Falcon Heavy Cross-Feed

[FuseUpHereAlone]

There's been a lot of talk about Falcon Heavy (FH) using cross-feed to move propellant from the boosters to the core stage, but not a lot of discussion as to how this could be accomplished.  The advantages of cross-feed are tremendous; feeding propellant from the boosters to the core allow for the for the core engines to be fired from ground to altitude, yet allow for the core to be nearly full when the boosters are jettisoned.  This is unlike Delta IV Heavy (D4H) which relies upon throttling down the core engine to conserve fuel consumption, and allowing for some propellant to be left in the core stage after booster jettison.  I've heard that Boeing/ULA have looked at incorporating cross-feed into D4H, but decided that it was not worth the investment (in fact upgrading RS-68 was chosen over cross-feed).  So this begs the question, how can SpaceX incorporate cross-feed into FH?

So what are the challenges of cross-feed?  The biggest one that comes to mind is that there are 9 Merlin engines burning fuel and oxidizer, so propellant feeding into the core needs to be nearly equivalent to what's being drained.  Assuming that each Merlin-D engine burns about 483 lb/s of fuel (based on 150K Thrust @ 310s Isp @ vacuum), then core stage drains about 4347 lb/s at 100% thrust, and about 3043 lb/s at 70%.  This implies that each booster needs to provide a minimum of 1500 lb/s of propellant to the core stage.  So how can this done?

A couple of ideas that come to mind:

    -Dedicated turbopump on each booster to move propellant to the core.  Not sure though if the head requirements are the same as on the engine; the pressure is not as high but the mass flow is much highier.  May require a new turbopump (and gas generator, exhaust duct, and plumbing).

    -Pressuring propellant in the booster.  If the ullage pressure in the booster is higher than the core, then it could be possible to flow propellant into the core without a turbopump.  Of course this assumes that the propellant tanks were already designed to with stand higher ullage pressure or a slight vacuum.

[Jim]

You are making harder than it is.
See heritage Atlas.
No additional turbopumps. Just valves and large interconnect ducts.  Isolate the core tanks from launch until right before staging. Then open core valves and then isolate boosters via valves in interconnect ducts.  Shutdown both boosters when one shows dry.

[FuseUpHereAlone]

Perhaps I am...but I thought that SM-65 Atlas had one fuel and one oxidizer tank for the booster and sustainer engines.  At altitude, the booster engines were jettisoned, but not the entire stage (hence 1.5 staging).  What I'm curious about is how is it possible to move propellant from one tank to another.

[Jim]

Perhaps I am...but I thought that SM-65 Atlas had one fuel and one oxidizer tank for the booster and sustainer engines.  At altitude, the booster engines were jettisoned, but not the entire stage (hence 1.5 staging).  What I'm curious about is how is it possible to move propellant from one tank to another.

The turbopumps suck.

it doesn't have to go from tank to tank but from feedline to feedline.

[sdsds]

For my own clarification I want to expand on this a bit.  I think it relies on the "nature abhors a vacuum" observation.  During boosted ascent, all three sets of tankage are feeding their own engines.  The two boosters are replacing the propellant they drain with pressurant gas, while the center core is replacing its propellant with fluid propellant taken from the boosters..

So yes, the core's turbopumps are sucking propellant into the core tankage.

[EDIT:  Ah.  Nevermind.]

[Jim]

[Hauerg]

But isn't the FH more a kind of "side feed" configuration, where 3 of the core engines (NOT the tanks) are supplied from the tanks of the left booster, 3 more from the right booster and only the 3 engines in the middle are fed from tanks on the central core??

[go4mars]

Yes.  My impression is also that it's only 6 engines in the core stage that participate. 

[sdsds]

Yes.  My impression is also that it's only 6 engines in the core stage that participate. 

And so at booster sep, what percentage of propellant has been drained from the core tankage?

[FuseUpHereAlone]

OK...Jim's picture makes a lot of sense to me.  Funny how complicated things can be if you look at the wrong problem.

[edkyle99]

Yes.  My impression is also that it's only 6 engines in the core stage that participate. 

And so at booster sep, what percentage of propellant has been drained from the core tankage?

If core propellant continuously feeds the center three engines and is not replaced, it would burn one-third of its propellant if the engine throttle settings are the same on all engines.  Less if not.  An ideal cross-feed setup would leave the core full, or nearly so, at booster shutdown.

 - Ed Kyle

[Jim]

OK...Jim's picture makes a lot of sense to me.  Funny how complicated things can be if you look at the wrong problem.

It also means the downcomers (the ducts from the tanks to the sump) have to larger on the boosters than the core.

[LegendCJS]

Yes.  My impression is also that it's only 6 engines in the core stage that participate. 

And so at booster sep, what percentage of propellant has been drained from the core tankage?

If core propellant continuously feeds the center three engines and is not replaced, it would burn one-third of its propellant if the engine throttle settings are the same on all engines.  Less if not.  An ideal cross-feed setup would leave the core full, or nearly so, at booster shutdown.

 - Ed Kyle

Not quite.  Assuming same tank volumes and mass flow/ throttle rates, the booster tanks would be feeding 12 engines each while the core fed three.  This means 1/4th of the core tank will be used up at the time the boosters run dry.

[go4mars]

Yes.  My impression is also that it's only 6 engines in the core stage that participate. 

And so at booster sep, what percentage of propellant has been drained from the core tankage?

If core propellant continuously feeds the center three engines and is not replaced, it would burn one-third of its propellant if the engine throttle settings are the same on all engines.  Less if not.  An ideal cross-feed setup would leave the core full, or nearly so, at booster shutdown.

 - Ed Kyle

Not quite.  Assuming same tank volumes and mass flow/ throttle rates, the booster tanks would be feeding 12 engines each while the core fed three.  This means 1/4th of the core tank will be used up at the time the boosters run dry.

And only 1/8th if the middle 3 are throttled down to 50%...

[RocketmanUS]

Would it be possible for each booster to feed 4 engines each and the core only one engine?

[Robotbeat]

Would it be possible for each booster to feed 4 engines each and the core only one engine?

Anything is "possible."

[ChefPat]

Would it be possible for each booster to feed 4 engines each and the core only one engine?

How many Angels can dance on the head of a pin?

[dunderwood]

How many Angels can dance on the head of a pin?

42

[RocketmanUS]

Would it be possible for each booster to feed 4 engines each and the core only one engine?

Anything is "possible."

So most likely not practical?

[edkyle99]

Yes.  My impression is also that it's only 6 engines in the core stage that participate. 

And so at booster sep, what percentage of propellant has been drained from the core tankage?

If core propellant continuously feeds the center three engines and is not replaced, it would burn one-third of its propellant if the engine throttle settings are the same on all engines.  Less if not.  An ideal cross-feed setup would leave the core full, or nearly so, at booster shutdown.

 - Ed Kyle

Not quite.  Assuming same tank volumes and mass flow/ throttle rates, the booster tanks would be feeding 12 engines each while the core fed three.  This means 1/4th of the core tank will be used up at the time the boosters run dry.

Right.  Thanks for that correction. 

 - Ed Kyle

[Robotbeat]

Would it be possible for each booster to feed 4 engines each and the core only one engine?

Anything is "possible."

So most likely not practical?

That depends on detailed trade studies. The only such analysis we have is that they are leaning toward the core doing 3 engines (I believe).

[cuddihy]

Can anyone explain why FH s0 is antipated to be 12/3/12 rather than 13/1/13 or 13.5/0/13.5?

Is it too much head loss in the cross feed piping or concerns about flow instabilities if the side boosters are cross connected?

[Robotbeat]

Can anyone explain why FH s0 is antipated to be 12/3/12 rather than 13/1/13 or 13.5/0/13.5?

Is it too much head loss in the cross feed piping or concerns about flow instabilities if the side boosters are cross connected?

I think the plan is that after the boosters stage off, only the engines fed by the core will stay lit (that way they don't have to switch propellant supplies). If you have too few engines burning at that point you have very high gravity losses, which is bad (obviously). Also, engine-out capability may still be desired.

[cuddihy]

Can anyone explain why FH s0 is antipated to be 12/3/12 rather than 13/1/13 or 13.5/0/13.5?

The Dec 2012 date is for hardware delivered to VAFB. Most likely for hardware comparability tests. When the FH gets launched depends on the readiness of the pad at SLC-4E and how the Cassiope flight goes.

Oops. Was talking about engine feed schemes, not dates. 12 right booster/3 core /12 left booster. Etc

[RDoc]

I think the plan is that after the boosters stage off, only the engines fed by the core will stay lit (that way they don't have to switch propellant supplies). If you have too few engines burning at that point you have very high gravity losses, which is bad (obviously). Also, engine-out capability may still be desired.

What is the major risk factor in doing crossfeed?

It would seem, on the face of it, to be valves failing to open and/or close, but that sounds like something that could be heavily ground tested. While I have no doubt that cryogenic valves have their own idiosyncrasies, it doesn't seem like a high risk item, but I have zero experience in the area.

If valve reliability is high, then wouldn't it be relatively safe to use 6 valves (2 fuel and 2 LOX for the outboard tanks, and another 2 for the core tanks)? Then all 9 core engines could be used. I'm assuming there already are valves in the outboard boosters to shut off the fuel and LOX, but perhaps there need to be 2 more each at the booster couplings?

[grakenverb]

How many Angels can dance on the head of a pin?

42

Sorry, the correct answer is 41, you are forgetting that this is a leap year.

[RocketmanUS]

Would the boosters separate around 127 seconds after lift off? ( That is for 12 engines feeding on a booster ).

Would only three engine on the core burning give it enough thrust for that velocity and altitude after booster separation?

[dcporter]

Oops. Was talking about engine feed schemes, not dates. 12 right booster/3 core /12 left booster. Etc

13/1/13 tipped me off. Even on a leap year.

I think the plan is that after the boosters stage off, only the engines fed by the core will stay lit (that way they don't have to switch propellant supplies). If you have too few engines burning at that point you have very high gravity losses, which is bad (obviously). Also, engine-out capability may still be desired.

Whoa FH is gonna run on three engines after booster sep? Any chance you remember where you heard that?

[charliem]

I have my own theory of how cross-feed works, made from the little info we have.

Spacex's website says that FH's mass and thrust at lift off is 1,400 mt and 3,800,000 lbf (1,724 mtf) (22 Merlin-1D at full thrust is the absolute minimum to elevate 1,400 mt).

It also says: "Propellant cross-feeding leaves the center core still carrying the majority of its propellant after the side boosters separate".

Can't find where I saw it but I remember to have read about a 90% propellant remaining in the core tanks right after boosters cut off.

In the video looks like all 27 engines are started at the same time, and we know that a Merlin-1D can only be throttled down to a 70% of its nominal thrust.

Easy to calculate that even with 3 engines at 70% more than a 10% of the core fuel would be gone when the booster tanks reach the empty mark.

So, I think it could be like this: Feed valves are set as Jim depicted. Before lift-off the core tanks valves are closed and all 9 core engines are fed from both booster tanks.

This mode lasts until the booster tanks are near 10%, when the core tanks valves are switch to open, and the valves that communicate core turbopumps and booster tanks are closed.

After that core and boosters keep burning exclusively its own fuel until the booster tanks are depleted, triggering separation.

P.E. And no, three engines are not enough to keep FH in flight after separation.

[jimvela]

That depends on detailed trade studies. The only such analysis we have is that they are leaning toward the core doing 3 engines (I believe).

I have no knowledge of SpaceX' plans, but it also depends on how simple they want to make a side-fed design (note I didn't say cross-fed, and neither should any of you when referring to the present concept).

The easiest would seem to be to use three mostly identical cores, with the outer cores each driving two engines in the inboard core, and shutting down those engines at separation.  Lets call that 11/5/11.  I think this also preserves the most likely engine-out capability through the flight profile.  It may not provide the best outcome as the center core may be overpowered and under-fueled after booster separation.

If they want to stretch the outer cores (and possibly SHRINK the center core), then something like 12/3/12 might make sense- again depending on the trades.  Without the stretch/shrink, the center core may end up under-powered and over-fueled after separation.   I can't help picturing this as looking Titan-esq with big (liquid) side boosters, and a big honking fairing not far above the booster nose cones.  I'd find that pretty satisfying considering the VAFB pad heritage. 

Either of the side-fed options might benefit from being able to drop those side engines away with the host core, but that added complexity may not be worth the cost/reliability trades given the lighter Merlins coming online.

The next BIG leap is to actually do cross-feed, in which case you want to draw as little fuel fromthe center core as possible while burning the outer cores, so that at separation the launcher is as close to fully fueled and fully powered as possible, maximizing the benefit of having boosters in the first place.

Cross-fed is a much tougher problem to solve, not just for plumbing and valve concerns, but also because you have to separate, and designing separable plumbing that is both reliable while mated and reliable in separation is a non-trivial task compared to just feeding some engines in the other core that you shut down when you sever the feeds.

Oh, to be a fly on the walls in Hawthorne... 

[jimvela]

I have my own theory of how cross-feed works, made from the little info we have.

That's a pretty good theory!

[SpacexULA]

Whoa FH is gonna run on three engines after booster sep? Any chance you remember where you heard that?

No 18 Engines at start 3 engines on left side of core feed off left strap-on.  3 engines on right side of core feed off Right Strap on. (So each of the strap-on are actually feeding 12 engines instead of 9, and core is only feeding 3)

After low fuel on a strap-on Staging happens.  Now all 9 core engines are feeding off a mostly full Center Core (only fed 3 engines for as long as it takes 12 engines to suck down a stage).

Does that make since?

[jimvela]

You are making harder than it is.
See heritage Atlas.
No additional turbopumps. Just valves and large interconnect ducts.  Isolate the core tanks from launch until right before staging. Then open core valves and then isolate boosters via valves in interconnect ducts.  Shutdown both boosters when one shows dry.

Isn't the first hard part of heritage Atlas that the valves below the center core are subject to all manner of bad things?  (e.g. boiling cryo/vapor lock, some oddity where you get surges in the RP supply line when you first open it, etc?)

High-speed liquid turbopumps don't like to ingest gasses or to have wildly varying inlet fuel conditions.

Isn't the second hard part of heritage Atlas getting the severable part of the feeds from each core to reliably sever/separate cleanly?

I can see an education in my near-term future 

[jimvela]

Whoa FH is gonna run on three engines after booster sep? Any chance you remember where you heard that?

No 18 Engines at start 3 engines on left side of core feed off left strap-on.  3 engines on right side of core feed off Right Strap on. (So each of the strap-on are actually feeding 12 engines instead of 9, and core is only feeding 3)

After low fuel on a strap-on Staging happens.  Now all 9 core engines are feeding off a mostly full Center Core (only fed 3 engines for as long as it takes 12 engines to suck down a stage).

Does that make since?

It would seem that if they do cross-feed at all, it's an all-or-nothing proposition as the other Jim points out. 

