Commercial Crew Hearing
Hearing: Update on NASA Commercial Crew Systems Development
“10 a.m. EST, the U.S. House Committee on Science, Space, and Technology will hold a Subcommittee on Space hearing titled An Update on NASA Commercial Crew Systems Development. The purpose of the hearing is to examine the development of NASA’s two commercial crew systems, being built by Boeing and SpaceX, to service the International Space Station.”
– Watch live
– Hearing charter
NASA will require seven flights of the Block 5 version of the Falcon 9 rocket before certifying it for humans, apparently. "Not a totally random number," Patricia Sanders says. Humans will go on second flight of SLS rocket.
— Eric Berger (@SciGuySpace) January 17, 2018
Prepared statements:
– Cristina Chaplain (GAO)
“Both Boeing and Space Exploration Technologies (SpaceX) are making progress toward their goal of being able to transport American astronauts to and from the International Space Station (ISS). However, both continue to experience schedule delays. Such delays could jeopardize the ability of the National Aeronautics and Space Administration’s (NASA) Commercial Crew Program to certify either company’s option–that is, to ensure that either option meets NASA standards for human spaceflight–before the seats the agency has contracted for on Russia’s Soyuz spacecraft run out in 2019.”
– Patricia Sanders (ASAP)
– Hans Koenigsmann (SpaceX)
– John Mulholland (Boeing)
– William Gerstenmaier (NASA)
– Rep. Smith
– Rep. Babin
– Rep. Bera
– Rep. Johnson
The “schedule delays” are caused by failing to fully fund commercial crew by congress and changing overly conservative legacy requirements by NASA.
And the lack of any commercial customers to share costs is due to the lack of any commercial destinations and the added expense from NASA overly conservative legacy requirements by NASA.
Bigelow Aerospace isn’t going to launch a commercial space station until there are commercial transports to get there. But, Bigelow Aerospace isn’t about to help fund commercial crew if NASA is going to do it for them.
Exactly. I find it annoying when folks complain that there is no where else for CC to go except the ISS when Bigelow has been waiting for years for CC before they could move further forward. The industry has been holding its breath waiting for actionably priced, NASA-certified people vehicles. Boeing has already lined up Bigelow as a customer. SpaceX has some customers lined up for a spin around the moon in a year or two. Both of these things will kickoff more activity. Demand for this has always been a chicken or the egg sort of thing and these providers will be launching more than just once per year almost right out of the box.
Except the FAA CST is responsible for licensing commercial systems, not NASA.
They do license the launches, yes. I was talking about connecting their reputation to the design.
NASA is responsible for working with industry to advance aeronautics and aerospace technology for commercial applications. Virtually all major industries and all major countries have arrangements for public-private collaboration for research and development which benefits the nation through development of new commercial industries.
Well put.
I see NASA’s aid to commercial cargo and commercial crew as similar to the US Government’s Air Mail contracts during the early years of aviation. A lot was learned by flying regular Air Mail routes. The lessons learned were put into the designs of newer aircraft. It really helped mature the industry in terms of safety and reliability.
That was why he offered the America Prize some years ago, but there were no takers afte NASA rolled out COTS. So he was forced to go into a holding pattern. The question is still out if CC will be affordable after NASA is done with it.
They should be since both providers plan on reusing capsules. SpaceX isn’t going to at first due to NASA’s request that only new capsules be used (at least for now). Hopefully neither capsule will cost a lot to refurbish. But, the devil is in the details and we won’t know for sure until they actually do reuse them.
I suspect some of the delays are due to a requirement for fewer than one fatal accident in 270 flights. For a program planning on one umanned and one manned flight (i.e. no meaningful test program) that’s impossible to prove. So I expect doing the impossible is taking longer than expected. Along these lines, does anyone know how many flights a new aircraft makes (either civilian or military) before it goes into service? I’d guess the answer is more than two.
Agreed. Even the space shuttle’s brief test flight program (5 orbital flights with two crew members in full pressure suits sitting in ejection seats, if memory serves) is better than what commercial crew is going to do. Well, except for the fact that the space shuttle’s first flight had two crew on board. 🙁
Four OFT missions down from the originally planned six. STS 5 had ejection seats but they were deactivated…4 crew.
Word on the street was that TK Mattingly pushed hard to fly an RTLS Abort profile during OFT but he was over-ruled.
