SpaceX Falcon Heavy Discussion (Thread 6)

[Jim]

For all intents and purposes, it is a stock Delta IV Heavy.  The third stage is just a payload.

[Zed_Noir]

For all intents and purposes, it is a stock Delta IV Heavy.  The third stage is just a payload.

So in theory a Falcon Heavy with a large hypergolic or solid kick stage, would the kick stage be counted as part of the payload?

[woods170]

For all intents and purposes, it is a stock Delta IV Heavy.  The third stage is just a payload.

So in theory a Falcon Heavy with a large hypergolic or solid kick stage, would the kick stage be counted as part of the payload?

If Jim does apples-to-apples he would answer this question with a "Yes".

Much like Star-48 kick stage is not part of a stock Delta IV-Heavy a kick-stage is not part of a stock Falcon Heavy.
In both cases the added kick-stage would be part of the payload.

[TripleSeven]

For all intents and purposes, it is a stock Delta IV Heavy.  The third stage is just a payload.

So in theory a Falcon Heavy with a large hypergolic or solid kick stage, would the kick stage be counted as part of the payload?

yes, its just FAlcon and Falcon heavy need a high energy upper stage...to boost "that payload" that includes a solid kick stage..to somewhere

[abaddon]

yes, its just FAlcon and Falcon heavy need a high energy upper stage

No, you completely missed the point.  Falcon Heavy does not need a high energy upper stage, it is competitive with Delta IV Heavy to high-energy trajectories as is.  AV551, FH, and DIVH all need an additional kick stage (e.g. Star48 family) to serve any of these payloads such as New Horizons or Parker Solar Probe.

Falcon 9 is a different matter, but nobody is seriously comparing Falcon 9 single-stick high-energy performance with Atlas V 551 or Delta IV Heavy.

[LouScheffer]

Just for completion. FH-R.

From these curves, FH could have launched New Horizons and/or the Parker Solar Probe, but would have required the expended configurations to do either one.  (PSP had a mass of about 2700 kg at separation at a C3 of 60 km^2/sec^2.  New Horizons had a somewhat lower C3 and mass.)

I got a request for estimate of New Horizons C3 after the second stage but before the kick burn.   Here'a a back of the envelope calculation.

We know the C3 at separation was 157.74 km^2/m^s.   So the energy, per kg, is C3/2 = 78.9  MJ/kg (at infinity)

Next we need to adjust the energy since the kick burn was not at infinity.  It was low, but not quite LEO since the Centaur burn took a few minutes.  So let's guess an altitude of 629 km since that makes the radius from Earth center an even 7000 km.  So we add mu/r, where mu for Earth is 3.986e14.   Then the energy per kg was 136 MJ/kg, hence V (earth relative) = 16,481 m/s at kick burnout.

How much delta V from the kick burn?  NH mass = 478kg, Star48 initial mass = 2141 kg, burnout mass = 131 kg, ISP=292.  So initial mass is 2619 kg, final mass 609 kg, and dv = 292*9.8*ln(initial/final) = 4,174 m/s.   So before this burn V was 12,307 m/s.

Now what is this in terms of C3?  1/2*m*v^2 to get energy per kg, subtract mu/r to get energy at infinity, which is 18.8 MJ/kg.  C3 is twice this value, or 37.6 km^2/sec^2.

So the second stage of the Atlas pushed about 2619 kg to a C3 of 37.6 km^2/sec^2.   This is just below the Atlas 551 curve for high energy trajectories shown in the thread above, so it all makes sense.

EDIT:  Fix typos and wording

[John Alan]

Seems to me... 
...a known satellite maker (like say SSL) could clean up in the planetary probe marketplace (such as it is) by designing a high capacity Hypergolic (not cryo) '3rd stage' that uses a SpaceX's dragon trunk perimeter mount to attach to a FH S2 (for lighter weight and efficient load transfer) and this new S3 "stage" is topped with a smaller Ruag fairing (as probes tend to be smaller anyway)...

SpaceX gets contracted to launch ~50 tonnes to LEO with Boosters to ASDS's and Core expended...
S2 saves enough prop to deorbit also... SpaceX gets $95mil plus for the trouble and expenses...

