I guess it’s time for some idle speculation. Should we start a pool on what got worn?
They might have a problem with the turbines getting a temperature spike on shutdown because the mixture is drifting toward stoichiometric, but I doubt they’d have their valve timings off enough to let that happen, and in any event the spike should be extremely brief.
Or they could have a problem with low flow on the CH4 pump during shutdown. When they close the main CH4 valve, which is on the downstream side, the flow drops but the turbine is still being driven by the highly pressurized CH4 which was circulating through the engine, which is like a big hot compressed fluid reservoir driving the turbine. There’s no throttling valve prior to the turbine, and the compressed CH4 vents through the turbine and into the combustion chamber.
Running a centrifugal pump at low or zero flow conditions can cause cavitation, pitting, and bearing wear from radial offset loads and thrust loads, as pressure waves goes back and forth through the pump. They don’t have a recirculation valve that would just let the LCH4 go round-and-round in a loop as the pump spins down. But I don’t know how long a pump would need to run in a zero-flow condition for wear to become a problem.
Or the brief startup-shutdown cycles could be causing some thermal stresses, though I don’t know where that would directly lead to wear.
In any event, lessons learned will feed back into the Raptor design until they’re in line with Elon’s goal of thousands of maintenance-free flights, but they’re certainly not quite there yet.
While we’re on speculation, what does this portend for operation in vacuum? Will we see stranded Starships, or big kabooms that contaminate orbital space? Stay tuned folks. It would be highly ironic that the system designed for rapid access to space provides the means to prevent access to space….
I’m not sure how you go from a few week delay in the early testing of a very ambitious new rocket to “Will we see stranded Starships, or big kabooms that contaminate orbital space?” You wouldn’t happen to work for ULA or BO, would you?
Well pure speculation is just that. Sometimes it leads to ironic possibilities not probabilities.
You wouldn’t happen to work for ULA or BO, would you?
All I can say is that I try to avoid BO when I’m at work.
“The problem with speculation is you make a speck out of you and some guy named Lation.”
– Dr. Gregory House
It’s an easy progression for certain types of passive-aggressive personalities. It’s a lot easier for people with a sizable degree of crab-bucketism in their psychologies, but no stomach for the full-throated hatred and bile-spewing of, say, a Gary Church, to just make mountains of molehills or conjure up exotic future calamities based on transient hiccups in an aggressive test program that was designed with breakage in mind.
Crab-bucketism, thanks for that. I had to look it up. Interesting behavior. But if your talking about me, I try not to be too crabby.
Idle speculation is merely that. Let’s try not to read too much into the mind or character of the contributor. It’s an interesting *possibility* definitely not a *probability*. In fact highly unlikely bordering on the impossible (denial of access to space). A stranded Starship, perhaps not impossible but still highly unlikely. A stranded crewed Starship even less so, given the testing program on-going.
Before I also get accused of fan-boisim, my biggest interest in SpaceX as a practical matter is in Starlink as an alternative / backup to my current monopoly ISP. If SpaceX/Starlink accomplishes just that only, I’d still consider it a great success.
“a temperature spike on shutdown because the mixture is drifting toward stoichiometric”
Seems a simple fix for that, bias valve timings to err the other direction on shutdown.
“When they close the main CH4 valve, which is on the downstream side, the flow drops but the turbine is still being driven by the highly pressurized CH4 which was circulating through the engine”
A bit harder, as a hardware change is involved, but a valve to bypass the turbine seems an obvious remedy. Or a valve downstream of the turbine.
Those three back-to-back restarts are roughly similar to the multiple restarts on the flight profile for SN8 and SN9. I wonder if they got any telemetry back from SN8 that indicated a possible problem. When they finally land one of these (hopefully SN9), those Raptors will certainly get a full tear-down and inspection.
Meanwhile, BE-4’s first flight will be on a paying customer.
My personal guess is that the issue ir related to resonance vibrations. When they shut down, there’s a very loud, deep sound on the audio, somewhat akin to a large water main valve closing. It’s been heard before on most of the Boca Chica Raptor static fires.
So, my guess is there’s an oscillating combustion effect, perhaps in the turbopumps, as the Raptor shuts down, causing vibration stresses.
That noise, BTW, is very audible on the live feeds I’ve seen. I’m not positive it’s related to the problem, but I am positive that it occurs.
I downloaded “SpaceX Boca Chica – SN8 First Ever 3 Raptor Static Fire” from Youtube, separated the audio and ran it through Sony SoundForge to get a spectrum on the scream or moan during the shutdown.
It had three resonant peaks that didn’t shift all that much throughout the event.
fundamental at 128 Hz and -31 dB, second harmonic at 257 Hz and -25 dB, and fourth harmonic at 524 Hz and -23 dB.
