The future of booster landings

[nicp]

Apologies if covered elsewhere, couldn't find anything.
Obviously SpaceX have demonstrated landing boosters on land and at sea, quite reliably.
But they recently (28th August 2024) lost a booster and Falcon 9 was briefly grounded (for two days) by the FAA.

Does that mean that every other U.S company that tries and fails to land a booster will be similarly grounded? Or could they get away - initially at least - with calling flights 'test flights' with landing as a secondary objective not requiring FAA intervention in the case of a landing failure?

And that question prompts another - how successful do people expect other countries or companies initial efforts to be?
SpaceX took a number of attempts at sea to get landing to work and I wonder if other companies or countries may be over optimistic having seen SpaceX's long string of landing successes.

Of course, some (BO) are not I believe going for a 'hoverslam' (which might make the problem easier), they will hover at least briefly. But that will impact the overall performance of the booster because all that lovely propellant is being thrown away just to hover.

Will other entities skip landing legs and go for a catch system like Starship?

Best Regards,
Nic
EDIT: Added no hoverslam makes it easier.

[kenny008]

I think some people are making too big a deal about the FAA grounding the booster due to a landing failure.  They needed to verify it wasn't an issue that ALSO affected normal booster flight.  For example, if the landing failure was due to FOD in the tankage (it wasn't), then it wouldn't have been a pure "landing" failure.  It would have been a process failure that COULD have affected initial flight and public safety.

In this case, they determined that it was in fact only a landing failure, and they cleared the booster for flight in only ~2 days.  This is exactly how I would expect a regulator to operate.  In fact, I think they worked the issue very efficiently. 

So yes, IMO, I think we should expect "landing" failures to trigger the same response, and I also expect the FAA (with the right data) to quickly make the same determination on future flights.

[AmigaClone]

I think some people are making too big a deal about the FAA grounding the booster due to a landing failure.  They needed to verify it wasn't an issue that ALSO affected normal booster flight.  For example, if the landing failure was due to FOD in the tankage (it wasn't), then it wouldn't have been a pure "landing" failure.  It would have been a process failure that COULD have affected initial flight and public safety.

In this case, they determined that it was in fact only a landing failure, and they cleared the booster for flight in only ~2 days.  This is exactly how I would expect a regulator to operate.  In fact, I think they worked the issue very efficiently. 

So yes, IMO, I think we should expect "landing" failures to trigger the same response, and I also expect the FAA (with the right data) to quickly make the same determination on future flights.

I suspect that the FAA might not be as quick to allow an orbital launch vehicle to launch again after being grounded due to a landing failure since at least at this point no other launch vehicle has the same number of successful consecutive landings. I personally would expect the timing to be closer to the time between the second and third test flights of Starship.

[sdsds]

In fairness the FAA can't take a long flight history as a predictor of future flight success, because processes in preparing vehicles for flight could have changed. (A supplier of goods or services used to prepare a vehicle for flight could have changed, or some other quality control issue might have crept into the process.) So launch providers with a long history of flight success need to point to a root cause of a failure and a remedy, and so do launch providers with no history of flight success.

On the more general question:
- now that we know a burn during re-entry protects a vehicle from re-entry stresses, and
- since everyone has access to the compute power to do a hover-slam, and
- since catch presents greater risk to launch infrastructure, and
- since only Starship has excess launch towers to burn,

I predict we'll see many launch providers doing leg-based hover-slam recoveries of their boosters.

Many, as in maybe 4 or 5 in the next decade.

[daedalus1]

This should have been only the concern of SpaceX, nothing is relevant for the FAA to be involved.

[litton4]

This should have been only the concern of SpaceX, nothing is relevant for the FAA to be involved.

Not if it was an issue that could have affected the uphill journey, especially with crewed launches coming up.

One thing to note is that the landings have permitted SpaceX to inspect boosters post-flight and identify impending issues that could (if they had been worse) affect a launch, and take preventative measures.
Previously this was only available for Shuttle (which had too many, sadly).

I'm sure this has contributed to their outstanding success record for both F9 S1 and S2.

[daedalus1]

This should have been only the concern of SpaceX, nothing is relevant for the FAA to be involved.

Not if it was an issue that could have affected the uphill journey, especially with crewed launches coming up.

One thing to note is that the landings have permitted SpaceX to inspect boosters post-flight and identify impending issues that could (if they had been worse) affect a launch, and take preventative measures.
Previously this was only available for Shuttle (which had too many, sadly).