Yeah, John Young was opposed to flying an RTLS. Something about you don’t practice bleeding.
Thing is, if shuttle could have been flown without a crew, RTLS could have been demonstrated without risking anyone’s life (like STS-1 had to do). One of the flaws in the space shuttle design was requiring crew on every flight. How much of this was out of necessity, or out of convenience (faster, cheaper development), or out of the politics of the astronaut office, who knows. I’m just glad that none of the crew capsules under development requires a crew on board for test flights.
I’m not sure. There is a discussion of testing RTLS aborts in the CAIB report (in a chapter on the history of the program.) I don’t have my copy handy, but I believe there was a significant risk that a RTLS abort would wreck the orbiter. Even a if the crew could walk away. Even if an unmanned abort test had been possible, would they have risked an orbiter?
That’s the problem with a vehicle designed to be flown by a human pilot. The risk of losing the orbiter would have been higher if it was automated versus piloted.
For example, STS-1 revealed that strange control system problem during reentry and landing that might have led to loss of orbiter if it were not piloted.
It’s a difficult balancing act between risking the life of the test pilot(s) and/or risking loss of the vehicle. The names of the streets at Edwards Air Force Base are a reminder of this grim fact.
Today’s NASA certainly seems far more risk averse (in general) than 1960’s NASA. How much of that is fear of LOM and how much is fear of LOC, I don’t really know.
Many in the program had serious doubts about RTLS. Legend had it that CDR Young used to sing “This will never work!” during RTLS runs in the simulator; never heard him do so myself.
Does the success of F9 stage recoveries back at the Cape retroactively suggest shuttle RTLS viability? I think to some degree it does…but not much beyond theory.
I asked a friend who used to be an engineer at Boeing about this. In his response he provided a link to a presentation by Boeing on this topic.
“Good question… I don’t remember the number of flight hours, but the actual flight testing takes a year or so. It usually includes 3-4 airplanes. One or two are fully instrumented, flight test only airplanes and the rest are intended for customers. The instrumented airplanes remain part of the flight test fleet after certification. They continue to be used to certify optional equipment, etc. At some point dedicated airplanes are no longer needed and they are turned into beer cans or donated to museums. The customer airplanes are refurbished and delivered to airlines. The customers in question get a pretty hefty discount for allowing the use of their airplanes in a flight test program.”
https://www.aa.washington.e…
That and NASA’s seemingly arbitrary rules. Some of their requirements are quite strict, but only for commercial crew. For example, Falcon 9 Block 5 must have flown successfully 7 times before flying a crew on top. SLS is going to fly crew on top on its second flight, but it will also be the very first flight for the upper stage. Ugh.
Jeff; I thought there was one other FH launch between the test vehicle on the pad now and the ‘Lunar Tourist’ one?
I think it’s two Falcon Heavy launches, one for the Air Force and one for ArabSat. But I thought the comment was about seven Falcon 9 block 5 flights, not additional Falcon 9 flights.
Falcon 9 Block 5 is the launch vehicle Dragon 2 will use for ISS missions. NASA wants 7 successful Falcon 9 Block 5 flights before putting a crew on top.
Falcon Heavy is not needed for manned ISS flights.
Perhaps he meant that it would be possible (but pricey/unlikely) to test two (or three) block 5 units at a time using a Heavy. Wasteful- but would speed certification. Wonder what the Space X assembly line looks like this year vis a vis block 5?
Perhaps, but NASA wouldn’t see it that way. They’re certifying Falcon 9 Block 5, the launch vehicle, for crew flights. They’re not certifying Falcon Heavy for crew flights. The two vehicles may share some common components, but they aren’t the same.
In particular, by saying Block 5 flights, NASA is saying the difference between Block 4 and 5 is too great for Block 4 to be relevant. Since Falcon Heavy involves more than just bolting Falcon 9 first stages together, that almost certainly means it’s as different as a Falcon 9 Block 4.
the 7 flights are pretty arbitrary, but I would say the LOC rate while difficult is sensible, the hope has always been that Dragon 2 and CST- 100 wont just be flying to ISS after all
I’m very glad one off the Congressmen pointed that out. If the program had been adequately funded from the start, instead of Congress carving big chunks out of it to feed snacks to SLS for two years, these systems would be flying by now. Complaining that the programs might slip past 2019, after starving their already record low funding, is just rich.
