Of the eight total orbital crewed spacecraft operated historically, six have used some form of crew escape, with two using an ejection seat system similar to aircraft and the rest using full capsule escape, the latter demonstrated successfully with crew onboard in 1983.
Of the two remaining vehicles, Voskhod, supported only two flights, and the Space Transportation System or space shuttle, conducted 133 successful flights with one launch failure. Lack of any form of crew escape during launch was among criticisms of the space shuttle, with safety being one of the primary drivers for ending the program. Launch abort systems have been baselined by NASA for both its Orion spacecraft and the Commercial Crew program in the post-shuttle era.
Given this historical precedent, currently adopted standards and emphasis on launch escape system development in all significant current spacecraft development programs, the lack of an escape system in the SpaceX Interplanetary Transport System is particularly glaring. Some comments by Musk in a post-conference press Q&A do allude to the issue:Musk: spaceship can serve as own abort system from booster, but on Mars, either youíre taking off or youíre not. #IAC2016
Tweet by Jeff Foust, 12:26 AM, 28 September 2016Asked about abort modes for launcher @elonmusk said "make it very reliable ... you do not have parachutes for commercial airliners" #IAC2016
- NASAWatch, 12:27 AM, 28 September 2016Musk: Aborting 100 people safely is not feasible. Instead, focusing on safety of vehicle. @SpaceX #IAC2016
Tweet by Brendan Byrne, 12:26 AM, 28 September 2016
While it would be reasonable to assume that some internal discussion of this topic has occured within SpaceX, these statements are uncorroborated and remain especially egregrious in comparing vehicle reliability with that of airliners. Failure rates for the shuttle were estimated at 1/100 000, and likely influenced the design of safety procedures and equipment; sober reappraisal indicated that the actual figure was closer to 1/100. While launch vehicle reliability can undoubtedly be improved, launch vehicles are not aircraft, and reliability rates approaching those of aircraft should not be regarded as a trivial feat.
Furthermore, no independent studies of the viability of large scale launch vehicle passenger abort systems have been conducted by the FAA, NASA or other organization with concerns in the aerospace safety field. Given the lack of study, determination of the infeasbility of such a system is doubtful, especially considering the effect of differing design decisions and assumptions on the outcome of a study.Comparative safety values
Launch vehicle success rates average 91.92% overall and 93.32% within the last 20 years, with highly successful vehicles (Delta II, Soyuz-U, Ariane 5 etc) having success rates of 97-99%. This implies a failure once every 30-100 flights for modern vehicles.
For comparison, fatalities per journey are given as about 1/8 500 000 for aircraft. This is a four order of magnitude difference. It seems doubtful that ITS will approach a comparable reliability rate, especially within its early operational lifetime.
While it is possible to accept a high level of risk from an ideological perspective, it is necessary to operate within what is deemed professionally, legislatively and socially acceptable. This is especially pertinent in reference to the ITS being a commercial product that must attract a market of users.Intrinsic danger of launch vehicles and airliners
Launch vehicles and airliners do not differ solely in reliability statistics, but in the overall danger presented by a failure in the event that one does occur. Launch vehicles contain significantly more chemical energy, potential mechanical energy, and operate within environments less hospitable to human life.
Usable fuel capacity of a 737-800 is 20 894 kg, roughly half the operating empty weight of the aircraft. This amount of kerosene contains ~0.98 terajoules of chemical energy. The stated propellant capacity for the ITS launch stack is 8 650 000 kg, roughly 20 times the dry mass of the vehicle. Assuming a mixture ratio of 3.8, this represents ~1 802 000 kg of methane or ~100 terajoules of chemical energy. This is equivalent to nearly 24 kilotons of TNT, although it should be noted that in practice, total explosive yield is significantly lower than the TNT equivalent of the total chemical energy available. Chemical energy of methane in the upper stage alone is ~22.5 terajoules.
Rocket engines present unprecedented pressures, mass flow rates, volumetric power density and specific power compared to typical turbofan engines. While the figures shown at the IAC talk indicate some form of protective shield around each engine, the ability to contain any catastrophic failure remains undemonstrated, particularly given the high potential energy contained in an engine of this type.
