Parametric Design MARYLAND

Size: px

Start display at page:

Download "Parametric Design MARYLAND"

Anissa Roberts
6 years ago
Views:

Parametric Design The Design Process Earth

Vehicle Trade Studies Program Reliability Analysis

Akin - All rights reserved http://spacecraft.ssl.

1 Parametric Design The Design Process Earth Orbital/Lunar Orbital Mission Architectures Launch Vehicle Trade Studies Program Reliability Analysis U N I V E R S I T Y O F MARYLAND 2007 David L. Akin - All rights reserved Parametric Design Principles of Space Systems Design

2 Overview of the Design Process Program Objectives System Requirements Vehicle-level Estimation (based on a few parameters from prior art) Basic Axiom: Relative rankings between competing systems will remain consistent from level to level Increasing complexity Increasing accuracy System-level Estimation (system parameters based on prior experience) Decreasing ability to comprehend the big picture U N I V E R S I T Y O F MARYLAND System-level Design (based on disciplineoriented analysis) Parametric Design Principles of Space Systems Design

3 Akin s Laws of Spacecraft Design - #3 Design is an iterative process. The necessary number of iterations is one more than the number you have currently done. This is true at any point in time. U N I V E R S I T Y O F MARYLAND Parametric Design Principles of Space Systems Design

4 UMd Vision for Space Exploration Design a system to return humans to the moon within a decade for the minimum achievable cost.

5 Lunar Mission ΔV Requirements To: From: Low Earth Orbit Lunar Transfer Orbit Low Lunar Orbit Lunar Descent Orbit Low Earth Orbit km/sec Lunar Transfer Orbit km/sec km/sec Low Lunar Orbit km/sec km/sec Lunar Descent Orbit km/sec Lunar Landing km/sec km/sec Lunar Landing km/sec km/sec

6 Akin s Laws of Spacecraft Design - #9 Not having all the information you need is never a satisfactory excuse for not starting the analysis. U N I V E R S I T Y O F MARYLAND Parametric Design Principles of Space Systems Design

7 Mission Scenario 1 What can be accomplished with a single Delta IV Heavy payload (23K kg)? Assume δ=0.15, Ve=3200 m/sec Direct landing LEO-lunar transfer orbit ΔV=3.107 km/sec Lunar transfer orbit-lunar landing ΔV=3.140 km/ sec Lunar surface-earth return orbit ΔV=2.890 km/ sec Direct atmospheric entry to landing

8 Scenario 1 Analysis Trans-lunar injection r T LI = e V T LI Ve = e = m T LI = m 0 (r T LI δ) = 23, 000( ) = 5260 kg Lunar landing r= m LS =1182 kg Earth return r= m ER =302 kg This scenario would work for a moderate robotic sample return mission, but is inadequate for a human program.

9 Mission Scenario 2 Assume a single Delta IV-H payload is used to size the lunar descent and ascent elements Delta payload performs landing and return Something else performs TLI for shuttle payload Assume δ=0.15, Isp=320 sec Direct landing (unchanged from Scenario 1) LEO-lunar transfer orbit ΔV=3.107 km/sec Lunar transfer orbit-lunar landing ΔV=3.140 km/ sec Lunar surface-earth return orbit ΔV=2.890 km/ sec Direct atmospheric entry to landing

10 Scenario 2 Analysis Lunar landing (23,000 kg at lunar arrival) r= m LS =5170 kg Earth return r= m ER =1320 kg Earth departure Payload mass still too low for human spacecraft. EDS gross mass of 77,600 kg is too large for any existing launch vehicle. r= m 0 = m T LI (r T LI δ) = 23, 000 ( ) Mass of Earth departure stage m EDS = m 0 m T LI = 77, 600 kg = 100, 600 kg

11 Mission Scenario 3 Assume three Delta IV-Heavy missions carry identical boost stages which perform TLI and part of descent burn Unique Delta IV-H payload completes descent and performs ascent and earth return Boost module inert mass fraction = 0.1 All other factors as in previous scenarios

12 Scenario 3 - Standard Boost Stage m o =23,000 kg m i =2300 kg m p =20,700 kg LEO departure configuration is three boost stages with 23,000 kg descent/ascent stage as payload m LEO =92,000 kg

