Mars Science Laboratory Project

1
Mars Science Laboratory Project
Entry, Descent, and LandingSystem Engineering
Challenge Adam SteltznerPhase Lead and
Development Manager,Entry, Descent and Landing
November 10, 2009NASA GSFC
2
Overview

• MSL is a NASA flagship mission to the surface of
  Mars
• MSL will delivery the largest rover ever to the
  surface of another planet
• MSL was scheduled to launch in last month (Oct.
  2009)
• Launch slipped to November 2011
• The MSL Entry, Descent and Landing system design
  has not proven itself yet
• But it has successfully navigated a challenging
  development program (575-950)
• This presentation is focused on the system
  engineering challenges of the design and
  development process for MSL EDL
• How to find the right architecture
• Given capabilities driven performance
  requirements
• How to partition the effort (requirements) across
  the team
• What design features and processes allow the
  system to be re-tuned as the development process
  unfolds

3
The Mars Challenge

• Large diverse planet, not well characterized.
• Gravity
• 3/8ths of Earth gravity
• Atmosphere
• 1/100th of Earth atm. density, mostly C02
• Topography
• Surface elevation between -1.5 km and 2.5 km
• Terrain
• Rocky, cratered, sandy surface features

4
Martian Entry, Descent and Landing

• Energy management
• All EDL architectures strive to manage the
  disposal of arrival KE
• Atmosphere is a great dump site
• Aerodynamics drag is rarely enough

Entry
Parachute Descent
99 of KE
Powered Descent
0.9 of KE
0.1 of KE
3e-06 of KE
Landing or Touchdown
5
25 kg
175 kg
950 kg
6
Martian Entry, Descent and Landing

• Touchdown System
• Larger rover than previously contemplated
• How can we take out that last, most difficult KE

Entry
Parachute Descent
Powered Descent
Landing or Touchdown
7
EDL Touchdown System Trade Space

• The touchdown system must perform three tasks
• Remove kinetic energy remaining from powered
  descent condition
• Land safely on uncertain terrain
• Allow rover to egress or drive away from the
  landed state
• Four major families of touchdown system exist
• Airbags, Legs, Pallet, and Direct Placement

Touchdown Systems
Closed-loop 6-DOF Propulsion 1-3 m/s vert., lt1
m/s hori.
Open Loop 1-DOF Propulsion 10-20 m/s vert., 10
m/s hori.
Airbags
Pallet
Legs
Direct Placement
8
Viking
MPL/PHX
9
Legged Landers

• Description
• Rover top mounted or bottom mounted
• Landing legs plastically absorb touchdown energy
• Stability augmenting outriggers for slopes
• Ramps (top-mounted) or short bridle deployment
  (bottom-mounted) used for egress
• Pros
• Exploits Viking and Apollo landing technology
• Single body control at all times
• Cons
• Ground/plume interaction
• High CG and post engine shut-off free fall reduce
  stability
• Touchdown sensing and high rate engine shut-off
• Validation of terrain interaction difficult
• Egress system mass and development
• Observation
• Family of architectures potentially feasible for
  use on MSL
• Landing stability, touchdown sensing,
  ground/plume interaction are challenges

10
Direct Payload Placement (e.g. Sky Crane)

• Description
• Propulsion module with bridle suspended rover.
• Rover placed directly in mobile configuration
• Pros
• Reduced ground-plume interaction
• Slower touchdown and lower CG allows greater
  stability and hazard tolerance
• Utilizes rovers inherent terrain interaction
  capabilities
• Touchdown signature is persistent and unambiguous
• Rover does not need to egress from lander
• Validation can be decomposed into surface
  interaction testing (rover) and closed loop
  propulsion/GNC simulation (descent stage)
• Cons
• New architecture
• Additional pendulum and multi-body dynamics must
  be addressed
• Observations
• Architecture is feasible for MSL
• Significant advantages for this architecture

11
Direct Payload Placement (e.g. Sky Crane)
Analyzable
Testable
12
(No Transcript)
13
Skycrane Event Timeline
Descent Stage commanded to follow Reference
Trajectory VVertical 0.75 m/sec VHorizontal
0.0 m/sec
One-Body Phase Duration 2 sec
Deployment Phase Duration 8 sec
Post-Deploy Settling Phase Duration 2 sec
Ready for Touchdown Phase Duration 0-8 sec
Touchdown Phase Duration lt 2 sec
11
14
(No Transcript)
15
Epilogue of a Trade
Simplicity!?

• Make Airbags Work!
• Land a much bigger rover and make the changes
  needed
• Slow the velocity of impact down to 2-5 m/s
• Closed-loop throttled system, IMU, Terrain
  Radar..