An airliner can perform a gear-up landing or water landing, with all persons aboard able to disembark safely in fortunate situations. It is difficult to see how a vertically oriented vehicle will land in a stable fashion on water, without landing gear deployed, or with partial gear deployment. Historical examples demonstrate that the propellant tanks of a tipping vehicle will promptly rupture on impact followed by rapid deflagration of the propellant within, even when near-empty. Despite decades of flight experience, with thousands of gear extensions and retractions daily, mechanical landing gear failure does still occur in airliners
If a vehicle does land safely, but must be evacuated rapidly, the over 25 meter height between the crew compartment and the ground represents a significant challenge.
Comparing the lack of an escape system in a launch vehicle to the lack of passenger parachutes in airliners is a non-sequitur. Not only are airliners an impressively reliable form of transport, but pose less danger when they do fail. A launch vehicle escape system is not comparable to per-passenger parachutes onboard airliners, which have significant practical issues and questionable safety benefit. A parachutable passenger cabin would be a more appropriate analogy, but again the situation differs from the launch vehicle case significantly- the passenger cabin represents a significant portion of the overall mass and volume of an aircraft, and the amount of hardware required to bring it to the ground is potentially not much less than that already used in the normal operation of the aircraft.Launch Escape/Crew Escape systems comparison
The claim that the ITS upper stage can function as an escape system from the booster can be evaluated using supplied figures.
Dry mass: 150 000 kg
Propellant mass (total?): 1 950 000 kg
Payload mass: At least 100 000 kg
Total mass assumption: 2 200 000 kg
3x SL Raptor (3050 kN) in kgf: 933 000 kgf
6x Vac Raptor (3500 kN) kgf: 2 141 400 kgf
T/W at sea level: 0.42
Acceleration capability at sea level: 4.16 m/s^2
Total thrust (all engines), vacuum: 30 855 kN, 3 146 300 kgf
Total acceleration capability, vacuum: 14.03 m/s^2
Dry mass, capsule: 6400 kg
Dry mass, trunk (assumed): 1000 kg
Propellant mass (total?): 1388 kg
Payload mass: 3310 kg
Total mass assumption: 12 098 kg
Total thrust (incl cosine losses*), N: 563 560 N
Total thrust (incl cosine losses*), kgf: 57 467 kgf
T/W, sea level: 4.75
Total acceleration capability, sea level: 46.58 m/s^2*15 degrees assumed.
It is apparent that with only sea level engines operating, the vehicle is unable to abort from a stationary pad with zero dynamic pressure. While all engines operating in a vacuum exceed a T/W of one, this acceleration is very low compared to historic launch escape systems. Even assuming full vacuum performance on all engines and no exhaust flow separation (an unrealistically generous assumption), effective acceleration from a static start would be 4.22 m/s^2, taking roughly 5 seconds to accelerate 50 meters. With continued booster acceleration over 14 m/s^2, or significant dynamic pressure, it would seem that the second stage is unable to function as an escape system even under this generous assumption. This would negate discussion of all other issues, such as engine start response time or response to a major failure in the upper stage itself.
Offloading vehicle propellant would increase acceleration capability, but impair vehicle performance. Lessened useful payload to orbit is the most obvious criticism of requiring an escape system. Using an offloaded upper stage as an escape system maintains or worsens this impact while providing inferior abort performance.Crew Escape on Mars
Conventional wisdom would assert that abort capability would be useless on Mars. However, this holds true only in an expeditionary context where no resources are available to effect post-landing rescue operations. In the context of a significant human presence, such as the growing (and ultimately self-sustaining) colony envisioned in the ITS architecture, this does not necessarily hold true, and a Martian settlement may be able to dispatch ground or air vehicles for rescue purposes.
After an on-pad or shortly post-launch abort, a landed crew compartment will be located nearby the launch site, easily within the range of foreseeable vehicles. During launch, the Instantaneous Impact Point (IIP) of a launch vehicle remains at a relatively short geographic distance from the launch site for much of the flight, with downrange values exceeding 1/4th the planetary circumference occurring for a brief period shortly before reaching orbit.
Crewed electric vehicles have been operated on the surface of other planetary bodies, and researched at length by NASA, with widespread ongoing proliferation of terrestrial use, and ranges of several hundred kilometers demonstrated. For journeys longer than the range of electric vehicles, or which electric vehicles would not be able to make in time, suborbital flight is a possibility. A base or colony able to support orbital vehicles can necessarily support suborbital transport, presuming that the orbital vehicles can handle the different flight envelope, or that specialized vehicles are present.
