SpaceX Starship : First Flight : Starbase, TX : 20 April 2023 - DISCUSSION

[ChrML]

So that’s exact what I expect for 28 engines: difficulty at first but long-term no problem. Is this unreasonable?

I think it's reasonable. Every problem that occurs has some fix, and after a while all common problems will be eliminated and the vehicle is reliable. Raptor is a lot more complex than Merlin, but not impossible.

If it's a software problem, they add more theories to the software test setup to cover the problem, and that same problem will never happen again in a future iteration.

If it's a hardware problem I trust they have some other way of addressing it while making sure it doesn't happen again.

[OTV Booster]

I wonder what the startup sequence would be like for a 28 engine cluster?

A fascinating engineering problem!

Perhaps something like starting opposing pairs together with additional pairs coming online separated by milliseconds. Probably the critical gimbal engines first then the outer ring.

Anyone familiar with how the Falcon Heavy lights?

Sounds like you just outlined a smart pre-flight test program. Maybe a series of one at a time static fires followed by lighting off combos and building to a full up static fire.

[OTV Booster]

The N-1's biggest issues were effects of vibration and heating on the closely clustered engines.  Certainly, computer simulation has made great strides in predicting these interactions, but I will point out that the interaction of 28 such powerful engines, their exhaust impingements, etc., gets closer and closer to chaotic.  And chaos theory is not a friendly taskmaster.

I guess I'd be more likely to believe that the first SH stages will fly just fine had there been any 28-engine cluster test firings on the ground.  Heck, the F-1 engines on the Saturn V were designed to handle the radiant heat and vibration from their neighboring engines, but you'll notice that a kludge -- thick thermal batting -- was plastered onto the outside of those engine bells after initial clustered test firings indicated that the extra protection and thermal coating was required to bring the engines within desired safety parameters. Again, yes, there has been more than 50 years of advancement in computer simulations.  But the simulations are never any better than one's assumptions, and it was the sets of assumptions, and not failure of simulations, that caused the N-1 problems, and required a kludged-on layer of protection to be added onto the F-1s.

Besides, with the rather extreme replacement rate these Raptors seem to undergo during checkout and after static fires, once attached to the prototype Starships, it seems that the Raptors are also still in fairly early prototype stages themselves.  So, the performance assumptions for the current version of the Raptor may be even more poorly defined than you might think.

Personally, I'd rather see SH stages sit on that orbital test mount and fire up their clusters for full-duration tests before bothering to mount a Starship on top.  If for no other reason than to iron out all of the poorest assumptions before risking the loss of the R-Vacs that will likely be installed on any SS that gets put on top of an SH.  Of course, that would take several months and endanger his build/launch/test site unnecessarily, and it seems Musk is in enough of a hurry that he doesn't want to go through that whole process.  Still... even SpaceX might get surprised at the backlash if an SH blows up one or two km into the air.  Because that would be, shall we say... a significant RUD. 

The marginal risk-reduction value of static fire time decreases significantly as the test gets longer. They would probably get 80% of the set of data from a 5-second test that they would from a 2 minute test. On the other hand, the work to build a test stand and flame trench that can take a 2-minute firing is massively larger than building a pad for a 5 second test.

Also, this booster probably isn't being reflown, and the next one will be built in a few months. At some point, it's faster and simpler to fly it than to test it longer on the ground. That way they get flight data too.

OTOH, a rocket engine hits steady state in 2-3 seconds. From there on it's diminishing returns. Individual static tests on all the engines can all be done in one 60-80 second test. Purpose: manifold characterization. Are there anomalies that match up with specific manifold positions? Next up, get a sense of the interactions with different startup combinations and increased manifold flow rates. Validate the models. CFD is hard.


At the same time build a methodology for fast testing iteration. Try to get 2-3 test sequences per day. Assuming no show stoppers testing might only take 2-3 weeks. Then replace some concrete and go for orbit.

[OTV Booster]

Do we know that the first Superheavy booster flight will feature all 28 engines?

No, we do not know.

Depending if they're trying to actually land this thing on Phobos/Deimos, that's a lot of Raptor engines to dump in the ocean

Opinion follows.


