totm nov 2023 SpaceX Discussion Thread

[tater]

OTW

Landed. SECO. Nominal orbit

2 down, 1 to go.

[tater]

Next launch 42 min

weather really looks lousy though. Storm is probably OK heading inland, but WIND.... not so great

They have a placeholder for 8:47 Pacific now, maybe later in window?

https://twitter.com/i/broadcasts/1kvJpvQLaMkKE

 

[Spaceception]

With this launch, SpaceX will be 21% towards their goal of 148 launches for the year. The year is almost a quarter of the way through though, so we'll see if they can catch up. I really like how these double and triple headers are becoming more common.

It's giving us a glimpse into the future of spaceflight when these aren't occasional events, but daily expectations. Especially when so many rockets are coming up with reusability and/or high cadence built in. I wouldn't be surprised if we hit roughly 1 launch/day in the next few years in the US, with bursts of multiple launches in the same day from different providers (and in some cases, the same provider, but probably just SpaceX for awhile) every week or so.

[Hotel26]

5 minutes ago, Spaceception said:

wouldn't be surprised if we hit roughly 1 launch/day

Shades of Gattaca...

[tater]

 

[tater]

 

[tater]

 

[darthgently]

3 hours ago, tater said:

 

At some point as frequency of launch increases, rocket designs and agency anxieties will adapt to the point that rockets like F9 will be able to launch in any weather a commercial airline flight takes off lands during.  Imagine an F9 launch and landing during a thunderstorm!  It will happen

Edited March 31 by darthgently

[Superluminal Gremlin]

Launch maybe. Landing is a hard pass into the dirt for me. The booster is too light with too little wind resistance.  A dangerous combination.

[CBase]

1 hour ago, Superluminal Gremlin said:

The booster is too light with too little wind resistance.

This will improve with scaling of starship. Falcon 9 booster has 152 m² side area and 25t dry mass, SH 639 m² and estimated 200t dry mass.

For commercial planes size makes it easier in bad weather as well.

[Superluminal Gremlin]

Also, planes are powered. That mayyy come in useful

Falcon 9: 6:1 ratio

SH: 3:1 ratio

Certainly a lot better

[AckSed]

I've already seen a video of the Soyuz launching in the middle of heavy snow. Not too far off.

[Superluminal Gremlin]

Thats Soyuz. A relatively imprecise, Parachute landing system. That is the size of a van. Light years from the pole of a Falcon 9 landing to within 10m of a target, retrorepulsivley.

(Arguably the parachute makes it harder, but it all it does is makes go all over the place)

[darthgently]

Still, consider advances in doppler radar and a steady stream of ground station generated 3D realtime data of wind vectors in the landing path volume being fed into landing AI as the booster descends. 

Maybe instead of RTLS on an open pad, land in a large open topped box as tall as the booster that would provide a wind break during the final moments and afterwards.  Drone ship landings during hurricanes will likely remain intractable problems however

[tater]

 

[CatastrophicFailure]

3 hours ago, darthgently said:

Still, consider advances in doppler radar and a steady stream of ground station generated 3D realtime data of wind vectors in the landing path volume being fed into landing AI as the booster descends. 

Maybe instead of RTLS on an open pad, land in a large open topped box as tall as the booster that would provide a wind break during the final moments and afterwards.  Drone ship landings during hurricanes will likely remain intractable problems however

Or just land it in a massive underground silo like a proper Bond villain. 

But speaking of all this, sort of, interesting thread here from a very astute person who IFT-3 Superheavy's return. Counter to Falcon 9 boostbacks, which actually glide a good ways, it basically came straight down. That could certainly explain why it didn't appear to slow down nearly enough, and points to grid fin issues again:

 

[darthgently]

20 minutes ago, CatastrophicFailure said:

Or just land it in a massive underground silo like a proper Bond villain. 

Yes!

 

20 minutes ago, CatastrophicFailure said:

Counter to Falcon 9 boostbacks, which actually glide a good ways, it basically came straight down

Isn't that what I posted after the flight?  I remember thinking that all it did was separate, flip, and cancel horizontal and thought I posted as much.  I'm thinking guessing safety concerns.  What if something went wrong during the boostback and it got a little too close to the coast?  I'm thinking it only had enough fuel to cancel horizontal and do a suicide burn on purpose 

Edited March 31 by darthgently

[Exoscientist]

On 3/29/2024 at 1:46 PM, sevenperforce said:

Appreciated a short synopsis. There's no way I was listening to a whole big post.