13 First Boost Module Mtotal=23,000 kg Mprop=20,700 kg Minert=2300 kg Isp=320 sec

14 Scenario 3 TLI Performance Boost stage 1 ( ) m0 m prop V 1 = V e ln V TLI remaining=2291 m/sec Boost stage 2 r=0.70; ΔV 2 =1141 m/sec V TLI remaining=1150 m/sec Boost stage 3 m 0 r=0.55; ΔV 3 =1913 m/sec = 816 m sec Residual ΔV after TLI=763 m/sec Total booster performance=3870 m/sec

15 Alternate Staging Trade Study Three identical stages Serial staging (previous chart) ΔV=3870 m/ sec Parallel staging (all three) ΔV=3597 m/sec (penalty is 273 m/sec) Parallel/serial staging (2/1) ΔV1=1913 m/sec ΔV2=1913 m/sec ΔVTotal=3826 m/sec (penalty is 44 m/sec) Pure serial staging is preferred

16 Ascent/Descent Performance 763 m/sec of lunar descent maneuver performed by boost stage 3 Remaining descent requires 2377 m/sec r= m i =3450 kg m LS =7493 kg Earth return r= m i =849 kg Return vehicle mass is still significantly below that of the Gemini spacecraft - need to examine other numbers of boost vehicles m ER =1913 kg

17 Effect of Number of Boost Stages Earth Return Payload Number of Boost Modules

18 Creating a Baseline Need to use Scenario 3 calculations to provide more than 3500 kg of Earth return mass (Gemini-class spacecraft) Select 6 boost modules based on trade study performed (mer=3719 kg) Establish as an early baseline: something to use as a standard, vary parameters to identify promising modifications Often termed strawman design It won t last!!! Design iteration will continue

19 Baseline System Schematic TEI Prop. TEI Prop. TEI Prop. TEI Prop. TEI Prop. Landing Prop. Landing Prop. Descent Prop. Descent Inert Ascent Prop. Ascent Inert Inert Mass Inert Mass Inert Mass Inert Mass Inert Mass Inert Mass Crew Cabin

20 Initial Configuration Sketch

21 Akin s Laws of Spacecraft Design - #20 (von Tiesenhausen's Law of Engineering Design) If you want to have a maximum effect on the design of a new engineering system, learn to draw. Engineers always wind up designing the vehicle to look like the initial artist's concept. U N I V E R S I T Y O F MARYLAND Parametric Design Principles of Space Systems Design

22 Transport Architecture Options Transportation node at intermediate point (low lunar orbit, Earth-Moon L1) Leave systems at node if not needed on the lunar surface, e.g.: TransEarth injection stage Orbital life support module Entry, descent, and landing systems Will increase payload capacities at the expense of additional operational complexity, potential for additional safety critical failures

23 Akin s Laws of Spacecraft Design - #10 When in doubt, estimate. In an emergency, guess. But be sure to go back and clean up the mess when the real numbers come along. U N I V E R S I T Y O F MARYLAND Parametric Design Principles of Space Systems Design

24 Variation 1: Lunar Orbit Staging Assume same process as baseline, but use lunar parking orbit before/after landing Additional ΔV s required for LLO stops Lunar landing additional ΔV=+31 m/sec Lunar take-off additional ΔV=+53 m/sec Low lunar orbit waypoint changes baseline payload to 3579 kg (-3.8%) Not useful without taking advantage of node to limit lunar landing mass

25 Lunar Orbit/Landing Trade Study Delta V for lunar landing = 2706 m/sec Delta V for launch to lunar orbit = 2334 m/ sec Assumptions: inert mass fraction = 0.15, exhaust velocity = 3200 m/sec 3.58 kg of mass in lunar orbit is required to land 1 kg of payload on the surface 10.8 kg of mass in lunar orbit is required to land 1 kg of payload on the surface and return it to orbit

26 Variation 2: Lunar Orbit Rendezvous What don t we need on the lunar surface? Trans-Earth injection stage Heat shield Parachutes Landing systems Structure stressed for Earth launch Leave unnecessary stuff parked in LLO Only take to the surface what needs to be on the surface

27 Earth Departure Configuration ΔV1=391 m/sec ΔV2=455 m/sec ΔV3=542 m/sec ΔV4=671 m/sec ΔV5=882 m/sec Initial Mass=176,400 kg

28 Variation 2 System Schematic TEI Prop. Inert Mass TEI Prop. Inert Mass TEI Prop. Inert Mass TEI Prop. Inert Mass TEI Prop. LOI Prop. Inert Mass LOI Prop. Inert Mass Descent Prop. Descent Inert Ascent Prop. Ascent Inert Lander Cabin Descent Prop. Descent Inert TEI Prop. ER Cabin TEI Inert

