Mu
lti-
Dis
cip
linary
Op
tim
isati
on
for
Pla
neta
ry E
ntr
y,
Desc
en
t an
d L
an
din
g S
yst
em
s
- 1 -
2nd UK Workshop on Optimisation in Space Engineering
Cambridge, 19 Mar 2014© 2014 Deimos Space UK Ltd. - www.deimos-space.com
Multi-Disciplinary Optimisation for Planetary Entry, Descent and
Landing Systems
David RileyDeimos Space Ltd., UK
Davide BonettiDeimos Space S.L.U., Spain
Mu
lti-
Dis
cip
linary
Op
tim
isati
on
for
Pla
neta
ry E
ntr
y,
Desc
en
t an
d L
an
din
g S
yst
em
s
- 2 -
2nd UK Workshop on Optimisation in Space Engineering
Cambridge, 19 Mar 2014© 2014 Deimos Space UK Ltd. - www.deimos-space.com
Table of Contents
• Introduction: The Problem Domain
• The Optimisation Process
• Tools for Numerical Trade-offs
• Optimization results
• Conclusion
Mu
lti-
Dis
cip
linary
Op
tim
isati
on
for
Pla
neta
ry E
ntr
y,
Desc
en
t an
d L
an
din
g S
yst
em
s
- 3 -
2nd UK Workshop on Optimisation in Space Engineering
Cambridge, 19 Mar 2014© 2014 Deimos Space UK Ltd. - www.deimos-space.com
INTRODUCTION:
THE PROBLEM DOMAIN
Mu
lti-
Dis
cip
linary
Op
tim
isati
on
for
Pla
neta
ry E
ntr
y,
Desc
en
t an
d L
an
din
g S
yst
em
s
- 4 -
2nd UK Workshop on Optimisation in Space Engineering
Cambridge, 19 Mar 2014© 2014 Deimos Space UK Ltd. - www.deimos-space.com
The problem
• Top-level requirement
– Bring a spacecraft safely to rest on the surface of another planet
• Challenges: Mars
– Entry velocity ~ 3 km/s
– Very thin atmosphere• equivalent to 35 km altitude on Earth
– Gravity around 1/3 of Earth’s
– Want to maximise the payload delivered
to the surface
– Subject to constraints in cost, risk,
geographical sourcing, development
timescale…
Image: NASA/JPL
Mu
lti-
Dis
cip
linary
Op
tim
isati
on
for
Pla
neta
ry E
ntr
y,
Desc
en
t an
d L
an
din
g S
yst
em
s
- 5 -
2nd UK Workshop on Optimisation in Space Engineering
Cambridge, 19 Mar 2014© 2014 Deimos Space UK Ltd. - www.deimos-space.com
Entry, Descent and Landing Systems
• Entry: aerodynamic deceleration using a heatshield
• Deploy a parachute to slow down faster
• Retro-rockets for final slowdown and fine control
• Airbags for touchdown
Entry Descent A Descent B Airbag inflation Free fall
FS Release RA Lowering Powered Descent Touchdown
Mu
lti-
Dis
cip
linary
Op
tim
isati
on
for
Pla
neta
ry E
ntr
y,
Desc
en
t an
d L
an
din
g S
yst
em
s
- 6 -
2nd UK Workshop on Optimisation in Space Engineering
Cambridge, 19 Mar 2014© 2014 Deimos Space UK Ltd. - www.deimos-space.com
Entry, Descent and Landing Systems
• Some parts are common:
– Heatshield for entry
– Parachute for deceleration
Entry Descent A Descent B Airbag inflation Free fall
FS Release RA Lowering Powered Descent Touchdown
Entry Descent A Descent B Airbag inflation Free fall
FS Release RA Lowering Powered Descent Touchdown
Entry Descent A Descent B Descent C Airbag inflation
FS Release Free fall Touchdown
• Others vary:
– One vs two parachute stages
– Whether to have a retro stage
– Touchdown on airbags, landing
legs, crushables…
Mu
lti-
Dis
cip
linary
Op
tim
isati
on
for
Pla
neta
ry E
ntr
y,
Desc
en
t an
d L
an
din
g S
yst
em
s
- 7 -
2nd UK Workshop on Optimisation in Space Engineering
Cambridge, 19 Mar 2014© 2014 Deimos Space UK Ltd. - www.deimos-space.com
Driving requirements
• Mass
• Height loss and verticalisation
• Thermo-mechanical loads
