SpaceX, rapid reuse, and vertical rocket stacking (integration)

[Jim]

I said before, an expendable arrives horizontally anyway, and does not require the kind of engine processing that a reusable does after a flight.

And you are still thinking it will be nine engines and the current vehicle.  Since you are fantasizing about a one launch per day, that is so far in the future that these will also happen and they will support my conop.

a.  It will be 1-4 engines.
b.  Deep throttling is not problem
c.  Methane or similar will be use eliminating coking and other issues.
d.  The engines will be more like jet engines and don't need inspection/removal after each flight.  Just like an airliner, if there is a need for maintenance, then it taken offline facility for work (airline: ramp vs hangar; rocket:  launch prep/assemble facility vs maintenance facility)
e.  Since we are fantasizing about high flight rates, it isn't going to be just one vehicle.  Over the course of the years starting from now, the need for more launches will be met more vehicles flying and support by more pads.    The point is before we get to one vehicle flying every day, there will be some ramping up of flight rate over the years, like the equivalent of 7 vehicles flying once a week .
My point with all this?  There can multiple launch prep/assemble facilities will minimalistic like the original Spacex hangar and equivalent to an airport ramp.  And there can be a maintenance facility for those vehicles that have an issue and are taken out of rotation and worked on.

If you are going to employ willing suspension of disbelief to support your once per day vehicle flight rate, then don't apply it to the current vehicle.

Then again take it further and , there is no second stage and the vehicle is fueled and launched from the landing pad and is an SSTO RLV.

But as long as it is going to be multistage with a separate payload in fairing, then horizontal wins out. 

[meekGee]

I said before, an expendable arrives horizontally anyway, and does not require the kind of engine processing that a reusable does after a flight.

And you are still thinking it will be nine engines and the current vehicle.  Since you are fantasizing about a one launch per day, that is so far in the future that these will also happen and they will support my conop.

a.  It will be 1-4 engines.
b.  Deep throttling is not problem
c.  Methane or similar will be use eliminating coking and other issues.
d.  The engines will be more like jet engines and don't need inspection/removal after each flight.  Just like an airliner, if there is a need for maintenance, then it taken offline facility for work (airline: ramp vs hangar; rocket:  launch prep/assemble facility vs maintenance facility)
e.  Since we are fantasizing about high flight rates, it isn't going to be just one vehicle.  Over the course of the years starting from now, the need for more launches will be met more vehicles flying and support by more pads.    The point is before we get to one vehicle flying every day, there will be some ramping up of flight rate over the years, like the equivalent of 7 vehicles flying once a week .
My point with all this?  There can multiple launch prep/assemble facilities will minimalistic like the original Spacex hangar and equivalent to an airport ramp.  And there can be a maintenance facility for those vehicles that have an issue and are taken out of rotation and worked on.

If you are going to employ willing suspension of disbelief to support your once per day vehicle flight rate, then don't apply it to the current vehicle.

Then again take it further and , there is no second stage and the vehicle is fueled and launched from the landing pad and is an SSTO RLV.

But as long as it is going to be multistage with a separate payload in fairing, then horizontal wins out.

Again, maybe.   

I think the line where things change is over 1/wk, since that makes individual stages of the operation be roughly 1 day.

1-4 engines - I think the rule is either 1 or many.  not 2, not 3, not 4.    Simple math shows that if failures are independent, then a large cluster is the best option, followed by a single large engine.  But not a small cluster.

How large?  Not large enough that the options of 2 independent bad engines is statistically possible, and most importantly - large enough so that the lost thrust of 1-out-of-N still gives you engine-out capability at lift-off.

9 is working great for SpaceX.  You are the one that always complains that they're lying about engine-out at T+0.
But the more engines you have, the easier it is to achieve that.  So I don't get the drive to 4 engines.


Deep throttling - sure - if it comes without a cost.   1-out-of-9 gives you insta-11% without any performance cost, and then you can add throttling on top of that.

Methane - hey, I'm all for it.   But I still think the engines will have to be inspected.  This is NOT a jetliner, there's no graceful return-to-base.  You have a single engine-out capability (or none at all if you're using a single engine) and so you'd better take a good look at them before you go.  Methane or not.


I agree about having more than one vehicle.  You can take the numbers I gave as per vehicle, not per the entirety of SpaceX.