I'm sure this has contributed to their outstanding success record for both F9 S1 and S2.

All other launchers collect zero data after main engine cut off. There return is over an area of hundred of square mile and a speed of hundreds of miles per hour.
Falcon 9 first stage worked perfectly to a landing within a couple of feet and a velocity only slightly above zero. A bit too much for the landing legs.
You are telling me that the FAA should be concerned about Falcon 9 but the others are OK?

[laszlo]

All other launchers collect zero data after main engine cut off. There return is over an area of hundred of square mile and a speed of hundreds of miles per hour.
Falcon 9 first stage worked perfectly to a landing within a couple of feet and a velocity only slightly above zero. A bit too much for the landing legs.
You are telling me that the FAA should be concerned about Falcon 9 but the others are OK?

The operations and warning zones for expendables are designed to accommodate the deliberate uncontrolled destruction of the boosters. Nor is there any possibility of them being used again so they will never affect another flight. For them the mission is over at MECO. For SpaceX the mission doesn't end until the landing is complete. With greater performance comes greater responsibility.

[litton4]

This should have been only the concern of SpaceX, nothing is relevant for the FAA to be involved.

Not if it was an issue that could have affected the uphill journey, especially with crewed launches coming up.

One thing to note is that the landings have permitted SpaceX to inspect boosters post-flight and identify impending issues that could (if they had been worse) affect a launch, and take preventative measures.
Previously this was only available for Shuttle (which had too many, sadly).

I'm sure this has contributed to their outstanding success record for both F9 S1 and S2.

All other launchers collect zero data after main engine cut off. There return is over an area of hundred of square mile and a speed of hundreds of miles per hour.
Falcon 9 first stage worked perfectly to a landing within a couple of feet and a velocity only slightly above zero. A bit too much for the landing legs.
You are telling me that the FAA should be concerned about Falcon 9 but the others are OK?

No, just that there is vastly more data available for F9, which allows the FAA and SpaceX to look more deeply when things don't go to plan.
The others are clearly not OK, at a 100% level, otherwise there wouldn't be any failures.

SpaceX has had a massively successful run, we don't know what changes they made along the way, based on the data collected or exactly how those changes contributed to that success, by nipping failures in the bud, or how they helped improve the performance (ie mass to orbit).

Other organisations will no doubt have telemetry (and other) data that they can use to infer when things aren't quite right, but there is no substitute for getting your hands on the thing after use, to inspect the hell out of it.
For them, often there is a failure and they see the signature in the data, so go back over past data and see other occurrences (maybe less severe) of similar data, and that can help spot past near-misses and lead to improvement.

The fact that SpaceX have so much data has meant the recent stand-downs have been extremely short.
One was actually caused by a change and we don't yet know the reason for the other.
Their other advantage is that they have an internal customer, for whom it's not a disaster if a mission goes astray, since even 20 satellites is a small percentage of the total payload required.
So, they can test changes out before trying them out on other Customers (big 'C').
They can also incrementally make small changes due to the high flight rate, rather than having to batch them up, so that when something goes wrong,  the answer to the question "What changed?" is easily answered.

[daedalus1]

I don't agree that the FAA should have grounded the fleet, unless SpaceX found something that could effect the launch.
The data could be given to the FAA if requested. Grounding however short was way over the top, especially in relation to other rockets.

[Jim]

All other launchers collect zero data after main engine cut off. There return is over an area of hundred of square mile and a speed of hundreds of miles per hour.

that is wrong. The upperstages are monitored until all systems are depleted.  booster stages, strap on motors and fairing are monitored to make sure they remain in the impact zone.  All active vehicles are monitored to make sure they follow the planned trajectory and remain intact.  Because they are active, the restrictions are tighter

[Jim]

This should have been only the concern of SpaceX, nothing is relevant for the FAA to be involved.

wrong

[kenny008]

I don't agree that the FAA should have grounded the fleet, unless SpaceX found something that could effect the launch.
The data could be given to the FAA if requested. Grounding however short was way over the top, especially in relation to other rockets.

Are you saying that we should assume a landing issue is not a safety issue unless / until proven otherwise?

Are you saying that no problems with booster landing (anything after staging) should have resulted in a grounding?  Even if the problem could have also affected the main boost flight and could have affected public safety (such as an engine failure)? 