• Landing site accuracy
• Volume
• Reliability / robustness
• Cost
• Development timescale
Mu
lti-
Dis
cip
linary
Op
tim
isati
on
for
Pla
neta
ry E
ntr
y,
Desc
en
t an
d L
an
din
g S
yst
em
s
- 8 -
2nd UK Workshop on Optimisation in Space Engineering
Cambridge, 19 Mar 2014© 2014 Deimos Space UK Ltd. - www.deimos-space.com
Trade-offs
Smaller parachute / bigger retros:
– Lighter parachute system
– Final speed is higher
– More retro fuel required
Entry Descent A Descent B Airbag inflation Free fall
FS Release RA Lowering Powered Descent Touchdown
Single-stage vs two-stage parachute:
– For low final speeds, two-stage
generally is lower mass
– Two-stage may take more
altitude to reach final speed
– Single-stage is simpler
Mu
lti-
Dis
cip
linary
Op
tim
isati
on
for
Pla
neta
ry E
ntr
y,
Desc
en
t an
d L
an
din
g S
yst
em
s
- 9 -
2nd UK Workshop on Optimisation in Space Engineering
Cambridge, 19 Mar 2014© 2014 Deimos Space UK Ltd. - www.deimos-space.com
Optimisation problem
• Problem includes discrete choices, integer values and continuous values
• Multi-disciplinary approach required
– Parachute design, retro system design, airbag design, trajectory, …
• Can’t split the problem into separate pieces - different parts of the design
affect each other
– Speed reached under heatshield must be safe for parachute deployment
– Each stage must carry the subsequent stages• e.g. parachute must slow down the retro fuel and airbags as well as the payload
• Need the trajectory to be self-consistent
– Test through simulation: enough time and altitude to stop before hitting ground
Entry Descent A Descent B Airbag inflation Free fall
FS Release RA Lowering Powered Descent Touchdown
Entry Descent A Descent B Descent C Airbag inflation
FS Release Free fall Touchdown
Mu
lti-
Dis
cip
linary
Op
tim
isati
on
for
Pla
neta
ry E
ntr
y,
Desc
en
t an
d L
an
din
g S
yst
em
s
- 10 -
2nd UK Workshop on Optimisation in Space Engineering
Cambridge, 19 Mar 2014© 2014 Deimos Space UK Ltd. - www.deimos-space.com
High Precision Landers: overview
• Mission concept
– Technology Research Study within the ESA MREP
(Mars Robotic Exploration Preparatory) Programme,
preliminary to Mars Sample Return
– Rover and return vehicle must land near each other
– Target is launch in late 2020’s
• Main study objective
– Design an optimum and robust EDL/GNC
configuration for MSR lander
– Project led by Airbus Defence & Space (Les Mureaux)
• Main requirements
– Achieve landing accuracy of at least 10 km, ideally
high precision (3 km) or even pinpoint (100 m)
– Mass at entry ~ 800 kg
– Use European technology as far as possibleImages: NASA/JPL
Mu
lti-
Dis
cip
linary
Op
tim
isati
on
for
Pla
neta
ry E
ntr
y,
Desc
en
t an
d L
an
din
g S
yst
em
s
- 11 -
2nd UK Workshop on Optimisation in Space Engineering
Cambridge, 19 Mar 2014© 2014 Deimos Space UK Ltd. - www.deimos-space.com
Small Mars Landers: overview
• Mission concept
– Technology Research Study within the ESA MREP
(Mars Robotic Exploration Preparatory) Programme,
preliminary to INSPIRE
– Target is launch in 2026 or 2028
• Main study objective
– Design an optimum and robust EDL/GNC
configuration for a Mars Network Science mission
– Project led by Deimos Space (Spain)
• Main requirements
– Three identical probes separating from the same
carrier and landing in different sites