29 Orbital vs. Landed Mass Trade Study Lunar Landing Cabin Mass (kg) Earth Return Vehicle Mass (kg)

30 The Next Steps From Here Perform parametric sensitivity analyses Inert mass fractions Specific impulse More detailed studies of crew cabins to determine minimum masses for lunar and Earth entry spacecraft Perform more trade studies, e.g. 2 crew cabins vs. 1 Different launch vehicles Reach decision(s) on revision of baseline design Configuration design

31 Case Study: Utilizing On-Orbit Assembly and Servicing to Enable Minimum-Cost Space Mission Architectures David L. Akin Mary L. Bowden AIAA Space 2005 Conference Long Beach, CA August 31, 2005

32 NASA Plan (Monolithic Architecture) Launch entire mission vehicle on single heavy-lift vehicle Launch crew in CEV on human-rated vehicle Earth orbit rendezvous docks crew and CEV to mission spacecraft Lunar orbit staging site leaves CEV in orbit while crew descends to lunar surface Lunar orbit rendezvous for crew to return to CEV CEV departs lunar orbit and travels back to earth (direct atmospheric entry)

33 Modular Architecture Multiple boost modules launched on EELVs and docked together Lunar landing/ascent vehicle launched on EELV and docked to boost module stack Launch crew in CEV on human-rated EELV Earth orbit rendezvous docks crew and CEV to mission spacecraft Lunar orbit staging site leaves CEV in orbit while crew descends to lunar surface Lunar orbit rendezvous for crew to return to CEV CEV departs lunar orbit and travels back to earth (direct atmospheric entry)

34 So What s the Argument? Both approaches are ELOR (Earth and lunar orbital rendezvous) Both approaches use CEV and dedicated lunar landing vehicle Both approaches use components from existing launch systems Both approaches have identical safetycritical rendezvous and docking operations

35 What are the Issues? Pros Cons Monolithic Minimize orbital operations Simpler operations Develop new large launch vehicles and associated ground infrastructure Modular Maximize use of existing assets Minimize nonrecurring costs Multiple docking operations increase odds of mission failure

36 Lunar Program Assumptions 2 lunar missions/year First lunar mission lunar missions total CEV entry vehicle mass 6000 kg Lander cabin mass 3000 kg ELOR mission with CEV as return craft LOX/LH2 Isp=450 sec Storables Isp=320 sec Inert mass fraction δ=0.1 except 0.15 for landing stage All launch vehicles asymptotically approach 97% reliability Rendezvous and docking reliability 99%

37 Lunar Mission ΔV Requirements To: From: Low Earth Orbit Lunar Transfer Orbit Low Lunar Orbit Lunar Descent Orbit Lunar Landing Low Earth Orbit km/sec Lunar Transfer Orbit km/sec km/sec km/sec Low Lunar Orbit km/sec km/sec Lunar Descent Orbit km/sec km/sec Lunar Landing km/sec km/sec

38 Candidate Cargo Launch Vehicles Delta IV Heavy 23K kg to LEO Operational Unmanned Representative of current large EELVs In-line SDLV 125K kg to LEO Conceptual Manned heritage

39 In-line SDLV Assumptions $8.4B nonrecurring (published estimate) 6 year development cycle $400M first unit production (shuttle parallel) 10 units at 85% learning curve $285M average flight cost

40 Shuttle-Derived CEV Assumptions $2B nonrecurring (NASA SVLCM estimate for second stage alone) 6 year development cycle $200M first unit production (shuttle parallel) 10 units at 85% learning curve $144M average flight cost

41 Delta IV Heavy Assumptions RDT&E amortized $2B nonrecurring for human rating $250M first unit 85 vehicle block buy and 85% learning curve yields $92M average cost (includes learning for 255 CBCs) 50% production surcharge for 11 human rated units ($138M)

42 Monolithic Launch Operations $429M average launch recurring cost Average amortized launch cost $1.45B 93% probability of individual mission initiation Probability of N missions initiating successfully 49% 10/10 85% 9/10 97% 8/10

