Starship!

[JayUtah]

Wasn't that  effect why the later WW I airplanes were really difficult to fly, because not only did the propeller rotate, the WHOLE ENGINE rotated with it? Having a big lump of metal equalling a large percentage of the plane's mass rotating, offset the stick and rudder imput by 90 degrees.

Yes, the old rotary engines.  Another mechanical engineering nightmare.  But we owe those early engineers a debt of gratitude.  The techniques they developed for a lubricant system that would operate with the engine in any orientation live on in today's jet engines and rocket motors.  To be sure, the rotary engine didn't recover its lube oil at all.  But the more traditional engines such as in the "Jenny" had to deal with the fact that the oil sump wouldn't supply oil to the pump if the airplane was inverted or in otherwise extreme maneuvers.

[JayUtah]

There's been some rough third party estimates of Raptor's operating conditions...pressures get quite high, but temperatures are surprisingly low:
https://en.wikipedia.org/wiki/File:Raptor_Engine_Unofficial_Combustion_Scheme.svg

Yeah, I saw that.  Temperature really ends up being the problem because the whole point of any preburner is to generate pressure which can be converted to velocity to drive turbines.  You can't get rid of pressure because you want the pressure.  But the best ways we have of producing pressure also produce high temperatures.  The vocabulary of materials we can apply to contain pressure at very high temperatures is severely limited.  So lowering the temperature is a win.  I need to see if there is a later edition of Sutton and Biblarz (the quintessential authors on rocket propulsion for engineers) that deals with full-flow engines.

Another win that struck me this morning as I was putting the bacon in the oven was that if you have two separate pumps, the rotor in each case is less massive.  The gyroscopic effects that complicate bearing design increase in proportion to mass, so a lighter rotor won't incur as much reaction during gimbaling.  The F-1 rotor was massive.  And since the rotor is shorter, you can place the bearings closer together toward the center.  A shorter rotor is naturally stiffer too.  Depending on the actual parameters, you could conceivably get away with one large bearing at the center of the rotor.  That would be a major win.  But still, most engineers' gut feelings are that you need two bearings in all cases.  Or rather, two small bearings slightly separated from the center point of the rotor is still better than one.  But the farther you get from the center, the more susceptible a bearing will be to eccentric loads created by engine gimbaling.

[jfb]

Thanks, Jay.  There's a reason I'm a code monkey and not a real engineer, but I appreciated and enjoyed the details you presented. 

[Zakalwe]

SN11 now rolled out onto the pad. This one has the Raptors already installed, so possibly moving into cyro tests early?

[VQ]

....The more popular choice is fluid bearings....

In some similar but stationary applications (turbine compressors and chillers) I see magnetically levitated bearings, which are very efficient. Are those too heavy to scale for use in rocket turbines?

[JayUtah]

Too heavy, but also susceptible to thermal degradation.  Magnets function poorly when they get very hot.  Also, it generally requires an electricity source to energize the magnet.  That presents a startup problem.  But yes, magnetic bearings are generally really cool.

[molesworth]

Well that didn't quite go as planned...  (SN11 launch that is.)

No info yet on what went wrong, but from the camera streams on NASA Spaceflight and Everyday Astronaut it looked like some big chunks of debris came down in and around the tank farm, which isn't a good thing.

[bknight]

Maybe launching in the fog wasn't a good idea.

[Peter B]

Maybe launching in the fog wasn't a good idea.

Obviously the fog made it hard to see what went wrong, but surely this sort of spacecraft couldn't be vulnerable to fog could it?

What I was wondering about was a diffuse orange flame with what appeared to be small but very bright points of origin from about 25 to 30 seconds after launch in the piping of the left hand engine.

[smartcooky]

Maybe launching in the fog wasn't a good idea.

Falcon 9 boosters land on a barge 500 to 1000 km out in the Atlantic; the whole thing is autonomous from the time the rocket goes into "start-up" at T-60 seconds to landing around 11 minutes later. Starship test vehicles are also autonomous. Almost all of the information they get comes from telemetry. Fog will have little if any impact.

[Count Zero]

Almost all of the information they get comes from telemetry.