– Each probe mass at entry ~350 kg
– Payload: 130 kg, 1150x355 mm (cylinder)
– Use European technology as far as possible
Image: ESA
Mu
lti-
Dis
cip
linary
Op
tim
isati
on
for
Pla
neta
ry E
ntr
y,
Desc
en
t an
d L
an
din
g S
yst
em
s
- 12 -
2nd UK Workshop on Optimisation in Space Engineering
Cambridge, 19 Mar 2014© 2014 Deimos Space UK Ltd. - www.deimos-space.com
Small Mars Landers: overview
10-3
10-2
0
0.2
0.4
0.6
0.8
1
1.2
1.4
1.6
1.8
2x 10
4
Pth. 99.5% CI Up density (kg/m3)
altit
ude
(m)
Global Model
Elysium1Elysium2
Elysium3
One site
Global model• Challenges of a multi-probe mission
– Several studies, only one launched (crashed)
– Target entry mass quite different to previous
single or multiple landers solutions
• Wide environment variability
– Triple landing site, identical probes
– Altitude below 0 MOLA, -15º < Latitude < 30º
– Global atmosphere models based on statistics
of European Mars Climate Database calls
Mu
lti-
Dis
cip
linary
Op
tim
isati
on
for
Pla
neta
ry E
ntr
y,
Desc
en
t an
d L
an
din
g S
yst
em
s
- 13 -
2nd UK Workshop on Optimisation in Space Engineering
Cambridge, 19 Mar 2014© 2014 Deimos Space UK Ltd. - www.deimos-space.com
THE OPTIMISATION PROCESS
Mu
lti-
Dis
cip
linary
Op
tim
isati
on
for
Pla
neta
ry E
ntr
y,
Desc
en
t an
d L
an
din
g S
yst
em
s
- 14 -
2nd UK Workshop on Optimisation in Space Engineering
Cambridge, 19 Mar 2014© 2014 Deimos Space UK Ltd. - www.deimos-space.com
Overall Process
Mu
lti-
Dis
cip
linary
Op
tim
isati
on
for
Pla
neta
ry E
ntr
y,
Desc
en
t an
d L
an
din
g S
yst
em
s
- 15 -
2nd UK Workshop on Optimisation in Space Engineering
Cambridge, 19 Mar 2014© 2014 Deimos Space UK Ltd. - www.deimos-space.com
Overall Process
Mu
lti-
Dis
cip
linary
Op
tim
isati
on
for
Pla
neta
ry E
ntr
y,
Desc
en
t an
d L
an
din
g S
yst
em
s
- 16 -
2nd UK Workshop on Optimisation in Space Engineering
Cambridge, 19 Mar 2014© 2014 Deimos Space UK Ltd. - www.deimos-space.com
Preliminary EDL & GNC Trade-offsTRADE-OFFS SUMMARY
Ph Id Trade-off Options
Arrival
A1 Arrival conditions
Direct Escape GTO constrained GTO 2022 unconstrained GTO 2024 unconstrained
Separation &
Coasting
SC1 SED Militar/Civil Heritage Beagle-2 Cassini/Hyugens
SC2 Entry type Prograde Retrograde
SC3 Separation times ESAT dimensions [25:0.5] days before EIP
Entry
E1 Aeroshell concept
70º cone + sphere 60º cone + sphere IBD Parashield
E2 Nose Radius 0.25 diameters 0.45 diameters
E3 TPS material
SLA-561V Norcoat-Liege AQ60 / Prosial ALESTRASIL PICSIL DO31 / SPA ASTERM
E4 Entry Control system Spinning 3-axis RCS damping
E5 FPA ESAT dimension [-18:-10.5]ºE6 Capsule diameter ESAT dimension [1.4:2] mE7 Capsule backshell angle ESAT dimension [30:50]º
Descent and Landing
DL1 Parachutes type
DGB Ringslot Ringsail Hemisflo Cruciform
DL2Number of parachute stages
Single Two stage
DL3 Use of retrorockets
No retros Retro-assisted parachute phase Fire retros after parachute separation Skycrane
DL4 Controllability Fixed Thrust - non pulsed Fixed Thrust – pulsed Throttleable (variable thrust)
DL5Velocity reduction by means of thrust
None Only Vertical Only Horizontal Vertical and Horizontal
TRADE-OFFS SUMMARY Ph Id Trade-off Options
Descent and
Landing
DL6 Use of lowering system No Yes
DL7 Landing system