So the higher frequencies were more intense, but that could be influenced by distance, mechanics, and the microphone. Assuming the mic is flat, in terms of sound (air) pressure, the three would’ve run about 40% at 128 Hz, 80% at 257 Hz, and 100% at 524 Hz.
Interesting. Those three measurements correspond pretty closely to C-flat or C in the 3rd, 4th and 5th octaves. Like brass wind instruments, rocket engines have bells and require gas flow to operate. The Raptor is far from alone in producing a “shriek” or “honk” upon shutdown. Some form of residual gas transient seems to produce characteristic “notes” or “chords” at shutdown.
But I suspect the minor damage Elon’s tweet referred to is unconnected to the transient shutdown “honking” of the engines or to anything else internal to the engine mechanisms.
SpaceX seems to have done a decent job of repairing previous damage to the concrete immediately below the test stands, but the three static fires still produced a bit of fresh spallation which was visible in the video coverage. SN9 apparently had its engine bay “up-armored” to at least the same degree as SN8’s was after some quite serious static fire spallation-related damage occurred pursuant to one of SN8’s static fires. I’m guessing the first round of engine bay up-armoring has proven to be not quite sufficient.
The high-pitched “chirp” after the SN9 static fires sounded much like the similar chirps we often (and a bit embarrassingly) had on the 4K14 engine on the X-Racer. The chirp on the XCOR engine was caused by the check valves in the cascaded purge chattering as they normally do, and Raptor may be making a similar honk.
This leads me to believe that the existing GN2 purge supply was refilled with GHe, and the CH4 header tank’s press system was then tapped from that. That is probably the quick and dirty solution at the trivial cost of buying more helium. That the purge tone went up from SN8 to SN9 also suggests the GHe substitution.
On a somewhat related note, NASA tested the SLS and I think the burn lasted for about 1:15 instead of the scheduled 8 minutes. Of course it was pretty much hailed as a success if you listen to NASA TV.
Well MCF was not a call this ascent flight director ever wanted to hear: Major Component Failure is detected by the SSME controller.
Someone else noted that Artemis 2021 is off the table because that schedule required the core stage to be shipped to Florida in February, and that is now unlikely.
They’re using some of the most tested engines ever made, but it’s been almost ten years since they’ve really run multiple RS-25’s, and I’m wondering if changes introduced for the SLS had an issue, or if knowledge, skills, and familiar procedures have gotten a bit rusty.
IMHO, you’re right on your speculation as to the causes, though my hunch is it’s not either/or, it’s both.
We have an interesting conjunction here; Starship had some engine problems during its 3 static fires, and requires two engines to be changed out. SLS had an engine problem, and now, I’m guessing, requires one engine to be changed out. In both cases, they have to figure out what went wrong, etc.
So, any guesses as to which program, SLS or Starship, will get the engine changeout done first and get on with their test program? I think I know which way I’d bet, and it won’t even be close. 🙂
Why does SLS have an eight minute test, when Starship and Falcon 9 get tested for a few seconds at a time?
And, do you-all think it’s significant that SpaceX had problems with two of three engines? Are these new ones or were they used before? Are they very different from the Falcon 9 engines that have been doing so well? Just trying to sort this out.
The purpose of the Green Run is to simulate the performance of the vehicle through MECO, which is about eight minutes. SpaceX is just making sure that they don’t have any propulsion issues.
Raptor is not Merlin, and the have much less experience with it (the SN8 flight was the highest altitude it’s ever flown). But they can make rapid adjustments, which NASA can’t. Any hope of the first flight of SLS this year is almost certainly gone.
Thanks!
Having gotten used to SpaceX’s cadence, I really can’t get my head around the idea that anyone expects that SLS is going to work. Work the first time, basically untested, after a decade of purely theoretical development?
Nuts.
Brian Barnes tweeted (in response to Wayne Hale)
They talked about this during the presser. 5 days to procure enough propellant, 10-15 days to replace the whole engine if necessary. At most, a month or 2 to completely recycle.
And that’s without tying everyone up in meetings called by a new administration looking to put their stamp on everything and chart a new direction for a program that’s never going to go anywhere.
And at Boca Chica, SpaceX has already replaced the two Raptors, and is looking at doing a new static fire tomorrow. Slightly faster than the SLS recycle measured in months (and I’ll bet it turns out to be more than two). 🙂
a program that’s never going to go anywhere.
Well, to paraphrase another administration, that all depends on how your definition of ‘go’ goes. As a means of producing reviews and studies, SLS may be the premiere program of all time. In terms of pages of documentation produced, I wonder how SLS compares to Apollo? Where are we in terms of money spent vs Apollo, after correction to 2021 dollars? That may be premiere as well. So you see, the definition of go, going, and gone all depends on your go-als.