Partial cross-feed is MORE complicated, and has LESS benefit!

[RocketmanUS]

You are making harder than it is.
See heritage Atlas.
No additional turbopumps. Just valves and large interconnect ducts.  Isolate the core tanks from launch until right before staging. Then open core valves and then isolate boosters via valves in interconnect ducts.  Shutdown both boosters when one shows dry.

So need to control the flow of propellant from each booster to core to keep a balance from each. If so can that be done with variable flow valves?

[pippin]

So need to control the flow of propellant from each booster to core to keep a balance from each. If so can that be done with variable flow valves?

What for? Again: don't make it more complicated than it is, these are communicating vessels:
http://en.wikipedia.org/wiki/Communicating_vessels

flow-control valves would actually only make things worse.

Sometimes physics can be so simple

[Jim]

So need to control the flow of propellant from each booster to core to keep a balance from each. If so can that be done with variable flow valves?

What for? Again: don't make it more complicated than it is, these are communicating vessels:
http://en.wikipedia.org/wiki/Communicating_vessels

flow-control valves would actually only make things worse.

Sometimes physics can be so simple

Only if the pressure in the boosters are kept equal

[phred]

So, considerng the additional hardware required (which sounds fairly minimal to my civilian eyes), what would be a realistic performance enhancement estimate for the FH?

[RDoc]

So need to control the flow of propellant from each booster to core to keep a balance from each. If so can that be done with variable flow valves?

There's already flow control to keep the engines running at proper thrust and mixture.

For lots of reasons, one tank is going to run out first, or get down to the level needed for stage recovery if that's planned.

[RDoc]

I have no knowledge of SpaceX' plans, but it also depends on how simple they want to make a side-fed design (note I didn't say cross-fed, and neither should any of you when referring to the present concept).

The easiest would seem to be to use three mostly identical cores, with the outer cores each driving two engines in the inboard core, and shutting down those engines at separation.  Lets call that 11/5/11.  I think this also preserves the most likely engine-out capability through the flight profile.  It may not provide the best outcome as the center core may be overpowered and under-fueled after booster separation.

Does that work well?

With 5 engines running isn't there barely enough thrust to lift the core if it were pointing straight up? If it's pitched over at all, doesn't it require at least 6 or 7 engines to avoid losing altitude and/or speed? Plus, would there be any engine out capability at all for some time after staging?

I don't have any way to calculate take off trajectory trade offs, but wouldn't it be best to keep both climbing and accelerating?

[douglas100]

So, I think it could be like this: Feed valves are set as Jim depicted. Before lift-off the core tanks valves are closed and all 9 core engines are fed from both booster tanks.

This mode lasts until the booster tanks are near 10%, when the core tanks valves are switch to open, and the valves that communicate core turbopumps and booster tanks are closed.

After that core and boosters keep burning exclusively its own fuel until the booster tanks are depleted, triggering separation.

P.E. And no, three engines are not enough to keep FH in flight after separation.

This makes most sense to me. It is the simplest way of doing it.

The fact that we have 27 engines to play with makes it tempting to speculate which engines in the core would be fed from the boosters, and which wouldn't. Why would you want to consume any of the central core propellant during cross feed operations? Chef Pat's angel comment is spot on.

If anyone has concrete information that SpaceX is proposing to operate FH in a mode different from what Jim or charliem describe, would they please post it.

[pippin]

Does that work well?

Yes. You don't need a T/W of more than 1 after staging. Ariane 5 doesn't have it, too, and works very well indeed.

[kirghizstan]

probably not going to happen, but what about airstarting the middle 3 thus lifting off with 24 engines and not drawing from the center core until those engines are lit

[RocketmanUS]

So need to control the flow of propellant from each booster to core to keep a balance from each. If so can that be done with variable flow valves?

What for? Again: don't make it more complicated than it is, these are communicating vessels:
http://en.wikipedia.org/wiki/Communicating_vessels

flow-control valves would actually only make things worse.

Sometimes physics can be so simple

Only if the pressure in the boosters are kept equal

So the pressure in the boosters would possible keep things balanced?

[Robotbeat]

probably not going to happen, but what about airstarting the middle 3 thus lifting off with 24 engines and not drawing from the center core until those engines are lit

That is, of course, possible, though it'd add some complexity. Usually, you do that if the center stages have a bigger nozzle on them and are optimized for high altitude, but there's hardly any room down there for bigger nozzles, so that doesn't seem terribly likely to me.

More likely would be to start the center engines at full thrust on lift-off, then throttle them down before max-q and keep them throttled down  (though high enough throttle for good Isp) until right before the boosters stage off. It's hard to know without doing more detailed analysis what the optimum is if you take into account gravity losses. A compromise is probably likely.

But yeah, you don't technically need nearly as much T/W on an upper stage as you do on lift-off, and the core stage may be a little shorter, too, for all we know (and some of its propellant will have been used up already, making it significantly lighter). Besides that, rocket engines produce significantly more thrust at near-vacuum conditions.

[FinalFrontier]

Seems like this would be the preferred option of doing the cross feed to the side feed idea.

[FinalFrontier]

probably not going to happen, but what about airstarting the middle 3 thus lifting off with 24 engines and not drawing from the center core until those engines are lit

Not a chance.


I'd be worried simply on the basis of complexity doing something like that.....


That and it would add quite a bit of extra weight in the form of additional hardware, you would probably end up losing any performance gains you get out of it. 

[dcporter]

Whoa FH is gonna run on three engines after booster sep? Any chance you remember where you heard that?

No 18 Engines at start 3 engines on left side of core feed off left strap-on.  3 engines on right side of core feed off Right Strap on. (So each of the strap-on are actually feeding 12 engines instead of 9, and core is only feeding 3)

After low fuel on a strap-on Staging happens.  Now all 9 core engines are feeding off a mostly full Center Core (only fed 3 engines for as long as it takes 12 engines to suck down a stage).

Does that make since?

You describe what I understood to be happening. I must have understood RB.

[arnezami]

If anyone has concrete information that SpaceX is proposing to operate FH in a mode different from what Jim or charliem describe, would they please post it.

I just thought it would be helpful to post Elon's comments on the crossfeed:

the crossfeed between the outer cores to the center is pretty helpful. It gives roughly 20-30% more payload. And we have an advantage here in that, because there are nine engines, we can control flow to those engines individually. So the way we do crossfeed is to basicly draw propellant for the engines that are adjacent to the core from the... so the engines that.. so the center core engine is pulling from its side core, on either side. And then we can shut prevalves to those engines, so that its not sucking down from the side core. You can't do this if you've got one engine... you can do it if you have multiple/nine engines, it's a lot easier.

I've highlighted some things, which, I think, is causing the speculation about feeding less than all 9 engines.

Here the video (starts at 8:45):



I'm not sure how to interpret his comments, but these are the "facts" we have. It seems clear though that (prevalves to) individual engines can be controlled, not the flow from entire tanks. And I assume it's "easier" that way, because the valves are smaller. Not sure though.

[MP99]

arnezami,

Spent quite some time trying to find that yesterday without any success. Great job finding that.

Cheers, Martin

[douglas100]

I'm not sure how to interpret his comments, but these are the "facts" we have. It seems clear though that (prevalves to) individual engines can be controlled, not the flow from entire tanks. And I assume it's "easier" that way, because the valves are smaller. Not sure though.

I'm not sure how to interpret his comments either, after one listen. Thanks for posting that.

[Nascent Ascent]

So, if SpaceX is going to go with the reusable concept I am assuming that the outer boosters wouldn't be emptied but have enough fuel and oxidizer left in order to return?

If this is correct, would this change the side/cross feed strategy?

-NA

[arnezami]

I was trying to picture how the (O2 and Kerosine) ducts in the side cores would be "split" and be connected to the inner core. Not easy to imagine it.

Here is a pic of the current tank ducts: they are "flexible" somewhat. Two inlets (for tanking), 18 outlets, right? Maybe somebody else has ideas how this could work.

[Jim]

My idea would be to just add a duct to interconnect the "head" of the "octopus"

[Pedantic Twit]

Just a reminder of what the octopus looks (or looked) like without any legs. (from here)

[douglas100]

Regardless of how they carry out cross feed, it looks like that area will have to be considerably redesigned.

[Jim]

Yes, because there is only one "octopus".  The other tank uses direct feedlines.

[MP99]

I just thought it would be helpful to post Elon's comments on the crossfeed:

the crossfeed between the outer cores to the center is pretty helpful. It gives roughly 20-30% more payload. And we have an advantage here in that, because there are nine engines, we can control flow to those engines individually. So the way we do crossfeed is to basicly draw propellant for the engines that are adjacent to the core from the... so the engines that.. so the center core engine is pulling from its side core, on either side. And then we can shut prevalves to those engines, so that its not sucking down from the side core. You can't do this if you've got one engine... you can do it if you have multiple/nine engines, it's a lot easier.

<snip>

I'm not sure how to interpret his comments, but these are the "facts" we have. It seems clear though that (prevalves to) individual engines can be controlled, not the flow from entire tanks. And I assume it's "easier" that way, because the valves are smaller. Not sure though.

Assuming that the outriggers will each feed prop to three core engines, it seems that either their octopuses will need to have three additional outlets, or the three core-side feedlines would have increased diameter to feed two engines each. (Latter seems like it has the potential for unwanted interaction between adjacent core & outrigger engines).

Alternatively, if the core will transition to internal operation before the outriggers burnout, the crossfeed doesn't need to pull from the very bottom part of the tank (especially if the outriggers retain recovery prop). Perhaps the crossfeed could have it's own independent drain from the tanks.

cheers, Martin

[jimvela]

Regardless of how they carry out cross feed, it looks like that area will have to be considerably redesigned.

Some more thoughts about side-fed and cross-fed:

PT's picture shows the current core has one feed line that becomes a decapus with a head source, which presumably is feeds for nine engines plus a fill/drain interface.  The other is still a decapus, but the head is the tankage immediately above- no feed line.

There isn't a convenient place in that picture to attach feed line Ts and valves, which makes sense because it is mass optimized for use as a single stage.

One option is to just pay a mass penalty, lengthen both feeds slightly and install large Ts and valves as Jim suggests.  I still don't think the center core feeds can be "off" while in flight because of problems with flow when we want to slam those valves all the way open at booster separation.  This is probably the easiest and best option.

One alternate option I see for either side-feed or cross-feed use would be to change the design to be a hexadecapus.  The center core would have two pair of 3-into-1 Y with a valve facing the outrigger on each side.  The outrigger would have a corresponding pair of 3-into-1 Y with a valve facing the core.  The problem with this approach is there isn't anywhere to put a valve to limit flow from the center core.  Maybe that isn't really necessary if the core stage tank pressure is kept lower than the outriggers.

For a side-fed application, I still think 11/5/11 is the configuration that makes the most sense. 

With either side-fed or cross-fed, there have to be thrust structure mods because plumbing needs to connect the stages to the engines being side-fed or for core-to-core cross-feed.  With thrust structure mods, modify the "talons" and push the outer four merlins in the center stage away from the centerline. 

For side-fed, this would allow the outer two engines of the core to be a part of the thrust structure that hangs under the outrigger. 

This has a mass penalty, but it also means that the four outside engines could separate away with the outriggers.  This probably isn't great for stage reuse as it would be a weirdly shaped stage coming back.

Now consider that if you use side-fed engines, the outriggers still have extra legs from the hexadecapus that could be plumbed over to the center core.

Again, if we keep the outriggers at a higher tank pressure than the center core, we again reduce draw from the center core by keeping the outriggers at a higher tank pressure.  The outriggers would be pressure-referenced to each other.

in this configuration, we want to deliberately burn up some of the fuel in the center core as we have only a 5-engine thrust structure in a "+" at booster sep.  Perhaps with the added space of the modified thrust structure, even use a slightly more vacuum optimized nozzle extension on those 5 engines.

I'm sure this thinking is full of holes, not the least of which is it isn't simple.  It's a great mental exercise, though.

[arnezami]

One option is to just pay a mass penalty, lengthen both feeds slightly and install large Ts and valves as Jim suggests.  I still don't think the center core feeds can be "off" while in flight because of problems with flow when we want to slam those valves all the way open at booster separation.  This is probably the easiest and best option.

I was thinking of not using Ts, but changing it from a 20 to 26 (or 28) tubes configuration. So a somewhat bigger octopus, with 3-4 more tubes attached to it. And on the tank itself (right above the octopus) a bigger "ring", with 3-4 more tubes attached also. This is only for the side cores of course. The center core octopus and tank attachments would stay the same.

Then two valves (or "shutters"?) for each cross tube would have to be added between the cores. One for sealing the center core tube, one for sealing the side core line (if you want to restart it for re-usability).

Does that make sense?

[go4mars]

Elon seems to imply that the fuel comes from the "adjacent engine". 

Would it be practical to have each octopus mouth feed 2 pipes on the side core? 

Multi-splitter like the octopus, but with less outputs:

Each pipe coming off this "Hexapus" is big enough to feed 2 engines.  So on the side boosters, 6 lines come off from each Hexapus.  Each of these lines has a splitter (like a Y-connector for your garden hose).  The 6 outermost engines on a booster are fed from 3 split lines.  The three engines on the side facing the middle core are fed in the same way, but (like the garden hose Y) there is a valve on the pipe going to a quick-disconnect system on the outer edge from the other split inthe Y's on this side.  These hook up to counterparts on the middle core to feed the "adjacent" engines. 

In the middle core, there is the traditional F9 style "octopus", but with valved branches facing the side cores on the feedlines for the 6 relevant engines. 

SpaceX had mentioned at one point that they hope to be able to completely fuel-up for a launch in less than 15 minutes IIRC.  Having these big fat valves available might help speed that up. 

Criticism welcome.     Thanks.

[RDoc]

Wouldn't it be simpler to just run a single line each RP1 and LOX from the outboard tanks to the core, then distribute the fuel/lox from there? That would only require two couplings rather than six on each side. Perhaps with valving, the current fill/empty lines could be expanded a bit and used?

Looking at the octopus (decapus?) jpg, and guessing the LOX tubes are about 8" in diameter, and a burn rate of about 4 cu ft/sec, I come up with a flow velocity of around 12 ft/sec. I would have guessed faster, but to keep the resistance low I suppose it's possible. Is that kind of velocity believable?

If so, the instabilities in switching from one feed to another wouldn't be that extreme as there would only be a few hundred lbs moving relatively slowly in the lines. Perhaps more importantly, I wouldn't think it very hard to model and test.

[jimvela]

If so, the instabilities in switching from one feed to another wouldn't be that extreme as there would only be a few hundred lbs moving relatively slowly in the lines. Perhaps more importantly, I wouldn't think it very hard to model and test.

Answer this: what happens to a cryogen in a feed line while the valve is closed and it sits there warm soaking?  When that valve is opened, what does the inlet side see from the valve?

[RDoc]

Answer this: what happens to a cryogen in a feed line while the valve is closed and it sits there warm soaking?  When that valve is opened, what does the inlet side see from the valve?