Legs Vented Airbags Non-Vented airbags Crushable (as stand-alone system, not part of legs or airbag system)
DL8Backshell avoidance manoeuvre
No Yes
DL9Parachute avoidance manoeuvre / flyaway
No
Yes
GNC
G1 Entry Guidance None Hypersonic guidance
G2 Descent Guidance None Guided parachute
G3 Landing Guidance Apollo derived G-turn None
N1 Coasting & Entry Navigation timer+MU timer+IMU+sun sensor
N2 Descent & Landing Navigation
timer+MU timer+RA timer+IMU+RA timer+IMU+Camera timer+IMU+Camera+RA timer+IMU+Camera+RD
C1 Entry Control None Liquid rockets
C2Descent & Landing velocity Control
None Solid rockets Liquid rockets
C3Descent & Landing attitude control
None Liquid rockets
T1 Descent triggering IMU-based correlation Navigated velocity Navigated dynamic pressure
T2 Main parachute triggering None Timer
T3 RA activation None Timer from mortar
T4 Retros triggering None Altitude and velocity
T5 Landing triggering None Altitude and velocity
Mu
lti-
Dis
cip
linary
Op
tim
isati
on
for
Pla
neta
ry E
ntr
y,
Desc
en
t an
d L
an
din
g S
yst
em
s
- 17 -
2nd UK Workshop on Optimisation in Space Engineering
Cambridge, 19 Mar 2014© 2014 Deimos Space UK Ltd. - www.deimos-space.com
3 Configurations selected for ESAT numerical trade-offs Entry Descent A Descent B Airbag inflation Free fall
FS Release RA Lowering Powered Descent Touchdown
Entry Descent A Descent B Airbag inflation Free fall
FS Release RA Lowering Powered Descent Touchdown
Entry Descent A Descent B Descent C Airbag inflation
FS Release Free fall Touchdown
SIMPLE
ROBUST
TY
PE 1
TY
PE 2
TY
PE3
Sub-system TypeAeroshell 70º sphere-coneEntry BallisticDescent 2 stages (DGB+Ringsail)Retro-rockets NoneLowering NoneLanding Non vented airbagsGNC sensors Sun sensor, RA, IMU
Sub-system TypeAeroshell 70º sphere-coneEntry BallisticDescent 1 stage (DGB)Retro-rockets Solid (Vertical only)Lowering YesLanding Non vented airbagsGNC sensors Sun sensor, RA, IMU
Sub-system TypeAeroshell 70º sphere-coneEntry BallisticDescent 1 stage (DGB)Retro-rockets Solid (Vertical & Horizontal)Lowering YesLanding Non vented airbagsGNC sensors Sun sensor, RA, IMU, VBN
MER-LIKE
MPF-LIKE
Beagle2-LIKE(no lowering)
Mu
lti-
Dis
cip
linary
Op
tim
isati
on
for
Pla
neta
ry E
ntr
y,
Desc
en
t an
d L
an
din
g S
yst
em
s
- 18 -
2nd UK Workshop on Optimisation in Space Engineering
Cambridge, 19 Mar 2014© 2014 Deimos Space UK Ltd. - www.deimos-space.com
Overall Process
Mu
lti-
Dis
cip
linary
Op
tim
isati
on
for
Pla
neta
ry E
ntr
y,
Desc
en
t an
d L
an
din
g S
yst
em
s
- 19 -
2nd UK Workshop on Optimisation in Space Engineering
Cambridge, 19 Mar 2014© 2014 Deimos Space UK Ltd. - www.deimos-space.com
Overall Process
ESAT: EDL&GNC Sizing and Analysis Tool
Mu
lti-
Dis
cip
linary
Op
tim
isati
on
for
Pla
neta
ry E
ntr
y,
Desc
en
t an
d L
an
din
g S
yst
em
s
- 20 -
2nd UK Workshop on Optimisation in Space Engineering
Cambridge, 19 Mar 2014© 2014 Deimos Space UK Ltd. - www.deimos-space.com
TOOLS FOR
NUMERICAL TRADE-OFFS
Mu
lti-
Dis
cip
linary
Op
tim
isati
on
for
Pla
neta
ry E
ntr
y,
Desc
en
t an
d L
an
din
g S
yst
em
s
- 21 -
2nd UK Workshop on Optimisation in Space Engineering
Cambridge, 19 Mar 2014© 2014 Deimos Space UK Ltd. - www.deimos-space.com
ESAT Architecture
ESAT relies on a modular and generic infrastructure.
Users concentrate efforts on Settings and Wrapper Function to solve a given problem.