True that but...  There's a question on how fast and how profound the raptors change. SX is happy to scrap partially built ships if they see no reason to test them. I think we've only seen one used raptor cycled back into use. Even if they save the raptors after an orbital flight would we expect to see them reused?


Same for the BN. It'll be a throw away for a while. It'll face mods. Maybe major mods. No biggie. As SX approaches operational with a more mature and less changing design (engine and airframe), we should see more emphasis on reusability. They may even do recovery and still not reuse.

[rsdavis9]

Same for the BN. It'll be a throw away for a while. It'll face mods. Maybe major mods. No biggie. As SX approaches operational with a more mature and less changing design (engine and airframe), we should see more emphasis on reusability. They may even do recovery and still not reuse.

After all you don't actually need to reuse a booster to know if it is reusable.
Recover engines.
Inspect.
No need to use it again at first.

[Alberto-Girardi]

All 6 Raptors will ignite after staging.

Opinion or fact?     But I like the optimism and am pulling for you being right.

Simulations show that three sea level Raptors on the second stage are sufficient to perform the flight. Whether SpaceX will add the three vacuum Raptors, we do not know.

Thanks! I understood that all 6 raptors would be needed after separation, but this is wrong as you say. Are only 3 raptors needed only for this particular flight with no payload, with the operational launches needing 6 engines?

[Lars-J]

All 6 Raptors will ignite after staging.

Opinion or fact?     But I like the optimism and am pulling for you being right.

Simulations show that three sea level Raptors on the second stage are sufficient to perform the flight. Whether SpaceX will add the three vacuum Raptors, we do not know.

Thanks! I understood that all 6 raptors would be needed after separation, but this is wrong as you say. Are only 3 raptors needed only for this particular flight with no payload, with the operational launches needing 6 engines?

Operational launchers will certainly require all 6 engines. It is theoretically possible that a stripped down SN20 could make it to orbit with 3 engines only, but I personally consider this unlikely - for two reasons:
1) SN20 (or whatever will be the first orbital Starship) is unlikely to be as light as some people imagine
2) RVac engines appear to be deep into the testing phase, and why wouldn't they want to fire them on a real flight? They will want to retire as MUCH risk as possible on a first orbital attempt, especially if it is an expendable launch

[punder]

I wonder what the startup sequence would be like for a 28 engine cluster?

A fascinating engineering problem!

Perhaps something like starting opposing pairs together with additional pairs coming online separated by milliseconds. Probably the critical gimbal engines first then the outer ring.

Anyone familiar with how the Falcon Heavy lights?

Sounds like you just outlined a smart pre-flight test program. Maybe a series of one at a time static fires followed by lighting off combos and building to a full up static fire.

Agree, but a lot of the “combo” testing has taken place already, with SN8 and subsequent flights, with lessons learned and subsequent model refinement each time.

And more generically, SpaceX must be far ahead of the industry baseline for this sort of modeling, having flown so many 9- and 27-engine clusters over the years. They didn’t have that knowledge base when they worked the bugs out of F9–they created it by doing so. So they have knowledge that no one else has (Rocket Lab included—similar issues but much lower energies). NASA most likely has studied the data for validation, but can’t apply it to anything of their own or pass it on to other industry players.

Still... as Elon says, “Excitement guaranteed!” 

[Jansen]

STA granted for Starship orbital test flight

Period from 20 June 2021 to 20 Dec 2021

https://apps.fcc.gov/els/GetAtt.html?id=275103

[edzieba]

Crossposting:

More FCC documents here: https://fcc.report/ELS/Space-Exploration-Technologies-Corp-SpaceX/0748-EX-ST-2021

Some notable bits from the email chains:
- "Maximum flight altitude:380,160' AGL (0-72 miles)"
- 2GHz low bandwidth transceiver at base of Super Heavy, all other comms on Super Heavy and Starship are 2.4GHz (i.e. no Starlink)
- "This STA will expire as soon as launch has been completed or 10 December 2021, whichever occurs first."
- "One (1) or more of six (6) blackout zones (BOZs) MAY be imposed as follows: (1) 1500 nautical mile radius centered at 22N160W; (2) 1500 nautical mile radius centered at 33.25N119.57W; (3) 1500 nautical mile radius centered at 4.11N175.2W; (4) 1500 nautical mile radius centered at 57.46N152.38W; (5) 1500 nautical mile radius centered at 32.37N106.47W. (6) 1500 nautical mile radius centered at 57.34N7.35W. The final launch schedule for this SpaceX mission will ultimately determine which, if any BOZ will be implemented."

