SpaceX's Grasshopper

[Alpha_Geek]

"I would like to die on Mars; just not on impact."- Elon Musk

[publiusr]

gturner wrote:

---

It needs a memorable quote.

"That's one small hop for a rocket, one giant leap for ... um, cheaper rocketry."

Well, that's the internet generation for you, no long form reading.

How about this:

"X-number of years ago, Neil Armstrong made a giant leap for all mankind. In his memory...we have made another."

That's how its done.

[gturner]

newtype_alpha wrote:

---

They don't. Depending on the fuel loads, their delta-vee is around 400m/s: just enough to escape from an exploding rocket, or enough to slow to a stop from a terminal velocity of around 250m/s and then hover for a second or two just before touchdown. That, combined with Dragon's RCS thrusters, would give you about 700m/s, about a third of what you'd need for lunar deorbit and controlled landing (The Apollo LEM needed just under 2100m/s on its way down and the ascent stage needed 1500 on its way up).

I think we're saying "thrust" where we mean either impulse or delta-V. Since the SuperDracos are abort engines, they have enough thrust to accelerate the Dragon capsule away from an accelerating Falcon. The DragonRider has 8 SuperDracos, each with about 15,000 lbs thrust, but they are mounted at an angle so cosine losses drop the total to about 100,000 lbs, which on a 20,000 lb Dragon in launch configuration should give you about 5 G's acceleration. But there's not much fuel on board, so the mass ratio is very poor, as is the total delta-V.

The engines would probably have an ISP of around 300 (I assume they're tuned for a a sea-level expansion ratio instead of vacuum), but the cosine losses would again reduce this to about 260 or so. The full-thottle mass flow rate would be about 380 pounds of fuel per second.

Since the fuel-fraction is so low, I'll skip the rocket equation and just note that you'd get 10 seconds of burn time on 3800 pounds of onboard fuel, and 10 seconds at 5 G's gives you a delta-V of only 490 m/sec. You could of course double the amount of onboard fuel but it's eating into the payload, so the Dragon would definitely need a seperate descent stage (even if it was just an external tank), and possibly an ascent stage external tank.

The Apollo Descent Engine only had a tenth as much thrust as the Dragon abort engines, so the engines are already serious overkill. They need one SuperDraco (already 50% more thrust than the Apollo DPS) and they've got eight. The change they sould make is eliminating the cosine losses (which directly reduce the ISP) by realigning the engines or using a seperate vertical SuperDraco for ascent and descent.

[gturner]

publiusr wrote:

---

Well, that's the internet generation for you, no long form reading.

How about this:

"X-number of years ago, Neil Armstrong made a giant leap for all mankind. In his memory...we have made another."

That's how its done.

Ooo... Much better. And the Grasshopper is an awfully lot like the lunar lander trainer (the flying bedpost) that Neil flew.

When they actually return to the moon, though, it would be hard to top the Onion's version of Neil's statement from their book "Our Dumb Century."

HOLY FUCK! I'M ON THE FUCKING MOON!

Or something like that.

[Crazy Eddie]

gturner wrote:

---

I think we're saying "thrust" where we mean either impulse or delta-V. Since the SuperDracos are abort engines, they have enough thrust to accelerate the Dragon capsule away from an accelerating Falcon. The DragonRider has 8 SuperDracos, each with about 15,000 lbs thrust, but they are mounted at an angle so cosine losses drop the total to about 100,000 lbs, which on a 20,000 lb Dragon in launch configuration should give you about 5 G's acceleration. But there's not much fuel on board, so the mass ratio is very poor, as is the total delta-V.

First of all, I knew exactly what you were saying. I am saying that the Delta-v for the Superdracos in an abort configuration is a little over 400m/s. It could be considerably less for a controlled landing because terminal velocity near sea level is going to be quite a bit less than that, but for a range of plausible ISPs plus how much fuel they've got on board (not to mention what an abort thruster would have to be able to produce in order to be useful in that role) it's no less than 400 and no more than 600.

