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

[rst]

In modern systems (like jet engines), manual inspections are replaced by automated monitoring/logging of operating temps/pressures/vibrations/g forces... identifying trends, out of bounds conditions, and pending failures..

SpaceX has all this airline/aircraft experience and technology to draw from...

But the thresholds are themselves based on experience, and extensive testing of a sort that SpaceX hasn't had the chance to do yet.  A new airliner typically has six months to a year of test flights before going into service, in addition to non-flight testing on static structural test articles.  And those test programs are only as short as they are because of prior experience with previous generations of airliners, which have given people a reasonably good idea of what sorts of problems they need to be looking for.

And while airliners need that much testing in part because they go up in a much greater range of wind and weather conditions, none of them involve anything much like the physical stress involved in a rocket launch.  So, it really isn't clear how experience with airliners is supposed to generalize to reusable launch vehicles.  At best, it'll give SpaceX a good set of tools to evaluate data on rocket wear and expected lifetime once they have it --- but that'll be a while.

The problem is that no one yet has much experience with reusable liquid-fueled orbital rockets.  So, even if SpaceX had been shooting for gas-and-go reusability as much as they could while designing F9R, they'd be crazy to expect it to work.  (Or to tell anyone about it, because they wouldn't expect it to work.)  As with the first recovery try off Vandenberg, it's more likely that they're expecting some unpleasant surprise.  And for the first few at least, if not the first few dozen, they'll likely be disassembling, reassembling, and doing the full round of factory-floor checkouts and qualification tests and then some trying to find it.

[MP99]

I don't have Elon's numbers of course, but there's a caveat to what you said. (which is partly correct - it's just not the whole story)

Capital expense (the cost of the rocket) gets amortized over number of flights.   If you look at your rocket as a revenue generator, like a jet liner, then you want it idle as little as possible.  Airliners don't get twice as many airplanes so they can process them more slowly.  Rather, they do everything they can in order to get the maximum air time our of each airframe.

We've seen Elon talk this way about the Mars vehicle.  He wants it back in one launch window, since otherwise he can only use each vehicle every 3rd cycle - and he'd rather take the expense if fuel of rushing the trip than build 3x as many ships.

By your analogy, airframes could perform more flights per day by running at Mach  2 or 3.

But, like with huge dV for TMI, it would take huge amounts of fuel.

Cheers, Martin

[AncientU]

In modern systems (like jet engines), manual inspections are replaced by automated monitoring/logging of operating temps/pressures/vibrations/g forces... identifying trends, out of bounds conditions, and pending failures..

SpaceX has all this airline/aircraft experience and technology to draw from...

The problem is that no one yet has much experience with reusable liquid-fueled orbital rockets. 

Exactly zero, that is.

[rst]

In modern systems (like jet engines), manual inspections are replaced by automated monitoring/logging of operating temps/pressures/vibrations/g forces... identifying trends, out of bounds conditions, and pending failures..

SpaceX has all this airline/aircraft experience and technology to draw from...

The problem is that no one yet has much experience with reusable liquid-fueled orbital rockets. 

Exactly zero, that is.

Well, there is the Space Shuttle and SSME, but that's a different enough system that again it isn't clear what lessons transfer, if any.

[meekGee]

I don't have Elon's numbers of course, but there's a caveat to what you said. (which is partly correct - it's just not the whole story)

Capital expense (the cost of the rocket) gets amortized over number of flights.   If you look at your rocket as a revenue generator, like a jet liner, then you want it idle as little as possible.  Airliners don't get twice as many airplanes so they can process them more slowly.  Rather, they do everything they can in order to get the maximum air time our of each airframe.

We've seen Elon talk this way about the Mars vehicle.  He wants it back in one launch window, since otherwise he can only use each vehicle every 3rd cycle - and he'd rather take the expense if fuel of rushing the trip than build 3x as many ships.

By your analogy, airframes could perform more flights per day by running at Mach  2 or 3.

But, like with huge dV for TMI, it would take huge amounts of fuel.

Cheers, Martin

Actually, the time-in-air is a serious consideration.  If you were to build a SST that can hop the Atlantic in an hour and thus fly several times a day, this ability would factor heavily into its commercial viability (since it generates revenue maybe 6 times a day).   Of course designing an SST is such a radical departure from subsonic jets that the increase in cost proved (at least in the case of Concorde) to be too much.