Proven

• Sky Crane Landing System
• Once you have bought all the complexity to get to
  2-5 m/s, dont get lazy!
• Drive the velocity down to lt1 m/s and simplify
  the system design

16
Martian Entry, Descent and Landing
Entry
Parachute Descent
Powered Descent
Landing or Touchdown
17
Powered Descent Vehicle Configuration

• Single body powered descent (Viking heritage)
• Reduces velocity from 100 m/s at 2000 m altitude
  to lt1 m/s at 20 m altitude
• Utilizes all 8 engines firing
• Reduce to 4 engines firing at start of Sky Crane
  phase to maintain engine throttle settings above
  minimum
• Viking heritage throttleable engines
• Technology development complete

MLE (8) (Modified Viking engine)
Terminal Descent Sensor (Radar)
Descent Stage IMU (DIMU)
18
Martian Entry, Descent and Landing
Entry
Parachute Descent
Powered Descent
Landing or Touchdown
19
Parachute Design Choice

• Parachute Function
• Slow MSL scale vehicle from supersonic to
  subsonic powered descent start conditions
• From a worst case Mach 2.3 to Mach 0.6
• Challenge
• Viking size parachute has insufficient drag to
  take vehicle to the staging conditions in
  acceptable time
• Terminal velocity Mach 0.4
• Architecture trades looked at alternate solutions
• All propulsive
• Multi-stage parachutes
• All supersonic parachutes
• Larger supersonic chute chosen
• Trade space and choice reviewed and approved by
  EDL Review Board
• MSL Design
• 21.5 m Viking DGB

20
Martian Entry, Descent and Landing
Entry
Parachute Descent
Powered Descent
Landing or Touchdown
21
Guided Entry

• Entry Function
• Remove 99 of arrival kinetic energy
• Fly-out atmospheric and vehicle dispersions via
  guided flight
• Kinetic Energy
• PICA TPS
• Effort lead by ARC and LaRC personnel as integral
  part of project team
• Guided Flight
• Guided entry required to meet landing accuracy
  and altitude requirements
• Lifting entry configuration
• Viking, Apollo, Gemini, etc. heritage
• Produces a nominal L/D of .24 _at_ M 24
• Lift vector control utilized to achieve guidance
• Leverages simple, high heritage, robust design
  Apollo guidance heritage
• Control achieved by rolling lift vector about
  velocity vector
• Mature technology with extensive flight heritage,
  existing validated simulation tools
• Effort led by experienced JSC personnel as
  integral part of project team

Velocity vector
22
Putting it all together..
23
Event Timeline 1/3
Final Approach
Exo-Atmospheric
EDL Start
Do PEDL
TCM-5
EDL Parameter Update Nav Update 2 TCM-5x
EDL Param. Update Nav Update 3 Stop ATCM
TCM-6
Do EDL
HRS Vent
Cruise Stage Separation
Enable GNC(Despin, Detumble, Turn to Entry) T-0
Nav Point
Entry Interface(r 3522.2 km)
Separate CBM Switch to TLGA
E-1 day
E-10 min
E-9 min
E-15 min
E-5 days
E-2 days
E-6 hrs
E-2 hrs
E-1330 min
E-0 min
500bps
X-Band
Tones
8 kbps
UHF
24
Event Timeline 2/3
Parachute Descent
Entry
Entry Interface(r 3522.2 km)
Pressurize Prop. Sys.
Peak Heating
Peak Deceleration
SUFREBM SeparationVictory Roll
Deploy Supersonic Parachute
Heatshield Separation
Begin Using Radar Solutions
Prime MLEs
E274 s
E305 s
E245 s
E85 s
E96 s
E-0 min
E230 s
E279 s
X-Band
Tones
8 kbps
UHF
25
Event Timeline 3/3
Powered Flight Includes Powered Descent, Sky
Crane, Flyaway
Flyaway
Sky Crane
Powered Descent
Backshell Separation
Powered Approach
ConstantVelocity- 20 m/s
ConstantDeceleration- 20 m/s to 0.75 m/s
Rover Separation
Throttle Down to 4 MLEs
Mobility Deploy
Activate Flyaway Controller
Touchdown
1000 m above MOLA areoid
E347 s
E358 s
E309 s
X-Band
Tones
8 kbps
UHF
26
(No Transcript)
27
Requirements, Teaming and Risk LevelingMinimize
the maxima.
28
Requirements, Teaming and Risk Leveling

• The objective of any spaceflight engineering
  development effort is to perform the required
  function while minimizing (in ranked order)
• Risk
• Cost/Schedule
• Risk minimization can be approached using many
  different algorithms, MSL EDL used the minimize
  the maxima
• Requirements are generated to inform the delivery
  organizations of their obligations in this risk
  balance
• Requirements are imperfect, partial descriptions
  of the design
• Humans (team members) can be much more complete
  vessels for design intent
• As long as everyone is really on the same page
• Requirements can help that

29
Requirements, Teaming and Risk Leveling

• During Mars Pathfinder project development, it
  was concluded that the EDL challenge was great
  enough to warrant a Phase Organization to
  tackle this
• This phase organization was the first stitching
  together of E-D-L into a single working
  organization
• The phase organization involved multiple NASA
  centers (LaRC, ARC, JSC, and JPL)
• The EDL Phase Team was exercised on MER and PHX
  and is currently working MSL
• Team consists of multiple centers and systems,
  subsystems, contractors and domain experts
• The EDL Phase Team was the tool to develop and
  flesh out the appropriate requirements and level
  risk
• Open involvement from delivery subsystems was
  required (twice weekly)
• Empowerment of the subsystems leads as co-owners
  of the requirements generation process