Well if I were designing it, I'd look into allowing a small flow through the valve to keep everything cold.

[Robotbeat]

Just a reminder of what the octopus looks (or looked) like without any legs. (from here)

Thank you for posting that.

[RDoc]

Answer this: what happens to a cryogen in a feed line while the valve is closed and it sits there warm soaking?  When that valve is opened, what does the inlet side see from the valve?

Actually, thinking about this a bit more, if the valve was a single Y valve (open before close) located right at or just astern of the octopus, I'm not sure it would be subjected to significant warm soaking.

First, the flow from the running tank would chill the entire valve, and secondly, convection back up into the octopus body would tend to carry away warmed LOX. The acceleration would help the convection effect and it wouldn't be far to a large reservoir of cold LOX.

Again, this would need modeling and testing, but it still doesn't sound very hard to engineer.

[sdsds]

I wouldn't think it very hard to model and test.

Is there any indication SpaceX has ever (on a test stand) switched propellant tanks feeding an engine during a burn?

[RDoc]

Is there any indication SpaceX has ever (on a test stand) switched propellant tanks feeding an engine during a burn?

I'm unaware of any cross/side feeding live fire tests at all. Have there been any?

In any case, wouldn't they try it on a test bench without an engine, etc. first? If my estimates of flow rates are anywhere near close, it doesn't sound very hard to do some modeling and then do a bench test to see how good the model is.

[MP99]

One option is to just pay a mass penalty, lengthen both feeds slightly and install large Ts and valves as Jim suggests.  I still don't think the center core feeds can be "off" while in flight because of problems with flow when we want to slam those valves all the way open at booster separation.  This is probably the easiest and best option.

I'm no expert by any means, but are these sort of valves the only ones that might be used?

ISTM that, if possible, it would make much more sense to gently shut down the side-feeds while (or after) opening up the feeds from the core. That would avoid shocks to the engine being fed.

Taken to an extreme, the RP-1 & O₂ feeds to each of the side-fed engines could be switched individually - twelve small (but slow) transients instead of a couple of big (but slow) ones.

cheers, Martin

[MP99]

If so, the instabilities in switching from one feed to another wouldn't be that extreme as there would only be a few hundred lbs moving relatively slowly in the lines. Perhaps more importantly, I wouldn't think it very hard to model and test.

Answer this: what happens to a cryogen in a feed line while the valve is closed and it sits there warm soaking?  When that valve is opened, what does the inlet side see from the valve?

Would a cryo valve have to be closed on the ground? Once closed in-flight, I don't think it would need to be opened again?

cheers, Martin

[RDoc]

Would a cryo valve have to be closed on the ground? Once closed in-flight, I don't think it would need to be opened again?

cheers, Martin

IFF some of the engines were to be switched from the side tanks to the core tanks (as opposed to being shut down) the LOX main tank feeds to them would start out shut on the ground, then opened in flight so there's the question of keeping them chilled.
 

[cuddihy]

Wouldn't it be simpler to just run a single line each RP1 and LOX from the outboard tanks to the core, then distribute the fuel/lox from there? That would only require two couplings rather than six on each side. Perhaps with valving, the current fill/empty lines could be expanded a bit and used?

Looking at the octopus (decapus?) jpg, and guessing the LOX tubes are about 8" in diameter, and a burn rate of about 4 cu ft/sec, I come up with a flow velocity of around 12 ft/sec. I would have guessed faster, but to keep the resistance low I suppose it's possible. Is that kind of velocity believable?

If so, the instabilities in switching from one feed to another wouldn't be that extreme as there would only be a few hundred lbs moving relatively slowly in the lines. Perhaps more importantly, I wouldn't think it very hard to model and test.

Simpler conceptually, but much more expensive. And this discussion has pretty much answered my question for me. From looking at the pictures above, it's evident that although it's not that big a change to go to the 12/3/12 configuration, because it looks like it can be done by splitting the individual smaller feed lines, and without major mods to the tank feedthrough, no added Kerosene plenum, and minor mods to the thrust structure. Also it allows you to keep a lot more commonality between the core and side feed systems and thrust structures.

On the other hand if you go to main feed line cross connects you now have to extend the feedlines, create a kerosene plenum to allow for smooth flow through the cross connects, extend the thrust structure to accomodate the big feed line cross connects, and you have major differences between the side booster feed areas (with two feed cross connects each) versus the four in the core. Add instability questions from big feed valves shutting and it's easy to see why they wouldn't do it for just another 5 or 8% boost in payload when it's already higher than the top of the market.

[simonbp]

I might have missed it in all this discussion, but if the center tank is 1/4 full at separation assuming everything throttles the same, couldn't you throttle down the "center three" before sep to lower their flow rate, allowing the center to be more than 3/4 full after sep?

[Robotbeat]

I might have missed it in all this discussion, but if the center tank is 1/4 full at separation assuming everything throttles the same, couldn't you throttle down the "center three" before sep to lower their flow rate, allowing the center to be more than 3/4 full after sep?

The other way...

It's normally supposed to be ~3/4 full after sep because the outer tanks effectively each feed three of the core's engines. Throttling would allow you to increase that to, say, ~7/8s full.

[simonbp]

Yes, that's what I meant. Which would conform better to Shotwell's statements at Space Access last year that the center would "nearly full" at staging.

[baldusi]

What if the booster's tanks were held at higher pressure than the core tank? And in the feeds to the outboard engines with a Y connection. On the core side you put a poppet valve from the core tank and on the Booster side you put a one-way valve. Thus, as long as you have the booster feeding at higher pressure, it would keep the poppet shut by pressure differential and keep the one way open, thus feeding from the side. Before staging you simply lower the pressure on the boosters side, thus, the popper valve would open and the one way would shut.
If you keep the poppets on the octopus, for example, there would be no sublimation problems. But the one way might generate some cavitation.
May be they can design sort of a 5 way valve, with the actuation activated by pressure differential. In fact, you could handle that with the pressurizations system directly.

[kevin-rf]

What if the booster's tanks were held at higher pressure than the core tank? And in the feeds to the outboard engines with a Y connection. On the core side you put a poppet valve from the core tank and on the Booster side you put a one-way valve. Thus, as long as you have the booster feeding at higher pressure, it would keep the poppet shut by pressure differential and keep the one way open, thus feeding from the side. Before staging you simply lower the pressure on the boosters side, thus, the popper valve would open and the one way would shut.
If you keep the poppets on the octopus, for example, there would be no sublimation problems. But the one way might generate some cavitation.
May be they can design sort of a 5 way valve, with the actuation activated by pressure differential. In fact, you could handle that with the pressurizations system directly.

What is the head pressure on 180 feet of LOX at 4g? That is what you would need to increase the pressure in the outriggers by. Meaning beefed up heaver less mass efficient structures.

And how will the turbo pump react when the pressure suddenly drops and the center tanks begins feeding?

[baldusi]

What is the head pressure on 180 feet of LOX at 4g? That is what you would need to increase the pressure in the outriggers by. Meaning beefed up heaver less mass efficient structures.

I understand that the old tanks where pressurized at 50psi (3.4kg/cm˛) for flight stress support. And LOX is 1.141g/cmł. At 54.8m of column that's 7.73kg/cm˛ at Earth's gravity. Plus the 4g, that's 30.92 kg/cm˛+7.73kg/cm˛+3.4kg/cm˛=42.03kg/cm=617psi at the head, right?
You'll have to help me here, but isn't that the same pressure the Boosters have to support early in the flight?
Another issue is that poppet valves can be balanced easily. You don't need higher pressure on one side. You need higher than a threshold pressure on such side. This is a basic exercise of surface differentials. In fact, that's where it sort of becomes a 3/4/5 way valve.

And how will the turbo pump react when the pressure suddenly drops and the center tanks begins feeding?

The pressure drop of the boosters tanks wouldn't be affected. It would be falling either way. What you get when you open the poppet is a pressure spike. Beside the problem of the initial hammer punch (that's how you call it?), which can be solved with the poppet's opening speed, you'd get back to the situation when the boosters where at whatever the core is at staging (I guess around 80%?)

[RDoc]

Wouldn't it be simpler to just run a single line each RP1 and LOX from the outboard tanks to the core, then distribute the fuel/lox from there? That would only require two couplings rather than six on each side. Perhaps with valving, the current fill/empty lines could be expanded a bit and used?

Simpler conceptually, but much more expensive.

That certainly doesn't seem obvious.

Assuming the fill/empty lines can be enlarged and are used for the cross feed to 3 engines in the core, then single cross feed for fuel and LOX requires for each booster:

Probably enlarging the 2 fill/empty lines
2 T's
2 line separation system
2 1x3 manifolds in the core
Space and support for 2 lines

Running separate lines using the booster feed lines requires

Probably enlarging 6 feed lines
6 1x2 manifolds
6 line separation system
Space and support for 6 lines

[oldAtlas_Eguy]

The technical problems that crossfeed design must overcome as outlined so far in this thread: (If I have missed one or more please indicate what they are?)

1) Difference in pressure between the booster and center cores on the operation of the crossfeed engine during changeover. Can be tested on the ground with a single engine, although a problem, solutions can be thoroughly tested on the ground with modifications to the M1D single engine test chamber. A secondary feed with a pressure reducer can easily simulate this condition.

2) Booster separation while center core engines still firing. Atlas heritage propellant quick disconnect would be a starting point. But having six propellant lines coming from the booster would be a higher risk than a larger two prop lines setup during booster separation. Some testing can be done on the ground but the real test cannot be simulated on the ground, it takes an actual flight to prove out the design, too many small variables involved with separation. Test jigs can reduce the risk but not resolve all risks.

3) Testing cross effects of propellant flow changes on the 9 engines of the center core. This can mostly be tested on the ground but at only 1 g and not the g level seen in flight.

4) The boosters will have a different “octopus” propellant feed structure than the center core (the center core would have a normal F9 “octupus”).  Simulation of propellant feed effects for the booster can be done only at 1g on the ground and not at the flight g levels.

A summary list of the technical challenges and whether ground testing can thoroughly test or can only mostly be tested except for flight conditions such as g levels, should be a task of this thread so that SpaceX test facilities can be evaluated for whether what technical challenges posed by the crossfeed as proposed can possibly be addressed by these existing or planned facilities like the new tri core hot fire facility at McGregor.

[Dmitry_V_home]

Angara-5 cross-feed version 

[MP99]

Answer this: what happens to a cryogen in a feed line while the valve is closed and it sits there warm soaking?  When that valve is opened, what does the inlet side see from the valve?

Well if I were designing it, I'd look into allowing a small flow through the valve to keep everything cold.

Would a cryo valve have to be closed on the ground? Once closed in-flight, I don't think it would need to be opened again?

IFF some of the engines were to be switched from the side tanks to the core tanks (as opposed to being shut down) the LOX main tank feeds to them would start out shut on the ground, then opened in flight so there's the question of keeping them chilled.

Ah, of course.

I presume the main tank feeds could be left open until seconds before launch, so presumably can be kept chilled at this phase? However, I can see that could be a problem in the (presumably hot) environment during cross-fed flight. Re "small flow" - how much flow is required to maintain cryo temps?

cheers, Martin

[Jim]

Angara-5 cross-feed version 

That shows that the boosters have to be at a higher pressure than the core.

[cuddihy]

Wouldn't it be simpler to just run a single line each RP1 and LOX from the outboard tanks to the core, then distribute the fuel/lox from there? That would only require two couplings rather than six on each side. Perhaps with valving, the current fill/empty lines could be expanded a bit and used?

Simpler conceptually, but much more expensive.

That certainly doesn't seem obvious.

Assuming the fill/empty lines can be enlarged and are used for the cross feed to 3 engines in the core, then single cross feed for fuel and LOX requires for each booster:

Probably enlarging the 2 fill/empty lines
2 T's
2 line separation system
2 1x3 manifolds in the core
Space and support for 2 lines

and how do you go about enlargening the drain/fill line to that extent (take a look, the other feed lines are the same size and are in the way). In effect you would be creating a new RP1 plenum. Not so with LOX because of the big downcomer. But again you still would end up having to extend the thrust structure.

Running separate lines using the booster feed lines requires

Probably enlarging 6 feed lines
6 1x2 manifolds
6 line separation system
Space and support for 6 lines

[Dmitry_V_home]

That shows that the boosters have to be at a higher pressure than the core.

Not necessarily.

[RDoc]

and how do you go about enlargening the drain/fill line to that extent (take a look, the other feed lines are the same size and are in the way). In effect you would be creating a new RP1 plenum. Not so with LOX because of the big downcomer. But again you still would end up having to extend the thrust structure.

It appears that the flow velocity is pretty low, so taking a SWAG I'd say the flow rate was pretty proportional to the area. So, assuming the fill/empty lines are the same as the downcomers, to increase the flow by a factor of 3 would require an increase in diameter of one line by 1.73, so an increase of .73. To increase the area of 3 lines by a factor of two would be 3 * 1.41 = 4.2 so an increase of 1.2. For 10 inch lines, this would be 7.3 inches total, .73 inches per line vs 12 inches total, or 1.2 inches per line.

The fuel lines probably aren't a big deal since there seems to be plenty of room to tighten up the spacing. I don't see how this would be particularly expensive in either case.

The LOX line is a different matter though I'd think. My guess is the feed/fill line is the one on the lower left of the aft-most hub of the octopus. There seems to be plenty of room to increase that. However, to increase 3 LOX lines would require either re-spacing the forward LOX lines which seem pretty tightly spaced, or increasing the diameter of the plenum. Changing the diameter would change the geometry and might require increasing the fuel header circle diameter.

There's no doubt that any cross/side feed system would require changes to both the fuel and LOX lines, but I doubt these are off the shelf items that are bought at a bulk discount. It does seem though that enlarging a single line for each would be somewhat less disruptive to the existing geometries than doing 3 each. I don't see where the significant increase in cost would lie in the smaller change vs the larger.

It's certainly also true that either design would require changing the thrust structure, but again, accommodating two larger lines isn't obviously a lot more expensive than 6 smaller lines.

[Jim]

That shows that the boosters have to be at a higher pressure than the core.

Not necessarily.

Huh?  if not, the core will use its own propellants

[go4mars]

That shows that the boosters have to be at a higher pressure than the core.

Not necessarily.

Huh?  if not, the core will use its own propellants

I don't read Russian, but is that a valve labelled on the middle tank (just below where the green feedlines come over)? 

[aga]

That shows that the boosters have to be at a higher pressure than the core.

Not necessarily.

Huh?  if not, the core will use its own propellants

I don't read Russian, but is that a valve labelled on the middle tank (just below where the green feedlines come over)? 

if i got it right, and according to wiki, the thing in the middle is in english called "butterfly valve"... sorry, not familiar with english terminology in this area

[go4mars]

That shows that the boosters have to be at a higher pressure than the core.

Not necessarily.