I noticed that part of why they called the hearing was over new 2019 schedule slippage concerns, but both companies stood by 2018 launch dates.
I also noticed how Boeing said that they would prefer not to do anything else with Space Act Agreement contracting after this…but SpaceX said they loved it!
I also noticed that the Boeing guy didn’t have any numbers in front of him for Boeing development cost sharing percentage estimates, but SpaceX did.
Good hearing. Gerst said 1 year to launch crew test flight. Plenty margin left. Soyuz brings home last NASA crew in Fall 2019. Probably will not be certified, I guess they will go anyway. I don’t think Shuttle was ever certified.
Nor Apollo, or Gemini or Mercury. or even Soyuz.
NASA writes the certification rules, but they also have consistently waived rules which they could not meet.
Sanders statement is infuriating.
SpaceX replicated the fault that caused the explosion. SpaceX developed operational procedures to mitigate the fault. But, in the interest of safety, we need to redesign the system. If NASA doesn’t think the redesign is acceptable SpaceX will have to use a different tank which is heavier. Which will reduce operational performance. So it will only be flown for NASA human missions and that will increase risk.
Meanwhile: “Two SRBs? on SLS Sure, that’s tested tech!”
Do these people even listen to how absurd they sound?
But by not redesigning the COPV to handle the case where the liquid helium might cause LOX to freeze solid, you would be accepting a single point of failure which, if it happened, would cause loss of mission (boom) and possibly loss of crew (if the escape system didn’t work 100%, and nothing is 100%). The failure could be caused by the people responsible for running the filling operation or some failure of the ground support equipment. Either way, NASA would have to agree that the risk of such a thing happening was so small that a redesigned COPV wasn’t needed. They don’t appear to be willing to accept that at this time.
Just to attach some numbers, if the odds of another COPV/solid oxygen catastrophic failure were 6% and the odds of escape system failing were also 6%, that would account for the entire 1 in 270 (0.37%) risk of crew loss NASA is asking for. I’m not implying the odd of either event are 6%. I’m just saying it doesn’t take much to eat up that 1 in 270 risk budget. And proving 6% risk of failure in a single full-up test isn’t all that easy.
But look at it a different way. Requiring a redesign requires making untested changes, and not just to the COPV (if, as someone wrote, the change added enough mass to require other changes.) What are the odds of introducing a new, unanticipated risk in the process, and how do those odds compare to the risk they are trying to avoid? I suppose requiring more tests, or unmanned flights, might do. But that could put them into an infinite loop of finding, fixing, and retesting/reflying before they’d let an astronaut fly.
Or they could just throw up their hands and say, “We are not aware of any unidentified risks.”
“But that could put them into an infinite loop of finding, fixing, and retesting/reflying before they’d let an astronaut fly.”
In computer software, we call that “analysis paralysis”. The only way to get out of that is to accept the risk and move forward. In this case, one possibility is to fly with the current tanks, with a waiver, while working on “better” tanks to be used in the future.
Let’s just put this out in the open.
There have been about 20 flights since the unfortunate COPV incident. That is direct, verifiable evidence that the issue has been understood and dealt with to a high degree of certainty.
NASA is fighting the last battle. It’s time for them to look up and focus on the future.
It isn’t quite that simple. 20 flights with no failures means a 50/50 chance the risk is over 3.4% per flight, and SpaceX has just gotten lucky. With a requirement to keep risks (of fatalities) under 1 in 270 (0.37%), just having a 20 out of 20 record isn’t enough. The problem (and any possibility that someone wants to derail commercial crew) is that 1 in 270 requirement. It isn’t realistic to expect anyone to prove they satisfy that requirement, at least not without hundreds of actual flights. (Nor, in my opinion, is setting unrealistic requirements and then waiving them a good idea. That was one of the things that led to the Columbia disaster.)
Granted, it is possible to fly with a known problem that simply has a low probability. I can’t prove that the COPV tanks either fit or DO NOT fit that description (at least with the publicly available data).
Would you agree that it is possible that an engineering fix could eliminate a failure mode? Like the redesign of the O-rings on the shuttle. After the fix, it wasn’t getting lucky, it was a fix. I’m presuming the that COPV issue is like that … and the 20 successful flights are not-yet-significant proof of that.