Falcon Heavy - General Discussion

[Grendal]

An important comment made by Elon at the post launch conference that needs attention is that SpaceX is taking expendable launches off the table at current pricing. He specifically said that the reason FH is only $90 million is because the company can recover all the cores. So we know the next FH will be a Block 5 and that it will always be launched in reusable mode. That could change if someone wants to shell out some serious bucks to force expendability.

 

[KarenRei]

In context of a FSH--which is where this particular conversation started--the core, and specifically its ability to carry heavier and heavier loads becomes more and more of the limiting factor.

That's precisely my point - that it's a payload mass limit, not a fairing limit, for the vast majority of potential payloads. The main things we want super-heavies for today are not "unusually large objects", they're "unusually heavy" objects (particularly propellant as well as propulsion/power systems and their associated radiators) or "unusually numerous objects" (multiple payloads). And in the cases where a larger fairing is required - say, a large-mirror space telescope, it is not impossible to implement, as it's demonstrably been done before on many rockets. You can't have too high of a moment of inertia on the Z axis for the payload, but when you're talking about a space telescope, that's unlikely to be a problem unless the primary mirror is a high portion of the total system mass (Hubble's primary was only 7% of the telescope's total mass). And to reiterate: the more cores you have, the greater the stability of the first (high acceleration/high-Q) stage.

I'll reiterate my point that a larger core diameter and larger fairings would be nice, but they're not going to limit much except for "colony modules", and SpaceX has no plans to launch those on FH/FSH regardless. What FH/FSH brings to the table is the ability to launch highly capable probes with a lot of delta-V; heavy military satellites; whole constellation planes at once; etc. And that's a wonderful contribution. I'm very much a pessimist on BFR's timeline, and so I'm glad that FH (and possibly even FSH) will be an option over the next decade.

You'd be hard pressed to find a low orbit payload that a) maxes out the lift capacity, b) fits in a 5m fairing, and c) doesn't break the CLA.

You'd be hard-pressed to find a LEO payload today that requires that at all. What sort of single-object >4,6m 64k tonne LEO launches are you envisioning? Even whole ISS modules fit inside 4,6m. Now if we eliminate the "single object" constraint, there's a ton of things that fit the bill - most notably, multi-satellite launches. Is it even worth mentioning that SpaceX wants to launch a 11943-satellite constellation?

 

[Grendal]

As opposed to making a bigger rocket and the expense that would incur from that, how about multiple launches and combining pieces in orbit? SpaceX essentially does that now with the Dragon capsule and the ISS. Why can't you do one launch with a very heavy deep space satellite and a second launch with a booster/maneuvering payload? You park the satellite in orbit then bring up the booster payload, have it dock with the satellite, and two combined travel to wherever you want to explore. This seems a lot easier and a lot cheaper than the attitude of "we need the biggest rocket possible" to carry everything in a single payload. SpaceX now makes such a thing possible with their low prices and reusability. I also think that a booster maneuvering package is easier than "in orbit" refueling in certain circumstances and situations. Two FH launches is less than the price of one Delta IV Heavy. A lot less.

 

[mongo]

As opposed to making a bigger rocket and the expense that would incur from that, how about multiple launches and combining pieces in orbit? SpaceX essentially does that now with the Dragon capsule and the ISS. Why can't you do one launch with a very heavy deep space satellite and a second launch with a booster/maneuvering payload? You park the satellite in orbit then bring up the booster payload, have it dock with the satellite, and two combined travel to wherever you want to explore. This seems a lot easier and a lot cheaper than the attitude of "we need the biggest rocket possible" to carry everything in a single payload. SpaceX now makes such a thing possible with their low prices and reusability. I also think that a booster maneuvering package is easier than "in orbit" refueling in certain circumstances and situations. Two FH launches is less than the price of one Delta IV Heavy. A lot less.

Very interesting. If the sensor payload stays attached to the 2nd stage, the stage maneuvering systems could be used for attitude adjustment to allow docking with the booster/ propellant stage. Then there is no duplication of maneuvering thrusters. Thrust module could even be commonized. Need to add retro thrusters to slow the final approach.
On this setup, may make sense to launch thrust module first, since 2nd stage is not designed to hang out for days waiting in orbit. Also doesn't leave sensor section up there alone if there is an issue with second launch.

 

[bxr140]

As opposed to making a bigger rocket and the expense that would incur from that, how about multiple launches and combining pieces in orbit?

Yes, that's what I meant earlier when I said "Especially with successful 'smaller' products like F9 and FH that could stage super heavy lifts".

Its not a new concept...I mean, that's what space stations are, and that's what pretty much any deep space mission looks like.