SSL has literally tonnes of prop margin to play with... to push the paying customers probe out to it's destination...

Bottom line is SpaceX likely does not want to get directly into high energy C3 launches, as they are few and far between.
But if someone like SSL wanted to finish the job and make FH the 'go to, lowest cost to way out there' C3 king...
Then yeah... 

[docmordrid]

Seems to me... 
>
SpaceX likely does not want to get directly into high energy C3 launches, as they are few and far between.
>

Then why build BFR (BFB, BFS, BFS Tanker, BFS Chomper....)?

It's designed for high energy C3, but is adaptable to fill the other roles and fully reusable. That's why they've said it'll replace F9/FH once customers accept it.

[John Alan]

For sending probes and other payloads out into the system... There will still likely be a need for upper kick stages, even with BFS... 

That said... too bad the RS-72 never really found a user...
http://www.astronautix.com/r/rs-72.html
https://archive.is/E5gH6
ISP of 340 in a vacuum and 55.4kN thrust... (12.5 klbs thrust)...
Pump fed/boosted Hypergolic... 

Four of them and 40 tonnes of prop might make an interesting FH S3 setup... 

[docmordrid]

For sending probes and other payloads out into the system... There will still likely be a need for upper kick stages, even with BFS... 
>

Like this?

[Jim]

Seems to me... 
...a known satellite maker (like say SSL) could clean up in the planetary probe marketplace (such as it is) by designing a high capacity Hypergolic (not cryo) '3rd stage' that uses a SpaceX's dragon trunk perimeter mount to attach to a FH S2 (for lighter weight and efficient load transfer) and this new S3 "stage" is topped with a smaller Ruag fairing (as probes tend to be smaller anyway)...

SpaceX gets contracted to launch ~50 tonnes to LEO with Boosters to ASDS's and Core expended...
S2 saves enough prop to deorbit also... SpaceX gets $95mil plus for the trouble and expenses...

SSL has literally tonnes of prop margin to play with... to push the paying customers probe out to it's destination...

Bottom line is SpaceX likely does not want to get directly into high energy C3 launches, as they are few and far between.
But if someone like SSL wanted to finish the job and make FH the 'go to, lowest cost to way out there' C3 king...
Then yeah... 

not workable.  How does the third stage and fairing interface with Spacex GSE?
Why bother with any Dragon related?  where is the data to support the claim "for lighter weight and efficient load transfer?
Who is going to pay for all the analysis of this drastically different configuration?  Who is going to qualified it for users?
What expertise does SSL or other satellite maker have on making upperstages, launch vehicle guidance systems and flight planning?

This makes no sense.

[LouScheffer]

Seems to me... 
...a known satellite maker (like say SSL) could clean up in the planetary probe marketplace (such as it is) by designing a high capacity Hypergolic (not cryo) '3rd stage'

not workable.  How does the third stage and fairing interface with Spacex GSE?
Why bother with any Dragon related?  where is the data to support the claim "for lighter weight and efficient load transfer?
Who is going to pay for all the analysis of this drastically different configuration?  Who is going to qualified it for users?
What expertise does SSL or other satellite maker have on making upperstages, launch vehicle guidance systems and flight planning?

This makes no sense.

Agree this makes no sense.  If it was really important to do this, use an existing upper stage that already does what you need.  Two obvious choices are Fregat and Centaur.

Fregat would be much easier technically.  It uses storable propellants and is already designed for horizontal integration.  However it is relatively low performance.

Centaur would be much better performance for more work.  You would presumably need fairing/transporter support during horizontal integration and tilting.  You need to get hydrogen and LOX (plus likely power and helium) to the stage once it's vertical (fortunately the crew access arm is at the right height.  Maybe you could use that).  You might need support like the forward load reactor supplies all the way into orbit, since the second stage of FH can accelerate the Centaur pretty strongly. 

Performance would be great, though.  Take Europa Clipper - 6000 kg probe,  Centaur mass 23077 kg fueled,  2247 kg empty, ISP= 450.  Then the delta-V is about 5557 m/s, of the 6750 m/s required by a worst-case launch window (according to NASA's trajectory browser)
So the FH has to push 29,077 kg to about 1200 m/s more then LEO.   This should be no problem at all, since it's rated to put 26,000 kg to GTO, which is LEO + 2400 m/s.   So this combination could push Europa Clipper direct to Jupiter at a worst case window with margin to spare.