Huh?  if not, the core will use its own propellants

I don't read Russian, but is that a valve labelled on the middle tank (just below where the green feedlines come over)? 

if i got it right, and according to wiki, the thing in the middle is in english called "butterfly valve"... sorry, not familiar with english terminology in this area

Thanks.  That sounds right.  I assume it allows flow from the boosters and disallows flow from the middle core until the boosters are empty and gone (then it folds its wings and opens its back). 

[Jim]

The fuel tank on the core does not have an equivilent valve

[Dmitry_V_home]

I assume it allows flow from the boosters and disallows flow from the middle core until the boosters are empty and gone (then it folds its wings and opens its back). 

Exactly!

[Dmitry_V_home]

One more possible scheme of the cross-feed rocket:

[RDoc]

Am I correct in thinking that the biggest mechanical issue with cross/side feed would be the umbilical separation systems between the boosters and core?

The other components seem pretty standard, manifolds, valves and pipes. Even the systems for dropping the boosters has a lot of industry precedents, although perhaps SpaceX might use hydraulics to lock and release the boosters?

Presumably FH would use something like the Shuttle external tank system couplings with explosive bolts, low pressure shutoff valves, coupling retractors, and fairing doors on both the core and boosters. That all seems like a lot of complex moving parts, but perhaps it could be simplified with an integrated cam/lever mechanism actuated by the hydraulic system. I wouldn't think they'd want to have any pieces jettisoned at that point that could hit the core or boosters as they fell away, although probably some RP1 and LOX would be spilled.

In any case it seems a bit tricky to get right and reliable.

[Jim]

Am I correct in thinking that the biggest mechanical issue with cross/side feed would be the umbilical separation systems between the boosters and core?

The other components seem pretty standard, manifolds, valves and pipes. Even the systems for dropping the boosters has a lot of industry precedents, although perhaps SpaceX might use hydraulics to lock and release the boosters?

Presumably FH would use something like the Shuttle external tank system couplings with explosive bolts, low pressure shutoff valves, coupling retractors, and fairing doors on both the core and boosters. That all seems like a lot of complex moving parts, but perhaps it could be simplified with an integrated cam/lever mechanism actuated by the hydraulic system. I wouldn't think they'd want to have any pieces jettisoned at that point that could hit the core or boosters as they fell away, although probably some RP1 and LOX would be spilled.

In any case it seems a bit tricky to get right and reliable.

no, see heritage Atlas

[RDoc]

no, see heritage Atlas

That seems like a very different system.

[Jim]

no, see heritage Atlas

That seems like a very different system.

No, it would have similar disconnects.

[Jim]

1.  The other components seem pretty standard, manifolds, valves and pipes. Even the systems for dropping the boosters has a lot of industry precedents, although perhaps SpaceX might use hydraulics to lock and release the boosters?

2.  Presumably FH would use something like the Shuttle external tank system couplings with explosive bolts, low pressure shutoff valves, coupling retractors, and fairing doors on both the core and boosters. That all seems like a lot of complex moving parts, but perhaps it could be simplified with an integrated cam/lever mechanism actuated by the hydraulic system. I wouldn't think they'd want to have any pieces jettisoned at that point that could hit the core or boosters as they fell away, although probably some RP1 and LOX would be spilled.

In any case it seems a bit tricky to get right and reliable.

1.  no, too complex and unneeded.  See Titan solids and Delta IV boosters.  Explosive nuts are all that is needed with some sep motors

2.  Again, no doors or hydraulics, too complex and unneeded.  No coupling retraction needed.

See heritage Atlas.  LOX and RP-1 disconnects.  Separation is done after engine cutoff (no flow conditions).

[go4mars]

See heritage Atlas.  LOX and RP-1 disconnects.  Separation is done after engine cutoff (no flow conditions).

So a re-light is required for the boost-back phase?  I guess that could be lower risk.  I was picturing only 6 engines shutting off on the outside cores...3 just keep firing. 

These 1D's are made for re-starting easily anyways right?  ?     

?

[Jim]

See heritage Atlas.  LOX and RP-1 disconnects.  Separation is done after engine cutoff (no flow conditions).

So a re-light is required for the boost-back phase? 

What boost-back phase?

[go4mars]

What boost-back phase?

The boost-back phase after separation of course! 

Yes I realize that there will be a few years of practice launches using expendible versions. I doubt they are designing with obviously non-reusable solutions for important components on the first few iterations though.   But you would know better than I. 


What boost-back phase? 
For example:

These quotes, in the context of this article, makes it pretty clear imo.
http://www.popularmechanics.com/science/space/rockets/elon-musk-on-spacexs-reusable-rocket-plans-6653023

"Falcon Heavy is particularly amenable to reuse of the first stage—the two outer cores in particular, because they separate at a much lower velocity than the center one, being dropped off early in the flight."  -Elon


"Multiple flights per day for first stage and side boosters," -Elon

[Liryc]

Hi everyone,

not sure if that's the right place to ask.. but I'll give it a try.

I'm trying to figure out the mechanical system used for Falcon's quick-disconnect ssystem...
In this page we can se the cryogenic plug disconnects when the launcher lifts-off...

But when I look at other cryogenic feeding lines (like European or some in the US (Delta IV Heavy)) they open shortly before lauch.
This seems to be quite constraintful as you have to get back up there to reconnect the line if launch is aborted after separation..

Any idea on how it's beeing made ?
Why don't we do it on all our launchers ?

Thanks !
Cyril

[Jim]

But when I look at other cryogenic feeding lines (like European or some in the US (Delta IV Heavy)) they open shortly before lauch.

No, Atlas and Delta disconnect fill and drain lines at liftoff and not before.

Atlas only disconnects a He purge line before liftoff

[Liryc]

Ok, thanks for the correction.
I'll try to find some info on it.

[fatjohn1408]

Cool thread, learned that alot I took for granted about this subject was horribly wrong.

Most interesting post in the thread is this one, why aren't there more replies? Is this not feasible and superior to all other configurations proposed?

I have my own theory of how cross-feed works, made from the little info we have.

Spacex's website says that FH's mass and thrust at lift off is 1,400 mt and 3,800,000 lbf (1,724 mtf) (22 Merlin-1D at full thrust is the absolute minimum to elevate 1,400 mt).

It also says: "Propellant cross-feeding leaves the center core still carrying the majority of its propellant after the side boosters separate".

Can't find where I saw it but I remember to have read about a 90% propellant remaining in the core tanks right after boosters cut off.

In the video looks like all 27 engines are started at the same time, and we know that a Merlin-1D can only be throttled down to a 70% of its nominal thrust.

Easy to calculate that even with 3 engines at 70% more than a 10% of the core fuel would be gone when the booster tanks reach the empty mark.

So, I think it could be like this: Feed valves are set as Jim depicted. Before lift-off the core tanks valves are closed and all 9 core engines are fed from both booster tanks.

This mode lasts until the booster tanks are near 10%, when the core tanks valves are switch to open, and the valves that communicate core turbopumps and booster tanks are closed.

After that core and boosters keep burning exclusively its own fuel until the booster tanks are depleted, triggering separation.

P.E. And no, three engines are not enough to keep FH in flight after separation.

[Idiomatic]

Explosive bolts could possibly be skipped simply because SpaceX has a history of doing so. There are a number of events where normally explosive bolts would be used and SpaceX has chosen not to. (I still think they will go explosive though)

[dwightlooi]

I really don't know where the only whole concept of side boosters feeding 3 additional engines each came from. I wasn't able to find any SpaceX references to it formally or informally. The only thing that seems to pop up is Musk saying that because they have a large number of small engines, it is possible for an engine to be individually connected to and draw propellants from adjacent engine(s). And, that this makes crossfeeding easier to accomplish -- presumably because same interconnects and small valves are more easier to deal wit that huge pipes and gigantic valves.

The math that adds up to 53 metric tons to 7.8km/s with an Isp of ~310 secs dictates that the FH will probably run all its engines from the side boosters until moments before separation. The only reason the center core is not 100% full at separation is probably from the fact that you CANNOT ever have a propellant switch over at separation, it'll have to happen before that. The logical thing to do will be for the vehicle to switch from drawing fuel solely from the side boosters to three unconnected boosters each feeding their own 9 engines during the last 10~15 seconds of the side booster burn. It'll probably be a rotary valve rather than two on-off valves doing the work with each connection. There will be zero transient shock in the fluid because the transition is gradual and every bit of flow reduction from the side tanks is exactly matched by a flow increase from the center tank at all times.

The last 5~10 seconds will see essentially three independent F9 cores with no crossflow. At separation, flow would have stopped a while ago and the interconnecting pipes would have been purged so it'll be a dry separation. When the side boosters go MECO and fall away the center booster will have consumed about 5~10% of its fuel only from the fact that it had been operating with partial to no cross flow during the last 15 seconds prior to separation.

[SpacexULA]

Elon Mentioned it in one of 1000 videos.. heck if I remember which one.

[Jim]

I really don't know where the only whole concept of side boosters feeding 3 additional engines each came from. I wasn't able to find any SpaceX references to it formally or informally. The only thing that seems to pop up is Musk saying that because they have a large number of small engines, it is possible for an engine to be individually connected to and draw propellants from adjacent engine(s). And, that this makes crossfeeding easier to accomplish -- presumably because same interconnects and small valves are more easier to deal wit that huge pipes and gigantic valves.

The math that adds up to 53 metric tons to 7.8km/s with an Isp of ~310 secs dictates that the FH will probably run all its engines from the side boosters until moments before separation. The only reason the center core is not 100% full at separation is probably from the fact that you CANNOT ever have a propellant switch over at separation, it'll have to happen before that. The logical thing to do will be for the vehicle to switch from drawing fuel solely from the side boosters to three unconnected boosters each feeding their own 9 engines during the last 10~15 seconds of the side booster burn. It'll probably be a rotary valve rather than two on-off valves doing the work with each connection. There will be zero transient shock in the fluid because the transition is gradual and every bit of flow reduction from the side tanks is exactly matched by a flow increase from the center tank at all times.

The last 5~10 seconds will see essentially three independent F9 cores with no crossflow. At separation, flow would have stopped a while ago and the interconnecting pipes would have been purged so it'll be a dry separation. When the side boosters go MECO and fall away the center booster will have consumed about 5~10% of its fuel only from the fact that it had been operating with partial to no cross flow during the last 15 seconds prior to separation.

Purging the lines for dry separation is not needed, see heritage Atlas
Same goes for rotary valves

[modemeagle]

Here is how I handle cross feed with my simulation (not SpaceX approved)

1.  cross feed is 6 of 9 engines
Elon said during a Q&A that each side feeds the engines on that side, not sure whether it means 2,4,6 or 8 cross feeding.  In the past I found better payload with 6 vs all 9, I will run the numbers again and will edit if I find differently.
2.  cross feed ends when booster propellant mass reaches 5% (3.30 seconds before booster MECO)
3.  booster MECO is at 1.2 seconds of propellant remaining

When SpaceX releases the real numbers then I will adjust my simulation accordingly.

[kirghizstan]

Here is how I handle cross feed with my simulation (not SpaceX approved)

1.  cross feed is 6 of 9 engines
Elon said during a Q&A that each side feeds the engines on that side, not sure whether it means 2,4,6 or 8 cross feeding.  In the past I found better payload with 6 vs all 9, I will run the numbers again and will edit if I find differently.
2.  cross feed ends when booster propellant mass reaches 5% (3.30 seconds before booster MECO)
3.  booster MECO is at 1.2 seconds of propellant remaining

When SpaceX releases the real numbers then I will adjust my simulation accordingly.

Assuming f9 v1.1 what about cross feeding 8, what is the performance of that?

[modemeagle]

Here is how I handle cross feed with my simulation (not SpaceX approved)

1.  cross feed is 6 of 9 engines
Elon said during a Q&A that each side feeds the engines on that side, not sure whether it means 2,4,6 or 8 cross feeding.  In the past I found better payload with 6 vs all 9, I will run the numbers again and will edit if I find differently.
2.  cross feed ends when booster propellant mass reaches 5% (3.30 seconds before booster MECO)
3.  booster MECO is at 1.2 seconds of propellant remaining

When SpaceX releases the real numbers then I will adjust my simulation accordingly.

Assuming f9 v1.1 what about cross feeding 8, what is the performance of that?

This is the FH with the F9V1.1 core.
8 gives you 6.39% residual vs 7.42% with 6.
9 gives 5.62%

[MP99]

Here is how I handle cross feed with my simulation (not SpaceX approved)
...
2.  cross feed ends when booster propellant mass reaches 5% (3.30 seconds before booster MECO)

3.  booster MECO is at 1.2 seconds of propellant remaining

Booster MECO = BECO?



ISTM the two boosters won't hit their cutoffs at exactly the same point, due to natural variations between engines. Obviously, simplest is to cutoff both when the faster-burning booster reaches it's limit, and I'd think this would have a pretty small hit to performance.

However, I understand the single-stick F9 uses mixture ratio management to ensure one propellant isn't exhausted before the other - I believe there was a callout to that effect on F9 #003 launch. It would seem to be a natural extension to try to coordinate exhaustion of the two boosters - ie BECO when the slower-consuming booster hits it's limit. There seem to be two obvious ways to achieve that:-

1) Minor throttle back on the side that is consuming faster (perhaps just on a single engine, given this is likely to be pretty small). This takes a small hit on T/W, but the vehicle will remain in balance as the boosters empty at the same rate. Any engine on the vehicle centre-line could be throttled - whichever trims out the net thrust best, ie minimises engine array gimballing. Booster outboard engine if that booster is the "hotter", or the cross-fed core engine if thrust is similar but consumption is higher.

2) The side that is consuming faster switches off cross-feed slightly earlier. Maximises T/W, but the cores would have a growing mass & T/W imbalance until the first booster ends cross-feed, with the imbalance timed to zero out just as the second booster cuts it's cross-feed.

Both have the same burn time - ie until the slower-consuming booster is drained instead of the faster.

cheers, Martin

[modemeagle]

Here is how I handle cross feed with my simulation (not SpaceX approved)
...
2.  cross feed ends when booster propellant mass reaches 5% (3.30 seconds before booster MECO)

3.  booster MECO is at 1.2 seconds of propellant remaining

Booster MECO = BECO?



ISTM the two boosters won't hit their cutoffs at exactly the same point, due to natural variations between engines. Obviously, simplest is to cutoff both when the faster-burning booster reaches it's limit, and I'd think this would have a pretty small hit to performance.

However, I understand the single-stick F9 uses mixture ratio management to ensure one propellant isn't exhausted before the other - I believe there was a callout to that effect on F9 #003 launch. It would seem to be a natural extension to try to coordinate exhaustion of the two boosters - ie BECO when the slower-consuming booster hits it's limit. There seem to be two obvious ways to achieve that:-

1) Minor throttle back on the side that is consuming faster (perhaps just on a single engine, given this is likely to be pretty small). This takes a small hit on T/W, but the vehicle will remain in balance as the boosters empty at the same rate. Any engine on the vehicle centre-line could be throttled - whichever trims out the net thrust best, ie minimises engine array gimballing. Booster outboard engine if that booster is the "hotter", or the cross-fed core engine if thrust is similar but consumption is higher.