I've plotted those points on a map: https://www.google.com/maps/d/u/0/edit?mid=1AHv8RFOWWBvZB3JAtPSeeECCeLb1LuAF&usp=sharing

Point 1 is the expected Starship splashdown zone off of Kauai.
Point 3 is east of the Marshall Islands, and likely a contingency site that still has some EDL coverage from the RRBMDTS (Kwajalein range).

From there, things get interesting:
Point 2 is San Nicolas Island, off the coast of California, which hosts a USN base is and is part of the Pacific Missile Range.
Point 4 is the Pacific Spaceport Complex on Kodiak Island.
Point 5 White Sands missile range.
Point 6 is the RAF Deep Sea Range, South Uist lsland, off the north-west coast of Scotland.

[Giovanni DS]

Is it safe to leave those raptors on the sea floor? I imagine those would be worth fishing out for some, not necessarily friendly, entity.

[woods170]

Is it safe to leave those raptors on the sea floor? I imagine those would be worth fishing out for some, not necessarily friendly, entity.

It is safe to leave them on the sea bed.
Same goes for the RS-25, the F-1 and even the RD-180. Apart from Jeff Bezos nobody has ever bothered to pick them up from the ocean floor. Not even the Chinese.

Also, why do people always assume that wreckage from engines is enough to reverse-engineer them?
I can tell you that it is not nearly enough to make a working clone.

[edzieba]

Is it safe to leave those raptors on the sea floor? I imagine those would be worth fishing out for some, not necessarily friendly, entity.

It is safe to leave them on the sea bed.
Same goes for the RS-25, the F-1 and even the RD-180. Apart from Jeff Bezos nobody has ever bothered to pick them up from the ocean floor. Not even the Chinese.

Minor point, but many nations (including the US!) have indeed recovered spent ICBM stages to examine, and not because they want to replicate them (mainly to confirm models of adversary capability match actual adversary capability). Cost/benefit rarely works out, as 'benefit' is confirmation of existing estimates, and 'cost' is the potential of being caught in someone's territorial waters pilfering hardware, as well as high financial cost of deep-ocean search & recovery.

[Giovanni DS]

Also, why do people always assume that wreckage from engines is enough to reverse-engineer them?
I can tell you that it is not nearly enough to make a working clone.

Thanks, I understand your point about cloning, clearly cloning it is more about the production process than the engine itself.

I would look more for the detailed internal architecture, metals, etc. Not a rocket scientist but as a SW developer, having access to some code base for sure can speed up things even if you cannot use it as-is.

Raptor cycle is unlike any other engine as far I understood, perhaps there could be expensive lessons to be learned even from damaged engines.

[mark_m]

My searching has failed me, but back during the discussion of "orbital" vs. "3/4 of the way around the Earth" someone made a statement that anything that goes more than halfway around the globe has to be "orbital". (I'm assuming this claim is only talking about a ballistic segment of the flight.)

IANARS, but naïvely this makes some intuitive sense to me—any trajectory that would intersect the Earth/planet/whatever would do so less than halfway around, right? But I didn't see any followup discussion, and I was wondering if that is really the case. Thanks for any elucidation!

[rsdavis9]

My searching has failed me, but back during the discussion of "orbital" vs. "3/4 of the way around the Earth" someone made a statement that anything that goes more than halfway around the globe has to be "orbital". (I'm assuming this claim is only talking about a ballistic segment of the flight.)

IANARS, but naïvely this makes some intuitive sense to me—any trajectory that would intersect the Earth/planet/whatever would do so less than halfway around, right? But I didn't see any followup discussion, and I was wondering if that is really the case. Thanks for any elucidation!

Well it all depends on where the engines cutoff.
Currently the SECO second stage engine cutoff occurs over africa?
From that point you have to draw an ellipse around the earth.
The cutoff point can be the apogee or the perigee.

1. If the perigee then you have achieved orbit. Because it will come back around to the same point after achieving apogee 180 degrees around earth. (assumes cutoff is above the atmosphere)

2. If the apogee then it will have a perigee 180 degrees around earth and probably reenter because it will be further into the atmosphere than the cutoff point.