43 Earth Departure Configuration 8 launches and 7 dockings required to start mission Assume Plaunch=0.97 and Pdock=0.99 Pno failures= Plaunch 8 Pdock 7 =0.73 Pall boost modules= Plaunch 6 Pdock 5 =0.792 Pall boost modules= Pno failures + P1 failure = (1-Plaunch)Plaunch 6 Pdock 5 = = 0.935

44 Spares - The Big Picture Have to get 6 functional boost modules for each of 10 missions Have to get functional lunar vehicle and crew module for each mission Assume composite reliability =0.97(0.99) =0.96 P (n n) = p n P (n n + 1) = n(p n 1 )(1 p)(p) n(n 1) P (n n + 2) = (p n 2 )(1 p) 2 (p) 2 n! P (n n + m) = (n m)!m! (pn m )(1 p) m (p)

45 Effect of Fleet Spares on Program Program Success Probability flights 60 flights Number of Spares

46 Spares Strategy Selection VSE approach: 2 launches and 1 dock: P=(0.97) 2 (0.99)=0.931 Program reliability over 10 missions: =0.492 Goal: meet VSE program reliability 1 lander and 1 CEV spare - p= each 2 boost module spares - p= Program reliability: (0.9308) 2 (0.5464)=0.473 Alternate goal: 85% program reliability 2 lander, 2 CEV, 4 BM spares: (0.9893) 2 (0.8871) = lander, 1 CEV, 6 BM spares: (0.9308) 2 (0.9838) =0.852

47 Modular Launch Operations $829M average launch recurring cost (includes cost of 5 fleet spares) Average amortized launch cost $1.10B 73% probability of individual mission initiation (no spares) Probability of 10 missions initiating successfully 16% (no spares) 71% (2 spares) 88% (3 spares) 96% (4 spares) 99% (5 spares)

48 Head-to-Head Launch Comparison 2000 Nonrecurring cost ($M) 10, Average production cost per mission ($M) 1096 Average amortized cost per mission ($M) 85 Total production run 432 NPV discounted cost per mission ($M)

49 Sensitivity to Monolithic Costing $432M Baseline NPV discounted cost per mission $432M Development costs cut in half $432M Production costs cut in half $432M $432M $878M $508M $809M Production is free $740M All costs cut in half $439M

50 Discussion of Reliability Monolithic architecture loses a mission when a launch or docking fails 75% of modular architecture failures occur on a boost module launch Plan for ready alert spare launch vehicle and boost module Continue mission buildup LV commonality allows robust spares strategy (boost module, CEV, lander) Can work full sparing for monolithic system, but requires both launch vehicle types, pads, etc.

51 Discussion of Costs Cost benefits of modular systems: Learning curve effects for large production runs Minimum up-front nonrecurring costs Modular systems also benefit from other markets for the same launch vehicles Minimal market synergy for monolithic vehicles In-line SDLV too large (5x current largest vehicle) SRB-based vehicle offers few intrinsic advantages to commercial/dod payloads Lesson: spend your money flying, not developing new vehicles (suggested mantra: flight rate, flight rate, flight rate )

52 Additional Caveats Didn t consider costing of mission vehicles CEV and lander are comparable for both architectures Additional cost advantages to modular system for smaller size/large production of boost modules as compared to monolithic TLI stage Modular system is sensitive to docking reliability, but it primarily shows up in spares strategy (low marginal cost for larger production run) Could use modular approach with SRBbased CEV launcher - minor overall impact to cost

53 Comments on Modular Architecture Modular system is highly adaptive to new missions and mass growth (add/subtract modules) Standard boost modules provide infrastructure for aggressive on-orbit operations ( space tugs ) Even with SDLV heavy-lift vehicles, will have to adopt modular-type operations for Mars missions but it is inelegant, complex, and just plain ugly

54 What You Just Saw... Orbital Mechanics Parametric Design Trade Studies Reliability Analysis Cost Estimation Engineering Economics Engineering Graphics U N I V E R S I T Y O F MARYLAND Parametric Design Principles of Space Systems Design

55 Akin s Laws of Spacecraft Design - #1 Engineering is done with numbers. Analysis without numbers is only an opinion. U N I V E R S I T Y O F MARYLAND Parametric Design Principles of Space Systems Design

Preliminary Cost Analysis MARYLAND

Preliminary Cost Analysis MARYLAND Preliminary Cost Analysis Cost Sources Vehicle-level Costing Heuristics Learning Curves 2 Case Studies Inflation Cost Discounting Return on Investment Cost/Benefit Ratios Life Cycle Costing Cost Spreading