2) The side that is consuming faster switches off cross-feed slightly earlier. Maximises T/W, but the cores would have a growing mass & T/W imbalance until the first booster ends cross-feed, with the imbalance timed to zero out just as the second booster cuts it's cross-feed.

Both have the same burn time - ie until the slower-consuming booster is drained instead of the faster.

cheers, Martin

3.  Regulation of flow rates during cross feed to make sure each booster is emptying at the same rate. (minimum performance loss)

4.  Early BECO. When one booster is depleted, both are shutdown and jettisoned. (some performance loss)

[MP99]

Here is how I handle cross feed with my simulation (not SpaceX approved)
...
2.  cross feed ends when booster propellant mass reaches 5% (3.30 seconds before booster MECO)

3.  booster MECO is at 1.2 seconds of propellant remaining

Booster MECO = BECO?



ISTM the two boosters won't hit their cutoffs at exactly the same point, due to natural variations between engines. Obviously, simplest is to cutoff both when the faster-burning booster reaches it's limit, and I'd think this would have a pretty small hit to performance.

However, I understand the single-stick F9 uses mixture ratio management to ensure one propellant isn't exhausted before the other - I believe there was a callout to that effect on F9 #003 launch. It would seem to be a natural extension to try to coordinate exhaustion of the two boosters - ie BECO when the slower-consuming booster hits it's limit. There seem to be two obvious ways to achieve that:-

1) Minor throttle back on the side that is consuming faster (perhaps just on a single engine, given this is likely to be pretty small). This takes a small hit on T/W, but the vehicle will remain in balance as the boosters empty at the same rate. Any engine on the vehicle centre-line could be throttled - whichever trims out the net thrust best, ie minimises engine array gimballing. Booster outboard engine if that booster is the "hotter", or the cross-fed core engine if thrust is similar but consumption is higher.

2) The side that is consuming faster switches off cross-feed slightly earlier. Maximises T/W, but the cores would have a growing mass & T/W imbalance until the first booster ends cross-feed, with the imbalance timed to zero out just as the second booster cuts it's cross-feed.

Both have the same burn time - ie until the slower-consuming booster is drained instead of the faster.

cheers, Martin

3.  Regulation of flow rates during cross feed to make sure each booster is emptying at the same rate. (minimum performance loss)

(3) is a subset of my (1), presuming the cross-fed engines are independent of the core as expected. Cutting flow rate on one cross-feed will throttle at least one of the core engines. But the thrust may be better balanced by throttling an outside engine on the booster instead of a core engine.

4.  Early BECO. When one booster is depleted, both are shutdown and jettisoned. (some performance loss)

Obviously, simplest is to cutoff both when the faster-burning booster reaches it's limit, and I'd think this would have a pretty small hit to performance.

cheers, Martin

[aero]

This is not directly speculation on how SpaceX is building their crossfeed system, but rather points to the losses resulting from suboptimal crossfeed implementations.

Can anyone point to the dirivation of the rocket equation for a triamese rocket? That is the Falcon Heavy configuration, and Wikipedia says this.

Bimese & Triamese (Crossfeed)
 
Two or three similar stages are stacked side by side, and burn in parallel. Using crossfeed, the fuel tanks of the orbital stage are kept full, while the tank(s) in the booster stage(s) are used to run engines in the booster stage(s) and orbital stage. Once the boosters run dry, they are ejected, and (typically) glide back to a landing. The advantage to this is that the mass ratios of the individual stages is vastly reduced due to the way cross feed modifies the rocket equation. Isp*g*ln(2MR^2/MR+1) & Isp*g*ln(3MR^2/MR+2) respectively. With hydrogen engines, a triamese only needs an MR of 5, as opposed to an MR of 10 for a single stage equivalent vehicle.
 
A criticism of this approach is that designing separate orbiter and boosters, or a single vehicle that could do both, would compromise performance, safety, and possible cost savings. Compromising maximum performance to reduce cargo cost however, is the POINT of the triamese approach. Stacking two or three winged vehicles can also be challenging. Optimistically, the lower mass ratios would translate to lower overall R&D costs, even if two different stage designs. While many aerospace designs have successfully been modified far beyond the original designers intentions (Boeing's 747 is perhaps the best example) the slow and painful birth of the F-35 family demonstrates that it is not always a guarantee of such flexibility.
 
Crossfeed is to be an important part of SpaceX's Falcon Heavy - and one of the main reasons it will be able to lift ~5 times as much cargo to orbit as the standard Falcon 9.

http://en.wikipedia.org/wiki/Reusable_launch_system

I ran a spreadsheet on the FH using the Delta V equation above and derived some really large excess capacity for the vehicle. I know speculation is that FH will only achieve something like 40 to 50 tonnes to LEO, but my spreadsheet gave well over 53 tonnes using 9.8 km/s as the total required Delta V. Of course, that was for an expendable FH.

[MP99]

This is not directly speculation on how SpaceX is building their crossfeed system, but rather points to the losses resulting from suboptimal crossfeed implementations.

Can anyone point to the dirivation of the rocket equation for a triamese rocket? That is the Falcon Heavy configuration, and Wikipedia says this.

Bimese & Triamese (Crossfeed)
 
Two or three similar stages are stacked side by side, and burn in parallel. Using crossfeed, the fuel tanks of the orbital stage are kept full, while the tank(s) in the booster stage(s) are used to run engines in the booster stage(s) and orbital stage. Once the boosters run dry, they are ejected, and (typically) glide back to a landing. The advantage to this is that the mass ratios of the individual stages is vastly reduced due to the way cross feed modifies the rocket equation. Isp*g*ln(2MR^2/MR+1) & Isp*g*ln(3MR^2/MR+2) respectively. With hydrogen engines, a triamese only needs an MR of 5, as opposed to an MR of 10 for a single stage equivalent vehicle.
 
A criticism of this approach is that designing separate orbiter and boosters, or a single vehicle that could do both, would compromise performance, safety, and possible cost savings. Compromising maximum performance to reduce cargo cost however, is the POINT of the triamese approach. Stacking two or three winged vehicles can also be challenging. Optimistically, the lower mass ratios would translate to lower overall R&D costs, even if two different stage designs. While many aerospace designs have successfully been modified far beyond the original designers intentions (Boeing's 747 is perhaps the best example) the slow and painful birth of the F-35 family demonstrates that it is not always a guarantee of such flexibility.
 
Crossfeed is to be an important part of SpaceX's Falcon Heavy - and one of the main reasons it will be able to lift ~5 times as much cargo to orbit as the standard Falcon 9.

http://en.wikipedia.org/wiki/Reusable_launch_system

I ran a spreadsheet on the FH using the Delta V equation above and derived some really large excess capacity for the vehicle. I know speculation is that FH will only achieve something like 40 to 50 tonnes to LEO, but my spreadsheet gave well over 53 tonnes using 9.8 km/s as the total required Delta V. Of course, that was for an expendable FH.

FH is not complex to model. Can easily do it on Schillings, etc. [Edit 2: Actually, ISTR Schillings can model parallel core & boosters, but the principal below applies if you're modeling for yourself.]

Treat the boosters as first stage, just with the thrust of all 27 engines (instead of 18), [edit: and consuming booster prop plus whatever core prop is used].

Once the boosters are ejected, treat the core as a simple second stage with nine engines, and whatever core prop is left. One benefit - there's no gap in thrust between first & second stages, though [edit: my speculation] there may be a short throttle-back at staging.

Upper stage is a simple third stage.



There is only one mystery - what is the exact cross-feed scheme, and how much prop does that leave in the core at booster separation?

cheers, Martin

[modemeagle]

There is only one mystery - what is the exact cross-feed scheme, and how much prop does that leave in the core at booster separation?

cheers, Martin

We will probably first hear about how it's done when we hear about the 4 1/2 minute full duration core test.  The time they run it to will tell us how many engines are cross-fed.

My simulation only uses 9.1 km/s to get to orbit, just over 1.4 km/s for gravity losses, .116 km/s aero losses and 7.375 km/s for insertion (.409 km/s was free from the Earth's rotation).

Here is a quick sheet showing the math.

[aero]

There is only one mystery - what is the exact cross-feed scheme, and how much prop does that leave in the core at booster separation?

cheers, Martin

We will probably first hear about how it's done when we hear about the 4 1/2 minute full duration core test.  The time they run it to will tell us how many engines are cross-fed.

My simulation only uses 9.1 km/s to get to orbit, just over 1.4 km/s for gravity losses, .116 km/s aero losses and 7.375 km/s for insertion (.409 km/s was free from the Earth's rotation).

Here is a quick sheet showing the math.

I used your numbers and the single stage rocket equation, 3 times, and got these numbers. Delta V at BECO = 2702.326619, additional Delta V at MECO = 3788.807724 and additional Delta V at S-2 Burn-out = 2774.079887 for a total Delta V of the 53 tonne payload and S-2 dry of 9265.21423. If you'd like to do a check of those numbers, feel free to tell me your findings.

The problem though, is still in this application of the rocket equation. Wikipedia states without equivocation that the correct formulation of the rocket equation with cross-feed is Delta V = Isp * g ln(3MR^2/MR+2). I asked before if anyone could derive this formulation, so far with no takers. I've Googled it and found 2 derivations, neither of which I could follow, but with results that are clearly not the same as the single stage rocket equation.

The point is that the Triamese rocket equation results in much higher Delta V values than results from using the single stage rocket equation. This is the case even though with the Triamese rocket, there are only two stages, with the second stage being correctly treated with the single stage rocket equation.

Using your numbers and the Triamese rocket equation gives Stage 2 plus payload burn-out velocity of 12471.88613 m/s. This discrepency between the two approaches crys for resolution and I am not satisfied by saying that Wikipedia is known to have errors.

[dwightlooi]

It's not that complicated. Just treat the F9H as a three stage rocket and using the basic rocket equation.

1st Stage Starting Mass = Falcon Heavy published Launch Mass + 53 tons
1st Stage Ending Mass = Launch Mass - Fuel of 2 x Side boosters
1st Stage Mean Isp = 293 secs -- (275 + 311) / 2
1st Stage Starting Velocity = 0

2nd Stage Starting Mass = Falcon 9 1.1 published Mass + 53 tons
2nd Stage Ending Mass = Falcon 9 1.1 published Mass + 53 tons - Fuel of F1.1 core
2nd Stage Mean Isp = 311 secs
2nd Stage Starting Velocity = 1st stage burn out velocity

3rd Stage Starting Mass = F9 Upper Stage Mass + 53 tons
3rd Stage Ending Mass = F9 Upper Stage Mass + 53 tons - Fuel of Upper Stage
3rd Stage Mean Isp = 311 secs
3rd Stage Starting Velocity = 2nd stage burn out velocity

Total Gravity + Aerodynamic + staging losses = 15% (typical for orbital launch vehicles)

This all works out to about 7.8 km/s only when the core booster is practically full at staging.

[Robotbeat]

3rd stage mean Isp is WAY off. Much higher than 311s for the Vacuum-optimized Merlin.

[meekGee]

I have a question for the experts.

If the side core tanks are somewhat (10%?) taller than the center core tanks, the pressure at their bottoms will be higher.  A cross-connect will transfer propellant from them into the main tank, keeping it full - until the levels match.

This will allow the center core engines to work off of the side core tanks for the beginning of the flight, when you need all engines.

Once the level equalize, you can shut down (throttle down?) the center engines (3?  maybe 9?) and only relight them later after side core separation.

The nice thing about this is that you don't need to feed engines from other cores - just do tank-to-tank transfer.

Is this being done on some rockets?  if not, why not?

[MP99]

It's not that complicated. Just treat the F9H as a three stage rocket and using the basic rocket equation.

1st Stage Starting Mass = Falcon Heavy published Launch Mass + 53 tons
1st Stage Ending Mass = Launch Mass - Fuel of 2 x Side boosters
1st Stage Mean Isp = 293 secs -- (275 + 311) / 2
1st Stage Starting Velocity = 0

This "combined" stage also consumes some of the core prop, so subtract this also from the ending mass. This does mean the "combined" stage / high thrust phase lasts longer.

This means "2nd stage" also starts with less prop.

2nd Stage Starting Mass = Falcon 9 1.1 published Mass + 53 tons

Also, would F9 v1.1's published mass include it's payload (about 40t difference)?

cheers, Martin

[modemeagle]

It's not that complicated. Just treat the F9H as a three stage rocket and using the basic rocket equation.

1st Stage Starting Mass = Falcon Heavy published Launch Mass + 53 tons
1st Stage Ending Mass = Launch Mass - Fuel of 2 x Side boosters
1st Stage Mean Isp = 293 secs -- (275 + 311) / 2
1st Stage Starting Velocity = 0

This "combined" stage also consumes some of the core prop, so subtract this also from the ending mass. This does mean the "combined" stage / high thrust phase lasts longer.

This means "2nd stage" also starts with less prop.

2nd Stage Starting Mass = Falcon 9 1.1 published Mass + 53 tons

Also, would F9 v1.1's published mass include it's payload (about 40t difference)?

cheers, Martin

I questioned this once and was quickly corrected that the published lift off mass includes max LEO payload.

[dwightlooi]

I have a question for the experts.

If the side core tanks are somewhat (10%?) taller than the center core tanks, the pressure at their bottoms will be higher.  A cross-connect will transfer propellant from them into the main tank, keeping it full - until the levels match.

This will allow the center core engines to work off of the side core tanks for the beginning of the flight, when you need all engines.

Once the level equalize, you can shut down (throttle down?) the center engines (3?  maybe 9?) and only relight them later after side core separation.

The nice thing about this is that you don't need to feed engines from other cores - just do tank-to-tank transfer.

Is this being done on some rockets?  if not, why not?

That doesn't really work because that 10% taller tankage shared between will only amount to 6~7% the burn time of the side boosters. Basically, you'll have crossfeed only for 10~12 seconds. That isn't worth much, and certainly won't get you from 36 metric tons for a non-crossfed F9H to the 53 metric tons advertized.

[vigleik]

I used your numbers and the single stage rocket equation, 3 times, and got these numbers. Delta V at BECO = 2702.326619, additional Delta V at MECO = 3788.807724 and additional Delta V at S-2 Burn-out = 2774.079887 for a total Delta V of the 53 tonne payload and S-2 dry of 9265.21423. If you'd like to do a check of those numbers, feel free to tell me your findings.

The problem though, is still in this application of the rocket equation. Wikipedia states without equivocation that the correct formulation of the rocket equation with cross-feed is Delta V = Isp * g ln(3MR^2/MR+2). I asked before if anyone could derive this formulation, so far with no takers. I've Googled it and found 2 derivations, neither of which I could follow, but with results that are clearly not the same as the single stage rocket equation.