There's a lot of interest these days on the commercial payload side for persistent platforms too--same deal. Its also not a new concept but definitely something that's being enabled by cheaper [and ostensibly more reliable] launchers. Basically, you huck a spaceraft bus into orbit that has all the boring stuff like power, ACS, TC&R, etc, and then you bolt some number of unique payload modules onto commonized interfaces that facilitate future R&R with better/faster/cheaper modules. In the geocomm world you could have, for instance, one ~stock 15 year bus and cycle through a number of 5-7 year payload modules. Those modules can better keep up with technology AND no longer have to be designed to survive the defacto industry standard of 15 years EOL. Its a win for everyone. Theoretically. I'll let you know when someone buys one.

 

[bxr140]

That's precisely my point - that it's a payload mass limit, not a fairing limit, for the vast majority of potential payloads.

In context, the mass problem and a volume problem are quite correlated. Again, spacecraft are already pretty optimized for launch loads and even with current launchers there are still variables that often push up against the limits (like mass or CG or primary mode). There's not a whole lot of practical design space to go 'up' because you just make all of those variables worse, so you have to go out. The first step in "out" is a wider adapter in the existing fairing (like you'd find in a ISS module), but eventually the fairing needs to get wider so you can push the CG back down an build a compatible satellite structure. Of course, there's the design constraint on the 'too wide' side of that space also, but I guess I won't rehash that...

What sort of single-object >4,6m 64k tonne LEO launches are you envisioning?
Now if we eliminate the "single object" constraint, there's a ton of things that fit the bill - most notably, multi-satellite launches.

That's exactly right. Its the constellation launches that are the problem. F9 is a problem because of its mass limits, FH is a problem because of its fairing limits. You can fill a FH fairing with dozens of spacecraft to max out the mass, but that assembly is so tall and heavy and skinny that its really hard to make the whole thing survive the launch environment.

That's not hypothetical, BTW. That's every constellation these days. I'm not just making this *sugar* up. Honest.

You'd be hard-pressed to find a LEO payload today that requires that at all.

So what do I win?

 

[mongo]

In context, the mass problem and a volume problem are quite correlated. Again, spacecraft are already pretty optimized for launch loads and even with current launchers there are still variables that often push up against the limits (like mass or CG or primary mode). There's not a whole lot of practical design space to go 'up' because you just make all of those variables worse, so you have to go out. The first step in "out" is a wider adapter in the existing fairing (like you'd find in a ISS module), but eventually the fairing needs to get wider so you can push the CG back down an build a compatible satellite structure. Of course, there's the design constraint on the 'too wide' side of that space also, but I guess I won't rehash that...



That's exactly right. Its the constellation launches that are the problem. F9 is a problem because of its mass limits, FH is a problem because of its fairing limits. You can fill a FH fairing with dozens of spacecraft to max out the mass, but that assembly is so tall and heavy and skinny that its really hard to make the whole thing survive the launch environment.

That's not hypothetical, BTW. That's every constellation these days. I'm not just making this *sugar* up. Honest.



So what do I win?

Practically, other than Starlink with 4k+ satellites (in low orbits), how many sats per orbital plane do you want to have? Or, how much fuel per sat are you willing to expend to alter the plane?

 

[KarenRei]

That's exactly right. Its the constellation launches that are the problem. F9 is a problem because of its mass limits, FH is a problem because of its fairing limits.

No, it's simply not a problem. There's something like 180 cubic meters of volume inside a standard fairing (let alone a lengthened one - length is what you want for satellite dispensers, not diameter). An OG2 satellite, to pick one, takes up 0,5m³ (stowed and attached to a dispenser) and weighs 172kg - a perfect match for payload and volume. Iridium NEXT satellites 0,084m³ volume and 50kg, aka 70% more weight than could be launched if the fairing was completely utilized. I would expect SpaceX's constellation satellites to be even denser, although right now all we have to go on is an ambiguous 4m x 1,8m x 1,2m figure which doesn't explicitly state that that's the deployed volume (although it apparently doesn't count solar cells).

Of course you can't ever completely utilize the volume in practice because of shape constraints and the dispenser volume. But the dispenser also adds mass, too, not simply taking up volume; you'd have no hope of actually launching the hypothetical maximum number of satellites even if there were no fairing size constraints at all. You can expect SpaceX's satellites and their dispenser to be co-designed to maximize use of the fairing volume rather than just using standard ESPA rings not custom designed to the fairing and satellites. And to reiterate, we're discussing the standard fairing, not a lengthened one (SpaceX is currently rather unusual among launch providers in not offering a lengthened fairing option). If they need a longer fairing for their dispenser, they'll make it.

Large payloads are even easier than multi-satellite launches. Let's say you wanted to launch a Cassini equivalent to the Kuiper Belt. What do you have to add? One or more boost stages. If chemical propellant, your're talking about a cylinder with an average density of 1500-2000kg/m³ (standard fairing min "payload density" for full utilization to LEO = ~355 kg/m³). Ion propulsion will approximate a cylinder (foldable radiators required), and while its propellant will be lighter, its structure will be heavier (nuclear reactors are not low density objects).

