Hacker News Reader

Comment by cyberfart | Hacker News Reader

Regards to comments mentioning that this is three F9s strapped together: While this is somewhat true, the development of this program was apparently lot more difficult than just tying them together. I remember Musk mentioning that they had underestimated the problem, the two additional boosters changed the dynamics significantly, and the core booster had to be redesigned to accommodate the new load parameters.

Needless to say, I'm extremely excited to even see a static fire of this monster. Re-using boosters for your early development is pretty nice and efficient too I imagine.

View on HN

> Re-using boosters for your early development is pretty nice and efficient too I imagine.

It sounds good on paper but it's not easy in practice, see: the Russian N-1 moon rocket, which basically just took the R7 rocket and scaled it up with more engines. Turns out too many engines means too many points of failure, and the rocket never worked.

https://en.wikipedia.org/wiki/N1_(rocket)

Much like in computing, you can't substitute one oxen with a thousand chickens and expect the same outcome when you plow your field.

The N-1 failed for lots of reasons besides having so many engines. It's designer died before it was completed. It's primitive control systems were inadequate. It's funding was inadequate.

SpaceX has already shown that with modern control systems, you can bundle a bunch of chickens together to plow a field. Having 9 engines has already helped them complete missions where one failed. The Falcon Heavy problem isn't scaling from 9 to 27 per se, it's building a new structure to handle all the new stresses that will occur, from the launch sequence to aerodynamic.

And even the seemingly small things; like how to detach the side boosters, how to handle the vastly different aerodynamics when you don't have an interstage (nosecone instead), etc.

It uses the same engines and very similar hardware, but there are so many new/not-flight-proven things on this rocket that it will be exciting, one way or another!

Although, Boeing has proven the concept with Delta IV heavy, (and, I think there are a few other precursors, as well, including STS). So this isn't completely new territory.

The engines were also ablative and couldn't be tested together.

All supercomputing is now done with armies of chickens. It was a great saying when Cray said it, but the future was not what he imagined.

Armies of chickens today would be something like ARM server chips, which have been touted for a long time but have perpetually stalled out precisely because replacing one powerful core with a dozen weak cores doesn't work. Even for commodity tasks, but especially when you throw the need for interconnect scalability into the mix.

Don't make the mistake of thinking that a large system is implicitly using weaker processors. They're as powerful as we can make them, some tasks just need a lot of them.

If you can control the chickens, anything is possible.

Wow, that's my new motto!

I believe GP means literally re-using boosters that flew in space, not just the design.

Also, my understanding is that initially there was supposed to be a corss-feed system that would increase the total delta-V considerably, but it isn't ready yet

That was initially the plan. It proved too complex and has since been dropped from the design. It won't be coming back.

Which is too bad, because it would be really awesome if they had managed to make it work.

SpaceX's genius has never been getting the absolute maximum possible thrust out of engines. It's been designing them so they can be mass produced inexpensively, and re-used. They've proven that is the far more important skill.

So making the Falcon Heavy work as simply and well as possible is the important part. Getting another 10% by making crossfeed work, at the cost of making it more complex, harder to build and re-use, simply wouldn't be worth it.

I think it would be worth it if they were not already looking working on BFR.

Falcon Heavy can fly for a couple of years now and it can handle current demands.

They will open up to next level capability with a next level vehicle.

I bet if Elon could turn back the clock he'd skip F9H and going straight to BFR. F9's Payload to LEO numbers have been going up; first design: 10,450 kg, current: 22,800 kg. On paper when Heavy was designed it was 25,000 kg to LEO.

I'm not sure. There are some things that the FH can do, military and space tourism and maybe a couple other things. Might be worth doing before the BFR, we will have to see how often i launches.

We just don't have the numbers, I would love to see the internal accounting of these projects.

I'd argue that if they put the engineering effort into BFR that much earlier, we might have a vehicle by now that can meet those demands as well as be a much better mars vehicle.

Reproducible results is a skill set unto itself.

Are you quoting something that SpaceX said?

Do you have a source?

https://twitter.com/elonmusk/status/726561442636263425

That doesn't imply that cross-feed isn't coming back, only that it isn't needed for the capabilities Musk posted.

While that single statement might not imply that it's gone for good, everything Elon and SpaceX have publicly stated leads one to the conclusion that it isn't ever going to happen.

Heck, Falcon Heavy has turned out to be a much larger engineering problem than originally thought even without asparagus staging (cross-feed) due to issues with reinforcing the center core to handle the loads from the side cores.

Based on Musk's latest presentation regarding BFR, it seems that Falcon Heavy will be a somewhat short-lived launch platform.

Then again, with SpaceX, plans change early and often...

> It proved too complex and has since been dropped from the design. It won't be coming back.

SpaceX plans to do in-orbit refueling for its BFR, though. I may be oversimplifying, but it seems to me this is a fairly similar feat than cross-feeding. Couldn't the experience gained in this help for that?

Not really. In orbit, you can transfer your fuel slowly, gently, and the receiving tank doesn't have to be under pressure. During launch, you need the receiving tanks to be at 50-100 PSI for the rocket to work, which probably means the fuel being pumped at a pressure that is much higher.

I wonder if they throttle back the central booster to ensure its fuel lasts longer than the two outside boosters. This is likely far easier than the cross-feed system though it would lower the total thrust.

Bingo! http://www.spacex.com/falcon-heavy

"Shortly after liftoff the center core engines are throttled down. After the side cores separate, the center core engines throttle back up to full thrust."

Does anyone know how they keep it NOT moving when doing a static fire of something like this?

There are clamps in the launch pad that hold the rocket down.

This is easier then one would think because the rocket is fuelled and the engines already have a hard time living the rocket at all.

They use the same launch pad during launch and the clamps only release when the computer has made sure all engines have correctly stared and are running correctly.

Check out the video from RocketLab when they had a launch abort to see a example of this.

Reply on news.ycombinator.com