Hacker News Reader

Comment by modeless | Hacker News Reader

Any more news on the performance of the first stage recovery test? Elon Musk tweeted that it was successful but I haven't heard anything more detailed than that.

View on HN

The two things that were new this time:

* The 3-engine deorbit burn worked; previously it was screwed up by uncontrolled roll preventing the engines from getting fuel

* The 1-engine landing burn, previously separately demonstrated by the Grasshopper and F9R test vehicles, worked for a rocket coming down from orbit

The remaining items:

* 3-engine burn to take the rocket back to the Cape from downrange

* Making everything safe enough to dare landing at the Cape, without killing anyone or destroying all the expensive ground equipment that's nearby.

> "The 3-engine deorbit burn worked; previously it was screwed up by uncontrolled roll preventing the engines from getting fuel"

My impression was that with the CASSIOPE attempt, the 3-engine burn was successful but the following 1-engine burn was cut short by the roll.

http://en.wikipedia.org/wiki/CASSIOPE#Post-mission_launch_ve...

As for the last item, it's not as daunting as it seems. Range safety is actually fairly comparable in either direction. The trajectory of the booster won't naturally head over populated areas and it will have to be adjusted until it reaches back to land. If, for whatever reason, the trajectory adjustment goes to far and the booster would end up landing farther back on land then the stage can be signaled to be destructed, which will screw up its drag coefficient enough so that the debris will just fall safely into the water. And that's a worst case scenario.

Landing is also a way less serious range safety concern than taking off because the rocket is essentially empty of fuel, rather than full of fuel.

Another reason it's less daunting is that, by the time of the landing burn, there's very little fuel remaining in the rocket. On the way up, the rocket has to have enough fuel to power the 9 first-stage engines for 181 seconds, and the second stage engine for 412 seconds. (As well as fuel for the various return and landing burns.) So it will have over 2000 engine-seconds of fuel.

In the landing burn, it's powering 1 engine for maybe 10 seconds. So that's less than half a percent of the initial fuel mass - something that would make a potential failure a lot less dangerous.

Ten seconds seems like an incredibly liberal estimate. You need a pretty decent "margin of error" reserve. I'd wager we're looking at something more like 30-60s of fuel during the last stages of landing, at least. So let's call it something like 2-3% of the fuel.

I'd imagine it's nowhere near full power for most of that, as it's an empty shell of a craft and thus quite light.

The engine can only throttle down to about 70%. For the final landing, the first stage will only use the center engine (leaving the other 8 shut down).

Even with only one engine at 70%, the thrust to weight ratio is greater than 1. That means the rocket can't hover, they just have to time the burn perfectly so the speed reaches 0 at the moment it reaches the ground.

Why can they only throttle to 70%? Is that related to old-school turbopumps?

Most rocket engines don't throttle down well. The main reason isn't the turbopumps (although that can also be an issue) but combustion instabilities in the chamber and flow separation in the nozzle.

Nasa overview of liquid engine throttling:

http://ntrs.nasa.gov/archive/nasa/casi.ntrs.nasa.gov/2010003...

That's a very readable reference, thanks.

Exactly the answer I was looking for-- thanks!

SpaceX claims that Falcon-9 first stage has a pretty good mass ratio - meaning that the fuel weights many times more than the tanks and engines (and everything else).

However, even if we assume 30 as the ratio of loaded vs. dry mass, which is better, for example, than Titan II first stage, which has pretty good mass ratio, we need to add the mass of landing legs - something unique for the Falcon. Suppose legs halve the mass ratio, making it 15.

9 engines are capable of lifting the dry first stage, the fuel in it, and the second stage plus payload. That means 9 engines can lift much more than 15 times the dry mass of the first stage. Dividing the thrust by 9 makes the thrust much more than necessary to support the dry stage.

In other words, even the thrust of a single Merlin engine is more than enough to brake the dry weight of the first stage. That means when first stage lands the engine is throttled - spending less mass of fuel per second than it does when Falcon lifts off. The better mass ratio of the first stage, the smaller flow of fuel is needed to decelerate it - so less fuel is needed to brake. It's not enough to count seconds of thrust - how big the thrust is should also be taken into account.

According to Elon it landed:

"Data upload from tracking plane shows landing in Atlantic was good! Several boats enroute through heavy seas." -- https://twitter.com/elonmusk/statuses/457307742495993856

Which is pretty amazing. Apparently the telemetry continued for 6 seconds as it fell over.

"Flight computers continued transmitting for 8 seconds after reaching the water. Stopped when booster went horizontal." -- https://twitter.com/elonmusk/statuses/457311780943822848

Well, also according to Musk, its last reported velocity was 360 m/s. That's a pretty hard landing.

That message was sent while the vehicle was still high up and had not started its powered landing over the water. It was only its last reported velocity at that time, not its last reported velocity now, or even shortly after.

Ah, did not realize that. Thanks for clarifying.

That was when it dropped over the horizon from the Cape and they lost telemetry. It was still 8.5 km in the air at that point.

I wasn't really sure why SpaceX decided to launch when they knew that recovering the first stage would be almost impossible when the waves in the Atlantic were 20' high. I guess recovery of the stage was the tertiary reason for the launch (after the payload for the ISS and for the cubesats), however was it critical to launch at that point or could they have waited a few more days for the weather to improve?