External module:- multidisciplinary- single discipline
EDL&GNC Sizing and AnalysisTool
Mu
lti-
Dis
cip
linary
Op
tim
isati
on
for
Pla
neta
ry E
ntr
y,
Desc
en
t an
d L
an
din
g S
yst
em
s
- 22 -
2nd UK Workshop on Optimisation in Space Engineering
Cambridge, 19 Mar 2014© 2014 Deimos Space UK Ltd. - www.deimos-space.com
MDO Architecture
C: CoastingE: EntryD: DescentL: LandingP/L: PayloadQdyn: Dynamic pressureFS: FrontshieldVt: terminal VNSP: Network Science Probe
Different forType 1, 2 and 3
SIM
ULA
TIO
N C
OR
E
Overall it is a complex problem:
– Multiple phases
– Multiple combinations of worst cases
(aerodynamics, atmosphere,
events…) => robust solutions
Mu
lti-
Dis
cip
linary
Op
tim
isati
on
for
Pla
neta
ry E
ntr
y,
Desc
en
t an
d L
an
din
g S
yst
em
s
- 23 -
2nd UK Workshop on Optimisation in Space Engineering
Cambridge, 19 Mar 2014© 2014 Deimos Space UK Ltd. - www.deimos-space.com
Bi-level MDO Architecture
• Bi-level surrogate models with internal optimization
– Level1: MDO (mission & system objectives)
– Level2: Coasting, D&L (nested in level 1)
• Achieve efficient optimizations
PESDO (TESSELLA)
Mu
lti-
Dis
cip
linary
Op
tim
isati
on
for
Pla
neta
ry E
ntr
y,
Desc
en
t an
d L
an
din
g S
yst
em
s
- 24 -
2nd UK Workshop on Optimisation in Space Engineering
Cambridge, 19 Mar 2014© 2014 Deimos Space UK Ltd. - www.deimos-space.com
OPTIMIZATION RESULTS
Mu
lti-
Dis
cip
linary
Op
tim
isati
on
for
Pla
neta
ry E
ntr
y,
Desc
en
t an
d L
an
din
g S
yst
em
s
- 25 -
2nd UK Workshop on Optimisation in Space Engineering
Cambridge, 19 Mar 2014© 2014 Deimos Space UK Ltd. - www.deimos-space.com
Problem 1: Probe separation timing
• 3 probes, want to optimise the separation events
ESAT wrapper for Small Mars Landers
Coasting and separation overview
=> simplified ESAT example =>
Mu
lti-
Dis
cip
linary
Op
tim
isati
on
for
Pla
neta
ry E
ntr
y,
Desc
en
t an
d L
an
din
g S
yst
em
s
- 26 -
2nd UK Workshop on Optimisation in Space Engineering
Cambridge, 19 Mar 2014© 2014 Deimos Space UK Ltd. - www.deimos-space.com
Problem 1: Probe separation timing
• Look at results for sites 1 and 3
Fuel mass for retargeting EIP FPA dispersionPareto Front: Dominant solutions
10
10
10
10
10
10
10
20
20
30
30
20
20
20
40
40
50
50
30
30
30
75
75
40
40
40
100100
50
50
50
30
75
75
Tsep Elysium 1 [days]
Tse
p E
lysi
um
3 [d
ays
]
Total Fuel Mass [kg]
5 10 15 20 25 30
5
10
15
20
25
30
10
20
30
40
50
60
70
80
90
100
0.60.6
0.6
0.6
0.6
0.6
0.6
0.6
0.8
0.8
0.8
0.8
0.8
0.8
0.8
0.8
11
1
11
1
1
1.2
1.2
1.2
1.2
1.2
1.2
1.4
1.4
1.4
1.4
1.4
1.4
1.6
1.6
1.6
1.6
1.6
1.8
1.8
1.8
1.8
1.8
2
2
2
2
2
2.22.2
2.2
2.2
1.21.41.6
1.82
2.2
Tsep Elysium 1 [days]
Tse
p E
lysi
um
3 [d
ays
]
FPA dispersion @EIP [deg]
5 10 15 20 25 30
5
10
15
20
25
30
0.6
0.8
1
1.2
1.4
1.6
1.8
2
2.2
2.4
2.6
0 10 20 30 40 50 60 70 80 90 1000
0.25
0.5
0.75
1
1.25
1.5
1.75
2
Total Fuel Mass
FP
A d
isp
ers
ion
@E
IP
Mu
lti-
Dis
cip
linary
Op
tim
isati
on
for
Pla
neta
ry E
ntr
y,
Desc
en
t an
d L
an
din
g S
yst
em
s
- 27 -
2nd UK Workshop on Optimisation in Space Engineering
Cambridge, 19 Mar 2014© 2014 Deimos Space UK Ltd. - www.deimos-space.com
Problem 2: Minimisation of EDLS Mass