Again, if the problem was specifically limited to landing equipment (failed leg lock), then it should not result in fleet grounding.  But you can't know that until the data is released by SpaceX to the FAA.  Which is what happened here.  It was a VERY fast turn-around (about 2 days), and I don't think this is an unreasonable expectation.

[butters]

Regardless of the public safety rationale for grounding launch systems after stage recovery failures, it's clear that there was no regulatory grounding of F9 during the developmental period of repeated booster landing failures. It was only after booster recovery became routinely successful that the FAA added the license provision about unplanned loss of vehicle. And SpaceX is not enjoying the same developmental grace period, as it were, for the Starship system. It's reflective of a change in policy, for better or worse.

But, did the FAA ground Electron after their helicopter booster recovery failures? They investigated the upper stage ascent failure, but no record of regulatory action in response to failing to catch the booster, unless I missed it.

I understand the rationale arguments on both sides, but it's difficult to rationalize the inconsistency in how the policy is being applied, raising genuine questions about how New Glenn, Neutron, etc. will be regulated by the FAA.

[laszlo]

...But, did the FAA ground Electron after their helicopter booster recovery failures?

I thought that was in New Zealand and adjacent international waters. Does the FAA have jurisdiction there?

[butters]

...But, did the FAA ground Electron after their helicopter booster recovery failures?

I thought that was in New Zealand and adjacent international waters. Does the FAA have jurisdiction there?

Yes, Rocket Lab is a US-headquartered company. Which is why the FAA investigated the upper stage startup anomaly.

[deltaV]

I'd like to see one or both of the following changes in the FAA's mishap regulations (https://www.faa.gov/space/compliance_enforcement_mishap) to limit groundings to actual safety issues only.

First, "Unplanned substantial damage to property associated with the activity", "Unplanned permanent loss of the vehicle", and "Failure to complete a launch or reentry as planned" should be removed from the list of types of mishaps. If there's an actual safety issue it will be covered by the other types of mishap. With this change neither the Falcon upper stage failure or the Falcon landing failure would likely have been mishaps and hence wouldn't result in groundings.

Second, mishaps should only ground activities that the mishap potentially casts doubt on the safety of. For example a Falcon flight termination system malfunction would ground all Falcon flights, a crewed Dragon that's destroyed during launch would only ground crewed Dragon, not cargo Dragon or Falcon, and the recent Falcon upper stage failure and Falcon booster recovery failure would likely ground nothing. This is admittedly a bit vaguer than a regulation would need to be.

[edzieba]

But, did the FAA ground Electron after their helicopter booster recovery failures?

Electron does not ascend under parachute. No flight system on the booster failed (or was even a factor) in the helicopter releasing the snagged stage.


Falcon 9 lands using flight hardware: engines, avionics, tanks. If a failure in one of those caused the anomaly, then it needs to be ruled out that the same failure could no occur to the same systems in other phases of flight. That's exactly what occurred, and is why RTF occurred within days. It's the same reason why a failure during a static fire can ground a vehicle.

[kenny008]

Second, mishaps should only ground activities that the mishap potentially casts doubt on the safety of. For example a Falcon flight termination system malfunction would ground all Falcon flights, a crewed Dragon that's destroyed during launch would only ground crewed Dragon, not cargo Dragon or Falcon, and the recent Falcon upper stage failure and Falcon booster recovery failure would likely ground nothing. This is admittedly a bit vaguer than a regulation would need to be.

Since the booster landing uses launch and first-stage flight hardware for landing, this is exactly what happened.  The landing problem could have been an engine issue, which would possibly affect launch / public safety.  SpaceX gave the FAA the data, the FAA reviewed that data, and everyone agreed that it did not affect public safety.  They therefore allowed return to flight.  All within 2 days.

[Barley]

Using anomalies to determine the need for a formal third-party safety review is a blunt instrument.

If the rocket is designed to handle, say, 3-sigma waves then losing 1 out of 300 is to be expected.  Whether you even address that is an economic rather than a safety issue.  (Assuming ordinary precautions such as clearing the landing area)

On the other hand, a safe landing on the corner of the barge is perhaps a 6-sigma event and much more likely to indicate something has changed and needs investigation.

It's difficult for an outside regulator to deal with this cleanly.  They shouldn't be too involved in all the details all the time, so they probably want to wait until either the operator flags something to them or something dramatic happens, and then demand an explanation.

It seems to me the FAA got this one about right.

Edit: fixed off by one error.