But did they get usable telemetry?  It looked like they were having downlink problems (though I don't know how extensive they were).  One announcer said "we've lost all data," but he may have only been referring to signal loss after the vehicle was destroyed.

Telemetry can usually get you to the bottom of the problem, but visual tracking and recording can help a lot.  NASA had multiple ground cameras for all launches, not just the test flights.  In the Apollo days, they had multiple cameras trained on each stage.  On both fatal Shuttle launches, ground cameras provided early clues to what went wrong, which helped focus the investigations.

I was extremely surprised that SpaceX would think that visual tracking was optional/unnecessary.

[cjameshuff]

Almost all of the information they get comes from telemetry.

But did they get usable telemetry?  It looked like they were having downlink problems (though I don't know how extensive they were).  One announcer said "we've lost all data," but he may have only been referring to signal loss after the vehicle was destroyed.

Telemetry can usually get you to the bottom of the problem, but visual tracking and recording can help a lot.  NASA had multiple ground cameras for all launches, not just the test flights.  In the Apollo days, they had multiple cameras trained on each stage.  On both fatal Shuttle launches, ground cameras provided early clues to what went wrong, which helped focus the investigations.

I was extremely surprised that SpaceX would think that visual tracking was optional/unnecessary.

They specifically mentioned they were still getting good telemetry in the earlier video dropouts.

If they were doing something substantially different they might be pickier, but the part it'd be most useful with is the part of flight they've had no trouble with, and the part they're having trouble with will mainly show in telemetry, with little outward sign (seeing SN10 come down hard wasn't very informative, for example). Also, this is the last test flight of its series of videos, SN15's been waiting for quite some time with many major upgrades (including engine mount changes that meant they had to use older Raptors on SN11), and it might be some time before there's better weather.

[apollo16uvc]

Moderately surprised to not see a thread on Starship, especially after yesterday's flight.  Despite the kaboom at the end, that was a spectacularly successful flight.  A real-world test of a new engine cycle, a new fuel, a new mode of flight, and they almost pulled it off on the first try. 

And, my God, this shot:

https://twitter.com/SpaceX/status/1336849897987796992

That's real time.  That's not slowed down.  It just looks slow because you're looking at a 12-story building falling at you. 

SN9 is already built, there are at least 6 more prototypes in various stages of completion, the first booster is under construction - 2021 is gonna be an interesting year in South Texas.

Yeah, 90 degree rotation in about three seconds. That's going to be interesting for passengers.

Perhaps it is to tress-test all the design elements, engines, gimbaling and flaps.

If structural integrity holds up in those manoeuvres its set for less extreme ones.

[JayUtah]

As we discovered in the theater, water fog can do a number on radio communications in certain bands, even in very small amounts.  So it doesn't surprise me that the video might have dropped out while the telemetry (ostensibly on a more robust frequency, and with data integrity checks in place) soldiered on.

The engine anomaly looks like it starts at around T+18, although the sudden deflagration doesn't occur until a few seconds later.  It looks to me like a fitting might have opened up.  The bright spots around the solid piping indicates the flame is hitting with considerable velocity, which then indicates a leak in high-pressure flow.  Unclear whether that relates to the ultimate explosion.  As Musk says, they won't know until they pick through the "bits."  Even well into the 21st century, sifting through wreckage is still part of the job.

[Peter B]

As we discovered in the theater, water fog can do a number on radio communications in certain bands, even in very small amounts.  So it doesn't surprise me that the video might have dropped out while the telemetry (ostensibly on a more robust frequency, and with data integrity checks in place) soldiered on.

The engine anomaly looks like it starts at around T+18, although the sudden deflagration doesn't occur until a few seconds later.  It looks to me like a fitting might have opened up.  The bright spots around the solid piping indicates the flame is hitting with considerable velocity, which then indicates a leak in high-pressure flow.  Unclear whether that relates to the ultimate explosion.  As Musk says, they won't know until they pick through the "bits."  Even well into the 21st century, sifting through wreckage is still part of the job.

The bits of the rocket? Or the bits of the telemetry data?