• EDL&GNC Robust Optimization Problem– 7 variable dimensions covering EDL, GNC and environment aspects
Type Phase DescriptionConstraint L Landing Accuracy E Maximum heat flux at nose E Maximum heat load at nose E Maximum dynamic pressure E/D Maximum load factor E/D Maximum dynamic pressure at parachute deployment
E/D Minimum altitude at parachute deployment E/D/L Percentage of non feasible simulations
C Total fuel mass for coasting phase E/D/L Consistency of volumes and sizesFixed E/D/L ESAT EDL/GNC configurationInputs L Payload mass L Payload diameter L Payload height E Capsule nose radius / Capsule diameter E Capsule front cone angle E Capsule shoulder radius / Capsule diameter E Capsule back plate diameter/capsule diameter E TPS density (Norcoat Liege)Variable ESAT Dimensions
E/D Capsule diameterE Capsule back cone angleE Flight Path Angle at EIP
E/D Dispersion on Mach number at parachute triggering D Dispersion on Altitude at Powered Descent Initiation D/L Dispersion on Altitude at Free Fall start
L Winds scale factor
Minimize system mass
Maximize altitude at parachute dep.
Mu
lti-
Dis
cip
linary
Op
tim
isati
on
for
Pla
neta
ry E
ntr
y,
Desc
en
t an
d L
an
din
g S
yst
em
s
- 28 -
2nd UK Workshop on Optimisation in Space Engineering
Cambridge, 19 Mar 2014© 2014 Deimos Space UK Ltd. - www.deimos-space.com
Problem 2: Minimisation of EDLS Mass
• Configurations selection process: Query and Filtering– More than 100 performances have been
managed in 7 dimensions with ESAT
– 2D slices / N-dim data mining
– Statistics of 50000 different samples0 0.5 1 1.5 2
0
20
40
60
80
100
all checks [-]
Per
cent
ile
400 410 420 430 4400
20
40
60
80
100
EIP Mass [kg]
Per
cent
ile
700 800 900 10000
20
40
60
80
100
Max qdyn drogue dep. [Pa]
Per
cent
ile
62 64 66 68 700
20
40
60
80
100
All-up mass of D&L system [kg]
Per
cent
ile
0.8 1 1.2 1.4 1.6
x 105
0
20
40
60
80
100
End alongtrack disp (3sigma) [m]
Per
cent
ile
12.5 13 13.5 14 14.5 150
20
40
60
80
100
max load factor [gE]
Per
cent
ile
5 6 7 8
x 105
0
20
40
60
80
100
max convective heat flux [W/m2]
Per
cent
ile
2000 4000 6000 8000 100000
20
40
60
80
100
max dynamic pressure [Pa]
Per
cent
ile
3.5 4 4.5 5
x 107
0
20
40
60
80
100
max heat load [J/m2]
Per
cent
ile
4000 4500 5000 5500 60000
20
40
60
80
100
min altitude @ 1st parachute dep [m]
Per
cent
ile
-1 -0.5 0 0.5 10
20
40
60
80
100
Percentage of NaN, phase 3 [-]
Per
cent
ile
15 20 25 300
20
40
60
80
100
Total Fuel Mass [kg]
Per
cent
ile
0.1
0.2
0.3
0.3
0.4
0.4
0.5
0.5
0.6
0.6
0.7
0.7
0.8
0.8
0.9
0.9
1
1
0
Capsule diameter [m]
Bac
k co
ne a
ngle
[deg
]
All checks [-]
1.6 1.7 1.8 1.9 2 2.1 2.2 2.3 2.4 2.535
40
45
50
55
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
400
400
410
410
420
430
440
450
460
470
Capsule diameter [m]
Bac
k co
ne a
ngle
[deg
]
EIP Mass [kg]
1.6 1.7 1.8 1.9 2 2.1 2.2 2.3 2.4 2.535
40
45
50
55
400
410
420
430
440
450
460
470
TOOSMALLCAPSULE
TOOHEAVYCAPSULE
Mu
lti-
Dis
cip
linary
Op
tim
isati
on
for
Pla
neta
ry E
ntr
y,
Desc
en
t an
d L
an
din
g S
yst
em
s
- 29 -
2nd UK Workshop on Optimisation in Space Engineering
Cambridge, 19 Mar 2014© 2014 Deimos Space UK Ltd. - www.deimos-space.com
• Pareto Frontier: optimum solutions, = chosen one– Predicted and validated Pareto frontiers match with very good accuracy