However, as Jim pointed out, mundane aspects would rule this out.  Who would pay for it (both stage and GSE) and qualify it?  Why should the two competitors cooperate, short of a national mandate?  And, as far as I know, there are no other missions that require this level of performance.

[IainMcClatchie]

Seems to me... 
...a known satellite maker (like say SSL) could clean up in the planetary probe marketplace (such as it is) by designing a high capacity Hypergolic (not cryo) '3rd stage'

I think both hypergols and hydrolox for anything exoatmospheric have been obsoleted by ion engines at this point.

Seems to me they could make a standardized ion drive "3rd stage", intended to be mounted to a Falcon and delivered to 400 km LEO.  This would be something like a satellite bus, standardized to boost from LEO to GEO.  If I scale up from the XIPS-25 drive and the Juno solar panels, I get this:

LEO mass: 15500 kg (400 km 28.5 deg LEO orbit delivered by Falcon 9 or Atlas 531 or 541)
Empty weight: 4800 kg, 4000 kg of which is solar panels.
Solar panel power: 150 kW at Earth
Thrust: 5.5 N
Exhaust velocity: 34800 m/s
Prop flow: 0.158 g/s
Xenon capacity: 3600 kg

Destination              GEO      Mars    Jupiter   Saturn
Delta-V       m/s       5000      4600       7500     9250
  (manoever)  m/s       1000       100        500      500
Payload        kg       8626      8781       7695     7082
Propellant     kg       2074      1919       3005     3618
Burn time   months      5.06      4.68       7.34     8.83
Power there    kW        150        65        5.5      1.9
Isp equiv     sec       1211      1150       1522     1685

The nice thing is that after the boost is done, you have 150 kW of power, derated for your destination of course.  For many payloads that's plenty, so those payload numbers don't need to include much weight for power.

I sized the panels really large to make the GEO boost time reasonable.  These panel sizes make the power available at Jupiter (5.5 kW) and even Saturn (1.9 kW) enough to rendezvous and manoever with, given a second, smaller, lower power ion engine with an exhaust velocity around 10 km/s.

I put in a line which shows the Isp that a chemical rocket would need to match performance.  Note the chemical rocket doesn't leave you with 150 kW when it's done.

I figure this "third stage" would be attractive for geosync comsats, low-flying spysats, any planetary probes, and maybe even the odd sample return from NEO mission.  If you can wait five months to put your GEO bird into action, you get more mass on station than any chemical rocket.  If you can't wait that long, you could always put larger solar panels on it, although there is probably some point around 2-3 months where it becomes more efficient to use a rocket.

[speedevil]

I think both hypergols and hydrolox for anything exoatmospheric have been obsoleted by ion engines at this point.

Seems to me they could make a standardized ion drive "3rd stage", intended to be mounted to a Falcon and delivered to 400 km LEO.  This would be something like a satellite bus, standardized to boost from LEO to GEO.  If I scale up from the XIPS-25 drive and the Juno solar panels, I get this:

You seem to be assuming continuous thrusting, is this the case?
If so, there is a large penalty for this - are you taking it into account?
(Plus, the obvious fact you can't continuously thrust till you're out of earths shadow)

Wholly reusable launchers with propellant transfer in orbit can compete with this, to a large degree, though I agree it's an interesting concept.

[IainMcClatchie]

You seem to be assuming continuous thrusting, is this the case?
If so, there is a large penalty for this - are you taking it into account?
(Plus, the obvious fact you can't continuously thrust till you're out of earths shadow)

I did assume continuous thrusting and did not account for the penalty (it's a simple spreadsheet).  The burn times are significantly smaller than the total transfer time.  Quite a bit of the time, in all cases, is just getting out of Earth's gravity well.  I suspect the burn time can't get much larger than 1.5 years, because even if you start with a Venus flyby, you can't stay in the inner solar system while burning much longer than that.

