Blue Origin's BE-4 Engine

[Starshipdown]

Where in any way does that contravene what I wrote? Nowhere did I say that the issue was not looked at or dismissed out of hand.

Two ways:  In facts, it was not the engine that overperformed, but one of the pumps (and this will result in negligible overperformance of the engine).

And in attitude:  "Overperformance" sounds like a positive.  "Unintended part-to-part variation" sounds like a negative.  I was pointing out that both interpretations are possible.

And I hope that you do not make an athlete go to drug testing when they have the very best day of their career!

Most bicyclists would describe winning the Tour de France, or even a single stage of the Tour de France, as the very best day of their career.  But by the rules, "The winner of the day's stage and the overall race leader are automatically tested, as well as six or seven additional riders."

I would say that shows just how far we've fallen in recent years. Making the assumption that everyone's always cheating as a way to explain great feats as men and women push themselves to their limits.

Probably for the best that greats like Babe Ruth, Jesse Owens, Sir Don Bradman, Jackie Robinson, Ted Williams, Jim Thorpe, and more never lived to see this shameful time in our history.

[Robert_the_Doll]

"And I hope that you do not make an athlete go to drug testing when they have the very best day of their career!"

There's a reason the authorities routinely drug check winners of sporting events, both human and animal.

Did you ever hear the name Lance Armstrong, amongst others?

Yes, and I followed his case for a while. I would say that there has never been any hard laboratory-based evidence to indicate that he ever did any performance enhancing drugs. The closest that came was when samples taken from him in 1999 were retested in 2005 supposedly testing positive for EPO, but the lab work was found later found to be sloppy. His 2013 confession is likely due to him being put under so much duress and give him some chance of salvaging his life which was utterly destroyed because of the allegations.

And remember, he was tested many times throughout his career as an athlete. But remember, most of the "evidence" was from Floyd Landis' allegations to the United States Anti-Doping Agency (USADA). Landis, who, as you probably know, did twice fail a urine test.  The USADA's claims all boil down simply to "He and his team took low doses that would not show up in tests or used saline injections to mask hematocrit levels that would indicate use of EPO". A plausible, but also very convenient series of explanations because a person cannot defend against that since there is no way for them to do so years after the fact with more tests.

As far as BE-4 is concerned, the engine that had the 5% overperformance was disassembled and it was determined that there was no problem and testing has since resumed. Hopefully the qualification testing will be able to finish up in the weeks ahead.

[LouScheffer]

"And I hope that you do not make an athlete go to drug testing when they have the very best day of their career!"

There's a reason the authorities routinely drug check winners of sporting events, both human and animal.

Did you ever hear the name Lance Armstrong, amongst others?

Yes, and I followed his case for a while. I would say that there has never been any hard laboratory-based evidence to indicate that he ever did any performance enhancing drugs.

Interesting you mention Armstrong in the context of unexplained 5% performance improvements.

In most races, the winning time gradually shrinks as technique, training, and equipment improve.   However this pattern does not hold in bicycling.  Look at the winning times up the Alpe de Huez, for example.   In 1997-2006 the record was smashed by 5%, by Armstrong and others.   This is an *enormous* improvement in sports, equivalent to breaking the 100 meter record by 0.5 second, or the marathon record by 6 minutes.  Then came increased drug testing, and the best since then is 5% worse than in that period, and in fact only 6 seconds better (out of 40 minutes) than the record before this era.

So maybe Armstrong had the ride(s) of his life, along with a bunch of others, all during this time period, and later athletes are wimps that cannot come within 5% of the previous best.  But that's not the way I'd bet.

[DanClemmensen]

Colleagues, we are not proposing to do drug testing of the BE-4 and we do not think the engine will enter the tour de France. Please stay on topic.

[Starshipdown]

Colleagues, we are not proposing to do drug testing of the BE-4 and we do not think the engine will enter the tour de France. Please stay on topic.

I'd pay money to see BE-4, Raptor, Archimedes, and Aeon-R compete in the Tour de France. 