400 420 440 460 480 500 5200
1000
2000
3000
4000
5000
6000
EIP Mass [kg]
min
alti
tude
@ 1
st p
arac
hute
dep
. [m
]
Type 1Type 2Type 3
60% of max winds
50% of max winds
96% of max winds
435 440 445 450 455 460 465 4704000
4200
4400
4600
4800
5000
5200
EIP Mass [kg]
min
alti
tude
@ 1
st p
arac
hute
dep
. [m
]
Filtered query pointsPredicted Pareto FrontValidated Pareto Front
Problem 2: Minimisation of EDLS Mass
TYPE 2
(mass includes system margins)
Mu
lti-
Dis
cip
linary
Op
tim
isati
on
for
Pla
neta
ry E
ntr
y,
Desc
en
t an
d L
an
din
g S
yst
em
s
- 30 -
2nd UK Workshop on Optimisation in Space Engineering
Cambridge, 19 Mar 2014© 2014 Deimos Space UK Ltd. - www.deimos-space.com
Optimum configurations comparison
• Overall summary
Criteria Type 1 Type 2 Type 3 Probe mass at EIP, kg 404 440 500 Probe diameter, m 2.038 2.045 2.032 Winds close to ground, m/s < 21.3 < 13 < 10.6 Reachable landing sites, % 96.2 63.7 47.6 Performances Coasting Good Good Good
Entry Good Good Poor
Descent Good Good Fail
Landing Good Good Fail
Relative GNC complexity High Mid Low Good candidate solution? Yes Yes No
Mu
lti-
Dis
cip
linary
Op
tim
isati
on
for
Pla
neta
ry E
ntr
y,
Desc
en
t an
d L
an
din
g S
yst
em
s
- 31 -
2nd UK Workshop on Optimisation in Space Engineering
Cambridge, 19 Mar 2014© 2014 Deimos Space UK Ltd. - www.deimos-space.com
Payload mass Frontshield
Backshell Airbags
Parachutes Retro-rockets (RAD)
Thermal control Spin and ejection device
MLI Clevises, Brackets, Miscellaneous
GNC sensors Bioseal
Heatshield instrumentation
0 20 40 60 80 100 120 140
Payload mass Frontshield
BackshellParachutes
Retro-rockets (RAD+TIRS) Airbags
Spin and ejection device Clevises, Brackets, Miscellaneous
Thermal control MLI
GNC sensors Bioseal
Heatshield instrumentation
0.0 20.0 40.0 60.0 80.0 100.0 120.0 140.0
Configurations selected
• Type 1&2: Mass budget and Sizes
Type 1:BC ~ 73 kg/m2
Type 2:BC ~ 79 kg/m2
(mass includes system margins)
P/L = 32.2% of Total
P/L = 29.6% of Total
Element Type 1 Type 2 UnitsPayload volume 0.369 0.369 m3
Capsule diameter 2.038 2.045 mCapsule front cone angle 70.00 70.00 degCapsule back cone angle 44.290 44.809 degCapsule back cover diameter 1.019 1.022
m
Internal volume 1.383 1.381 m3
Sensors volume 0.003 0.002 m3
Parachute diameter 13.366 13.573 mParachute volume (stowed) 0.074 0.076
m3
RCS volume 0.012 0.012 m3
RCS diameter 0.124 0.124 mAirbags volume (stowed) 0.048 0.129 m3
Airbags thickness (stowed) 0.014 0.035
m
DL volume 0.134 0.216 m3
Ballistic coefficient 72.8 78.7 Kg/m2
kg
kg
Type 1: winds < 21.3 m/s
Type 2: Winds < 13 m/s
Mu
lti-
Dis
cip
linary
Op
tim
isati
on
for
Pla
neta
ry E
ntr
y,
Desc
en
t an
d L
an
din
g S
yst
em
s
- 32 -
2nd UK Workshop on Optimisation in Space Engineering
Cambridge, 19 Mar 2014© 2014 Deimos Space UK Ltd. - www.deimos-space.com
Configurations selected
• Type 1&2: GNC solutions
• Type 1: Modes– Events (blue)
– Sensors (grey)
– State vector (red)
Function Phase GNC solution Equipments
GUIDANCE EDL N/A N/A
NAVIGATIONInertial: Coasting & Entry
Ballistic mode Inertial kinematic
Sun sensor + IMU
Relative: Descent and Landing
Hybridized with lateral velocity estimation
RA + IMU + VBN