The point is that the Triamese rocket equation results in much higher Delta V values than results from using the single stage rocket equation. This is the case even though with the Triamese rocket, there are only two stages, with the second stage being correctly treated with the single stage rocket equation.

Using your numbers and the Triamese rocket equation gives Stage 2 plus payload burn-out velocity of 12471.88613 m/s. This discrepency between the two approaches crys for resolution and I am not satisfied by saying that Wikipedia is known to have errors.

The triamese rocket equation from wikipedia only works with zero payload. If M is the wet mass and m is the dry mass of each core and you treat it as a two stage rocket you get a mass fraction of 3M/(M+2m) for the first stage and M/m for the second stage. From that you can derive the triamese rocket equation by adding logs.

Since the payload is not zero, you have to treat it as a three stage system. There's really no way around it.

[aero]

Hey all you experts, why do you think the FH has a second stage? Run your numbers without the second stage. Looks to me like the boosted core stage can reach orbit by itself.

[modemeagle]

Hey all you experts, why do you think the FH has a second stage? Run your numbers without the second stage. Looks to me like the boosted core stage can reach orbit by itself.

My simulation get 13.125 tonnes to orbit not using a 2nd stage with 7.15% residual fuel in SI.  Does not make sense to spend 3 cores for the same payload of F9V1.1.

[aero]

Hey all you experts, why do you think the FH has a second stage? Run your numbers without the second stage. Looks to me like the boosted core stage can reach orbit by itself.

My simulation get 13.125 tonnes to orbit not using a 2nd stage with 7.15% residual fuel in SI.  Does not make sense to spend 3 cores for the same payload of F9V1.1.

The way I'm doing it, using your numbers, I get Delta V at MECO of 8.847 km/s with 53 tonne payload. That may not be enough velocity, but it is arguable.

All I did was zero out the wet mass of the second stage in the calculations I posted before using the normal version of the rocket equation.

[modemeagle]

Hey all you experts, why do you think the FH has a second stage? Run your numbers without the second stage. Looks to me like the boosted core stage can reach orbit by itself.

My simulation get 13.125 tonnes to orbit not using a 2nd stage with 7.15% residual fuel in SI.  Does not make sense to spend 3 cores for the same payload of F9V1.1.

The way I'm doing it, using your numbers, I get Delta V at MECO of 8.847 km/s with 53 tonne payload. That may not be enough velocity, but it is arguable.

All I did was zero out the wet mass of the second stage in the calculations I posted before using the normal version of the rocket equation.

You could probably get close to 30 tonnes (using the calculation spreadsheet I posted earlier), but the guidance system better be very good.  Mine is not very responsive to accurate insertion at very high acceleration (between 3.5 and 6 g)

[aero]

I used your numbers and the single stage rocket equation, 3 times, and got these numbers. Delta V at BECO = 2702.326619, additional Delta V at MECO = 3788.807724 and additional Delta V at S-2 Burn-out = 2774.079887 for a total Delta V of the 53 tonne payload and S-2 dry of 9265.21423. If you'd like to do a check of those numbers, feel free to tell me your findings.

The problem though, is still in this application of the rocket equation. Wikipedia states without equivocation that the correct formulation of the rocket equation with cross-feed is Delta V = Isp * g ln(3MR^2/MR+2). I asked before if anyone could derive this formulation, so far with no takers. I've Googled it and found 2 derivations, neither of which I could follow, but with results that are clearly not the same as the single stage rocket equation.

The point is that the Triamese rocket equation results in much higher Delta V values than results from using the single stage rocket equation. This is the case even though with the Triamese rocket, there are only two stages, with the second stage being correctly treated with the single stage rocket equation.

Using your numbers and the Triamese rocket equation gives Stage 2 plus payload burn-out velocity of 12471.88613 m/s. This discrepency between the two approaches crys for resolution and I am not satisfied by saying that Wikipedia is known to have errors.

The triamese rocket equation from wikipedia only works with zero payload. If M is the wet mass and m is the dry mass of each core and you treat it as a two stage rocket you get a mass fraction of 3M/(M+2m) for the first stage and M/m for the second stage. From that you can derive the triamese rocket equation by adding logs.

Since the payload is not zero, you have to treat it as a three stage system. There's really no way around it.

OK, that was simple enough. Thanks. Note that in the Wikipedia equation, there is a missing "(" in the denominator. The correction gives

Delta V = Ve * ln ( 3MR^2/(MR + 2))

As for lack of payload, you're right, this equation doesn't have the payload term. I'll look into it farther. I have discovered other forms of the rocket equation so maybe I can figure something out. Its good to realize that this equation is just a re-casting of the original for a specific purpose.

[Jason1701]

Hey all you experts, why do you think the FH has a second stage? Run your numbers without the second stage. Looks to me like the boosted core stage can reach orbit by itself.

It would burn out 5 minutes into flight and something else would have to provide circularization.

[aero]

Hey all you experts, why do you think the FH has a second stage? Run your numbers without the second stage. Looks to me like the boosted core stage can reach orbit by itself.

It would burn out 5 minutes into flight and something else would have to provide circularization.

Right, and that's true no matter how much excess delta V the three boosters generate. What does it take to circularize 53 tonnes of payload? A full second stage or something less? Maybe a restart of one of the core engines?

[meekGee]

I have a question for the experts.

If the side core tanks are somewhat (10%?) taller than the center core tanks, the pressure at their bottoms will be higher.  A cross-connect will transfer propellant from them into the main tank, keeping it full - until the levels match.

This will allow the center core engines to work off of the side core tanks for the beginning of the flight, when you need all engines.

Once the level equalize, you can shut down (throttle down?) the center engines (3?  maybe 9?) and only relight them later after side core separation.

The nice thing about this is that you don't need to feed engines from other cores - just do tank-to-tank transfer.

Is this being done on some rockets?  if not, why not?

That doesn't really work because that 10% taller tankage shared between will only amount to 6~7% the burn time of the side boosters. Basically, you'll have crossfeed only for 10~12 seconds. That isn't worth much, and certainly won't get you from 36 metric tons for a non-crossfed F9H to the 53 metric tons advertized.

Some of this can be solved.  Extend the outer tanks while shortening the center core.  Same total propellant, same total thrust, but now maybe 40% difference in height.  So now you have free propellant transfer to the center core until about 25% of your propellant is depleted, at which point you can shut down the center core.

Now you're flying on two side cores, and not depleting the center core.  Finally, you drop the side cores, relight the center core, and finish the first stage burn.

The downside of this scheme, aside from the extra ignition step, is that you sacrifice thrust, and so IIUC will suffer more gravity losses.

The upside is that you made the x-feed simpler.

It could be that this is just not a good enough trade and so was never used.

[aero]

How did they do the cross feed on the space shuttle? H2 from the external tank to the engines was like cross feed.

The main problem IMO is proving reliability, but triply redundant cross feed pipes seems foolish. If one fuel or oxidizer pipe breaks, it's LOM, no matter how many good pipes you have. Design it to make use of the existing turbopumps with highly reliable valves. Would turning a valve from one source of fuel to the other cause any engine transient problems?

[dwightlooi]

Some of this can be solved.  Extend the outer tanks while shortening the center core.  Same total propellant, same total thrust, but now maybe 40% difference in height.  So now you have free propellant transfer to the center core until about 25% of your propellant is depleted, at which point you can shut down the center core.

Now you're flying on two side cores, and not depleting the center core.  Finally, you drop the side cores, relight the center core, and finish the first stage burn.

The downside of this scheme, aside from the extra ignition step, is that you sacrifice thrust, and so IIUC will suffer more gravity losses.

The upside is that you made the x-feed simpler.

It could be that this is just not a good enough trade and so was never used.

If the tanks are simply connected with NO VALVING you won't be able to shut down the center engines and preserve the center booster fuel supply -- because they are all connected!

If you have valving, then it'll be better to simply switch the center engines' fuel supply between the center and side tanks.

[Jorge]

How did they do the cross feed on the space shuttle? H2...

and O2

... from the external tank to the engines was like cross feed.

Critical distinction: on the shuttle, MECO was prior to ET sep, so ET sep was not performed under acceleration, nor did they have to restart engines after ET sep.

[Jim]

http://forum.nasaspaceflight.com/index.php?topic=28424.msg876160#msg876160

[meekGee]

If the tanks are simply connected with NO VALVING you won't be able to shut down the center engines and preserve the center booster fuel supply -- because they are all connected!

If you have valving, then it'll be better to simply switch the center engines' fuel supply between the center and side tanks.

Yes I was not implying no valving...  Was trying to separate the x-feed valving from the engine block, since the x-connect can be higher.  But I think at the end the complexity savings are not that great, so nevermind.

[aero]

If the tanks are simply connected with NO VALVING you won't be able to shut down the center engines and preserve the center booster fuel supply -- because they are all connected!

If you have valving, then it'll be better to simply switch the center engines' fuel supply between the center and side tanks.

Yes I was not implying no valving...  Was trying to separate the x-feed valving from the engine block, since the x-connect can be higher.  But I think at the end the complexity savings are not that great, so nevermind.

A point. Cross feed, tank to tank requires two pipes and two valves. Cross feed tank to engine requires a number (nine?) pipes and valves. There is a mass and complexity trade in there somewhere, not to mention space for pipe runs.

[modemeagle]

The simplest setup is like Jim's diagram, 2 pipes from each booster terminating into a single manifold for each propellant and fed down to the core stage's engines.  They could shut down an upper valve (above manifold/octopus) and keep the central core pressurized, but not using propellants if full cross feed is used.  They could regulate the flow rate from each booster to make sure the booster's drain at the same rate for balance.  They could also skip the Core valve (one more thing to fail closed) and just pressurize the boosters higher than the core to keep the flow essentially zero.

Last year I sat down and wrote out how I would do cross feed and its similar to the above and diagram.

[Idiomatic]

The simplest setup is like Jim's diagram, 2 pipes from each booster terminating into a single manifold for each propellant and fed down to the core stage's engines.  They could shut down an upper valve (above manifold/octopus) and keep the central core pressurized, but not using propellants if full cross feed is used.  They could regulate the flow rate from each booster to make sure the booster's drain at the same rate for balance.  They could also skip the Core valve (one more thing to fail closed) and just pressurize the boosters higher than the core to keep the flow essentially zero.

Last year I sat down and wrote out how I would do cross feed and its similar to the above and diagram.

I do still worry about what that will do to pressures. Wonder what mechanism they will use to reduce the shock of switching streams.

[modemeagle]

The simplest setup is like Jim's diagram, 2 pipes from each booster terminating into a single manifold for each propellant and fed down to the core stage's engines.  They could shut down an upper valve (above manifold/octopus) and keep the central core pressurized, but not using propellants if full cross feed is used.  They could regulate the flow rate from each booster to make sure the booster's drain at the same rate for balance.  They could also skip the Core valve (one more thing to fail closed) and just pressurize the boosters higher than the core to keep the flow essentially zero.

Last year I sat down and wrote out how I would do cross feed and its similar to the above and diagram.

I do still worry about what that will do to pressures. Wonder what mechanism they will use to reduce the shock of switching streams.

1.  If there is a valve on the core above the manifold then open it before cutting the cross feed, then close the cross feed valves over a few second period to limit any shock.
2.  If using pressure differential then increase the pressure of the core over a few seconds to stop the cross flow.  Then close the booster valves over a few seconds to limit any shocks.

[Jim]

I do still worry about what that will do to pressures. Wonder what mechanism they will use to reduce the shock of switching streams.

Not a big deal, see heritage Atlas or any launch vehicle.   They all have valves that close to shut down engines.

[dwightlooi]

I do still worry about what that will do to pressures. Wonder what mechanism they will use to reduce the shock of switching streams.

Not a big deal, see heritage Atlas or any launch vehicle.   They all have valves that close to shut down engines.

Forget the Atlas... every single engine on the Falcon 9 has valves that shut down the engine. That's how they achieve MECO!

[Jim]

I do still worry about what that will do to pressures. Wonder what mechanism they will use to reduce the shock of switching streams.

Not a big deal, see heritage Atlas or any launch vehicle.   They all have valves that close to shut down engines.

Forget the Atlas... every single engine on the Falcon 9 has valves that shut down the engine. That's how they achieve MECO!

Didn't I say "or any launch vehicle"
What do you mean forget the Atlas?  It is the best example.  Shutting down some engines while others keep burning

[Idiomatic]

I do still worry about what that will do to pressures. Wonder what mechanism they will use to reduce the shock of switching streams.

Not a big deal, see heritage Atlas or any launch vehicle.   They all have valves that close to shut down engines.

Forget the Atlas... every single engine on the Falcon 9 has valves that shut down the engine. That's how they achieve MECO!

Didn't I say "or any launch vehicle"
What do you mean forget the Atlas?  It is the best example.  Shutting down some engines while others keep burning

FH crossfeed is a bit more difficult. As I understand it:

You have 3 cores running, fed by squid thing which are fed by the booster tanks. This is fine, pressure is steady though lower I would assume than the F9 1.1 since they get 2/3 of a tank each core.

When the boosters are low, then
- the central tank is opened into squid thing resulting in a sudden boost in pressure on the center or all cores
- the crossfeed shuts off. resulting in a drop in pressure on the main core and a rise in the side cores
- the side cores shut off, if this happens before crossfeed shutoff then it causes a pressure boost on the main core. it it happens after then... nothing ... so that's probably the way to do it.


So... What mechanism is in place to handle the change in pressure? Can the funny octopus things handle it? Will the valves have to handle it? Can the engines handle it? Or something else.

[cordor]

I think the plan is that after the boosters stage off, only the engines fed by the core will stay lit (that way they don't have to switch propellant supplies). If you have too few engines burning at that point you have very high gravity losses, which is bad (obviously). Also, engine-out capability may still be desired.

What is the major risk factor in doing crossfeed?

It would seem, on the face of it, to be valves failing to open and/or close, but that sounds like something that could be heavily ground tested. While I have no doubt that cryogenic valves have their own idiosyncrasies, it doesn't seem like a high risk item, but I have zero experience in the area.

If valve reliability is high, then wouldn't it be relatively safe to use 6 valves (2 fuel and 2 LOX for the outboard tanks, and another 2 for the core tanks)? Then all 9 core engines could be used. I'm assuming there already are valves in the outboard boosters to shut off the fuel and LOX, but perhaps there need to be 2 more each at the booster couplings?

spacex has long track record of valve problems.

[kevin-rf]

spacex has long track record of valve problems.

Huh? I thought they had a long history of having the software constraints set to tightly causing last minute aborts.

[Lobo]

Not sure if this question has already been answered, as I've not read all the pages on this thread, but I was wondering the reasons for doing crossfeed, rather than simply throttling down the central core during ascent to preserve it's fuel, and then after booster separationthere is still fuel left in the core.  I guess this would lead to higher staging, as the boosters aren't drainging their propellant into the central core, but wouldn't it accomplish roughly the same thing?

or not?
And if not, why not?