Second of all, when you start running the rocket equation to figure these things out, the only thing that really matters is the amount of propellant on board the craft. Even an increase or decrease in ISP doesn't make a huge difference unless you double or triple it. The cosine losses don't count for much; the only way to get a lot of movement for the SuperDracos in terms of performance is to change the fuel they're using and increase the combustion chamber pressure, and once you've done that, you've basically developed a whole new engine.

The engines would probably have an ISP of around 300

They wish. 270 is probably pushing it, 210 is more realistic considering the superdracos won't have comparatively large nozzles.

Since the fuel-fraction is so low, I'll skip the rocket equation

I didn't skip the rocket equation, I spent a couple of days last year going through these calculations.

The Apollo Descent Engine only had a tenth as much thrust as the Dragon abort engines

But the descent MODULE had ten times the fuel capacity and a slightly higher ISP, thus a larger overall delta-v. Significantly: even if you used the DragonRider with half of its superdracos firing and throttled way down to about 1.6m/s^2 (enough to hover over the surface of the moon) they would only be able produce that low acceleration for a finite period of time (250 seconds, with a D-V of 400m/s).

The thing is, the Dragon is coming to a landing on the lunar surface from orbital velocity of about 1600m/s. That means that even if all eight of those superdracos fire at maximum thrust and burn until they're bone dry, Dragon will still be moving way too fast to land safely. Even more importantly, if you somehow added enough fuel to cancel its orbital velocity entirely, you still need another minute or so of hover time to give your pilots time to zero in on a landing site, and another 1000 or more in reserve if you have to abort your landing and jump back into orbit.

The change they sould make is eliminating the cosine losses (which directly reduce the ISP) by realigning the engines or using a seperate vertical SuperDraco for ascent and descent.

That would change nothing, since even a 70% increase in ISP -- way more than you'd get from the cosine losses -- would still leave the Dragon about 1000m/s short of a safe landing.

What they actually need is about 7 times more propellant than the Dragon can carry internally; the lunar descent stage carried 8 tons of hypergolic propellants to the Dragon's 1200km and the Dragonrider's (probable) 2000kg. So Dragon either needs to add some gigantic propellant tanks under its heat shield (like the Soviet lunar landers; their ascent/descent engine was on the capsule and the landing stage was just fuel and legs) or it needs to get about 90% lighter.

[gturner]

We're in agreement, arguing the same points. The SuperDracos have vastly more than enough through for the mission, but the Dragon doesn't carry nearly enough fuel to give a mass ratio to carry it out without adding a lot of external tankage. The reason the cosine losses are significant is that it's much more efficient to just reorient a nozzle than add all the extra fuel mass to compensate for the inefficient mounting, in both the descent stage and the ascent stage, which itself would require a larger descent stage tank.

And on top of those losses, as you mentioned, the SuperDraco's nozzle is probably underexpanded even for sea-level operation.

What the Dragon needs for a lunar landing is a dedicated, fully expanded SuperDraco (or cluster of Dracos) aligned with the thrust vector, along with the removal of the eight unnecessary SuperDraco abort engines and the heat shield, the addition of landing legs, an airlock, vastly more fuel, a high-gain antenna, and either windows or external cameras for visibility during the landing phase.

We've probably done something like this before, slightly altering the Apollo command module into the LM by removing the GNC computer and some solenoid valves and attaching them to a different vehicle.

[sojourner]

So, basically, rip off everything and start with the inner pressure vessel from a Dragon. Might as well do it right and just clean sheet it.

[gturner]

Yep. That seems like the best approach in terms of schedule, perhaps leaving the attitude thruster system completely intact just because it's already adequate (I think) and thus would save lots of mounting and software changes. However, if the thruster exhaust is going to impinge on the tanks then maybe they'd repeat Grumman's idea of mounting the attitude thrusters on struts. If the schedule isn't critical then further optimizing is probably a good idea (visibility, entry/exit, internal tankage, small breakfast nook, etc).