[meekGee]

I think vertical processing will be faster, and less expensive

you have nothing relevant to back up that "think"

Jim - I presented this as an opinion, based on my engineering analysis and my common sense. Of course it could be wrong.  I don't have inside info from SpaceX. But even though you disagree, it's not like I stated it without anything to back it up.  There are lots of posts upthread that back it up.  You just happen to think they're wrong.

What I don't understand is how you're so absolutely certain.

We had a similar argument about whether F9 1.1 will have reusability features designed into it.  You were equally ferocious.  But it turned out that it did.   We had a similar one about whether SpaceX is developing a Methane engine.   We don't have the inside info - all we can do is make best guesses based on what we see, and they're always going to be uncertain.  This would be a better foundation for such conversations.  I state upfront that this is simply my prediction, it can be wrong, and I care more about what the final outcome is than whether I was right about it or not.

[Jim]

We had a similar one about whether SpaceX is developing a Methane engine. 

That is still too early to call.
But you did get me on the re-usability.  I was looking at old data. 

[Jim]

What I don't understand is how you're so absolutely certain.

Because I know what is takes to process a vehicle (ELV or reusable) and everything points to horizontal.   

[Jim]

1-4 engines - I think the rule is either 1 or many.  not 2, not 3, not 4.    Simple math shows that if failures are independent, then a large cluster is the best option, followed by a single large engine.  But not a small cluster.

 nine engines were not selected based on reliability.  It was an easy way to get to EELV class with an existing engine.  There was going to be a Falcon 5.  And Spacex matra was one engine per stage.  They just went to engine out because of back pedaling 

[meekGee]

1-4 engines - I think the rule is either 1 or many.  not 2, not 3, not 4.    Simple math shows that if failures are independent, then a large cluster is the best option, followed by a single large engine.  But not a small cluster.

 nine engines were not selected based on reliability.  It was an easy way to get to EELV class with an existing engine.  There was going to be a Falcon 5.  And Spacex matra was one engine per stage.  They just went to engine out because of back pedaling

I remember the F5, but I think the F9 was planned even then.

I don't think SpaceX knew for sure where they are headed, though Elon thinks pretty far ahead.  So maybe the reason for the engine sizing was that a) it's what they could comfortably do, b) it took them from F1 to larger rockets, and c) it gave them the most flexibility in deciding how big a rocket to build next.

But irrespective of the rationale, 9 engines is working out well.  It makes it easy to get engine-out from as early as possible in the flight (obviously T+0 engine-out depends on how much payload is on top), it makes it easier to land, etc.

So even if the initial rationale no longer applies, my money is that they'll stay with 5/7/9 engines when they move to Raptor and a heavy lifter.  With extra money of 7/9.  (5 is just weaseling out...)

[guckyfan]

7 sounds perfect if it gives the desired thrust. Equal spacing of the engines allows for maximum space for the engine bells and for gimballing. The throttling range needs to allow for landing but that should be achievable.

[Jim]

I remember the F5, but I think the F9 was planned even then.

Nope, F5 was first, they were going after Delta II and F9 came later

[meekGee]

I remember the F5, but I think the F9 was planned even then.

Nope, F5 was first, they were going after Delta II and F9 came later

I seem to recall the F5 as the "next rocket we'll build" and the F9 as the "far off next-next rocket we'll build".  Then everything shifted up.

But the point was - nevermind the initial rationale.  We don't know what Elon was thinking of.   What matters is that the 9-engine cluster is working out well today, whether it's due to premonition or blind dumb luck. 

Today, 9-engine works well for the 11% down throttle, whether they intended this from the beginning or not.  It works well for 88% of remaining thrust on engine-out.

So given what we know today, I don't think they'll go to a small cluster or single engine with Raptor.

[meekGee]

We had a similar one about whether SpaceX is developing a Methane engine. 

That is still too early to call.
But you did get me on the re-usability.  I was looking at old data.

And Jim - I appreciate the first-hand knowledge that you bring in.  I picked up a lot from these arguments, though usually I have to back you into a corner before you cough up the "why" that I'm looking for...

It's just that knowing the details of how things work doesn't immediately translate into seeing what SpaceX is planning.  They have some surprises left in them, I'm pretty sure, and almost nothing is cast in stone.