Wholly reusable launchers with propellant transfer in orbit can compete with this, to a large degree, though I agree it's an interesting concept.

Reusable + transfer can get awesomely huge payloads, but it seems to me they will always have higher cost of launch.  I fully agree that there are missions for which bigger payloads or timeliness are much more important than cost/kg of launch.

But, for instance, cargo to Mars would benefit from an ion drive third stage.  And cargo is most of what's going there.

[speedevil]

Reusable + transfer can get awesomely huge payloads, but it seems to me they will always have higher cost of launch.  I fully agree that there are missions for which bigger payloads or timeliness are much more important than cost/kg of launch.

But, for instance, cargo to Mars would benefit from an ion drive third stage.  And cargo is most of what's going there.

Note that Xenon costs $1000/kg or so, and the global market is a few tens of tons.
One 150 ton payload launch to Mars would use easily the whole global capacity for Xenon, and bringing online more capacity isn't obviously straightforward.

Xenon has an effective ISP of 1200 or so.

Clearly, there are alternatives to Xenon, but at least for 'small' delta-v of 4-8km/s or so, ion engines are not an obviously clear win against cheap propellant in orbit.

At 4km/s delta-v, if the Xenon costs $1000/kg or so, LOX/methane may beat it pricewise if it costs under $250/kg or so.
FH in non-reusable S2 form cannot get launch costs quite this low.

One of the things that would make a platform like the above described very interesting is if there is any water ice in shadowed craters on phobos or deimos.

[IainMcClatchie]

Note that Xenon costs $1000/kg or so, and the global market is a few tens of tons.

There are a bunch of startups looking at alternative propellants for ion engines right now (e.g. molten salt, and I think metals).  The ESA has recently launched a satellite to test an ion engine that uses oxygen and nitrogen scooped from the atmosphere as it's propellant.  I'm not sure about the scooping as a viable technology, but if oxygen can be used as propellant for a year long burn, then ion drive propellant is not a problem.

Clearly, there are alternatives to Xenon, but at least for 'small' delta-v of 4-8km/s or so, ion engines are not an obviously clear win against cheap propellant in orbit.

4-8 km/s delta-V is the interesting range of exoatmospheric boost.  The ion drive 3rd stage I roughed out looked like an obvious clear win across the range, assuming you can wait months to get where you are going.  I totally understand that for manned missions it's not a good match.

If oxygen as propellant works, I'd be tempted to suggest that trips both to AND FROM Mars would benefit from (a) multiple tanker trips to a gradually fuelled-up tanker in orbit and (b) that tanker in orbit gathers energy by using an ion drive to boost itself to a high orbit.  When it's time for people to go from e.g. Mars to Earth, they lift off, rendezvous with the fully fuelled tanker in high orbit, and then use a chemical burn to get to Earth.  Boosting chemical fuel most of the way out of a gravity well with an ion drive looks like a pretty awesome way to bank energy that you want to use quickly months later.

[Herb Schaltegger]

This is all intellectually fascinating but decidedly off-topic for Falcon Heavy per se.

[LouScheffer]

[OK, now that we have the performance curve, who can calculate the exact  Falcon9/FH Upper stage dry mass from these numbers?

Here we go.  There is an analysis in L2, but it neglects several factors.  First, the second stage has to get to LEO, so the on-orbit mass will depend on the payload.  Second, the payload mass will affect the first stage, at least somewhat.  A quick guess shows this effect is 10-20 m/s per 1000 kg of payload, so it cannot be ignored.

Since the second stage has to get both to LEO, and do LEO to escape, there is no need to consider LEO at all.  So you can just back calculate the speed at orbital altitude that corresponds to a given C3.  Then the first stage and second stage combined must offer that.  This simplifies the calculation.

Next, performance of the first stage is hard to estimate.  So I just represent it as the speed with 0 payload on top of the second stage, plus a simple linear derating - so many less m/s for each additional kg of payload.

For raw data I went to the NASA LSP web site, and typed in C3 from 0 to 100 by 10, and recorded the FH payload for each C3.