[Starshipdown]

https://twitter.com/jeff_foust/status/1636086726811590656

Bruno: still confident in a Vulcan launch in May. A little more than halfway through qualification testing of BE-4; likely the pacing item for launch.

We don't have a solid date for exactly when Qual testing started, but we know from past Tory tweets that 'pre-Qual testing' was underway mid-August (2022) and Qual testing was underway mid-September (2022), so being ~6 months into Qual testing and "little more than halfway through" does not inspire the greatest confidence for a rapid completion of Qual testing in the near future.

I'm not quite sure where you get this. The very earliest possible test firing of one of the two qual engines was caught by an imaging satellite in late November:
https://twitter.com/Harry__Stranger/status/1598106301363781632?ref_src=twsrc%5Etfw%7Ctwcamp%5Etweetembed%7Ctwterm%5E1598106301363781632%7Ctwgr%5E29e5a2e81b847add7145dffae144860fc11e7418%7Ctwcon%5Es1_&ref_url=https%3A%2F%2Fforum.nasaspaceflight.com%2Findex.php%3Ftopic%3D39674.2120

The Qual-1 engine firing seen in the recent video took place a little over two weeks later on December 9 (a rare instance of the timestamps being left in the video).

It's very possible that there were qual engine firings well before then, but we have little evidence. We only know by late September or early October that development work was done per the official press release, and the spin up to starting the qual testing began, presumably with cold flow tests, maybe an ATP for each engine to ensure they'd work, and then full on qual testing by November. We also don't know how the qual testing here is structured and what exact objectives are being looked for with each firing.

February 23rd we were told that 6 more weeks was needed:

https://twitter.com/SciGuySpace/status/1628894884982579201?ref_src=twsrc%5Etfw%7Ctwcamp%5Etweetembed%7Ctwterm%5E1628894884982579201%7Ctwgr%5E2e73094a8e56cadf274eb82d5310fdba857b116a%7Ctwcon%5Es1_&ref_url=https%3A%2F%2Fforum.nasaspaceflight.com%2Findex.php%3Ftopic%3D39674.2220

So, it's about 3 weeks or so later, and we should be hearing in about two to three weeks if everything is finished or if more time is needed.

[WindnWar]

https://twitter.com/jeff_foust/status/1636086726811590656

Bruno: still confident in a Vulcan launch in May. A little more than halfway through qualification testing of BE-4; likely the pacing item for launch.

We don't have a solid date for exactly when Qual testing started, but we know from past Tory tweets that 'pre-Qual testing' was underway mid-August (2022) and Qual testing was underway mid-September (2022), so being ~6 months into Qual testing and "little more than halfway through" does not inspire the greatest confidence for a rapid completion of Qual testing in the near future.

I'm not quite sure where you get this. The very earliest possible test firing of one of the two qual engines was caught by an imaging satellite in late November:
https://twitter.com/Harry__Stranger/status/1598106301363781632?ref_src=twsrc%5Etfw%7Ctwcamp%5Etweetembed%7Ctwterm%5E1598106301363781632%7Ctwgr%5E29e5a2e81b847add7145dffae144860fc11e7418%7Ctwcon%5Es1_&ref_url=https%3A%2F%2Fforum.nasaspaceflight.com%2Findex.php%3Ftopic%3D39674.2120

The Qual-1 engine firing seen in the recent video took place a little over two weeks later on December 9 (a rare instance of the timestamps being left in the video).

It's very possible that there were qual engine firings well before then, but we have little evidence. We only know by late September or early October that development work was done per the official press release, and the spin up to starting the qual testing began, presumably with cold flow tests, maybe an ATP for each engine to ensure they'd work, and then full on qual testing by November. We also don't know how the qual testing here is structured and what exact objectives are being looked for with each firing.