Note also that we're discussing launching to LEO here. Anything over the minimal takes extra dV and reduces your available payload mass, but has no bearing on fairing volume.

 

[bxr140]

Practically, other than Starlink with 4k+ satellites (in low orbits), how many sats per orbital plane do you want to have? Or, how much fuel per sat are you willing to expend to alter the plane?

I'm going to dodge that a bit because, like, that's what I do, and instead I'll cop out with the less than useful 'it depends' answer. It really is all about the constellation closing the business case, and the reason you don't currently see a bunch of high performing constellations doing that is because of a) current space technology and b) current launcher capacity. That's simply the reality of the space business today.

On one end of the spectrum, tiny spacecraft--say, 50-100kg--are limited in power/dissipation and propulsion (and thus performance and maneuverability), so while you can efficiently package them for launch, you need a ton of them. The other massive problem is the industrialization effort required to build that kind of thing because you need thousands of units on orbit. IRR is a real thing.

On the other end of the spectrum bigger spacecraft--say, 500-100kg--those issues mostly invert (though from a practical perspective you're always limited by power/dissipation) with two major benefits. Frist, you can stuff performance in that bigger package that you simply can't beat by from summing up smaller packages--you can't put a v8 in a volkswagen beetle, and two volkswagen beetles don't equal a v8...so to speak. The other is that your supply chain may only need to jump one order of magnitude instead of two. I've babbled about this problem in this forum before, but a space supply chain that can support constellations with reliable products is a major topic of discussion in industry right now. When you pareto the total cost of a constellation, supply chain is way above everything else.

Oneweb is about the practical low end limit of what a really useful spacecraft can do with current and near term technology, and they're struggling. Starlink is super cool because it doesn't use traditional technology and they're being pushed by a company that doesn't have to deliver on quarterlies. Unfortunately, if history is any judge of Elon time, we won't see a practical network for quite a bit longer than whatever they're advertising these days...

 

[bxr140]

No, it's simply not a problem.

Lol. If it were simple....

I honestly do appreciate and share your optimism. And you’re right to push—there’s definitely a lot of FUD and conservatism in the industry, which makes it imperative to question. No joke, “it’s not a problem, make it work” is practically my go-to when I get things like “the packaging doesn’t work” or “the structure won’t survive” for these kinds of missions.

Alas, while dreaming about what could be and waxing philosophic about space stuff on the Internet is fun, I accept that ultimately I am still a slave to reality (and someone else’s money).

So...

No, it's simply not a problem.

I respectfully [if not unfortunately] disagree.

 

[Grendal]

A big twitter discussion on FH by Elon, Doug Ellison, and Scott Manley with lots of important details about FH:

Elon Musk on Twitter

Elon:
The performance numbers in this database are not accurate. In process of being fixed. Even if they were, a fully expendable Falcon Heavy, which far exceeds the performance of a Delta IV Heavy, is $150M, compared to over $400M for Delta IV Heavy.

When asked about improving the second stage:
Under consideration. We’ve already stretched the upper stage once. Easiest part of the rocket to change. Fairing 2, flying soon, also has a slightly larger diameter. Could make fairing much longer if need be & will if BFR takes longer than expected.

Both exhaust velocity (Isp) and mass ratio drive the rocket equation. Also thrust/mass matters a lot for Oberth effect. Delta upper stage Isp is good, but mass ratio and thrust are not.

Elon discusses FH expending the center core:
Side boosters landing on droneships & center expended is only ~10% performance penalty vs fully expended. Cost is only slightly higher than an expended F9, so around $95M.

Discussing comparative rockets cost when told a Delta IV Heavy is less than $400 million:
That was three years ago, before ULA cancelled all medium versions of Delta IV. Future missions have all Delta fixed costs piled on, so their cost is now $600M+ for missions contracted for launch after 2020. Nutty high.

When someone mentions Vulcan as competition:
Maybe that plan works out, but I will seriously eat my hat with a side of mustard if that rocket flies a national security spacecraft before 2023.

Tory Bruno commented:
Hey @elonmusk , congrats again your heavy launch. Clarification: Delta IV Heavy goes for about $350M. That’s current and future, after the retirement of both Delta IV Medium and Delta II. She also brings unique capabilities, At least until we bring Vulcan on line.

Here you go:

 

[KarenRei]

Favourite part:

Under consideration. We’ve already stretched the upper stage once. Easiest part of the rocket to change. Fairing 2, flying soon, also has a slightly larger diameter. Could make fairing much longer if need be & will if BFR takes longer than expected.

Volume is not a problem.