With respect to the booster engine landing burn relight: once grid fin control was lost, the booster return was expected to be a failure. If you'll recall, the CRS-16 mission suffered a similar fate when one of its grid fins entered a hydraulic stall and stuck hard-over. There, the single engine tried valiantly to correct for the problem, but it was already off-nominal. This is the same situation. Engine relight depends on a number of factors, including vehicle stability, and so once you are spinning out of control you don't expect engine relight to work.

He has an obvious error at item 4 -- "orbital" velocity. There is no indication whatsoever that Starship failed to reach its intended velocity or trajectory.

He then says that the payload door demo was not successful because the vehicle started to spin once the payload bay vented to space. That doesn't make any sense. It's clear that Starship lost attitude authority (likely due to frozen propulsive vents); whether or not it lost attitude authority doesn't impact the success of the payload door demo. He subsequently says "we have no indication that the test took place" and claims that this is because we did not see any graphical change in the LOX levels, which is also nonsensical; there was no indication that the GUI was supposed to show this. This should be a questionmark, not a failure.

Finally, he claims that because the re-entry time was three minutes different from the estimate from some random person on the internet, this meant it did not re-enter where it was supposed to re-enter. I shouldn't have to explain how silly THAT is.

 

He has an obvious error at item 4 -- "orbital" velocity. There is no indication whatsoever that Starship failed to reach its intended velocity or trajectory.

 That is a debatable point because Elon did say there was an 80% chance of the Starship reaching orbit on this flight:

https://youtu.be/lCe8a7XcG8o

 But later SpaceX said they were planning a flight just under orbital velocity. So in that infographic from TonyBela.com it might be Bela was taking the reaching orbital velocity objective in item #4 from what Elon said.

 But there still remains the question of why didn’t IFT-3 reach orbital velocity? In the case of IFT-2 they vented LOX reducing the velocity it was capable of, and Elon even said if they had a payload and had not vented LOX they would have reached orbit.

 But they didn’t vent LOX on IFT-3 and SpaceX said they had a full propellant load and from the view of the propellant gauges the propellant was virtually expended in both stages. So why were they not able to reach orbit even though carrying 0 payload?

  Bob Clark

[magnemoe]

14 hours ago, Superluminal Gremlin said:

Launch maybe. Landing is a hard pass into the dirt for me. The booster is too light with too little wind resistance.  A dangerous combination.

And its part of the launch contract for SpaceX I believe, you have an very fixed launch window: important geo satellite down and we need an replacement ASAP, or an strategic need for it up now and you pay an premium as we will likely to loose the first stage. 

[Exoscientist]

On 3/14/2024 at 2:07 PM, AckSed said:

All right, let's say Exoscientist's suspicions are justified, and the Raptors, even now, are prone to exploding. Presume they are also a pain to relight and require ullage and settling of the propellant before they'll behave.

Can you make a staged combustion engine more reliable? Peter Beck of Rocket Lab says you can: by building it to withstand and run at extremes, then under-driving it, you end up in the same level of reliability as a gas-generator rocket engine. Neutron's Archimedes is ox-rich, not full-flow, but building to run at max, then under-running at a more comfortable level is a solid path to good reliability.

Can they program Starship and/or Superheavy to under-run the engines at the cost of payload? I say yes.

Will they, or is Elon cowboying ahead with demands for, "More pressure! MORE thrust! More payload! Stuff blowing up! Boom!" and laughing maniacally? I don't have that insight, but they are complying with FAA regulations and NASA requirements, and I do believe Musky-boy knows when not to push his engineers and the physics to make things blow up, now the basic stack mostly works.

On relight: How many flight tests would it take to uncover all the quirks in handling the stack, and implement the necessary hardware and software changes to make relight reliable? We have at least four more planned, because that's how many boosters are being built right now. There are more on the way.

SpaceX has been using Falcon 9 Starlink launches to hone reuse parameters, learn more about the airframe and where its margins are e.g. jettisoning the fairings half a second earlier each flight. It's clear they intend to do the same with Starship/Superheavy. They will have at least four more attempts to relight engines in orbit and on landing, and test heatshields further.

Starships and Superheavies can and have been modified, or scrapped entirely, thanks to their stainless steel construction. The engines are the most expensive part and apparently problematic, but I have outlined a path to making them more reliable. If the engine design is fundamentally flawed and cannot relight at all without a forest of proper ullage thrusters they may have to add them. And that's fine. They will do that.

When I look at SpaceX and what they've achieved, I'm reminded of Parson Gotti, in the webcomic Erfworld: "We try things. Occasionally they even work."

 Thanks for that. Do you have a link where Beck said this? I’ve suggested that SpaceX intentionally ran the Raptors on IFT-2 and IFT-3 at reduced power, i.e., thrust, to improve reliability. 