Finally, I put this in a spreadsheet where you enter the apogee of the injection burn (166 km for the NASA numbers), the first stage contribution and penalty, then the second stage starting and ending mass.  From this it calculates the total delta-V compared to the NASA delta-V.  Then you minimize the differences, helped by the Excel solve function.

This simple model fits the data ASTONISHINGLY well.   For each payload mass, it predicts the final velocity with less than 1.5 m/s of error, out of about 12,000 m/s.  That's about one part in 10,000.

So what do we get?   The first stage (side + core) must provide about 5252 m/s.  This includes Earth rotation, but is still way above any recoverable speed, as expected.  The first stage + side boosters provide about 16 m/s less per extra tonne of payload.

The second stage starts at 109 tonnes + payload, and ends at 5537 kg + payload.   The 5537 kg includes the dry mass, the residuals, and the payload adapter.  Both of these are consistent with earlier estimates.  Musk talked about the first stage lifting 125 tonnes, which could be 109 (stage) + 12 (payload) + 4 (fairing).  Also, the stage has about 100 tonnes of fuel, so 1% residuals would be 1 tonne, so the dry mass would be about 4.5t.   So all is consistent.

In terms of sensitivities, you can go down to 100t start, 5451 kg end, and still get a quite good fit (3 m/s error). Likewise you can go up to 116t start, 5598 kg end, with a similar error.  I don't think you can go outside these values and still be consistent with Musk's 125t total for the second stage + payload.  So 5450-5600 kg at end of burn seems like the plausible range.

[envy887]

[OK, now that we have the performance curve, who can calculate the exact  Falcon9/FH Upper stage dry mass from these numbers?

Here we go.  There is an analysis in L2, but it neglects several factors.  First, the second stage has to get to LEO, so the on-orbit mass will depend on the payload.  Second, the payload mass will affect the first stage, at least somewhat.  A quick guess shows this effect is 10-20 m/s per 1000 kg of payload, so it cannot be ignored.

Since the second stage has to get both to LEO, and do LEO to escape, there is no need to consider LEO at all.  So you can just back calculate the speed at orbital altitude that corresponds to a given C3.  Then the first stage and second stage combined must offer that.  This simplifies the calculation.

Next, performance of the first stage is hard to estimate.  So I just represent it as the speed with 0 payload on top of the second stage, plus a simple linear derating - so many less m/s for each additional kg of payload.

For raw data I went to the NASA LSP web site, and typed in C3 from 0 to 100 by 10, and recorded the FH payload for each C3.

Finally, I put this in a spreadsheet where you enter the apogee of the injection burn (166 km for the NASA numbers), the first stage contribution and penalty, then the second stage starting and ending mass.  From this it calculates the total delta-V compared to the NASA delta-V.  Then you minimize the differences, helped by the Excel solve function.

This simple model fits the data ASTONISHINGLY well.   For each payload mass, it predicts the final velocity with less than 1.5 m/s of error, out of about 12,000 m/s.  That's about one part in 10,000.

So what do we get?   The first stage (side + core) must provide about 5252 m/s.  This includes Earth rotation, but is still way above any recoverable speed, as expected.  The first stage + side boosters provide about 16 m/s less per extra tonne of payload.

The second stage starts at 109 tonnes + payload, and ends at 5537 kg + payload.   The 5537 kg includes the dry mass, the residuals, and the payload adapter.  Both of these are consistent with earlier estimates.  Musk talked about the first stage lifting 125 tonnes, which could be 109 (stage) + 12 (payload) + 4 (fairing).  Also, the stage has about 100 tonnes of fuel, so 1% residuals would be 1 tonne, so the dry mass would be about 4.5t.   So all is consistent.

In terms of sensitivities, you can go down to 100t start, 5451 kg end, and still get a quite good fit (3 m/s error). Likewise you can go up to 116t start, 5598 kg end, with a similar error.  I don't think you can go outside these values and still be consistent with Musk's 125t total for the second stage + payload.  So 5450-5600 kg at end of burn seems like the plausible range.

Keep in mind that the dry mass calculated from LSP numbers will include residuals to allow the some performance reserve to meet LSP margins. SpaceX's published figures are probably for a burn to depletion. I think that better explains the ~1,000 kg difference between this method and other methods for backing out the mass of the upper stage.