Blue Origin Thread (merged)

[StrandedonEarth]

As far as RCS countering tilt vs Cg goes, there's also the wind factor, which was blowing pretty briskly. I bet that was a factor to. The lower the Cg, the more leverage the wind has to work with. All other factors being eliminated, I wonder how much wind it would take to knock it over? How does one counter wind-caused drift with a TVC?

Would it be easier to sail the barge with the wind, to lower the relative windspeed? Of course, then Falcon has to hit a moving target, but if it's a steady course that shouldn't be tooooooo hard.

[Kryten]

The Falcon has no idea where the barge actually is, there's no communication between the two and no sensors other than a radar altimeter. It's programmed to head towards a certain spatial position which just happens to be where the barge is.

[totalitor]

Falcon needs tyres just like office chairs. Lateral speed would not be a problem...

[Brotoro]

...until it gets to the edge of the barge.

[YoetoJoe]

I wonder if it would be easier for them to land on something like an oil rig.

[TMS]

I wonder if it would be easier for them to land on something like an oil rig.

lol... what could possibly go wrong?

[tater]

I think he meant something that is settled on the bottom, instead of floating.

[Albert VDS]

Video from the barge:

https://vid.me/i6o5

[sgt_flyer]

Wow - looks like all 4 legs were correctly deployed, but 2 of them broke upon landing - the rocket then fell onto the engines, bending several nozzles before body slamming the droneship - quite impressive approach though much more controlled than last time

Yet, it seems that the rocket could not finetune it's suicide burn (because of the thrittle static friction problem) and had still significant vertical velocity in addition to the horizontal velocity.

Edited April 16, 2015 by sgt_flyer

[TMS]

Amazing footage.

YouTube alternative:

[Nuke]

i like how the rcs engines are trying so hard to stop the rotation. it looked like one of the landing struts collapsed on landing.

[bigdad84]

it looked like one of the landing struts collapsed on landing.

Well, that was quite the impact that they took.

[PakledHostage]

The Falcon has no idea where the barge actually is, there's no communication between the two and no sensors other than a radar altimeter. It's programmed to head towards a certain spatial position which just happens to be where the barge is.

Certainly PPS GPSs would be accurate enough for the purpose, but I am still a bit surprised to read that?

Also, what do you know about how the control system compensates for wind? Does it attempt to account for wind drift in its control inputs (feed forward) or is it purely reactionary? There seems to be some evidence of breaking waves in the videos I've seen. That would suggest winds were not negligible at the time of the landing attempt. Presumably feeding near real-time wind speed vs. altitude data of the near surface winds to the rocket's control system would help preclude some of the over controlling that we see in the videos (if it was aggravated by wind drift, as I assume it was).

[Nibb31]

It has GPS plus inertial data. That's more than enough for precise positioning. It doesn't need weather or wind data, it just compensates automatically.

When you're driving your car on the motorway, you don't need wind predictions to keep your car on the road. If your car is pushed sideways by the wind, you just steer it back on the road to compensate.

[PakledHostage]

When you're driving your car on the motorway, you don't need wind predictions to keep your car on the road. If your car is pushed sideways by the wind, you just steer it back on the road to compensate.

I don't think that is a good analogy. Your car's wheels on the road don't have as much latency in control response as SpaceX's rocket does. And because (as I understand it) it has too much thrust to hover, it has to zero out vertical and horizontal velocity simultaneously, just as it touches down.

[Nibb31]

The vertical velocity problem seems to be solved. It's just a matter or timing the suicide burn so that the vertical speed and altitude curves intersect at zero. It's a pretty easy mathematical problem and the landing looked ok from that point of view.

What looked wrong here, was that the rocket was coming in at an angle, with too much horizontal velocity, and then tried to right itself by overcompensating. It might be due to the wind, but it looks like a PID/oscillation problem to me. Ideally, any horizontal velocity should have been dealt with before the final descent.

The cause of the tipover seems to be that a leg collapsed or failed to deploy. They need to investigate whether that was due to a fragility in the design or the result of the messy landing, or a bit of both.