[meekGee]

spacex has long track record of valve problems.

Huh? I thought they had a long history of having the software constraints set to tightly causing last minute aborts.

Well, at least they're now being given the credit of "long track record"...
One could even say that they had valve problems on 1/3 of their F9 launches.

[krytek]

You can do that, but that's probably not as efficient. Crossfeed is essentially like adding a another stage since the core is almost full at separation.

[modemeagle]

Not sure if this question has already been answered, as I've not read all the pages on this thread, but I was wondering the reasons for doing crossfeed, rather than simply throttling down the central core during ascent to preserve it's fuel, and then after booster separationthere is still fuel left in the core.  I guess this would lead to higher staging, as the boosters aren't drainging their propellant into the central core, but wouldn't it accomplish roughly the same thing?

or not?
And if not, why not?

Comparison using my simulator.  The only change is shut down cross feed and activate throttle down.  Even the initial pitch is the same.

EDIT:  Updated sheet to make MECO with ~300 m/s residual prop.
Edit again:  Updated with data for cross feed between 1 and 8 engines.

[notsorandom]

Not sure if this question has already been answered, as I've not read all the pages on this thread, but I was wondering the reasons for doing crossfeed, rather than simply throttling down the central core during ascent to preserve it's fuel, and then after booster separationthere is still fuel left in the core.  I guess this would lead to higher staging, as the boosters aren't drainging their propellant into the central core, but wouldn't it accomplish roughly the same thing?

or not?
And if not, why not?

Comparison using my simulator.  The only change is shut down cross feed and activate throttle down.  Even the initial pitch is the same.

I am not seeing much of a difference at all. Am I looking at it incorrectly?

[modemeagle]

Not sure if this question has already been answered, as I've not read all the pages on this thread, but I was wondering the reasons for doing crossfeed, rather than simply throttling down the central core during ascent to preserve it's fuel, and then after booster separationthere is still fuel left in the core.  I guess this would lead to higher staging, as the boosters aren't drainging their propellant into the central core, but wouldn't it accomplish roughly the same thing?

or not?
And if not, why not?

Comparison using my simulator.  The only change is shut down cross feed and activate throttle down.  Even the initial pitch is the same.

I am not seeing much of a difference at all. Am I looking at it incorrectly?

I don't see much of a difference either, but my simulator does not model the change is ISP as the engine is throttled down to 70%.  Maybe later I will try to integrate that into the model.

[MP99]

Not sure if this question has already been answered, as I've not read all the pages on this thread, but I was wondering the reasons for doing crossfeed, rather than simply throttling down the central core during ascent to preserve it's fuel, and then after booster separationthere is still fuel left in the core.  I guess this would lead to higher staging, as the boosters aren't drainging their propellant into the central core, but wouldn't it accomplish roughly the same thing?

or not?
And if not, why not?

Well, you do start off with the same prop in the three first stage tanks, but if you throttle down the core, you reduce the thrust when the vehicle is heaviest. That increases gravity losses (but does assume the vehicles' structure can take all the thrust).

Also, if speculation is correct that the outriggers feed twelve engines each (their own, plus 3x core engines), then the outriggers will drain 4x faster than the core. At outrigger burnout, the core should have at least 75% of it's prop load remaining.

If you were to do the same via throttling the core, you'd need to throttle down to 25% (and lose a lot of T/W). M1Ds seem to throttle only down to 70%.

cheers, Martin

[aero]

Not sure if this question has already been answered, as I've not read all the pages on this thread, but I was wondering the reasons for doing crossfeed, rather than simply throttling down the central core during ascent to preserve it's fuel, and then after booster separationthere is still fuel left in the core.  I guess this would lead to higher staging, as the boosters aren't drainging their propellant into the central core, but wouldn't it accomplish roughly the same thing?

or not?
And if not, why not?

Well, you do start off with the same prop in the three first stage tanks, but if you throttle down the core, you reduce the thrust when the vehicle is heaviest. That increases gravity losses (but does assume the vehicles' structure can take all the thrust).

Also, if speculation is correct that the outriggers feed twelve engines each (their own, plus 3x core engines), then the outriggers will drain 4x faster than the core. At outrigger burnout, the core should have at least 75% of it's prop load remaining.

If you were to do the same via throttling the core, you'd need to throttle down to 25% (and lose a lot of T/W). M1Ds seem to throttle only down to 70%.

cheers, Martin

IMO, using crossfeed benefits by a big reduction in gravity loss. Gravity loss is t*g*((re+H)^2)/re^2 in units of velocity. H is altitude. For Low Earth Orbits, H is small compared to re (radius of Earth) so Gravity loss is almost equal to g * t. (In the vertical direction.) F = m*a, so in terms of force, gravity loss becomes (mass-mdot)*g.

Yes, you could design your rocket many different ways, but the objective is to reach orbital altitude quickly and to get rid of mass quickly. Using crossfeed, the rocket burns the fuel from the booster tanks more quickly, so that the boosters can be jettisoned sooner. Then the core rocket engines do not need to carry that mass further. That is, both the fuel mass and the tank/engine mass are expended at the earliest possible time.

[Lobo]

Not sure if this question has already been answered, as I've not read all the pages on this thread, but I was wondering the reasons for doing crossfeed, rather than simply throttling down the central core during ascent to preserve it's fuel, and then after booster separationthere is still fuel left in the core.  I guess this would lead to higher staging, as the boosters aren't drainging their propellant into the central core, but wouldn't it accomplish roughly the same thing?

or not?
And if not, why not?

Well, you do start off with the same prop in the three first stage tanks, but if you throttle down the core, you reduce the thrust when the vehicle is heaviest. That increases gravity losses (but does assume the vehicles' structure can take all the thrust).

Also, if speculation is correct that the outriggers feed twelve engines each (their own, plus 3x core engines), then the outriggers will drain 4x faster than the core. At outrigger burnout, the core should have at least 75% of it's prop load remaining.

If you were to do the same via throttling the core, you'd need to throttle down to 25% (and lose a lot of T/W). M1Ds seem to throttle only down to 70%.

cheers, Martin

Which leads me to my 2nd question, would it be cheaper/easier to develop  M1d with that deep of throttling?  Or the extra complexity of crossfeed?

I don't know, just curious. 

[vigleik]

Not sure if this question has already been answered, as I've not read all the pages on this thread, but I was wondering the reasons for doing crossfeed, rather than simply throttling down the central core during ascent to preserve it's fuel, and then after booster separationthere is still fuel left in the core.  I guess this would lead to higher staging, as the boosters aren't drainging their propellant into the central core, but wouldn't it accomplish roughly the same thing?

or not?
And if not, why not?

Comparison using my simulator.  The only change is shut down cross feed and activate throttle down.  Even the initial pitch is the same.

I am not seeing much of a difference at all. Am I looking at it incorrectly?

I agree that the difference is smaller than expected, but it's there. The residual fuel for all three stages is less in the throttle down scenario. My BOE calculation shows this to correspond to approximately a 25m/s difference in delta V, or a 500kg difference in payload, if you want to keep the residual fuel constant. (I don't have exact numbers so this is a rough estimate only.)

[Lobo]

Not sure if this question has already been answered, as I've not read all the pages on this thread, but I was wondering the reasons for doing crossfeed, rather than simply throttling down the central core during ascent to preserve it's fuel, and then after booster separationthere is still fuel left in the core.  I guess this would lead to higher staging, as the boosters aren't drainging their propellant into the central core, but wouldn't it accomplish roughly the same thing?

or not?
And if not, why not?

Comparison using my simulator.  The only change is shut down cross feed and activate throttle down.  Even the initial pitch is the same.

I am not seeing much of a difference at all. Am I looking at it incorrectly?

I agree that the difference is smaller than expected, but it's there. The residual fuel for all three stages is less in the throttle down scenario. My BOE calculation shows this to correspond to approximately a 25m/s difference in delta V, or a 500kg difference in payload, if you want to keep the residual fuel constant. (I don't have exact numbers so this is a rough estimate only.)

Is that assuming just M1d's 70% planned throttling?

Hmmm...if that's true, 1/2mt doesnt seem like a lot of lost payload on a 50mt-ish LV.  52.5mt instead of 53mt? 
To save the extra cost of designing and building a crossfeed system?

Dunno...doesn't seem like a bad trade if that's even close to accurate...

[vigleik]

Is that assuming just M1d's 70% planned throttling?

Hmmm...if that's true, 1/2mt doesnt seem like a lot of lost payload on a 50mt-ish LV.  52.5mt instead of 53mt? 
To save the extra cost of designing and building a crossfeed system?

Dunno...doesn't seem like a bad trade if that's even close to accurate...

This is assuming modemeagle's simulator is accurate, I don't know exactly what assumptions he used. My estimate is simply based on the small difference in residual fuel for each stage.

I'm not actually a rocket scientist, but my gut feeling is that you lose much more than 0.5t. The fact that S1 staging happens at a much lower altitude (114.5km versus 149.2km) seems odd, maybe that's a clue.

According to the simulation it seems that the upper stage doesn't care so much about the staging altitude. But if that's true, then, in the cross-feed scenario, we should be able to take advantage of that by adjusting the pitch. Does that make sense?

[MP99]

Not sure if this question has already been answered, as I've not read all the pages on this thread, but I was wondering the reasons for doing crossfeed, rather than simply throttling down the central core during ascent to preserve it's fuel, and then after booster separationthere is still fuel left in the core.  I guess this would lead to higher staging, as the boosters aren't drainging their propellant into the central core, but wouldn't it accomplish roughly the same thing?

or not?
And if not, why not?

Well, you do start off with the same prop in the three first stage tanks, but if you throttle down the core, you reduce the thrust when the vehicle is heaviest. That increases gravity losses (but does assume the vehicles' structure can take all the thrust).

Also, if speculation is correct that the outriggers feed twelve engines each (their own, plus 3x core engines), then the outriggers will drain 4x faster than the core. At outrigger burnout, the core should have at least 75% of it's prop load remaining.

If you were to do the same via throttling the core, you'd need to throttle down to 25% (and lose a lot of T/W). M1Ds seem to throttle only down to 70%.

Which leads me to my 2nd question, would it be cheaper/easier to develop  M1d with that deep of throttling?  Or the extra complexity of crossfeed?

I don't know, just curious. 

modemeagle's FH simulations above show a liftoff T/W of 1.320.

If you throttled the nine central core engines to 25% at liftoff, then T/W would drop to 0.99, ie the vehicle just wouldn't lift off the ground at all!

That throttling to 25% was rather arbitrary, but it makes the point. DIVH throttles its central core, but doesn't throttle until well after liftoff, and nowhere near 25%. From first principles I'd assume similar restrictions on an FH that just used core throttling instead of cross-feed.

However, that does raise an interesting question - SpaceX quote over 40t for the non-crossfed config. Does that assume a DIVH-like central-core-throttled profile (which drops the outriggers before core burnout, just not as early as cross-feed). The all-three-cores-throttled-the-same profile should have less payload, but avoids a staging event. The three possible profiles would be:-

53t  full FH with crossfeed
??t  FH with throttling instead of crossfeed
??t  FH with all three cores throttled same (no outrigger separation required)

cheers, Martin

[ArbitraryConstant]

I agree that the difference is smaller than expected, but it's there. The residual fuel for all three stages is less in the throttle down scenario. My BOE calculation shows this to correspond to approximately a 25m/s difference in delta V, or a 500kg difference in payload, if you want to keep the residual fuel constant. (I don't have exact numbers so this is a rough estimate only.)

Are you taking gravity losses into account? That seems low.

If you throttled the nine central core engines to 25% at liftoff, then T/W would drop to 0.99, ie the vehicle just wouldn't lift off the ground at all!

That seems like pretty significant gravity losses.

[Idiomatic]

500kg loss seems extremely low. First, you aren't getting the most out of your engines by throttling down, this causes gravity losses. Second, you are carrying two boosters significantly further than you would be normally. Pure waste, they are like half the dry weight of the whole craft, and 2/3rds of the drag.

[modemeagle]

I updated the sheet in my previous post. 
http://forum.nasaspaceflight.com/index.php?topic=28424.msg925279#msg925279
Changes made:
1.  Added calculations to adjust ISP to account for any throttling at any altitude.
2.  Added another sheet to compare no cross feed.
3.  Adjusted payload to leave ~ 300 m/s residual in SII so the caparison is more apples to apples.
4.  Added residual delta-v to output sheet.

Remember, just like the Shilling's calculations this is not exact. The guidance SpaceX uses will not be the same guidance I use.

[dudealus]

I find the small difference very suprising as well. Ow well, maybe it is true . Though what stands out for me as well are the G-loads in the non crossfeed versions, they seem rather harsh... Maybe the difference would be alot bigger if you would constrain all 3 versions to 4 G's?

[modemeagle]

I find the small difference very suprising as well. Ow well, maybe it is true . Though what stands out for me as well are the G-loads in the non crossfeed versions, they seem rather harsh... Maybe the difference would be alot bigger if you would constrain all 3 versions to 4 G's?

The throttle point was set to 6.5g and peaked at 6.928 3.0 seconds after min throttle was hit.  The second stage may be a little heavier on the real one which would prevent the spike in acceleration at the end of the booster phase.

To lower the acceleration I would have to cut engines which will lead to higher gravity losses.  If SpaceX is limiting the rocket to a lower acceleration then this may be why their non cross feed is lower then mine.  Waiting for a falcon heavy user manual for answers.

[dudealus]

Sounds about right. Didn't really know what kind of G-loads are usual in launch vehicles, so looked it up for Atlas V (max 4.8 ), Delta 4 (6.0) and Falcon 9 B2 (6.0 max). So I would suggest setting the throttle point at 6g, they aren't throttling down the Falcon 9 for nothing i guess

[vigleik]

I think I understand now. Let's assume the drag losses are the same for all the flight profiles we consider, and let's ignore the fact that the high g-forces we see in some profiles might not be acceptable. Then there are two considerations:

1. Gravity loss. We want to spend the fuel as quickly as possible to minimise gravity loss. Hence the no cross-feed flight profile has the lowest gravity loss, while full cross-feed has the highest gravity loss.

2. Staging. The no cross-feed flight profile does not take advantage of the "third stage", while full cross-feed takes maximal advantage.

Now consider varying the amount of cross-feed, between the two scenarios. As we increase the amount of cross-feed the gravity loss increases while the advantage of staging increases. Consider payload to LEO as a function of the amount of cross-feed used. This will almost certainly be a concave function. I thought the maximum would be at maximum cross-feed, but it's clear from the numbers that the maximum is going to be somewhere in the middle.

Also, if we can use any amount of cross-feed we want then it's always better to use that than to throttle down. To be specific, use as much cross-feed as required to leave the core with the same amount of fuel at BECO as with the throttle down. Then we get the same benefit from staging, but because the thrust is higher the gravity loss will be lower.