30
EDL Phase Team

• EDL design is largely capabilities driven, with
  both bottoms-up design and top-down design
• Hardware and software limitations drive design
  performance and risk
• The EDL design team is a collection of system and
  subsystem implementers that inform the
  capabilities limitations and guide the design
  process

VV
Entry Guidance
GNC
Thermal
Powered Flight
TPS
Propulsion
Aerodynamics
EDL Round Table
Mechanical
Telecom
Atmospheres
Avionics
Flight Dynamics and Simulation
JPL Subsystems
JSC/LaRC/ARC
JPL Systems
Management
EEDCS
Software
Systems
Terrain Interaction
Power
Aerothermal
MDNav
EDL Lead
31
Designing in Margin

• In an EDL design, the high coupling of the
  performance can create the inability to close the
  design
• To close and maintain a stable design margin has
  to be placed within the system consciously
• Overlapping requirements is a technique to
  establish margin
• An eye toward the most brittle outcome must be
  used
• Graceful vs brittle failure
• Requirements overlap must be understood and
  tracked at a system level but not at a subsystem
  level
• Subsystems must march off to their best to meet
  requirements
• Example Wrist mode, backshell separation and
  on-chute damping

32
Parachute Wrist Mode

• Under the parachute the spacecraft can experience
  an oscillations
• There are two transverse modes of oscillations
• Pendulum mode
• Shape Parachute and S/C move together
• Frequency 0.05 Hz.
• Wrist Mode
• Shape S/C rotates about CG parachute doesnt
  move much
• Frequency 0.5 Hz.
• Wrist mode generates highest angular rate for
  the S/C
• Function of frequency
• Wrist mode is driven by two possible sources
• Initial parachute deployment transient
• Straighten up and fly right employed to reduce
  this
• Disturbances after parachute deployment
• MER flight reconstruction indicates that we do
  not fully understand the mechanisms for
  excitation of the wrist mode at Mars

33
Wrist Mode Flight Observations
Viking BLDT AV-4 Flight Data

• Wrist mode observed in Viking BLDT
• Data shows initial attitude rates at parachute
  deploy, which are then damped
• Initial rate caused by angle of attack and
  parachute inflation
• MER A and B flight experienced unexpected
  behavior
• Initial disturbance is very small
• Growth in the wrist mode oscillation occurs
  after parachute deployment
• Simulations predicted damping however, build
  up and decay of oscillations observed

MER-A Flight vs. Pre-Flight Prediction
Pre-Flight Prediction
MER-A Flight
34
Wrist Mode Concerns

• Wrist mode motion is a risk
• Reduce parachute performance
• Degrade TDS performance above 50-70 deg/sec
• Result in separation recontact at backshell sep
• Team needed design response to ensure robustness
  in face of increased wrist mode uncertainty
• Design fix
• Overlapping requirements

Low Risk
Graceful Degradation
Brittle
35
Margin Design Response 1 On Chute Rate Damping

• Objective 1 Do No Harm
• No parachute plume aerodynamics instability
• No hydrazine parachute issues
• No degradation of parachute drag performance
• Objective 2 Prove Effective
• Demonstrate energy dissipation of order greater
  than the scaled MER input

Rigid Body Mode driven by Cn, Ct
50 m
Flow-field lt200 m/s
Wrist Mode driven by Capsule Inertias
5 deg
36
Margin Design Response 2 Overlapping
Requirements

• Requirements are overlapped because of
  uncertainty in driving physics and performance
  criticality
• GNC damping dead bands allow later risk leveling
  as fuel is used to solve various problems
• Increase dead bands if fuel is needed elsewhere

37
Risk Re-levelingThe best laid plans.
38
MSL Development Challenges (Greatest Hits)

• The road from PMSR through ATLO Readiness has had
  its bumps (partial list)

625, 675, 725, 775, 825, 850, 875, 900, 925, 915,
940, 950
Grow Fuel Tanks to Ma (19-23x26x)
Rover mass increase
Increase parachute size
Decrease altitude perf.
TDS near field 6m vs 1m
Redesign antenna farm
Heat shield TPS
Mobility dep. to rover dep.
Change TPS to PICA
BUD bearing loads
Move mobility dep later in Sky Crane timeline
Mobility settling time
Increase rover dep timeline
Consume more fuel
PDV first mode 16Hz not 20Hz
Redesign controller
Rover CG location
Ballast rover?
or
Separation recontact prob
Redesign controller?
Relax requirements?
39
SummaryNo one can guarantee success in war,
but only deserve it. Winston Churchill
40
Summary

• MSL EDL architecture resulted from extended trade
  study
• Approach is an evolutionary outgrowth of past
  Mars missions
• Architecture and most designs have survived the
  development cycle
• Performance allowed some elasticity
• Margin within the original design (overlapping
  requirements) allowed the rest of the needed
  elasticity
• The balanced design was and is the product of an
  active multi-disciplinary team
• Delivery subsystems as systems engineers
• Expert participation from other centers
• Flat open team culture
• Summer of 2012 will be the test of this
  challenging engineering effort and the team will
  take every step to ensure that victory is deserved