It does illustrate that vertical landing is a complex problem, and it might be a while before it can be done confidently enough to reuse stages or carry humans. In a sense, I'm happy that Dragon V2 will be landing on parachutes for the foreseeable future.

I also wonder if it wouldn't have been easier to design the Falcon 9 with a wider but shorter core stage (like the DC-X or Rombus concepts) instead of this long thin top-heavy design. Even with most of the weight at the bottom, there's still a limit to how low you can get the CG, especially as the Falcon 9 now has all the grid fin mechanisms, the helium tanks and the RCS thrusters on top.

Edited April 16, 2015 by Nibb31

[Nuke]

Well, that was quite the impact that they took.

yea they definitely voided the warranty.

[Rakaydos]

I also wonder if it wouldn't have been easier to design the Falcon 9 with a wider but shorter core stage (like the DC-X or Rombus concepts) instead of this long thin top-heavy design. Even with most of the weight at the bottom, there's still a limit to how low you can get the CG, especially as the Falcon 9 now has all the grid fin mechanisms, the helium tanks and the RCS thrusters on top.

But then they couldnt use the same assembily lines for the falcon heavy- which is basically 3 Falcon first stages strapped together.

[sgt_flyer]

The fuel tanks are a bit special though - and the basic structure layout is shared between both stages (so one machine can build the structural elements for both)

Add to the fact that they are so lightweight in construction that they need to be kept pressurized to prevent collapse. (The strongback arm holds the rocket prior to pressurization) - for such a lightweight structure, having both stages walls lined up helps cut down on structural supports - and thus on weight. (Cylinder shapes are used for a reason in rocketry ) if the 1st stage was larger, it would need additional structure on top to help transmit the stress forces - which would add to the rocket's dry weight.

[Kibble]

I also wonder if it wouldn't have been easier to design the Falcon 9 with a wider but shorter core stage

Unfortunately, not only is short and fat bad for aerodynamics, but changing tank diameter means you have to build entirely new tooling. It would bee almost the same as designing the rocket from scratch. You can stretch a tank vertically with the same tooling, people do that all the time, to all kinds of rocket stages - the Falcon 9 first and second stage tanks even share tooling to minimize costs. But changing the diameter is not so easy.

[Armchair Rocket Scientist]

It does illustrate that vertical landing is a complex problem, and it might be a while before it can be done confidently enough to reuse stages or carry humans. In a sense, I'm happy that Dragon V2 will be landing on parachutes for the foreseeable future.

I also wonder if it wouldn't have been easier to design the Falcon 9 with a wider but shorter core stage (like the DC-X or Rombus concepts) instead of this long thin top-heavy design. Even with most of the weight at the bottom, there's still a limit to how low you can get the CG, especially as the Falcon 9 now has all the grid fin mechanisms, the helium tanks and the RCS thrusters on top.

Dragon V2 has a MUCH lower center of gravity and is capable of hovering IIRC. The SuperDraco is deeply throttleable and, being pressure-fed, could shut down and restart during the final approach if it had to. Even a relatively high-speed landing like the F9 attempt would be survivable.

After the Falcon 9 tests as well as the pad and in-flight abort tests of the Dragon V2, I would be surprised if it doesn't soft-land successfully the first time they try it.

[Streetwind]

I also wonder if it wouldn't have been easier to design the Falcon 9 with a wider but shorter core stage (like the DC-X or Rombus concepts) instead of this long thin top-heavy design.

The Falcon 9 is one of the thinnest, tallest rockets ever flown. It is in fact thinner (relatively speaking, in terms of width/height ratio) than the space shuttle solid rocket motors. And that is fully intentional, because the Falcon 9 is limited in its diameter by one critical number that cannot be avoided:

The maximum legally allowed diameter of objects transported on U.S. highways. Because the Falcon 9 stages are transported by truck from California to Texas, and then from Texas to the launch site (Florida and California, and soon Texas). That is literally the only reason it is as tall and thin as it is. (For comparison: Delta IV cores, for instance, are transpoted by ship.)

SpaceX's next generation rocket is tentatively planned to be much wider. I can only assume that they will either transport it in a different fashion, or assemble and test it directly at the launch site.

