Gwynne Shotwell on The Space Show 3/21

[hrissan]

Also remember that When Shotwell was talking reusability, she appears to be only talking about first stage reusability. 2nd stage reuse is a bit down the road.

I'm not sure that's accurate. I got the impression with FH she was talking about full IE all stage recoverability, and that with full resue it could get the biggest GEO comm sats to their orbit and still bring all stages back.

How they will get those upper stages (especially the 3rd stage) will be recovered is of course anyone's guess. 

When she was speaking about FH reusability she said "numbers for heavy are with first stages reuse", I assume first stages are 3 FH cores, but not upper stage.

[Hyperion5]

Listening to Ms. Shotwell cleared up a misconception I had about payload capacity v.s. reusability. I had assumed that implementing reusability would reduce payload capacity from those posted on the SpaceX website. It turns out I was wrong. Falcon 9 and Falcon 9 Heavy have approximately 30% more payload capacity than what is posted. She said that all that additional capacity is reserved for recovery and reuse requirements. She said that the posted capacities are for the reusable launch vehicles, not the expendables. She also said that the only thing that would change once reuse becomes standard is a lowering of the prices for launch. Wow.

Before you put that belief on the books I encourage you to run the numbers through the rocket equation and ask yourself what really makes sense.  Even assuming full expendability, the advertised numbers require unprecedented or nearly unprecedented mass ratios.  Squeezing reuse out of that I can't see without payload reduction.

 - Ed Kyle

It wouldn't take that much effort to get unprecedented or nearly unprecedented mass ratios with all-kerolox rockets, Ed.  It's not like the competition is exactly fierce when it comes to weight-saving designs out there.  There are exactly zero all-kerolox Russian or Chinese LVs with common bulkheads on all stages to the best of my knowledge, and both the Atlas V and Ariane 5 use common bulkheads only on their hydrolox stages.  Add in lithium-aluminum alloy construction (rarer than it should be), friction stir-welding, computer-aided design, a composite PLF, and engines with unprecedented t/w ratios, and you've got a fantastic recipe for great mass ratios.  Spacex can thus achieve "unprecedented" mass ratios simply because the competition has not pushed very hard when it comes to saving weight on all-kerolox rockets.  This is a big part of the reason how Elon Musk can brag about the 30:1 mass ratio on the Falcon Heavy boosters, which he rather unfairly compares with the Delta IV Heavy's boosters' mass ratio. 

[Coastal Ron]

Listening to Ms. Shotwell cleared up a misconception I had about payload capacity v.s. reusability. I had assumed that implementing reusability would reduce payload capacity from those posted on the SpaceX website. It turns out I was wrong. Falcon 9 and Falcon 9 Heavy have approximately 30% more payload capacity than what is posted.

I know she said that about Falcon 9, but I don't remember her stating it for Falcon Heavy.  Falcon Heavy has a lot of variations that they can implement with reusability - and we still have not heard any updates on cross-feed that I know of, so I'm not sure how that still fits in.

[DavidH]

  This is a big part of the reason how Elon Musk can brag about the 30:1 mass ratio on the Falcon Heavy boosters, which he rather unfairly compares with the Delta IV Heavy's boosters' mass ratio.

Can you expand on this? Are you saying the extra size of the H2 tank makes the MR higher?

[Jim]

Add friction stir-welding, computer-aided design, a composite PLF,

Those exist on all US boosters

[Hyperion5]

Add friction stir-welding, computer-aided design, a composite PLF,

Those exist on all US boosters

You know, Jim, I was trying to emphasize the common bulkheads and lithium-aluminum alloy construction, which is why I mentioned "add" in my statement.  The point was that some of the foreign competitors lack even those bits in their design and construction, which is partly why it is not particularly hard to best their mass ratios.  I wasn't just comparing the Falcon 9 to the Antares, Atlas V or the Delta IV family. 

[Hyperion5]

  This is a big part of the reason how Elon Musk can brag about the 30:1 mass ratio on the Falcon Heavy boosters, which he rather unfairly compares with the Delta IV Heavy's boosters' mass ratio.

Can you expand on this? Are you saying the extra size of the H2 tank makes the MR higher?