I think the burn lasted for about 1:15 instead of the scheduled 8 minutes.
Scott Manley – “um, well unlike everything else in the SLS program, the test ended ahead of schedule.”
I guess it’s time for some idle speculation. Should we start a pool on what got worn?
They might have a problem with the turbines getting a temperature spike on shutdown because the mixture is drifting toward stoichiometric, but I doubt they’d have their valve timings off enough to let that happen, and in any event the spike should be extremely brief.
Or they could have a problem with low flow on the CH4 pump during shutdown. When they close the main CH4 valve, which is on the downstream side, the flow drops but the turbine is still being driven by the highly pressurized CH4 which was circulating through the engine, which is like a big hot compressed fluid reservoir driving the turbine. There’s no throttling valve prior to the turbine, and the compressed CH4 vents through the turbine and into the combustion chamber.
Running a centrifugal pump at low or zero flow conditions can cause cavitation, pitting, and bearing wear from radial offset loads and thrust loads, as pressure waves goes back and forth through the pump. They don’t have a recirculation valve that would just let the LCH4 go round-and-round in a loop as the pump spins down. But I don’t know how long a pump would need to run in a zero-flow condition for wear to become a problem.
Or the brief startup-shutdown cycles could be causing some thermal stresses, though I don’t know where that would directly lead to wear.
In any event, lessons learned will feed back into the Raptor design until they’re in line with Elon’s goal of thousands of maintenance-free flights, but they’re certainly not quite there yet.
While we’re on speculation, what does this portend for operation in vacuum? Will we see stranded Starships, or big kabooms that contaminate orbital space? Stay tuned folks. It would be highly ironic that the system designed for rapid access to space provides the means to prevent access to space….
I’m not sure how you go from a few week delay in the early testing of a very ambitious new rocket to “Will we see stranded Starships, or big kabooms that contaminate orbital space?” You wouldn’t happen to work for ULA or BO, would you?
Well pure speculation is just that. Sometimes it leads to ironic possibilities not probabilities.
You wouldn’t happen to work for ULA or BO, would you?
All I can say is that I try to avoid BO when I’m at work.
“The problem with speculation is you make a speck out of you and some guy named Lation.”
– Dr. Gregory House
It’s an easy progression for certain types of passive-aggressive personalities. It’s a lot easier for people with a sizable degree of crab-bucketism in their psychologies, but no stomach for the full-throated hatred and bile-spewing of, say, a Gary Church, to just make mountains of molehills or conjure up exotic future calamities based on transient hiccups in an aggressive test program that was designed with breakage in mind.
Crab-bucketism, thanks for that. I had to look it up. Interesting behavior. But if your talking about me, I try not to be too crabby.
Idle speculation is merely that. Let’s try not to read too much into the mind or character of the contributor. It’s an interesting *possibility* definitely not a *probability*. In fact highly unlikely bordering on the impossible (denial of access to space). A stranded Starship, perhaps not impossible but still highly unlikely. A stranded crewed Starship even less so, given the testing program on-going.
Before I also get accused of fan-boisim, my biggest interest in SpaceX as a practical matter is in Starlink as an alternative / backup to my current monopoly ISP. If SpaceX/Starlink accomplishes just that only, I’d still consider it a great success.
“a temperature spike on shutdown because the mixture is drifting toward stoichiometric”
Seems a simple fix for that, bias valve timings to err the other direction on shutdown.
“When they close the main CH4 valve, which is on the downstream side, the flow drops but the turbine is still being driven by the highly pressurized CH4 which was circulating through the engine”
A bit harder, as a hardware change is involved, but a valve to bypass the turbine seems an obvious remedy. Or a valve downstream of the turbine.
Those three back-to-back restarts are roughly similar to the multiple restarts on the flight profile for SN8 and SN9. I wonder if they got any telemetry back from SN8 that indicated a possible problem. When they finally land one of these (hopefully SN9), those Raptors will certainly get a full tear-down and inspection.
Meanwhile, BE-4’s first flight will be on a paying customer.
My personal guess is that the issue ir related to resonance vibrations. When they shut down, there’s a very loud, deep sound on the audio, somewhat akin to a large water main valve closing. It’s been heard before on most of the Boca Chica Raptor static fires.
So, my guess is there’s an oscillating combustion effect, perhaps in the turbopumps, as the Raptor shuts down, causing vibration stresses.
That noise, BTW, is very audible on the live feeds I’ve seen. I’m not positive it’s related to the problem, but I am positive that it occurs.
I downloaded “SpaceX Boca Chica – SN8 First Ever 3 Raptor Static Fire” from Youtube, separated the audio and ran it through Sony SoundForge to get a spectrum on the scream or moan during the shutdown.