[Nibb31]

Unfortunately, not only is short and fat bad for aerodynamics, but changing tank diameter means you have to build entirely new tooling. It would bee almost the same as designing the rocket from scratch. You can stretch a tank vertically with the same tooling, people do that all the time, to all kinds of rocket stages - the Falcon 9 first and second stage tanks even share tooling to minimize costs. But changing the diameter is not so easy.

I agree, but this is where SpaceX's approach reaches its limits. In engineering, you design something to meet a set of requirements. You make trade-offs on some points because other points are more important. Yet the requirements for the Falcon 9 as a low-cost mass-produced launcher and as a robust reusable launcher are contradictory. The Falcon 9 was designed from scratch and optimized to be cheap and mass-produced. It has relatively cheap and light tanks and the factory is geared to mass-produce up to 400 engines per year, which is nearly one new Falcon 9 rocket every week.

Now, they are bolting additional stuff onto a cheap disposable rocket to make it reusable, which it wasn't optimized for from the start. The result is the bolt-on legs, the high CG, stretched tanks, and additional hardware added at the top of the stage that makes it even more top heavy. The other result is manufacturing overcapacity that is going to idle and lose money if they start massively reusing core stages. The Falcon 9 is a very clever design as cheap launch vehicle, and it has done a lot to lower the cost of access to orbit, but I'm not sure that it's the most robust design to base a reusable rocket on.

If you were designing a reusable booster from scratch, you would integrate all of this stuff into a design that is optimized for different trade-offs than a disposable booster. You could afford to make it sturdier and more expensive, because you would only build them in small numbers. You could afford more expensive tooling for a conical shape because the trade-off in cost would be worth it. You wouldn't need to constrain the width to road tunnels for to lower mass-transportation costs, because transport from the factory to the launch site would be exceptional. It would be designed from the start with landing in mind, with a wide footprint and low CG, instead of bolting legs onto tall thin tank.

In the end, this is a bit like turning a bicycle into a motorbike. Sure, it can be done if you bolt enough extra stuff onto it, but it will never be as good as a motorbike designed from scratch as a motorbike. Yes, it's hard, but I think they might be making it harder than it needs to be.

Edited April 17, 2015 by Nibb31

[Streetwind]

In the end, this is a bit like turning a bicycle into a motorbike. Sure, it can be done if you bolt enough extra stuff onto it, but it will never be as good as a motorbike designed from scratch as a motorbike. Yes, it's hard, but I think they might be making it harder than it needs to be.

I like to look at it this way: perhaps they're bolting things to a bicycle in order to learn how to build a motorbike from scratch? The reuse technologies SpaceX is shooting for did not exist prior. So they built a cheap mass launch vehicle that makes them money. With that money, they can then develop these technologies. But what to test them on? ...Hey, why not that cheap mass launch vehicle they already have!

Once the technologies are fully developed and proven, they can then build a new launcher that's designed from the ground up to make use of them.

[AngelLestat]

I agree, but this is where SpaceX's approach reaches its limits. In engineering, you design something to meet a set of requirements. You make trade-offs on some points because other points are more important. Yet the requirements for the Falcon 9 as a low-cost mass-produced launcher and as a robust reusable launcher are contradictory. The Falcon 9 was designed from scratch and optimized to be cheap and mass-produced. It has relatively cheap and light tanks and the factory is geared to mass-produce up to 400 engines per year, which is nearly one new Falcon 9 rocket every week.

They are cheap because is a private company which really cares about how to optimize cost..

In a goverment company is like: ok, I will use the best of the best and make 100 test on each part from my area so if something happens they dont point to me.. The money comes from above.. so there is not problem.. This same thing is applied to the politicians, there are only in danger if they force to someone to work with less budget than the average. So every body just spent a lot to avoid responsabilities.

I remember you that the structural safety margins in falcon 9 are 40% above flight loads, higher than the 25% margins of other rockets.

So when you said cheap tanks or cheap things.. it has nothing to do with the actual quality of the components. It has to do with good production procedures.