David, I believe you meant to say, "Are you saying the extra size of the H2 tank makes the MR lower?" The answer is yes, the extra size makes the mass ratio lower.  There is one other reason, besides tank size, that causes the big difference in mass ratios.  Namely the fact that hydrogen is a "deep cryogen" that likes to be 20 degrees above absolute zero while kerosene can be kept at room temperature.  This difference means that you also have to add a large amount of insulation to hydrogen tanks to keep them from turning into a gas.  This adds mass that you won't find on a kerosene tank, which helps the kerolox rocket achieve a better mass ratio. 

However, when it comes to tank size, there's one thing that really matters: impulse density (or the amount of bang you get out of a given volume of propellants).  Generally speaking, hydrolox rockets have terrible impulse density compared to kerolox rockets like the Falcon 9.  Presuming you were able to design a kerolox and a hydrolox rocket with an identical payload capacity, the kerolox rocket's average propellant density, or bulk density, would be dramatically higher than that of the hydrolox rocket.  Here's a good site for getting a better understanding: http://settlement.arc.nasa.gov/Nowicki/SPBI1LF.HTM. 

While I may not be a rocket scientist like Jim, some basic math makes it easy enough to figure out why a hydrolox rocket has an inferior mass ratio to a kerolox rocket.  In the example on the site provided, the bulk density, or the average density of the propellants, is provided for both a hydrolox rocket and a kerolox rocket.  I've converted down to liters, or 1/1000th of a cubic meter, in order to make the math thereafter easier to follow. 

Hydrolox rocket (6.0 O/F): .358 kg/liter (10 cm^3)
Kerolox rocket(2.5 O/F): 1.026kg/liter (10 cm^3)

So basically, if you were to measure them against one another, the kerolox rocket would mass 2.866X as much per unit of propellant tank volume as the hydrolox rocket of identical capability.  That would be fine for the hydrolox rocket IF it could compensate via increased efficiency.  Here's the relative efficiency of these two rockets' engines (each with identical chamber pressure). 

Hydrolox rocket Specific Impulse: 455.9 seconds
Kerolox rocket Specific Impulse: 354.6 seconds

We multiply Isp by the Earth's gravitational pull to get the exhaust velocity (v_e):
455.9x9.807 m/s=4471 m/s
354.6x9.807 m/s=3477.6 m/s

We then need to multiply these exhaust values by the bulk densities of each engine's propellant mixture to get the impulse density:

Impulse Density Id = v_e(exhaust velocity)*d_p(bulk density)
1600.6 Ns/L = 4471 m/s*.358 kg/L (hydrolox rocket)
3568 Ns/L = 3477.6 m/s *1.026 kg/L (kerolox rocket)

So you see, despite all of the hydrolox rocket's added efficiency, the kerolox rocket will still pack 2.23X as much delta-v into a liter of propellants.  This means your hydrolox rocket must have tanks at least 2.23X as large to compensate, and more if you factor in the added mass of the tanks and their insulation.  This is what I meant about the comparison between the Delta IV Heavy boosters' mass ratio and the FH boosters' MR ratio not being fair.  If you're using hydrogen, then clearly you aren't using it because it will give a good mass ratio.  You're using it because it will let you build a much lighter, more efficient rocket to lift the same amount of payload to orbit.  It should also be noted that due to their greater efficiency, mass ratios are not as critical for hydrolox rockets as they are for kerolox rockets.  The Delta IV Heavy can afford to have boosters with only an MR ratio of 10, while the Falcon Heavy needs an MR ratio of 30 (and a few other tricks) in order to push its full-size payloads to orbit. 

[CuddlyRocket]

Falcon 9 and Falcon 9 Heavy have approximately 30% more payload capacity than what is posted.

43% more! (Reusability is said to reduce payload capacity by 30% from the expendable figure, so if you start from the reusable figure the expendable payload capacity is a 43% increase.)