This has broader implications than just the Raptor question:

Can running a rocket engine at reduced thrust extend lifetimes?

Can someone in rocket propulsion answer if this fact about jet engines also holds for rocket engines?

If so, increasing a turbopump rocket engine power just 10% to 15% cuts engine life in half. And conversely, decreasing it by 10% to 15% doubles engine life. And would this still work if we repeated the concept multiple times? If we reduced the thrust by .9^5 = .60, i.e., to 60%, which most turbopump engines can manage, then we could increase the lifetime by a factor of 2^5 = 32 times? Then a Merlin engine with a lifetime of, say, 30 reuses by running it only 60% power could have its lifetime extended to 1,000 reuses? 

 Is this a known fact about turbopump rocket engines their lifetimes increase radically by a relatively small decrease in their thrust levels?

  Bob Clark

[Ultimate Steve]

 

46 minutes ago, Exoscientist said:

If so, increasing a turbopump rocket engine power just 10% to 15% cuts engine life in half. And conversely, decreasing it by 10% to 15% doubles engine life. And would this still work if we repeated the concept multiple times? If we reduced the thrust by .9^5 = .60, i.e., to 60%, which most turbopump engines can manage, then we could increase the lifetime by a factor of 2^5 = 32 times? Then a Merlin engine with a lifetime of, say, 30 reuses by running it only 60% power could have its lifetime extended to 1,000 reuses? 

Material lifetime curves are unpredictable and vary based on lots of things (material, temperature, stress, amplitude, cycle count) and are usually empirically tested for rather than calculated. I don't think you can extrapolate rules of thumb like that.

 

I'm not the most knowledgeable about this (Haven't even finished Aero undergrad, I guess you could say I technically specialize in command and data handling and spacecraft communication and not propulsion or structures), but I don't think rules of thumb can be applied like that to something so complex.

There's going to be three types of wear that I can think of. Creep, fatigue, and wear.

Creep, deformation under constant load, is like if you took a blob of silly putty and hung it from something and watched it stretch. Any well designed propulsion system designed for long term reuse, jet, rocket, or otherwise, will (if I am being sensible and know what I'm talking about) not operate in the creep regime. You can operate above the creep limit for a short period of time and have things not break. Non reusable rocket engines such as the RL-10 will do this to enhance performance, as they only need to run for a few minutes. In this case, reducing power can drastically prolong engine life, taking it from a few minutes to practically forever (or until something else fails first). I can't imagine that Merlin has this problem. From what I remember, for a given temperature there's a sharp cutoff where creep doesn't occur. Increasing stress beyond this point will quickly decrease time until failure.

The engines that do this, I'm pretty sure only do this with the nozzle/maybe combustion chamber and not the turbopump, as you can't have a precision component like that changing size throughout the burn without causing major problems.

I could foresee a future in which "emergency power" is available to reusable rockets, where an engine can slightly increase in power for a few minutes to compensate for other losses, at the expense of the rest of that engine's operational lifetime.

Then there's fatigue, weakening due to repeated cycles of stress, like if you took a paper clip and bended it back and forth repeatedly until it broke. Some metals are susceptible to this at all stresses, some metals have an endurance limit where you can do an infinite number of cycles as long as you remain below a certain amplitude. The alloys in jet engine and rocket engine turbines are crazy stuff and the graphs for them probably aren't available online, but the factors in the equation are stress amplitude and number of cycles. Rocket engines haven't yet gone through a lot of macro cycles (startups and shutdowns), and many graphs aren't even available for numbers below 1000, so macro cycles probably won't be an issue.

Micro cycles, like a slightly unbalanced turbine spinning, or something vibrating, could be an issue if the stress is high enough or they are using a material without an endurance limit, as those things go through a lot of cycles. Fatigue failure of rapidly spinning jet engine components has been a major cause of aircraft crashes, but reducing power would indeed dramatically prolong the life of the engine... However, this timeline is generally measured in years, today's reusable rockets have total runtimes measured in hours.

And then there's wear, either by friction or slow burning chemical reactions with the hot propellants, or through other means. I'm not well versed here, and I'm not sure if any rocket engine has run long enough to properly characterize these phenomena in these environments.

The above phenomena generally manifests over very long engine runtimes (think months), reusable rocket engines currently have runtimes measured in single digit hours. It isn't impossible for it to run fine for 10 minutes and then fail on minute 11 due to fatigue or creep, but I would be very surprised if that is what's happening. It would require either devious new failure modes almost nobody has run into before (e.g. flammable titanium, solid lox) or unbelievable stupidity for the leading experts in reusable rocketry to aim for something that runs for days with thousands of cycles to fail in 5 minutes 2 or 3 cycles in (excluding those tests where they push it further to see how high they can go and where it fails, like those tests where they push Raptor to ridiculous chamber pressures).

