for human rated GSLV MK-III
while the re-design is being attempted.
This lesson is one of the important one we learned during the process of detailed design.
19. CIRCULAR SHAPED FLSC DESIGN
Circular shaped FLSC was adopted for severing the links in the S200 separation system. This
system is common for the forward bolt as well as the aft links. The diameter of the FLSC here is
about 100 mm. The design was adopted form FLSC used in PSLV. The charge density was
increased from 7g/m length to about 14 g /m length to meet the cutting requirements. All tests in
the linear modes were successful. However when the curved shaped FLSC is made, failure to cut
the thickness is observed.
When this was analysed, it is observed that the cone angle of the FLSC is increasing while the
FLSC is rolled to circular shape from linear shape. This change in cone angle reduces the
efficiency of the cord.
The cone angle was then reduced and the processing of the cord was changed to swaging form
rolling. This has brought improvements in the performance of the FLSC.
There were other lessons learned in this design. These include the toughness specification of the
material to be defined in addition to strength. The sensitivity of initiation at the head instead of
end was also brought out by a failure in one test.
The lessons learned in the development of FLSC system is very important and is typical of any
such new developments. These are to be catalogued for the future work.
20. C25 LOX TANK WITH CROWN CONE AND NEW THRUST FRAME
The C25 LOX tank was configured with a crown cone which is integral to the LOX tank. The
engine would be attached to a machined interface here. No separate thrust frame was envisaged.
To manufacture the crown cone a large integral forging of AA 2219 material was required. The
realisation of such a forging was faced with difficulties. The conical shape was also found not
suitable for the tank design. Thus a decision was taken to design a separate thrust frame for C25.
This decision helped along with the decision for common domes for C25 and L110 helped in
adopting stretch forming of all dome petals for GSLV MK-III tanks instead of pressing/
spinning.
The lesson learned is that feasibility study was not conducted in the first phase while the crown
cone arrangement was chosen, and all impacts assessed.
14
HSP SYSTEM ENGINEERING - LESSONS LEARNED
21. CONFIGURATION OF CREW ESCAPE SYSTEM
The basic challenge in the configuration design of CES is to ensure its aerodynamic static
stability (c.g of vehicle to be ahead of the aerodynamic centre of pressure) throughout its flight
regime (from subsonic to M=6). It must be remembered that the massive Crew Module weighing
~ 4 t is at the base of the vehicle pushing the c.g to almost down to 70 % of its length. Now
placing the aerodynamic centre of pressure in the remaining 30 % of the vehicle length (to
ensure static stability), is the difficulty. The original external configuration conceived to perform
the task is shown in Figure.
Figure .Original CES
It was thought that the conical portion would provide required stability to place centre of
pressure below c.g. It definitely does, as expected from our initial assessments, but only when
the jets issuing out of the 4 reverse flow nozzles are off. To our surprise, under jet on conditions,
we found that the entire lift force generated by the conical portion is lost as the jet completely
surrounded the cone.
Lesson 1 : Need additional lift producing surfaces at the aft of the vehicle to improve stability
under jet on conditions.
Solution 1 : After studying many options, 4 grid fin configuration was finalised. This
configuration provides stability for the vehicle under jet on conditions only up to M=2.
Lesson 2 : Grid fins failed to provide the required stability beyond M=2. The escape motor
nozzle exit pressure is ~ 6 bar. This pressure is significantly high compared with the external
aerodynamic pressure at M=2 (altitude ~ 17 km). Jet flow dominates external aerodynamics and
therefore stability contribution from conical and grid fins are overpowered by the jets. Attempts
were made to circumvent the problem by using high burn rate propellants in order to reduce the
motor operating pressure and nozzle exit pressure to ~ 1 bar. But this was not practicable at that
point of time as this will demand development and testing of new propellants.
Solution 2 :
15
Examined escape system configuration of other vehicles (Apollo, Soyuz and Shenzhou).