[john smith 19]

It wouldn't take that much effort to get unprecedented or nearly unprecedented mass ratios with all-kerolox rockets, Ed.  It's not like the competition is exactly fierce when it comes to weight-saving designs out there.  There are exactly zero all-kerolox Russian or Chinese LVs with common bulkheads on all stages to the best of my knowledge, and both the Atlas V and Ariane 5 use common bulkheads only on their hydrolox stages.  Add in lithium-aluminum alloy construction (rarer than it should be), friction stir-welding, computer-aided design, a composite PLF, and engines with unprecedented t/w ratios, and you've got a fantastic recipe for great mass ratios.  Spacex can thus achieve "unprecedented" mass ratios simply because the competition has not pushed very hard when it comes to saving weight on all-kerolox rockets.  This is a big part of the reason how Elon Musk can brag about the 30:1 mass ratio on the Falcon Heavy boosters, which he rather unfairly compares with the Delta IV Heavy's boosters' mass ratio.

It's a subtle point but I think FSW has made the use of AlLi alloys much easier. The issue with welded joints has always been weld "efficiency" IE weld strength /parent metal strength.  This is especially important when you you have to do rework and how thick the plate you have to start with (or if you have to add stiffener plates around the weld areas) in order to ensure you still have a strong enough joint after the second welding attempt (or possibly 3rd  ).

IIRC the Shuttle ET might have 100 inches of rework and NASA expected a welded joint to be 70% efficient OTOH IIRC Spacex have claimed FSW has required no rework. I recall that a recent biz jet design also went from a composite fuselage to an FSW Al alloy and cited no rework as well.

The other benefit is that FSW is claimed to deliver near parent metal strength and as the metal is not actually melted (I've always thought "welding is a bit of a misnomer  ) the crystal structure is much better preserved and (relatively) volatile elements are not vaporized out.

[clongton]

Falcon 9 and Falcon 9 Heavy have approximately 30% more payload capacity than what is posted.

43% more! (Reusability is said to reduce payload capacity by 30% from the expendable figure, so if you start from the reusable figure the expendable payload capacity is a 43% increase.)

I quoted Ms. Shotwell.

[Llian Rhydderch]

It wouldn't take that much effort to get unprecedented or nearly unprecedented mass ratios with all-kerolox rockets, Ed.  It's not like the competition is exactly fierce when it comes to weight-saving designs out there.  There are exactly zero all-kerolox Russian or Chinese LVs with common bulkheads on all stages to the best of my knowledge, and both the Atlas V and Ariane 5 use common bulkheads only on their hydrolox stages.  Add in lithium-aluminum alloy construction (rarer than it should be), friction stir-welding, computer-aided design, a composite PLF, and engines with unprecedented t/w ratios, and you've got a fantastic recipe for great mass ratios.  Spacex can thus achieve "unprecedented" mass ratios simply because the competition has not pushed very hard when it comes to saving weight on all-kerolox rockets.  This is a big part of the reason how Elon Musk can brag about the 30:1 mass ratio on the Falcon Heavy boosters, which he rather unfairly compares with the Delta IV Heavy's boosters' mass ratio.

It's a subtle point but I think FSW has made the use of AlLi alloys much easier. The issue with welded joints has always been weld "efficiency" IE weld strength /parent metal strength.  This is especially important when you you have to do rework and how thick the plate you have to start with (or if you have to add stiffener plates around the weld areas) in order to ensure you still have a strong enough joint after the second welding attempt (or possibly 3rd  ).

IIRC the Shuttle ET might have 100 inches of rework and NASA expected a welded joint to be 70% efficient OTOH IIRC Spacex have claimed FSW has required no rework. I recall that a recent biz jet design also went from a composite fuselage to an FSW Al alloy and cited no rework as well.

The other benefit is that FSW is claimed to deliver near parent metal strength and as the metal is not actually melted (I've always thought "welding is a bit of a misnomer  ) the crystal structure is much better preserved and (relatively) volatile elements are not vaporized out.

John, that was very useful input on the topic of friction stir welding and weld strength, and the economics of rework etc.  I'd like to read more about it.  Might you have any sources or links you could point interested readers to?

[Elmar Moelzer]

I'm not taking it to the bank just yet Ed. But what it does tell me is that SpaceX is not planning on loosing performance to LEO because of reusability. We'll see how that works out.

They are loosing performance. The numbers on their website have simply factored that loss in already. So the actual performance of an expendable F9 is 30% higher than given on the website. But I do wonder whether Shotwell means that the numbers include reuse of the first stage only and not the second stage (which would put everything a bit more in line with what we have heard and calculated previously). But it would be quite cool if that was for reuse of first and second stage.