One thought which had occurred to me was maybe they needed to get people working on the upcoming Orbcomm OG2 launch which is currently scheduled for May. When you think about all the time and energy which goes into fabricating the rocket along with all nine Merlin engines, it really hits home how significant of an achievement making reusable rockets will be when SpaceX pulls it off. You have to think that at some point in the not-too-distant future it would be absolutely crazy to launch a rocket where you knew you were going to lose the first stage.

SpaceX is getting free testing out of commercial payloads. That's already a huge boon - asking the commercial flights to delay their stuff so SpaceX can test better is pushing it a bit.

The flight was already delayed 10 times (April 6 2013, Sept. 30 2013, Nov. 11 2013, Dec. 9 2013, Jan. 15, Feb. 11, Feb. 22, March 1, March 16 and March 3) and scrubbed once. Given the importance of demonstrating a fundamentally new capability and potentially saving millions of dollars might be worth delaying for another couple of days. If NASA was willing to wait over a year for launch, presumably they'd be willing to wait a little bit longer. They've been tremendously helpful and supportive of SpaceX up to this point.

Even if NASA was fine with it, their other customers may not be. Don't look a gift horse in the mouth. They have managed to build a system where the customers basically pay them to test new products. That's a fantastic situation! They don't want to even hint at something that might cause the next customer to say, "We want to make sure you're focused on our mission, so kindly leave the landing legs in your factory for this one, thanks."

Another way to look at it though is the customer gets really, really cheap access to space (in this case, cheaper than anyone else including Orbital Sciences), but has to put up with some amount of tinkering. I guess there's only so much that any organization is willing to bear.

Regardless, I'm really glad that the mission went off mostly without a hitch. I can't wait to see when SpaceX glues all of the pieces together and has a truly reusable rocket platform.

I'll be more interested at the moment to see how recovered engines perform in "reused" state testing. Even the breakdown on the cost of refurbishment would be big news.

From what I understand, they had a window on Friday, a fallback window on Saturday. If they missed the Saturday launch, then the mission would go to Orbital.

Maybe they technically could have waited until Saturday, but there is no way they would scrap one of their windows for a tertiary objective.

In the post-launch press conference, NASA made it clear that ISS scheduling is tight, with a lot of traffic (Dragon, Cygnus, Soyuz) and onboard activities, and rescheduling was not going to be easy.

Yeah, this launch has been delayed and rescheduled over and over again for various reasons. Too many to count actually, the customer (NASA) will always be the primary concern and they did not even have to allow Spacex to run these tests while transporting their cargo.

Because recovering the first stage was not what they were being paid for. They are paid to move stuff to the Space Station.

That said, the ability to 'tag along' additional testing without jeopardizing the main mission is a huge win for SpaceX. The cost of the launch (about $100M according to their web site) is paid for by the contract. Getting flight test data as well is pretty priceless.

Finally, even if the seas were calm it would be unlikely they would put a ship in the area as it would put people at risk of having a chunk of rocket and fuel crashing into them. They might have gotten better video though. If I were there I would have tried to figure out how to borrow a predator drone to have it loitering right there to film the arrival.

I'm willing to bet that this is what everyone is most curious about.

During the post launch news conference (https://www.youtube.com/watch?v=8G2pRVQ1JhA) Elon said he was pessimistic about a successful recovery of the stage for this launch. However, he said SpaceX is "connecting the dots" and seemed optimistic that they are making good progress (for example, they successfully canceled the roll this time). The rough timeline he mentioned was that they will probably recover a stage sometimes this year, with fully successful recoveries (i.e. capture, refuel, and ready to launch again within hours) sometime early 2015.

The main reason he was worried about recovery at that point was that the booster had splashed down into something like 20-foot waves. That's high enough to be real trouble for some of the recovery boats they wanted to have out there (which may be one reason why we haven't heard much about how it finally wound up). And it could also very well sink a rocket stage that wasn't built to handle this particular kind of buffeting, and is made to be pretty lightweight.

As to longer term plans, it's not clear they will relaunch the first booster they get back, even if seawater exposure weren't an issue. More likely, they'll disassemble it, and see how the parts handled the stress of a full launch and re-entry, looking for wear that could endanger a second launch of the same booster. If there are any unpleasant surprises like that lurking, it's a lot cheaper to find out this way than in a launch with a payload at risk.

Oh, I doubt they ever intended to relaunch this one. At that point you have to pretty much rebuild everything, so you may as well just trash it and build a brand new one.

But they would sure like to know that it behaved properly after they lost contact, and they'd like to know if anything was damaged before it hit the water.

They have telemetry until it went into the water --- see tweets quoted immediately below. The last few seconds' worth of that were delayed because the rocket was over the horizon from the launch site, but they got it from chase planes.

Oh, I didn't realize that.

But still, they really would be better off if they had it to pull apart.

Nothing detailed about the first stage yet. They certainly have telemetry showing it hovered for a while but that is all the public has seen. During the launch they mentioned having video of the first stage retrofire but we haven't seen it.

Lack of news of recovery makes me think they lost the stage in rough seas. I think we would have heard if it was recovered.

Anyone know if they had video recording on the first stage? That'd be fun to watch.

At landing? Don't believe so. I think if they had that, they would have been able to say more than they did. And I think they mentioned not great weather?

It would be cool if they did :).

I don't expect them to have drones recording the landing from the water, but I would fully expect them to have mounted some cameras to the stage itself. But I can't find confirmation of that.

Reply on news.ycombinator.com