February 23rd we were told that 6 more weeks was needed:

https://twitter.com/SciGuySpace/status/1628894884982579201?ref_src=twsrc%5Etfw%7Ctwcamp%5Etweetembed%7Ctwterm%5E1628894884982579201%7Ctwgr%5E2e73094a8e56cadf274eb82d5310fdba857b116a%7Ctwcon%5Es1_&ref_url=https%3A%2F%2Fforum.nasaspaceflight.com%2Findex.php%3Ftopic%3D39674.2220

So, it's about 3 weeks or so later, and we should be hearing in about two to three weeks if everything is finished or if more time is needed.

The problem with those sat images is they don't capture how many times a day/week those test fires are happening. It leaves a lot of guess work and wiggle room in comparison to say Raptor or Merlin tests where there are basically live camera feeds watching it. So it is really difficult to tell what sort of test firing pace they are doing, based on does the dirt look a bit different from the last image taken X days ago.

So no way to tell when it was placed on the stands, if it's had to be removed at any point, etc. All the bits you need to be able to actually guess how the test campaign is actually going. But that is par for the course for Blue. At least when they start using the test stand near the Bama factory it'll be easier to know how and when test firings are done.

[Starshipdown]

The idea is not necessarily to show how many times the engine is firing per week, but instead to give us another data point in how far back the qualification firings started.

[WindnWar]

The problem being is we have no way to tell if that's a qual firing, or some other engine test, we can guess, but that's about al it is, a guess, the firings might have started in late November... or they could have started earlier and had a break in between.

Without site of the stand the only thing that can be said with any certainty is occasional an engine is test fired there. It's about as clear as mud beyond that.

[LouScheffer]

Without [sight] of the stand the only thing that can be said with any certainty is occasional an engine is test fired there. It's about as clear as mud beyond that.

It's clear as mud even with sight of the test stand(s).  With SpaceX we can see each firing in real time, but still don't know what they are for.   Presumably some are acceptance tests of completed engines, some are qualification tests, some are limit-finding, some are testing proposed changes, and so on.  Occasionally it's clear from the test (rapid restart, thrust vectoring), but usually it's not.  Plus we don't know which engine is being tested for each test.

So when the sightings are low-detail static snapshots once a week, it's hopeless to try to infer the details.

[Starshipdown]

The problem being is we have no way to tell if that's a qual firing, or some other engine test, we can guess, but that's about al it is, a guess, the firings might have started in late November... or they could have started earlier and had a break in between.

Without site of the stand the only thing that can be said with any certainty is occasional an engine is test fired there. It's about as clear as mud beyond that.

What other engine test would it be around that time frame?

I don't know how it could be anything else that wasn't qual testing related.

If you read the evidence I posted, the Blue Origin press release quotes Tory Bruno that development is over (this by early October).

Therefore, the satellite photos showing an engine test firing in late November are most likely qual tests or something very closely related to it, like a qual engine ATP firing. The November test firings are just mere weeks (23-29) before a known test seen in this video of the Qual-1 engine (December 9):

[LouScheffer]

The problem being is we have no way to tell if that's a qual firing, or some other engine test, we can guess, but that's about al it is, a guess, the firings might have started in late November... or they could have started earlier and had a break in between.

Without site of the stand the only thing that can be said with any certainty is occasional an engine is test fired there. It's about as clear as mud beyond that.

What other engine test would it be around that time frame?

We know the BE-4 version that will be used for New Glenn differs from that for Vulcan (air restart, if nothing else).   Presumably this needs to be tested, and it would make sense to do this in parallel with Vulcan qualification, rather than waiting until the last minute.

[Comga]

Where in any way does that contravene what I wrote? Nowhere did I say that the issue was not looked at or dismissed out of hand.

Two ways:  In facts, it was not the engine that overperformed, but one of the pumps (and this will result in negligible overperformance of the engine).

And in attitude:  "Overperformance" sounds like a positive.  "Unintended part-to-part variation" sounds like a negative.  I was pointing out that both interpretations are possible.

I was just watching Tim Dodd’s fascinating tutorial on “How to Start a Rocket Engine”.
It shows the amazingly complex interactions between all the parts of the RS-25, valves, turbines, igniters, coolant channels, etc. and all the ramp-ups, transients, oscillations, and dampening.  It showed things like moving a valve by a fraction of a degree at a precise moment to suppress a particular transient.
It would seem to be even harder to get right if the performance of individual subsystems like, say, a turbopump, vary by +/-5% due to manufacturing tolerances.