[R7]

Impulse Density Id = v_e(exhaust velocity)*d_p(bulk density)
1600.6 Ns/L = 4471 m/s*.358 kg/L (hydrolox rocket)
3568 Ns/L = 3477.6 m/s *1.026 kg/L (kerolox rocket)

So you see, despite all of the hydrolox rocket's added efficiency, the kerolox rocket will still pack 2.23X as much delta-v into a liter of propellants.  This means your hydrolox rocket must have tanks at least 2.23X as large to compensate, and more if you factor in the added mass of the tanks and their insulation.

It is not that simple. The formulas the your methane F9 family thread apply here too. You can't equate plain impulse density value directly to delta-v. Half of it (density) is inside the logarithm of rocket equation and you'd need more info to calculate actual desired delta-v. Or vice versa, you peg desired delta-v from which you calculate tank volume, or the handy "specific volume" (propellant volume divided by burnout mass)  for different propellants to compare.

For example if there's upper stage needing to do 6000m/s dv then your values give specific volume of 4.496L/kg for kerolox and 7.896L/kg for hydrolox, so volume ratio is 1.756

Lower stages become more complex because reduced upper stage weights need to be taken into account. A first stage doing 3500m/s needs "only" 1.462 times larger tanks using hydrolox with above upper stage example adjusted for its lesser wet mass, assuming it's 80% of first stage burnout mass. Without adjusting the ratio would be 1.961

Yes this is too simple and does not take into account tank mass per volume, insulation, different engine T/Ws etc. just wanted to point out it is not directly proportional to impulse density. Calculated numbers with too-messy-to-upload mathcad sheet.

This is so OT isn'it it 

[clongton]

They are loosing performance.

Ms. Shotwell specifically said they are not loosing performance.
Do you know something that she does not?

[PerW]

I think It was very interesting, especially that Bigelow is a focus area and not just a client among others. Looking forward to hear more about this.

[Elmar Moelzer]

Ms. Shotwell specifically said they are not loosing performance.
Do you know something that she does not?

Did you actually read my entire post, or are you just cherry picking parts to criticize later? If you had read the entire post, this would be a lot clearer to you.

[Owlon]

They are loosing performance.

Ms. Shotwell specifically said they are not loosing performance.
Do you know something that she does not?

They are most certainly losing performance, they just already have the reduced performance posted on the website. That's exactly what she said.

[Robotbeat]

SpaceX's Falcon 9 is the newest clean-sheet launch vehicle with a very, very recent almost complete redo of the rocket. They are using basically the densest propellants available and have the flexibility of building basically everything in-house. They are very well-equipped to get record-breaking mass ratios. And for a boost-back first stage landing on legs, mass ratio is almost the only thing that matters. If you had a 100:1 mass ratio (unrealistic, but throwing that out there) for a first stage, your performance penalty for boost back and landing would be incredibly low. Heck, you'd get better performance out of a first stage with a 100:1 mass ratio that does boost-back than you would out of a first stage with 10:1 mass ratio that is expendable.

SpaceX has publicly mentioned planning on subcooling propellants, too. That's not completely unprecedented (NK-33 does it for oxygen for other technical reasons), but it probably has not been combined with the extremely good mass ratios they're already getting for their Al-Li alloy tanks (with a very, very lightweight thrust structure... especially compared to v1.0's essentially boiler-plate thrust structure)

I'm fairly certain nearly all of Shotwell's statements were about core reuse, not upper stage reuse.

[Lobo]

Since Gwynne said she believed Pad 39A would be too small for the BFG doesn't that implicitly confirm that it'll be three cores?

I'd say so as I don't think they are planning to build another larger core to be there "BFR" after the 10 core.

[Lobo]

Since Gwynne said she believed Pad 39A would be too small for the BFG doesn't that implicitly confirm that it'll be three cores?

It could also be a 18 Raptor 15m single stick. Or 12 Raptor with increased thrust. I think those could also not launch at 39A. Given RTLS constraints with Heavy central cores such a configuration seems to make more sense to me than a Heavy. But what do I know?

Probably tri-core...

http://www.nasaspaceflight.com/2014/03/spacex-advances-drive-mars-rocket-raptor-power/