I don't think Raptor has a massive widespread reliability problem right now, we haven't seen many failures that weren't related to the fuel feed system. But if there is a widespread Raptor reliability problem, it almost surely isn't directly the fault of the design being too close to fatigue or creep limits.

Merlin did indeed have fatigue issues early on, a few came back with (presumably fatigue) cracks at the beginning of the reusability era, but I haven't heard anything about them in years (likely pre block 5), and to my knowledge, they have long since been fixed and didn't ever cause a failure (although IDK if we ever found out the exact reason behind the CRS-1 Merlin failure).

[RCgothic]

Ablation and corrosion are types of wear - material thickness or effective thickness reducing by physical or chemical action.

I've also always found it weird that "no endurance limit" is not actually a good thing.

[AckSed]

4 hours ago, Exoscientist said:

Do you have a link where Beck said this?

Here's my condensed transcript: https://forum.kerbalspaceprogram.com/topic/156515-rocketlab-discussion-thread/?do=findComment&comment=4339716

Quote

Archimedes development going well, but it remains the long pole in the tent.
Chose oxidiser-rich staged combustion because if you dial it back a bit from squeezing out every last bit of performance, you end up in really benign operation at the same level of performance as a gas-generator cycle, but "kind of bulletproof". Compared it to an airplane engine.
Ox-rich combustor couldn't be dialled back too much or it would extinguish; had to solve that in the process of building the most boring, unboring engine.

However, Archimedes is oxygen-rich staged combustion, which I encourage you to look over Everyday Astronaut's article on engine cycles to explain. Scroll down to "Full Flow Challenges". Very short summary: comparing the enthalpy of a hypothetical rocket engine with the same mass flow rate but different staged combustion cycles, oxygen-rich staged combustion has a moderate rise in temperature in its preburners, while full flow staged combustion has the lowest rise in temperature for the preburners while extracting the most amount of work, not only stressing the preburners less.

So my conclusion may have been fallacious. Sorry.

Funny story. Something I missed in the transcript is that Rocket Lab is also considering full-scale testing for Neutron's first flight, citing experience with recovering Electron. At 1hr 09 min:

Quote

Host: Are there any plans for like sort of hop tests with Neutron or is your first launch with Neutron just going to be sort of an all up, "We're sending it"?

Beck: Yeah no we'll just- we'll just bring it home. We'll try to- I mean look, if if we hadn't had the experience with electron, then you'd go, "Yeah no." It's- there's there's a lot of- a lot we need to learn there, but this is kind of- you know I think I've said it before it's if we hadn't gone through the process of of trying to reuse Electron, like trying to make a reusable launch vehicle with Neutron would just be an incredibly daunting task and you'd have so many unknowns um in Known Unknown and unknown knowns that it would be- you know it would be be really difficult and you'd have to kind of you know chunk the problem up and to start breaking it off into into various kind of different things.

You know it's it's not it's not the first time we've we've we've ridden the wave back through the Earth's atmosphere so you know- the learnings you know if if nothing like if nothing other than um that those learnings comes from electron's reusable program it will be 1,000% totally worth it: from materials, from control algorithms, to understanding how vehicles actually interact with the atmosphere, to Thermal Protection Systems um we've learned it all so- um well I'm sure there's more to learn but, you know we've learned enough that we'll just- we'll just have a crack at bringing it straight in.

Even cautious businessmen like Peter Beck are taking the 'just have a crack at bringing it straight in' approach, thanks to their experience with reentry on their previous, smaller booster.

Rocket Lab has 6 recoveries under their belt.

SpaceX has 290 successful landings and counting.

Your honour, members of the kangaroo science court, learned Kerbalists: this is strong evidence for full-scale prototyping and blown-up boosters being an attractive and, dare I say it, workable development process.

[tater]

SpaceX is targeting Monday, April 1 for a Falcon 9 launch of 22 Starlink satellites to low-Earth orbit from Space Launch Complex 4 East (SLC-4E) at Vandenberg Space Force Base in California. Liftoff is targeted for 7:30 p.m. PT, with backup opportunities available until 11:30 p.m. PT. If needed, additional opportunities are also available on Tuesday, April 2 starting at 7:30 p.m. PT.

[sevenperforce]

On 3/30/2024 at 4:35 PM, tater said:

Instead of trying to launch 3 rockets within <4.5 hours, they should do the prudent industry thing and spread these 3 launches over at least one Quarter. This just isn't "how it's done."

Departing from industry standard. Tsk tsk. And they expect to get government contracts?!