 

[dhanson865]

FH has multiple recoverability options

* all 3 cores Return to landing site (RTLS) - lowest payload capacity
* 2 side cores RTLS, center core lands on an Autonomous Spaceport Drone Ship (ASDS) - next higher capacity

both of the first two options will see use for varying cargo loads and destinations.

* all 3 cores expendable - not likely cost would be prohibitive, but it would give maximum capacity

somebody might pay for this, but it'll be rare and expensive, may not happen, but definitely part of the advertised price list.

Those are known guaranteed options, now the possible what could they do?

* 2 side cores land on ASDS - center core expendable - not likely would require more drone ships and would slow launch cadence and increase cost
* 2 side cores expendable - center core lands on ASDS - not much gain in cargo capacity vs the RTLS option, not worth the cost difference?

The 2 side cores expendable could be a way to use up old block 4 first stages since they have so many but I'd assume they would rather have those as insurance (falcon 9 production stops at some point and they need a stock pile of usable boosters for contract missions while switching over to BFR/ITS production. So maybe they just keep a few in reserve in case the transition lasts longer than they expect).

I think you'll only see the top two RTLS options covering 90% of the flights, none of the expendable variations are likely but if they go there I'd say it'd be for an all out everything goes launch. You either get 3 cores back or you send 3 cores out never to come back, nothing in between.

Go figure Elon contradicts me and the truth is:

Side boosters landing on droneships & center expended is only ~10% performance penalty vs fully expended. Cost is only slightly higher than an expended F9, so around $95M.


...a fully expendable Falcon Heavy, which far exceeds the performance of a Delta IV Heavy, is $150M


Fairing 2, flying soon, also has a slightly larger diameter. Could make fairing much longer if need be & will if BFR takes longer than expected.

A third is under construction ... A Shortfall of Gravitas

 

[bxr140]

Favourite part:
Volume is not a problem.

Yup, larger diameter.

 

[KarenRei]

Yup, larger diameter.

The larger diameter that I stated that they could make whenever they wanted?

And in the cases where a larger fairing is required - say, a large-mirror space telescope, it is not impossible to implement, as it's demonstrably been done before on many rockets.

Also wrote:

And to reiterate, we're discussing the standard fairing, not a lengthened one (SpaceX is currently rather unusual among launch providers in not offering a lengthened fairing option). If they need a longer fairing for their dispenser, they'll make it.

Was told that I wasn't being realistic, that the core was just too narrow for larger payloads. Hmm...

Back to my original point: pretty much the only thing that the current core diameter will prevent them from launching are large colony modules (combination of very high width plus high Z moment of inertia). There's no other reasonable payload that the FH stack can't launch. And SpaceX doesn't plan to launch colony modules on FH, so...

 

[bxr140]

Was told that I wasn't being realistic, that the core was just too narrow for larger payloads.

Too funny!

In the interest of not turning the corner at internet argument street, I encourage you to re-read what was said.

 

[KarenRei]

Was told that I wasn't being realistic, that the core was just too narrow for larger payloads. Hmm...

In the interest of not turning the corner at internet argument street, I encourage you to re-read what was said.

You mean:

Then of course you have the diameter problem to deal with--a falcon core is only 3.7m wide and, as noted earlier (or maybe another thread) that's not conducive to bolting a whole mess of stuff on top of it. People might put a mailbox on a 4x4 post, but they don't ask the UPS guy to balance that package on a 4x4 post...

FH is a problem because of its fairing limits. You can fill a FH fairing with dozens of spacecraft to max out the mass, but that assembly is so tall and heavy and skinny that its really hard to make the whole thing survive the launch environment.

Alas, while dreaming about what could be and waxing philosophic about space stuff on the Internet is fun, I accept that ultimately I am still a slave to reality (and someone else’s money).

Musk disagrees with you. His timing, however, was just ironic Hopefully he'll paint the higher-diameter, "much longer" fairings (atop the "easiest part of the rocket to change") like a UPS package balanced on a post for you

FH wasn't designed the way it was for giggles. SpaceX wasn't going to build a rocket which provides capability that they're unable to utilize.

 

[Electroman]

Karen was spot on..

 

[bxr140]

You mean

Something like this?

I’d say a 7m fairing (and more importantly, the payload it represents) is probably about the limit for a falcon core.

As noted, I encourage you read and try to understand everything that was said. Snipping out of context quotes doesn’t do anything for this forum.

 

[KarenRei]

If your point was not "they can't put a big enough fairing on it to make use of its mass launch capacity", then what exactly was your point?

I'll reiterate mine: the fairing is already big enough for the vast majority of payloads, and they can readily make it much more voluminous. And now we find out: they already plan to whenever there's a need for it.

I'll lastly reiterate:

FH wasn't designed the way it was for giggles. SpaceX wasn't going to build a rocket which provides capability that they're unable to utilize.