Low atmospheric density on Mars poses a significant disadvantage for aerodynamic deceleration, potentially precluding an abort compartment from using parachutes. However, the thinner atmosphere propagates blast waves less effectively and incurs less drag throughout all phases of ascent. This combined with the lower mass of the launch vehicle and lower gravity losses during a pad abort may allow an abort regime on Mars to spare delta V for propulsive landing.Technical notions- escape system design space
For an escape system to be a realistic option, the minimum necessary volume and hardware to support the crew during launch must be provided by the vehicle for the duration of launch and landing, i.e. seating, transfer passages, and short-term life support only.
Major design decisions involve the flight attitude, propulsion system and landing method of the escape compartment. An aft-first flight attitude would make the abort compartment comparable to typical capsule vehicles; a front- or side-first biconic or Corona capsule style attitude may allow the use of the normal vehicle TPS for reentry after a high altitude abort. A separate volume accessible from the escape section could provide extra volume during orbital flight; this would resemble the MOL, TKS and Big Gemini designs.
In recent years there has been significant interest in integrated or 'pusher' abort systems, some with liquid propellants. This is a significant departure from the paradigm of solid fuel tractor systems that must separate during each launch. While a helium pressure-fed hypergolic system would likely have the highest technological readiness, a variety of propellants and pumping mechanisms should be investigated to determine the best options for safety, reliability and performance of abort propulsion in reusable vehicles.
Landing is perhaps the biggest issue for very large abort compartments. Propulsive landing is highly scalable and may be an option depending on control, relight and mass ratio requirements. The scalability of parachute landing is questionable, however, the space shuttle SRBs (dry mass of over 80 tons) were landed under parachutes with an impact velocity of 23 m/s, very plausibly within the range of a short propulsive burst.Notes
1: Soyuz 7K-ST No. 16L fire in 1983, after roughly 300 previous successful flights of the Soyuz-U launch vehicle.http://www.russianspaceweb.com/soyuz_sas.htmlhttp://www.spacelaunchreport.com/r7suma3.txt
2: Commercial Crew Transportation System Certification Requirements for NASA Low Earth Orbit Missions, section 5.6, Crew Survival/Abort Requirementshttps://www.nasa.gov/pdf/504982main_CCTSCR_Dec-08_Basic_Web.pdf
3: "Report of the Presidential Commission on the Space Shuttle Challenger Accident". Appendix F: Personal observations on the reliability of the Shuttle.
4: Worldwide Orbital Launch Summary by Year, Space Launch Reporthttp://www.spacelaunchreport.com/logyear.html
5: The Risks of Travelhttp://www.numberwatch.co.uk/risks_of_travel.htm
6: Figure for usable fuel, B737-800, pg 13http://wayback.archive.org/web/20111027100527/http://www.boeing.com/commercial/airports/acaps/737sec2.pdf
7: Energy density of various substances. Kerosene, 46.2 MJ/kg, Methane 55.6 MJ/kghttps://neutrium.net/properties/specific-energy-and-energy-density-of-fuels/
8: IAC presentation slides, pg 31, 33, 36 http://www.spacex.com/sites/spacex/files/mars_presentation.pdf
9: Final Environmental Assessment for Issuing an Experimental Permit to SpaceX for Operation of the DragonFly Vehicle at the McGregor Test Site, McGregor, Texashttp://www.faa.gov/about/office_org/headquarters_offices/ast/media/DragonFly_Final_EA_sm.pdf
10: SpaceX Raptor- Spaceflight101 http://spaceflight101.com/spx/spacex-raptor/
11: One terajoule is equivalent to roughly 239 tons of TNThttp://www.unitconversion.org/energy/tons-explosives-to-gigajoules-conversion.html
12: Saturnís fury: effects of a Saturn 5 launch pad explosion; cites 1960s paper "Saturn 5 Booster Explosion Hazards and Apollo Survivability Analyses", with stated TNT equivalent 0.54 kT. http://www.thespacereview.com/article/591/1
13: An Empirical Non-TNT Approach to Launch Vehicle Explosion Modeling https://ntrs.nasa.gov/archive/nasa/casi.ntrs.nasa.gov/20150002957.pdf
14: DC-XA failure
15: Falcon 9 flight 21 landing
16: Tesla model S https://electrek.co/2016/08/24/tesla-model-s-p100d-first-all-electric-car-over-300-miles-range/