Does anyone here know the scale of subsystem variabilities in established engines like the RS-25 or Merlin, and how launcher programs accommodate these variations?

[Starshipdown]

The problem being is we have no way to tell if that's a qual firing, or some other engine test, we can guess, but that's about al it is, a guess, the firings might have started in late November... or they could have started earlier and had a break in between.

Without site of the stand the only thing that can be said with any certainty is occasional an engine is test fired there. It's about as clear as mud beyond that.

What other engine test would it be around that time frame?

We know the BE-4 version that will be used for New Glenn differs from that for Vulcan (air restart, if nothing else).   Presumably this needs to be tested, and it would make sense to do this in parallel with Vulcan qualification, rather than waiting until the last minute.

As far as we know, the difference between the NG BE-4 and the Vulcan BE-4 is "slight" as per Tory Bruno. The air restart capability is something that only one engine on NG will need, and also recall that Vulcan and its BE-4s have priority over NG's. With delays to NG, I doubt Blue Origin will be doing much qual work on the restart BE-4 until everything else is done.

[Jim]

Does anyone here know the scale of subsystem variabilities in established engines like the RS-25 or Merlin, and how launcher programs accommodate these variations?

There is a difference between a head start engine like the RS-25 and spin start engine like Merlin.   The complexity for a head start is evident in the video.  Spin start are easy.

[trimeta]

Does anyone here know the scale of subsystem variabilities in established engines like the RS-25 or Merlin, and how launcher programs accommodate these variations?

There is a difference between a head start engine like the RS-25 and spin start engine like Merlin.   The complexity for a head start is evident in the video.  Spin start are easy.

Do we know anything about how the BE-4 starts?

[xyv]

Does anyone here know the scale of subsystem variabilities in established engines like the RS-25 or Merlin, and how launcher programs accommodate these variations?

There is a difference between a head start engine like the RS-25 and spin start engine like Merlin.   The complexity for a head start is evident in the video.  Spin start are easy.

Do we know anything about how the BE-4 starts?

"Easy" is relative.  Spin or no spin it's still a whole lot of stuff to get up and keep stable through all of the transients.  We don't even know much about how Raptor starts ("...the secret sauce..." from another Everyday Astronaut video) and Blue is famously silent compared to Elon.  I am acutally surprised that showing start sequences doesn't start to cross into ITAR technical data concerns (not being able to watch the video yet).

[Jim]

"Easy" is relative.  Spin or no spin it's still a whole lot of stuff to get up and keep stable through all of the transients. 

The transients and interrelations are much less on a spin start.  Spin start are used on open cycle engines and there is less feed back

[trimeta]

"Easy" is relative.  Spin or no spin it's still a whole lot of stuff to get up and keep stable through all of the transients. 

The transients and interrelations are much less on a spin start.  Spin start are used on open cycle engines and there is less feed back

We may not know details about how BE-4 starts, but we do know it isn't an open-cycle engine.

[LouScheffer]

"Easy" is relative.  Spin or no spin it's still a whole lot of stuff to get up and keep stable through all of the transients. 

The transients and interrelations are much less on a spin start.  Spin start are used on open cycle engines and there is less feed back

Head start on BE-4 seems technically much harder than on a fuel rich preburner such as RS-25.  To do a head start they would have to let the LOX flow first, let it flash into vapor, and let the vapor start the turbine.  However since LOX is not used to cool the nozzle, there's a lot less surface area to boil the LOX, so the flow would be a lot less.  In addition this would mean an oxygen-rich main chamber start, which would need to transit through stochiometric to get to the steady state.  That seems dangerous.  In addition it would need to work for both a ground start (engine at room temperature) and when starting the landing burn (engine was running a few minutes earlier).  These seem like very different thermal states.

So I'd guess a spin start of some kind.  If so, what they use for initial spin power we don't know.