On the flip side, leaving the thermal protection and parachutes in place would give you a lifeboat useable all the way to splashdown, and if you did just add extras to an existing Dragon you could probably execute the mission with one craft, but with severe performance penalties because your Earth Orbit Insertion fuel would also have to ride all the way down to the moon and back up, unless left in lunar orbit, in which case you've not saved any lunar rendevous requirements.

Given the Falcon9H payloads, you're not going to execute the mission in one launch, so you might as well use most of the Apollo mission profile with multiple launches and LEO rendevous.

[YellowSubmarine]

sojourner wrote:

---

So, basically, rip off everything and start with the inner pressure vessel from a Dragon. Might as well do it right and just clean sheet it.

Not really. Of course that a Moon lander based on the Dragon will be a completely new vehicle. That goes without saying. I was contemplating the Dragon landing on the Moon as-is because it's good food for thought.

However, the know-how is there, and SpaceX have already done their homework in building such things. Designing a new moon lander for them will be significantly less difficult than building a capsule from scratch.

Guidance systems – done.
Major engineering hurdles – solved.
Propulsion – only little rework required.
Interior – I'm pretty sure that they'll nearly CC it.

I'm not sure about the figures, but if I remember correctly, designing Dragon costed $300-400M*, building one vehicle costs $100M, and designing the SuperDraco cost a couple hundred million at most. My bet is that the moon lander design will be in the SuperDraco price range. At worst.

I'm also willing to speculate that expansion to landing on Solar system bodies has been considered in the design and it might require a lot less work than we're thinking, but maybe I'm giving them too much credit here.

* Or was it $600M? I'm not sure if $600-700M was the price for the Dragon and Falcon 9 combined or for each one of those. I don't remember in which interview Elon said that to check.

[gturner]

If they're going to do a clean-sheet design then I think we'd all have some suggestions. Large solar panels would greatly extend surface missions, and the capability to have a crew of more than two would be nice (can you imagine what a millionaire would pay to walk ON THE MOON?!) as would a slightly more comfortable sleeping area. And in lunar gravity they can eat normal food, cooked normally, so a little stove (powered by the solar panels) where they could flip an egg in 1/6th G would probably be the most-watched egg flip in the history of live television.

Another thing I think we would all agree on is a better moon-buggy, perhaps with both fuel cells and solar/batteries so that it doubles as a tele-operated rover after the astronauts leave. Perhaps the greatest omission of Apollo was putting something as big and robust as the moon buggies on the moon, fully equiped with cameras and high speed uplinks, and not having each of them creep around for a few more years, covering hundreds of times more area than the astronauts reached.

[sojourner]

gturner wrote:

---

If they're going to do a clean-sheet design

Their not. Elon has no interest in the moon. So unless NASA contracts them to go to the moon, this is just spitballing.

[gturner]

Elon says he has no interest in just going to the moon, because we've already done that and reruns are never as good, but he does say a lunar base would be significant.

[Gary7]

If a space vehicle production company decides to launch one of their products, are they by law required to send them off from a government certified and operated installation?

[Crazy Eddie]

sojourner wrote:

---

So, basically, rip off everything and start with the inner pressure vessel from a Dragon.

Hell no. Just take the landing mechanism from the grasshopper (and probably adapt some of its control software too) and put nine tons of hydrazine on it. Then put four SuperDracos underneath it with gimbals and expansion nozzles, and attach that module to a Dragon's heatshield.

Of course, at that weight and that propellant level it wouldn't be able to take off again. I kind of expect that would be half the point.

[sojourner]

Gary7 wrote:

---

If a space vehicle production company decides to launch one of their products, are they by law required to send them off from a government certified and operated installation?

Launches are regulated by the FAA if over a certain altitude. Launch sites do have regulations that must be met, but do not have to be government run.

Example: SpaceX plans to launch Grasshopper (to a certain altitude) on test flights from it's McGregor, TX facility.