This past week, SpaceX had another SN24 engie test.
But just a startup and shutdown.

Considering that during the "iterative" testing early on, one of the
problems was the restart of the engine(s) for landing, shouldn't engine
tests at this point in time test the ability to start and restart
engines to emulate a flight ?

In a previous inteview, the PR person had explained that the engine
start is the most challenging part with a serious challenge to spin up
the turbo pumps at the right time/sync etc. (so that would explain why
previous tests had been for short burns, since they were only testing
engine start, not actual engine performance).

But as engine start remains a challenge, shouldn't restarts be tested?

I know it is totally diferrent, but when thgey built the Ottawa LRT, I
could see they knew of problems in advance because they were avoiding
testing stuff. Big snow storm? testing cancelled (when it should have
been done to test behaviour). Didn't test doors while trains were
traveling in empty/closed stations. By NOT finding problems during
testing, they were able to get paid by the city for meeting various
steps in the building of the system, so there was incentive to not find
problems (and later hide them under a carpet).

So it is in this context where I note how it seems to me that SpaceX
isnt REALLY testing its ships to the full extend that is possible on
ground to really debug. Looks to me like their goal is just to launch
and no care about landing, but that would mean that once they have done
their first orbital flight, they will have to restart testing to get the
landing to work and suspect that will take quite a biut of time before
they have the trust needed to let a landing happen at the tower.

And without the ability to target an "X" on a landing site (since
neither booster now Starship have landing legs), how can they test the
ship's ability to land precisely at a tower?

JF Mezei speculated:

> This past week, SpaceX had another SN24 engie test.
> But just a startup and shutdown.
>
>
> Considering that during the "iterative" testing early on, one of the
> problems was the restart of the engine(s) for landing, shouldn't engine
> tests at this point in time test the ability to start and restart
> engines to emulate a flight ?
>
> In a previous inteview, the PR person had explained that the engine
> start is the most challenging part with a serious challenge to spin up
> the turbo pumps at the right time/sync etc. (so that would explain why
> previous tests had been for short burns, since they were only testing
> engine start, not actual engine performance).
>
> But as engine start remains a challenge, shouldn't restarts be tested?

Restarts are tested, at SpaceX's McGregor site.

Note that with Falcon 9s, SpaceX doesn't test relight during a static
fire test.

[...]
> And without the ability to target an "X" on a landing site
> (since neither booster now Starship have landing legs),
> how can they test the ship's ability to land precisely at a tower?

With the water landings, they will have a chance to assess the
performance of the landing guidance, although they already have some
information from the belly-flop tests. Plus all the experience with
the F9 landings.

/dps

--
And the Raiders and the Broncos have life now in the West. I thought
they were both nearly dead if not quite really most sincerely dead. --
Mike Salfino, fivethirtyeight.com

On 2022-12-17 21:47, Snidely wrote:

> Restarts are tested, at SpaceX's McGregor site.

And that didn't prevent engine problesm for re-ignition during actual
flights since the problems were more of a system integration between
tanks and engines.

> With the water landings, they will have a chance to assess the
> performance of the landing guidance, although they already have some
> information from the belly-flop tests. Plus all the experience with
> the F9 landings.

F9 experience will be of help to the "static" booster that always comes
back empty and uses the similar "grid fins" for guidance and staying
upright. On the other hand, I have to wonder if they will have enough
data on G-force at time of landing which is something that would matter
when doinging on the "chopsticks" and which would be measured if it had
landing legs landing on hard surface.

But for Starship itself, the couple fo tests flights they had so far
resulted in them moving weight to the nose (the landing tanks), but what
happens when they start to fit the rocket with real payloads that are
returned (humans etc) ? The variability in how Starship will return will
change a lot in how the falling upject turns to vertical at last second
to land.

Remember when JF Mezei bragged outrageously? That was Sunday:
> On 2022-12-17 21:47, Snidely wrote:
>
>> Restarts are tested, at SpaceX's McGregor site.
>
> And that didn't prevent engine problesm for re-ignition during actual
> flights since the problems were more of a system integration between
> tanks and engines.

And now they have the data from those flights, and IIRC the re-ignition
problems were on SN8 and SN9, with SN10 and SN15 not having issues.
SN11 was probably Raptor 1 issues, since 2 engines were not at full
performance on the way up.

Also, there is some evidence that one or more headers tanks were filled
for SN24's latest static fire.

But keep those worry beads handy.

>> With the water landings, they will have a chance to assess the
>> performance of the landing guidance, although they already have some
>> information from the belly-flop tests. Plus all the experience with
>> the F9 landings.
>
> F9 experience will be of help to the "static" booster that always comes
> back empty and uses the similar "grid fins" for guidance and staying
> upright. On the other hand, I have to wonder if they will have enough
> data on G-force at time of landing which is something that would matter
> when doinging on the "chopsticks" and which would be measured if it had
> landing legs landing on hard surface.
>
> But for Starship itself, the couple fo tests flights they had so far
> resulted in them moving weight to the nose (the landing tanks), but what
> happens when they start to fit the rocket with real payloads that are
> returned (humans etc) ? The variability in how Starship will return will
> change a lot in how the falling upject turns to vertical at last second
> to land.

They already have some evidence as to how good their models worked.
(All 5 flown starships have belly flopped just fine, turned upright
just fine.) Elon has said the forward flaps are too big.

Turning upright is a relatively small part of the accuracy issue.
Getting the braking force right is crucial, of course, but remember
that Raptors have deeper throttling than Merlins, and Starship has some
slight ability to hover, unlike a Falcon 9 which has a hot slam
landing.

/dps

--
Maybe C282Y is simply one of the hangers-on, a groupie following a
future guitar god of the human genome: an allele with undiscovered
virtuosity, currently soloing in obscurity in Mom's garage.
Bradley Wertheim, theAtlantic.com, Jan 10 2013