Friday, March 5, 2021

Another Strike Against Boeing and the SLS

According to a news item in the Ars Technica weekly "Rocket Report,"  George Abbey, NASA veteran as former director of Johnson Space Center and an influential, long-time human spaceflight leader, in a report for the Biden administration (pdf warning) has said that the Space Launch System (SLS) should be reconsidered.  The goal of the document, from the Baker Institute, was to provide decision-makers "relevant and effective ideas" for supporting the nation's policy goals.
Launch costs should matter ... "In view of the current availability of a significant number of commercial launch vehicles with proven payload capabilities, as well as the industry's progress in providing a launch vehicle with significantly greater lift capabilities, the Biden administration should reconsider the need for the SLS during its annual budget review," writes Abbey, who is now a senior fellow in space policy for Rice University.
I know that it may seem that I'm picking on the SLS, but it seems to be the embodiment of what I consider "old space."  The program was started for the political benefit of spreading money around rather than for a specific mission, and was changed over and over again as projected missions changed.  Now, approaching five years late, the dates on their schedule appear wildly optimistic. 

Last January, 2020, NASA and Boeing delivered the first (partially) completed SLS core stage to Stennis Space Center, where an unused half-century-old test stand was refurbished for several hundred million dollars for a single major objective: the Green Run. 

The Green Run amounts to a full-duration (eight minute) static fire of the SLS core stage: the main liquid-fueled booster and its four repurposed Space Shuttle Main Engines (SSMEs).  Prior to that ignition and burn test, Boeing and NASA would put the rocket through a number of other basic tests for the first time, including power-on, modal analysis, and a wet dress rehearsal (a static fire without the ignition).  If you watch SpaceX at Boca Chica, this is what you see going on when a new Starship prototype goes to the test stand.  They take a month, but it's a much smaller vehicle.

At the time of arrival of the SLS core at Stennis, ever zealously overconfident NASA and Boeing officials talked about the Green Run like it was a small, moderately inconvenient hurdle set in front of their flawless SLS rocket.  A Boeing official estimated that the rocket would complete the test campaign and be ready to ship to Kennedy Space Center by the middle of last year!

Of course, that didn't happen.  A full year of apparently unanticipated issues with the vehicle led to January's static fire test that ended with an aborted test 62 seconds into a planned eight minute firing, about 13% of the desired run.  After an embarrassingly long time before they announced they would actually insist on a re-test, NASA and Boeing discovered an engine valve issue for the second time (the first was last November), requiring at least another 4-6 weeks to repair. That valve was fixed on March 3rd, setting up a second static fire attempt no earlier than March 16th.

SLS-1 Core at Stennis Space Center.  NASA photo.

Here's the big issue.  This SLS core is intended to fly the first, unmanned, orbital Artemis mission.  Hilariously, NASA and Boeing are still sticking to their schedule fantasy and continue to parrot the party-line that SLS's Artemis-1 launch debut with this prototype is still on track to launch in "November 2021."  That date was set almost a year before the SLS Green Run schedule delays ballooned by at least nine months.  If NASA and Boeing are lucky and there are no more failures that eat months at a time, Artemis-1 might fly by the fall of 2022.


  1. Heh....."Old Space".

    I think we're witnessing the beginning steps of "Space V2.0".

  2. The SLS needs to be canceled. It is pure boondoggle. Heck, SpaceX is further along that the ULA is even with all of the problems with the Starship test articles. SpaceX has spent much less money and gotten some real results.

  3. One core issue is that, at some point in the prehistoric past, politicians decided that NASA didn't actually need to do *development* on their new launch vehicle. Why, they had all these parts from the Space Shuttle left over, all NASA needed to do was use the parts to make a new vehicle. Rearrange them however, but NOTHING NEW. That would be expensive.

    One expression I heard was "never let a congressman tell you you'll save money buying a used rocket engine." It turns out that air and spacecraft aren't lego structures, and components aren't lego bricks.

    Anyway, there are only so many OMS engines and SSMEs left over from the shuttle program. (In fact, there aren't enough SSMEs to make the program work, even if all were in perfect condition.) This leads to an extreme reluctance to actually test these engines, even though test them they must. They might break them before the final vehicle is assembled, and then where would they be?