CONTROL
Entry N/A N/A
Velocity control: Descent and Landing
Vertical velocity control/reduction Lateral velocity
compensation/control
Solid rockets for vertical/lateral control
(RAD+TIRS)
Attitude control: Descent and Landing
N/A N/A
Only for Type 1 configuration
Mu
lti-
Dis
cip
linary
Op
tim
isati
on
for
Pla
neta
ry E
ntr
y,
Desc
en
t an
d L
an
din
g S
yst
em
s
- 33 -
2nd UK Workshop on Optimisation in Space Engineering
Cambridge, 19 Mar 2014© 2014 Deimos Space UK Ltd. - www.deimos-space.com
CONCLUSIONS
Mu
lti-
Dis
cip
linary
Op
tim
isati
on
for
Pla
neta
ry E
ntr
y,
Desc
en
t an
d L
an
din
g S
yst
em
s
- 34 -
2nd UK Workshop on Optimisation in Space Engineering
Cambridge, 19 Mar 2014© 2014 Deimos Space UK Ltd. - www.deimos-space.com
Conclusions: Mars Landers
• Identified optimal landing site sequences, configurations and component
trade-offs for the Small Mars Landers project
• Robustness is critical for a network Mars landers mission
• Higher GNC complexity is the price of adding flexibility to site selection
– Vision based navigation
– Lateral control
• ESAT allows the System, Mission and GNC engineers to perform high-
fidelity EDL-GNC architecture trade-offs relaying on high-fidelity and end-
to-end approach. It increases the reliability of the selected design solutions
with a reduced number of iteration loops (number and extent)
Mu
lti-
Dis
cip
linary
Op
tim
isati
on
for
Pla
neta
ry E
ntr
y,
Desc
en
t an
d L
an
din
g S
yst
em
s
- 35 -
2nd UK Workshop on Optimisation in Space Engineering
Cambridge, 19 Mar 2014© 2014 Deimos Space UK Ltd. - www.deimos-space.com
Conclusions: Design Optimisation Tools
• We have found the approach adds a lot to EDLS optimisation
• Need to choose an appropriate level of optimisation based on the level of
fidelity of the model
– No point fine-tuning something that exists in the model, not in reality
• Surrogate modelling tools like ESAT are valuable when the models are
complicated, slow to run
• Multi-disciplinary approach is vital – we’ve got as far as we can with tuning
each component separately, choosing handover conditions by guesswork
• Mathematical optimisation is effectively “yet another discipline”
– Important to add appropriate support not just extra complication
• ESAT is a useable front-end to a sophisticated optimisation approach
– Allows the user to focus on creating the “wrapper function”
Mu
lti-
Dis
cip
linary
Op
tim
isati
on
for
Pla
neta
ry E
ntr
y,
Desc
en
t an
d L
an
din
g S
yst
em
s
- 36 -
2nd UK Workshop on Optimisation in Space Engineering
Cambridge, 19 Mar 2014© 2014 Deimos Space UK Ltd. - www.deimos-space.com
Acknowledgements
• Deimos Space (Spain)
– Gabriele De Zaiacomo, Irene Pontijas Fuentes, Rodrigo Haya Ramos,
Gabriele Bellei, Jordi Freixa Mallol
• Airbus Defence and Space
– Timothee Verwaerde, Aurélien Pisseloup, Cédric Renault
• European Space Agency
– Eric Bornschlegl, Kelly Geelen, Alvaro Martinez Barrio, Thomas Voirin
Mu
lti-
Dis
cip
linary
Op
tim
isati
on
for
Pla
neta
ry E
ntr
y,
Desc
en
t an
d L
an
din
g S
yst
em
s
- 37 -
2nd UK Workshop on Optimisation in Space Engineering
Cambridge, 19 Mar 2014© 2014 Deimos Space UK Ltd. - www.deimos-space.com
Thank you!
Top Related