At one stage of the design we were studying more of Soyuz and Shenzhou than our own design
for HSP- CES. Later, we realised that Soyuz and Shenzhou have two sets of escape motors, one
for use in case of aborts at low altitudes and another set (with lesser thrust levels)for use at
higher altitudes. Surprisingly Apollo had only a single set of motors. Later after detailed survey
of Apollo literature, we learnt that Apollo escape system too is unstable beyond M=3 (this was
clear from the experimental data documents released by NASA). However, the risk was
acceptable to the management, and began accepting black points (points along Saturn vehicle’s
trajectory from where no escape is possible) in their eagerness to win over the Russians in the
race to moon.
Learning from Soyuz and Shenzhou : For HSP - Crew Escape System, 4 motors (HEM) to
provide escape at High altitudes were designed.
The final configuration of CES is shown in Figure
Figure Final CES configuration
22. CONFIGURATION OF CREW MODULE FOR MONO-STABLE NATURE
Modules of different shapes and dimensions have been designed and employed for
successful human spaceflight missions. For the design of the Crew module, it would require a
weighting and compromise of the various desired characteristics to produce a “best” set of
vehicle shape parameters. As in a normal design practice, different study configurations were
analysed and configuration 2 was chosen.
16
The one single feature of Soyuz and Shenzhou module that is strikingly different from
that flown by US (Mercury, Gemini and Apollo) is their mono-stable nature. This nature helps a
great deal during their Crew Escape missions, where a simpler abort sequence is adopted leading
to proper attitude of Crew module for parachute deployment (Compare this with Apollo abort
sequence where additional aerodynamic surfaces (Canard) are deployed to re-orient the Crew
module to the forward facing side to enable parachute deployment).
Lesson learned: Realised the need for monostable nature of Crew module.
We learnt that Soyuz and Shenzhou achieve this nature by use of external fixed aerodynamic
surfaces on the crew Module. Apollo too in their initial period of development attempted such
ideas and finally shelved it as it would demand additional thermal protection system.
HSP – Crew module by virtue of its external configuration presents monostability at Supersonic
and Hypersonic Mach numbers. However, it does not at subsonic Mach numbers. The fact that
the capsule is monostable at Hypersonic entry mach numbers is adequate.
Based on the case studies given above, it is important to note that thorough understanding
of system engineering and interface design aspects is essential for success of any mission.
Also it is important to imbibe following personal attributes for effective management of the
mission: -
Understanding team cultures across disciplines
Being open in communicating issues
Building and preserving a sense of ownership and creating passion
Training your team
BARRIERS TO LESSONS LEARNED
There are few barriers to lessons learned which are to be overcome to facilitate its usefulness to
on-going and successive programme:
Reluctance to share failure experiences
Neglecting lessons learned from past projects due to busy schedules
Overlooking a lesson which may not be widely known
APPLYING LESSONS LEARNED
Many lessons learned have common themes. Hence application of these issues learned should be
carried out by systematically infusing this knowledge into programs and reviews so that they are
not lessons forgotten. Some of these lessons are amenable for codifying. A systemized database
and archival of such lessons learned is a necessity for ongoing missions and projects.
17
CONCLUSION
Space engineers in ISRO will face similar issues during the various phases of the life cycle of
their projects. Hence it is important to raise awareness of the impact similar issues can have on
the overall technical-development effort in ISRO programs. This awareness is the first step
toward developing a mind-set that proactively scans the horizon for these threats, and learning
the skills and approaches that help the teams and in turn the organization to mitigate and address
them as they occur.
The more prepared a team or an individual is, to identify and address these issues as they arise,
the greater the likelihood that they can be dealt with before they significantly damage the project
or development effort.
“People learn by experience, we should profit by other’s experience”
*******
Acknowledgement
The author wishes to acknowledge the valuable inputs and suggestions provided by Shri Sudhir
G.Verghese, APD, GSLV Mk-III, Shri S.Somanath, PD, GSLV Mk-III and Shri S Unnikrishnan
Nair, PD, HSP towards preparation of this article
18