Some additional points:
(a) If cross-feed is not available, then throttle-down might still make sense, sacrificing some gravity loss for advantage of staging.

(b) If the dry weight of the boosters is higher, the benefit from staging is higher. This will tilt the balance towards more cross-feed.

(c) If we need to throttle down to limit the g-forces, that tilts the balance towards more cross-feed as well.

(d) If we're going to try to reuse the boosters, that tilts the balance towards more cross-feed, both because of (b) and for the obvious reason.

[modemeagle]

I think I understand now. Let's assume the drag losses are the same for all the flight profiles we consider, and let's ignore the fact that the high g-forces we see in some profiles might not be acceptable. Then there are two considerations:

1. Gravity loss. We want to spend the fuel as quickly as possible to minimise gravity loss. Hence the no cross-feed flight profile has the lowest gravity loss, while full cross-feed has the highest gravity loss.

2. Staging. The no cross-feed flight profile does not take advantage of the "third stage", while full cross-feed takes maximal advantage.

Now consider varying the amount of cross-feed, between the two scenarios. As we increase the amount of cross-feed the gravity loss increases while the advantage of staging increases. Consider payload to LEO as a function of the amount of cross-feed used. This will almost certainly be a concave function. I thought the maximum would be at maximum cross-feed, but it's clear from the numbers that the maximum is going to be somewhere in the middle.

Also, if we can use any amount of cross-feed we want then it's always better to use that than to throttle down. To be specific, use as much cross-feed as required to leave the core with the same amount of fuel at BECO as with the throttle down. Then we get the same benefit from staging, but because the thrust is higher the gravity loss will be lower.

Some additional points:
(a) If cross-feed is not available, then throttle-down might still make sense, sacrificing some gravity loss for advantage of staging.

(b) If the dry weight of the boosters is higher, the benefit from staging is higher. This will tilt the balance towards more cross-feed.

(c) If we need to throttle down to limit the g-forces, that tilts the balance towards more cross-feed as well.

(d) If we're going to try to reuse the boosters, that tilts the balance towards more cross-feed, both because of (b) and for the obvious reason.

Very good observations.  I originally found cross feeding 6 gave me the best performance.  I am going to run the sim doing zero through 9 and will post that for comparison.

I left the sim in 9 since it would reach orbit with 300 m/s residual which should be a decent margin.  I would like to hear from anybody, even in a PM, as to what kind of margin a real launcher would normally have.

EDIT:  I updated my sheet again to include cross feed between 1 and 8 engines.  Interesting to see what the best result was.

[ArbitraryConstant]

1. Gravity loss. We want to spend the fuel as quickly as possible to minimise gravity loss. Hence the no cross-feed flight profile has the lowest gravity loss, while full cross-feed has the highest gravity loss.

I assume you're saying this because the total fuel of the first stage + boosters is expended faster with 27 engines the whole time rather than 9 for part of the time. But aren't you also carrying the boosters for longer and to higher altitude? More seconds carrying the boosters means more total impulse received from gravity. I guess taking longer to relieve the vehicle of fuel loss does that.

But force times time is linear, the difference in kinetic energy received by the boosters is squared.

huh

Not that intuitive I guess. Rocket science, lol.

[Idiomatic]

"Not that intuitive I guess"

Imagine you launched both versions at the same time.

1. Both versions fire all engines at max expending the same amount of fuel.

2. At some point, the boosters will drop off of the core. At this point the crossfeed one will have 1 core and one tank of fuel. The non-crossfeed will have 3 cores and one tank of fuel spread across them. Non-crossfeed will be spending fuel 3 times faster and accelerating faster now but weigh more and have more drag (which wouldn't matter much at 50km or so with negligible atmosphere). Crossfeed will now weigh less but it will be accelerating slower and spend more time being effected by gravity.

The dry:launch weight ratio for the boosters is supposedly 30:1. So they add a few percentage points of mass to the rocket which includes a fully fueled core, 2nd stage, payload.

That said, a rocket cannot be launched like this. 7G is too much even for cargo. 4G is tops for people though F9 should be doing all the HSF missions afaik. Crossfeed keeps acceleration fairly even without being a large loss in efficiency like throttling down. As well, crossfeed should? show larger gains in BEO targets.

[Jim]

That said, a rocket cannot be launched like this. 7G is too much even for cargo. 4G is tops for people though F9 should be doing all the HSF missions afaik.

Huh?  Where are those requirements documented?

[Idiomatic]

That said, a rocket cannot be launched like this. 7G is too much even for cargo. 4G is tops for people though F9 should be doing all the HSF missions afaik.

Huh?  Where are those requirements documented?

Err... the Saturn V shut off engines to keep under 4g. The shuttle limited to 3. I don't think that SpaceX will be pushing any frontiers in the field of making astronauts unsafe. NASA would not let them anyways.

[Jim]

That said, a rocket cannot be launched like this. 7G is too much even for cargo. 4G is tops for people though F9 should be doing all the HSF missions afaik.

Huh?  Where are those requirements documented?

Err... the Saturn V shut off engines to keep under 4g. The shuttle limited to 3. I don't think that SpaceX will be pushing any frontiers in the field of making astronauts unsafe. NASA would not let them anyways.

Err... what?
Don't need to stutter here.  Either base conjecture on facts or don't make them.
You are making assumptions not based on reality.
Shuttle environments are not applicable.
This isn't a Saturn V
ELV's routine have 6's or more, so why is that too much for cargo?
Who says 6g's isn't safe for crew?  Gemini loads were higher.

Again, where are those requirements documented?
Here is where they are documented:
NASA-STD-3001, VOLUME 2

And one data point which negates your line of reasons.  7.5g's is allowed for 300 seconds, basically the whole ride into orbit.

[kirghizstan]

That said, a rocket cannot be launched like this. 7G is too much even for cargo. 4G is tops for people though F9 should be doing all the HSF missions afaik.

Huh?  Where are those requirements documented?

Err... the Saturn V shut off engines to keep under 4g. The shuttle limited to 3. I don't think that SpaceX will be pushing any frontiers in the field of making astronauts unsafe. NASA would not let them anyways.

Err... what?
Don't need to stutter here.  Either base conjecture on facts or don't make them.
You are making assumptions not based on reality.
Shuttle environments are not applicable.
This isn't a Saturn V
ELV's routine have 6's or more, so why is that too much for cargo?
Who says 6g's isn't safe for crew?  Gemini loads were higher.

Again, where are those requirements documented?
Here is where they are documented:
NASA-STD-3001, VOLUME 2

And one data point which negates your line of reasons.  7.5g's is allowed for 300 seconds, basically the whole ride into orbit.

5g at launch and almost 6 during second stage for gemini if i remember correctly

[dudealus]

@Jim, as you seem to have the most experience in this. Are most payloads built to cope with those kind of forces? I just imagined 6 g's was the maximum normal (off course you could get higher if your launch vehicle and payload can cope with it) as the user manuals of multiple launchers stated it as their max loads. I have no idea what most payloads are calculated on though, maybe you can clarify it a bit? thanks

[Idiomatic]

I stand corrected. Though I'm now confused as to why the g-forces were limited previously......... You'd think a billion dollar device wouldn't be giving up capability unless it were required.

[Downix]

Not sure if this question has already been answered, as I've not read all the pages on this thread, but I was wondering the reasons for doing crossfeed, rather than simply throttling down the central core during ascent to preserve it's fuel, and then after booster separationthere is still fuel left in the core.  I guess this would lead to higher staging, as the boosters aren't drainging their propellant into the central core, but wouldn't it accomplish roughly the same thing?

or not?
And if not, why not?

Well, you do start off with the same prop in the three first stage tanks, but if you throttle down the core, you reduce the thrust when the vehicle is heaviest. That increases gravity losses (but does assume the vehicles' structure can take all the thrust).

Also, if speculation is correct that the outriggers feed twelve engines each (their own, plus 3x core engines), then the outriggers will drain 4x faster than the core. At outrigger burnout, the core should have at least 75% of it's prop load remaining.

If you were to do the same via throttling the core, you'd need to throttle down to 25% (and lose a lot of T/W). M1Ds seem to throttle only down to 70%.

Which leads me to my 2nd question, would it be cheaper/easier to develop  M1d with that deep of throttling?  Or the extra complexity of crossfeed?

I don't know, just curious. 

modemeagle's FH simulations above show a liftoff T/W of 1.320.

If you throttled the nine central core engines to 25% at liftoff, then T/W would drop to 0.99, ie the vehicle just wouldn't lift off the ground at all!

That throttling to 25% was rather arbitrary, but it makes the point. DIVH throttles its central core, but doesn't throttle until well after liftoff, and nowhere near 25%. From first principles I'd assume similar restrictions on an FH that just used core throttling instead of cross-feed.

However, that does raise an interesting question - SpaceX quote over 40t for the non-crossfed config. Does that assume a DIVH-like central-core-throttled profile (which drops the outriggers before core burnout, just not as early as cross-feed). The all-three-cores-throttled-the-same profile should have less payload, but avoids a staging event. The three possible profiles would be:-

53t  full FH with crossfeed
??t  FH with throttling instead of crossfeed
??t  FH with all three cores throttled same (no outrigger separation required)

cheers, Martin

I did some calculations:

With separate core throttling: 31 metric tons
WIthout separate core throttling: 27 metric tons

[aero]

Did you take into account the loss of efficiency resulting from throttling the engines below their optimum (design) thrust level? Actually, throttled engines seem to perform quite well with less than 10% loss in Isp over a wide range of thrust. Above 80% thrust, Isp increases to 100% of the nominal value, but between ~30% to ~80% throttle, ~10% reduction in Isp seems to be typical. Thrust can be increased to compensate by throttling up, but fuel consumption would increase. Also, throttleable engines seem to be noteably more massive than the fixed thrust counterpart. Of course, 1% of engine weight is only 5 added kg  so maybe the added mass wouldn't be much in absolute terms.

Do a Google search on "rocket engine throttling," or something like that, there are a lot of references.

[93143]

I would expect Isp reduction due to throttling to be strongly dependent on ambient pressure, as well as on propellant type.  A hydrogen/oxygen rocket in vacuum almost doesn't care what the chamber pressure is, if I'm recalling my CEA experiments correctly...

[Jim]

@Jim, as you seem to have the most experience in this. Are most payloads built to cope with those kind of forces? I just imagined 6 g's was the maximum normal (off course you could get higher if your launch vehicle and payload can cope with it) as the user manuals of multiple launchers stated it as their max loads. I have no idea what most payloads are calculated on though, maybe you can clarify it a bit? thanks

Yes, they as built to take these loads.
current ELV's are around 6 g's

[Downix]

Did you take into account the loss of efficiency resulting from throttling the engines below their optimum (design) thrust level? Actually, throttled engines seem to perform quite well with less than 10% loss in Isp over a wide range of thrust. Above 80% thrust, Isp increases to 100% of the nominal value, but between ~30% to ~80% throttle, ~10% reduction in Isp seems to be typical. Thrust can be increased to compensate by throttling up, but fuel consumption would increase. Also, throttleable engines seem to be noteably more massive than the fixed thrust counterpart. Of course, 1% of engine weight is only 5 added kg  so maybe the added mass wouldn't be much in absolute terms.

Do a Google search on "rocket engine throttling," or something like that, there are a lot of references.

Yes I did, although I gave SpaceX the benefit of the doubt in that regards for how much is lost from the throttling, 8% loss. I've been doing this for awhile.

[cordor]

Not sure if this question has already been answered, as I've not read all the pages on this thread, but I was wondering the reasons for doing crossfeed, rather than simply throttling down the central core during ascent to preserve it's fuel, and then after booster separationthere is still fuel left in the core.  I guess this would lead to higher staging, as the boosters aren't drainging their propellant into the central core, but wouldn't it accomplish roughly the same thing?

or not?
And if not, why not?

Well, you do start off with the same prop in the three first stage tanks, but if you throttle down the core, you reduce the thrust when the vehicle is heaviest. That increases gravity losses (but does assume the vehicles' structure can take all the thrust).

Also, if speculation is correct that the outriggers feed twelve engines each (their own, plus 3x core engines), then the outriggers will drain 4x faster than the core. At outrigger burnout, the core should have at least 75% of it's prop load remaining.

If you were to do the same via throttling the core, you'd need to throttle down to 25% (and lose a lot of T/W). M1Ds seem to throttle only down to 70%.

cheers, Martin

Which leads me to my 2nd question, would it be cheaper/easier to develop  M1d with that deep of throttling?  Or the extra complexity of crossfeed?

I don't know, just curious. 

throttling is not the only option. they can turn some of the engine off.

[fatjohn1408]

Not sure if this question has already been answered, as I've not read all the pages on this thread, but I was wondering the reasons for doing crossfeed, rather than simply throttling down the central core during ascent to preserve it's fuel, and then after booster separationthere is still fuel left in the core.  I guess this would lead to higher staging, as the boosters aren't drainging their propellant into the central core, but wouldn't it accomplish roughly the same thing?

or not?
And if not, why not?

Comparison using my simulator.  The only change is shut down cross feed and activate throttle down.  Even the initial pitch is the same.

EDIT:  Updated sheet to make MECO with ~300 m/s residual prop.
Edit again:  Updated with data for cross feed between 1 and 8 engines.

Nice simulator. How do you program the circularization burn so nicely?
And wrt gravity loss. I always wondered wether that needs to take centrifugal force into account since the flightpathangle is not defined wrt an inertial reference frame. My apologies if i'm taking this too off topic.

[go4mars]

Don't apologize.  These are the kinds of questions that a lot of us can learn from.  Plus it's on topic.

[modemeagle]

Not sure if this question has already been answered, as I've not read all the pages on this thread, but I was wondering the reasons for doing crossfeed, rather than simply throttling down the central core during ascent to preserve it's fuel, and then after booster separationthere is still fuel left in the core.  I guess this would lead to higher staging, as the boosters aren't drainging their propellant into the central core, but wouldn't it accomplish roughly the same thing?

or not?
And if not, why not?

Comparison using my simulator.  The only change is shut down cross feed and activate throttle down.  Even the initial pitch is the same.

EDIT:  Updated sheet to make MECO with ~300 m/s residual prop.
Edit again:  Updated with data for cross feed between 1 and 8 engines.

Nice simulator. How do you program the circularization burn so nicely?
And wrt gravity loss. I always wondered wether that needs to take centrifugal force into account since the flightpathangle is not defined wrt an inertial reference frame. My apologies if i'm taking this too off topic.

In my simulator the centrifugal force is in the vertical acceleration equation subtracting the current gravity force.  Once the velocity is greater then orbital velocity then the centrifugal force is greater than gravity and the vertical velocity increases.

An issue I have is when the Apogee is above about 3000 km my simulator increases the apogee higher and higher.  I don't think it effects the powered flight part, but is effecting the coast periods.  Below 400 km the increase is only a few KM.