If it ends up that they remain with Horizontal processing, so be it...   But if one day we'll see the rocket pulled on its legs back to a vertical hanger, I'll have a drink to this thread....

[Jim]

I will use a never here.  They will never integrate the upperstage and payload and then attach them to the booster.  That is too disruptive to the flow.

[meekGee]

I will use a never here.  They will never integrate the upperstage and payload and then attach them to the booster.  That is too disruptive to the flow.

I hear you about the payloads.

The thread was mostly about first stage processing and stacking of the second stage - payload is a more specialized operation, and it differs from payload to payload.  So that's secondary.

So the main things I'm saying:
1. I think it will be advantageous, in terms of speed and probably per-unit cost as well, to process the first stage while vertical, and never lay it down.
2. I think loading the second stage onto the first stage is also best done while both are vertical.
3. I think the connection between the two can be simplifies to the point where it is practically only mechanical.
4. I think the first stage will be an independent flyer.

1 and 2 - they only become relevant with high flight volume.  My completely WAG is somewhere between 1/week and 1/day.

3 and 4 - I can't see why they wouldn't be doing it even now.  (I don't think they are, because we'd have heard about it, but we do know the capabilities to do so are already in the first stage)

[mgfitter]

How far downrange does the first stage land (or splashdown or crash) on these flights without the boost-back profile?

And (maybe this question should go in the Q&A thread, but I'll put it here for now), is there an easy way to calculate an approximate landing point for the first stage, assuming you know the vertical and horizontal velocities at separation?

Thanks!

-MG

[meekGee]

How far downrange does the first stage land (or splashdown or crash) on these flights without the boost-back profile?

And (maybe this question should go in the Q&A thread, but I'll put it here for now), is there an easy way to calculate an approximate landing point for the first stage, assuming you know the vertical and horizontal velocities at separation?

Thanks!

-MG

yeah, this belongs in the general thread - please re-post there.

[RocketmanUS]

I hear you about the payloads.

The thread was mostly about first stage processing and stacking of the second stage - payload is a more specialized operation, and it differs from payload to payload.  So that's secondary.

So the main things I'm saying:
1. I think it will be advantageous, in terms of speed and probably per-unit cost as well, to process the first stage while vertical, and never lay it down.
2. I think loading the second stage onto the first stage is also best done while both are vertical.
3. I think the connection between the two can be simplifies to the point where it is practically only mechanical.
4. I think the first stage will be an independent flyer.

1 and 2 - they only become relevant with high flight volume.  My completely WAG is somewhere between 1/week and 1/day.

3 and 4 - I can't see why they wouldn't be doing it even now.  (I don't think they are, because we'd have heard about it, but we do know the capabilities to do so are already in the first stage)

Payload could possibly use a payload canister similar to the one planned for the Venture Star. Fairing and 2nd stage could be one making the fairing reusable too. Would require the larger Raptor based vehicle for usable payload mass.

1. Why would it be a problem laying it down?
     Use a mobile transport to bring it back to the hanger horizontally.
     Go up to it's landing area and grab on to it ( multiple arms on a strong back ), then lower it to horizontal
     position.

2. If they are on horizontal rolling racks would the stages aline easy compared to 2nd stage hanging from
    a crane?

3/4. K-1 stages were to be independent flyer's. So with the 1st stage flying back it should have it's own
       systems already.

They will need several 1st and 2nd stages for rapid reuse.
Inspections stations.
integration stations.

The same 1st stage would not always launch with the same 2nd stage.

[meekGee]

I hear you about the payloads.

The thread was mostly about first stage processing and stacking of the second stage - payload is a more specialized operation, and it differs from payload to payload.  So that's secondary.

So the main things I'm saying:
1. I think it will be advantageous, in terms of speed and probably per-unit cost as well, to process the first stage while vertical, and never lay it down.
2. I think loading the second stage onto the first stage is also best done while both are vertical.
3. I think the connection between the two can be simplifies to the point where it is practically only mechanical.
4. I think the first stage will be an independent flyer.

1 and 2 - they only become relevant with high flight volume.  My completely WAG is somewhere between 1/week and 1/day.