Of course a second engine start to change the orbit is possible.

So probably what they are doing is 2. Without another engine start to raise the orbit reentry will occur less than 180 degrees around the earth from the engine cutoff. Is the marshall islands less than 180 degrees? Remember to come down in hawaii the reentry starts further westward.

[daedalus1]

My searching has failed me, but back during the discussion of "orbital" vs. "3/4 of the way around the Earth" someone made a statement that anything that goes more than halfway around the globe has to be "orbital". (I'm assuming this claim is only talking about a ballistic segment of the flight.)

IANARS, but naïvely this makes some intuitive sense to me—any trajectory that would intersect the Earth/planet/whatever would do so less than halfway around, right? But I didn't see any followup discussion, and I was wondering if that is really the case. Thanks for any elucidation!

If it reenters naturally after 3/4 of circumference, then it's suborbital. If it has to use engines to return (slow down), then it's orbital.

[mark_m]

My searching has failed me, but back during the discussion of "orbital" vs. "3/4 of the way around the Earth" someone made a statement that anything that goes more than halfway around the globe has to be "orbital". (I'm assuming this claim is only talking about a ballistic segment of the flight.)

IANARS, but naïvely this makes some intuitive sense to me—any trajectory that would intersect the Earth/planet/whatever would do so less than halfway around, right? But I didn't see any followup discussion, and I was wondering if that is really the case. Thanks for any elucidation!

If it reenters naturally after 3/4 of circumference, then it's suborbital. If it has to use engines to return (slow down), then it's orbital.

Let me restate my question. My understanding and paraphrasing of the claim is: If a body in free fall travels more than halfway around the globe without impacting, it will always make it all the way around (ignoring drag, etc.). My question is, is that a true statement?

[eriblo]

My searching has failed me, but back during the discussion of "orbital" vs. "3/4 of the way around the Earth" someone made a statement that anything that goes more than halfway around the globe has to be "orbital". (I'm assuming this claim is only talking about a ballistic segment of the flight.)

IANARS, but naïvely this makes some intuitive sense to me—any trajectory that would intersect the Earth/planet/whatever would do so less than halfway around, right? But I didn't see any followup discussion, and I was wondering if that is really the case. Thanks for any elucidation!

Well it all depends on where the engines cutoff.
Currently the SECO second stage engine cutoff occurs over africa?
From that point you have to draw an ellipse around the earth.
The cutoff point can be the apogee or the perigee.

1. If the perigee then you have achieved orbit. Because it will come back around to the same point after achieving apogee 180 degrees around earth. (assumes cutoff is above the atmosphere)

2. If the apogee then it will have a perigee 180 degrees around earth and probably reenter because it will be further into the atmosphere than the cutoff point.

Of course a second engine start to change the orbit is possible.

So probably what they are doing is 2. Without another engine start to raise the orbit reentry will occur less than 180 degrees around the earth from the engine cutoff. Is the marshall islands less than 180 degrees? Remember to come down in hawaii the reentry starts further westward.

A spacecraft can be injected at any point in its orbit, not just apogee or perigee. If you inject right "after" perigee in an elliptical orbit you will complete a large fraction of an orbit before reentering.

[daedalus1]

My searching has failed me, but back during the discussion of "orbital" vs. "3/4 of the way around the Earth" someone made a statement that anything that goes more than halfway around the globe has to be "orbital". (I'm assuming this claim is only talking about a ballistic segment of the flight.)

IANARS, but naïvely this makes some intuitive sense to me—any trajectory that would intersect the Earth/planet/whatever would do so less than halfway around, right? But I didn't see any followup discussion, and I was wondering if that is really the case. Thanks for any elucidation!

If it reenters naturally after 3/4 of circumference, then it's suborbital. If it has to use engines to return (slow down), then it's orbital.

Let me restate my question. My understanding and paraphrasing of the claim is: If a body in free fall travels more than halfway around the globe without impacting, it will always make it all the way around (ignoring drag, etc.). My question is, is that a true statement?

Someone might correct me on this, but my understanding is that in that circumstance the globe is centre of one of the two elliptical points. So it will not impact the surface.