However, here we're talking about designing a rocket and infrastructure for faster turn around times, but with the actual performance of the rocket remaining more or less the same.  If you could bring the turn around time from a month to a day, I don't think the vehicle will cost 30x more.  (Assuming reusability in both cases - the comparison to non-reusable rockets is of course much more extreme)

[MP99]

I don't have Elon's numbers of course, but there's a caveat to what you said. (which is partly correct - it's just not the whole story)

Capital expense (the cost of the rocket) gets amortized over number of flights.   If you look at your rocket as a revenue generator, like a jet liner, then you want it idle as little as possible.  Airliners don't get twice as many airplanes so they can process them more slowly.  Rather, they do everything they can in order to get the maximum air time our of each airframe.

We've seen Elon talk this way about the Mars vehicle.  He wants it back in one launch window, since otherwise he can only use each vehicle every 3rd cycle - and he'd rather take the expense if fuel of rushing the trip than build 3x as many ships.

By your analogy, airframes could perform more flights per day by running at Mach  2 or 3.

But, like with huge dV for TMI, it would take huge amounts of fuel.

Cheers, Martin

Actually, the time-in-air is a serious consideration.  If you were to build a SST that can hop the Atlantic in an hour and thus fly several times a day, this ability would factor heavily into its commercial viability (since it generates revenue maybe 6 times a day).   Of course designing an SST is such a radical departure from subsonic jets that the increase in cost proved (at least in the case of Concorde) to be too much.

The large increase in dV for rapid Mars transit seems to depend on very cheap prop delivery to LEO (or an expensive SEP development programme).

However, here we're talking about designing a rocket and infrastructure for faster turn around times, but with the actual performance of the rocket remaining more or less the same. 

Actually, we weren't. We were discussing whether the huge increase in dV to achieve 3 month Mars transit is worth the reduced time in air / increased utilisation of the craft.

If you could bring the turn around time from a month to a day, I don't think the vehicle will cost 30x more.  (Assuming reusability in both cases - the comparison to non-reusable rockets is of course much more extreme)

And then, the question becomes whether there exists 30x the demand.

Or,  more specifically, whether the price is then low enough to trigger 30x the demand through the price elasticity curve.

Cheers, Martin

[meekGee]

I don't have Elon's numbers of course, but there's a caveat to what you said. (which is partly correct - it's just not the whole story)

Capital expense (the cost of the rocket) gets amortized over number of flights.   If you look at your rocket as a revenue generator, like a jet liner, then you want it idle as little as possible.  Airliners don't get twice as many airplanes so they can process them more slowly.  Rather, they do everything they can in order to get the maximum air time our of each airframe.

We've seen Elon talk this way about the Mars vehicle.  He wants it back in one launch window, since otherwise he can only use each vehicle every 3rd cycle - and he'd rather take the expense if fuel of rushing the trip than build 3x as many ships.

By your analogy, airframes could perform more flights per day by running at Mach  2 or 3.

But, like with huge dV for TMI, it would take huge amounts of fuel.

Cheers, Martin

Actually, the time-in-air is a serious consideration.  If you were to build a SST that can hop the Atlantic in an hour and thus fly several times a day, this ability would factor heavily into its commercial viability (since it generates revenue maybe 6 times a day).   Of course designing an SST is such a radical departure from subsonic jets that the increase in cost proved (at least in the case of Concorde) to be too much.

The large increase in dV for rapid Mars transit seems to depend on very cheap prop delivery to LEO (or an expensive SEP development programme).

However, here we're talking about designing a rocket and infrastructure for faster turn around times, but with the actual performance of the rocket remaining more or less the same. 

Actually, we weren't. We were discussing whether the huge increase in dV to achieve 3 month Mars transit is worth the reduced time in air / increased utilisation of the craft.

If you could bring the turn around time from a month to a day, I don't think the vehicle will cost 30x more.  (Assuming reusability in both cases - the comparison to non-reusable rockets is of course much more extreme)

And then, the question becomes whether there exists 30x the demand.

Or,  more specifically, whether the price is then low enough to trigger 30x the demand through the price elasticity curve.

Cheers, Martin

The dV and time-to-mars were in a different thread...

This conversation (vertical processing) was about whether the LEO launcher (which Elon wanted to launch with a same-day turn-around, and now we understand why, as you point out) should be built to turn-around quickly on the ground, or rather be built for longer turn-around times but cheaper as was suggested up-thread.

I still think that fast turn-around is critical, even if it means the equipment/infrastructure costs more.  To a point, of course, but right now everything is weighted very far towards reducing cost of infrastructure (which makes sense for now)

I've been mulling the numbers for the whole system too.  It warrants a different thread.  The high dV requirement has many ramifications, beyond only explaining why Elon wants such a rapid launch rate of such large rockets.

Interestinger and interestinger