3 and 4 - I can't see why they wouldn't be doing it even now.  (I don't think they are, because we'd have heard about it, but we do know the capabilities to do so are already in the first stage)

Payload could possibly use a payload canister similar to the one planned for the Venture Star. Fairing and 2nd stage could be one making the fairing reusable too. Would require the larger Raptor based vehicle for usable payload mass.

1. Why would it be a problem laying it down?
     Use a mobile transport to bring it back to the hanger horizontally.
     Go up to it's landing area and grab on to it ( multiple arms on a strong back ), then lower it to horizontal
     position.

2. If they are on horizontal rolling racks would the stages aline easy compared to 2nd stage hanging from
    a crane?

3/4. K-1 stages were to be independent flyer's. So with the 1st stage flying back it should have it's own
       systems already.

They will need several 1st and 2nd stages for rapid reuse.
Inspections stations.
integration stations.

The same 1st stage would not always launch with the same 2nd stage.

Read upthread, but the short version is that I think processing a 9-engine stage is a lot faster (and more "natural" when it is vertical.
The actual laying down is not the main problem.

If the stage connection is simple, then doing it vertically should be (ready for it?) a snap. 

[Nomadd]

 Not sure why processing vertical is faster or more natural. The things were assembled horizontally and that's how they're used to working on it. Removing and replacing a component would be a more complicated movement if it was vertical. You'd need an entirely different procedure as opposed to mainly using the same one you did for assembly at the plant. Same for putting the 2nd stage on. Why would they do it differently than they do for the first flight?

[meekGee]

The premise was rapid reuse.  1-day turn-around or thereabouts.

I agree there's no point making any changes when you're only launching once a month.

The tail of a 747 is 65 feet, and you don't worry about it striking power lines or bridges between the terminal and the runway, right?  We're talking about a recurring operation here, not a one time exercise.

Center of gravity of an empty first stage - how high do you think that is?  I don't see why you'd have any problem with it, though I'd like to see the F9R with legs deployed first.

Not quite. In the case of the 747 there are airports it can not land at due to short runways, lack of equipement or places to store and so on.  As for 1 day turn arounds I would love to see that one day but at the moment the entire world only launches maybe about 100 or a little over 100 times a year not 365 times from a single space port and to enable this you would need to have nothing in the way from the landing pad to the processing facility.

The reason you could have problems with the center of gravity is becuase the stage is empty(Propellant eqauals weight). The stage is also lacking the wieght of the other stages on top of it and to add icing to the cake, I suspect the stage is likely bolted down on the pad or has an hold down system to keep it from moving before launch all of which would be missing at landing.

Also to move it the ground between the landing pad and the processing place is going to have to be flat becuase any bump, incline or decline could damage or tip it or cause unexpected loads in the structure. The shuttle moves in a vertical fashion from or to it's pad and it ain't quick.
 
You could also have problems with wind. There maybe a day when rockets only need to be refueled to fly again like cars, busses, trains, and planes but that is not soon.

Sorry missed this one earlier.

A reusable launch facility, if considering rapid reuse, is very similar to an airport.

You must have several landing pads (at least 3 for first stages and 1 for upperstage - but you probably want spares.

You then must have taxiways from them to the processing facility.   They will be level and clean enough for the stages to move on.

(Quoting GS again - "We don't want each flight to be like giving birth")

So just like in an airport where a 747 is operating out of, the taxiways are compatible with the airplane's dimension.

--

c.g. wise, lack of second stage and payload drops the center of gravity.
Lack of fuel in the tank does the same, since you have a heavy engine cluster at the bottom, and empty tank on top.
Look at a side-view of GH and try to paint where the c.g. is.  (granted, the legs are wider and it is a tad shorter, but still - it's not anywhere near in danger of tipping)

[meekGee]

Not sure why processing vertical is faster or more natural. The things were assembled horizontally and that's how they're used to working on it. Removing and replacing a component would be a more complicated movement if it was vertical. You'd need an entirely different procedure as opposed to mainly using the same one you did for assembly at the plant. Same for putting the 2nd stage on. Why would they do it differently than they do for the first flight?

It's all upthread.

For example here:  http://forum.nasaspaceflight.com/index.php?topic=33430.msg1129579#msg1129579

and then Jason's comment about how in the plant they do it in the vertical orientation.


From my experience - it's easier to slowly drop or raise a heavy (and delicate) components to place then to slide it sideways.