  4. All the old hand-drafted drawings of the components are from companies that no longer exist (Aerojet, for example (de facto)) in locations that no longer exist (the California Rocketdyne campus has been parted off and sold) by people that retired at least two decades ago, probably four. I remember management making a huge deal about finding one of the last people who had actually turned wrenches on the engines during the shuttle program and hauling him out of retirement so he could fill in the entire universe of tacit knowledge between the documentation (sparsely preserved) and the physical thing. Irreproducible is the world of engineering that went into making it. Panicked management shouting "we can't have a test program, WE DONT HAVE MONEY OR TIME", but there was endless money and time for trawling the documentation.

    NASA prefers to reference old test reports for tests conducted in the 90s, than conduct new tests: They might break something! That constitutes their due dilligence and QA. The few functional mockup tests they had conducted (with some discrete exceptions) were mickey-mouse caricatures not representative of the final system at all - that would be expensive and take time! (Airbus testing with water, for example, instead of the propellants, with standin valves, and standin orifices for the engines, not properly documenting pressure drops and flowrates, and checking the box by horking up a 1500 page pdf text "report" instead of their datasets.) Don't get me started on their computer "modeling" - this post is long enough as it is.

  5. The propellant chemistry for the in space engines is actually very complex and finicky, and they had no one on hand who knew anything about it. They asked us to get up to speed by having us read old NIST standards about the properties of MON and MMH. It turns out that MON (mixed oxides of nitrogen) is actually this witches brew of nitrous oxides, nitrogen tetroxide (you hope in majority), and other compounds formed from your tank walls in precarious equilibrium. It turns various pretty rust-red colors, or green, depending on what mood it's in. The history book Ignition was actually a little more informative about the entire program of development that had to be conducted to find out all the things you had to do to handle it properly.

    One test, some manager had the bright idea of reusing old MON leftover from the Apollo program becuase "cheaper!". And having someone make more in the modern world is a daunting proposition! It had been sitting in a stainless steel tank until it turned pale green. Per Ignition, this is a warning that it had leached the iron into it and could not be safely used. When fired, compounds called "fuel-oxidizer-reduced-products" would from from incomplete combustion - brown nicotine-stain colored jello substances which would violently explode on subsequent engine firings.

  6. **** anyway this is getting long ***

    You can't have an engine without an engine development project. It looks like an engine is just something that sits in a box that you can take out and plug in. But an engine is actually all the engineers that know how to make the engine, the industrial world that they live in (where they sit on the org chart - hint: not at the very bottom!), their vendors and suppliers, the test facilities and the people who actually conducted and watched the tests. The paper is dead - it isn't actually the knowledge that lets you use the engine. You can't construct a spacecraft by crawling around ancient documentation like Kafka's bug-man while the test stands rust away outside your office block.

    SpaceX slapped some Raptors on the Starship. To an idiot, this looks like an equivalent action to Boeing slapping an SSME on the core stage. (That old tale about the pacific island cargo cult comes to mind.) The difference is SpaceX had the engineers who actually built and tested the engine do the slapping-on. Those engineers had spent the previous year blowing up variants of the Merlin and Raptor on actual test pads. They knew their suppliers (themselves in many cases) and could call for replacement parts. There wasn't anything on the engine they couldn't modify if needed.

    1. Anon - thanks for the great perspectives.

      I'm reading this as another example of the problems with documentation vs. expertise of the engineers and other people doing the job. This is a problem that's everywhere in every industry. I firmly believe that's where the old joke that "the project doesn't fly until the paperwork stack is taller than the launch vehicle" comes from. The next problem is when the new crop of engineers and managers come along 40 years later, do they actually study that pile of documentation? Not just read it but study every sheet? Chances are that what they need to know is in that stack, but the important parts might be small paragraphs in obscure places. The experienced engineer knows those things but to the newcomer it's just another paragraph.

      But that's only meaningful if they encounter every possible problem and address them in the documentation.

      You might have come across some quotes from Elon Musk about the importance of manufacturing. His vision is factories turning out Raptors and Starships, in the quantities of the factories turning out cars. At "Class S" quality and reliability levels.

    2. It looks like NASA is Amtrak, run by Dilbert's pointy-haired boss.

  7. What we called "Tribal Knowledge" at Boeing. No matter how well the documentation is, if you can find one of the old test notebooks with the Test Conductors notes in it, and you can read between the lines a bit, you can gain as much knowledge as from the original documents.