It had three resonant peaks that didn’t shift all that much throughout the event.
fundamental at 128 Hz and -31 dB, second harmonic at 257 Hz and -25 dB, and fourth harmonic at 524 Hz and -23 dB.
So the higher frequencies were more intense, but that could be influenced by distance, mechanics, and the microphone. Assuming the mic is flat, in terms of sound (air) pressure, the three would’ve run about 40% at 128 Hz, 80% at 257 Hz, and 100% at 524 Hz.
Interesting. Those three measurements correspond pretty closely to C-flat or C in the 3rd, 4th and 5th octaves. Like brass wind instruments, rocket engines have bells and require gas flow to operate. The Raptor is far from alone in producing a “shriek” or “honk” upon shutdown. Some form of residual gas transient seems to produce characteristic “notes” or “chords” at shutdown.
But I suspect the minor damage Elon’s tweet referred to is unconnected to the transient shutdown “honking” of the engines or to anything else internal to the engine mechanisms.
SpaceX seems to have done a decent job of repairing previous damage to the concrete immediately below the test stands, but the three static fires still produced a bit of fresh spallation which was visible in the video coverage. SN9 apparently had its engine bay “up-armored” to at least the same degree as SN8’s was after some quite serious static fire spallation-related damage occurred pursuant to one of SN8’s static fires. I’m guessing the first round of engine bay up-armoring has proven to be not quite sufficient.
The high-pitched “chirp” after the SN9 static fires sounded much like the similar chirps we often (and a bit embarrassingly) had on the 4K14 engine on the X-Racer. The chirp on the XCOR engine was caused by the check valves in the cascaded purge chattering as they normally do, and Raptor may be making a similar honk.
This leads me to believe that the existing GN2 purge supply was refilled with GHe, and the CH4 header tank’s press system was then tapped from that. That is probably the quick and dirty solution at the trivial cost of buying more helium. That the purge tone went up from SN8 to SN9 also suggests the GHe substitution.
On a somewhat related note, NASA tested the SLS and I think the burn lasted for about 1:15 instead of the scheduled 8 minutes. Of course it was pretty much hailed as a success if you listen to NASA TV.
Wayne Hale (Twitter) had a few thoughts in his feed.
Someone else noted that Artemis 2021 is off the table because that schedule required the core stage to be shipped to Florida in February, and that is now unlikely.
They’re using some of the most tested engines ever made, but it’s been almost ten years since they’ve really run multiple RS-25’s, and I’m wondering if changes introduced for the SLS had an issue, or if knowledge, skills, and familiar procedures have gotten a bit rusty.
IMHO, you’re right on your speculation as to the causes, though my hunch is it’s not either/or, it’s both.
We have an interesting conjunction here; Starship had some engine problems during its 3 static fires, and requires two engines to be changed out. SLS had an engine problem, and now, I’m guessing, requires one engine to be changed out. In both cases, they have to figure out what went wrong, etc.
So, any guesses as to which program, SLS or Starship, will get the engine changeout done first and get on with their test program? I think I know which way I’d bet, and it won’t even be close. 🙂
Why does SLS have an eight minute test, when Starship and Falcon 9 get tested for a few seconds at a time?
And, do you-all think it’s significant that SpaceX had problems with two of three engines? Are these new ones or were they used before? Are they very different from the Falcon 9 engines that have been doing so well? Just trying to sort this out.
The purpose of the Green Run is to simulate the performance of the vehicle through MECO, which is about eight minutes. SpaceX is just making sure that they don’t have any propulsion issues.
Raptor is not Merlin, and the have much less experience with it (the SN8 flight was the highest altitude it’s ever flown). But they can make rapid adjustments, which NASA can’t. Any hope of the first flight of SLS this year is almost certainly gone.
Thanks!
Having gotten used to SpaceX’s cadence, I really can’t get my head around the idea that anyone expects that SLS is going to work. Work the first time, basically untested, after a decade of purely theoretical development?
Nuts.
Brian Barnes tweeted (in response to Wayne Hale)
And at Boca Chica, SpaceX has already replaced the two Raptors, and is looking at doing a new static fire tomorrow. Slightly faster than the SLS recycle measured in months (and I’ll bet it turns out to be more than two). 🙂
a program that’s never going to go anywhere.
Well, to paraphrase another administration, that all depends on how your definition of ‘go’ goes. As a means of producing reviews and studies, SLS may be the premiere program of all time. In terms of pages of documentation produced, I wonder how SLS compares to Apollo? Where are we in terms of money spent vs Apollo, after correction to 2021 dollars? That may be premiere as well. So you see, the definition of go, going, and gone all depends on your go-als.
I think the burn lasted for about 1:15 instead of the scheduled 8 minutes.
Scott Manley – “um, well unlike everything else in the SLS program, the test ended ahead of schedule.”
*snort*
+1