NASA just launched 8 satellites from a rocket dropped from a plane at 40,000ft | Hacker News Reader

NASA just launched 8 satellites from a rocket dropped from a plane at 40,000ft

theregister.co.uk

141 pointsandrewflnr72 comments

Lifting a rocket that high before firing it carries another benefit besides the (small) altitude gain and lower fuel requirements due to drag: Max Q, the maximum dynamic pressure that has to be sustained by the rocket usually happens at a lower altitude than that (depending on the mission).

This means your rocket has to sustain a far lower structural load, making it safer and lighter.

Max Q for the Peagus and Saturn V are actually at the exact same altitude. ~14-16km. And nearly identical pressures.

The thing is because all rockets follow the same trajectory to orbit, The gravity curve. So you hit similar speeds at similar altitudes. If you don't, you either aren't making it to orbit, or you're wasting fuel.

This is because the gravity curve is a straight line if you ask Einstein, just earth's gravity makes space-time look wibbly-wobbly.

Newton calls this the derivative of the balastic arc.

There isn't a magical trajectory to orbit that is cheaper.

I was disappointed to realize that "balastic arc" was just a typo. I thought I'd learned a new word. :)

Lighter? I'm not so sure. Most of the reduced mass has to be added back in again for Pegasus to have aerodynamic control during stage 1 flight.

The overall benefits of the platform are questionable, at best, especially considering its absurd launch cost.

Damn, $56M+, that is expensive...

The altitude gain is significantly less important than the horizontal velocity. An aircraft can easily launch from the equator and add 650+MPH.

Note that this launch wasn't from the equator, because this constellation needs to overfly hurricanes.

It's fairly uncommon for LEO satellites to want to be at the equator. GEO, extremely common.

Launching from the equator is useful for a wide range of orbits as a component of the final vector 1600km/h is really not that fast and your final orbit crosses the equator anyway. The only issue is if the ideal launch point is not over the ocean. Also, there is a wide band near the equator with close to optimal velocity.

Yeah, I suppose my comment was not the most clever. In this case they were only launching 1/2 the max payload so performance wasn't really an issue anyway.

It's about 5% of the final velocity, but that's way more than 5% fuel savings because you don't have to suspend million of pounds against gravity for a minute and also don't have to accelerate that fuel.

For example Saturn V used 40% of its total fuel to get up to this speed. That's a huge savings.

Unless I am reading the data incorrectly:

http://www.braeunig.us/apollo/SaturnV.pdf

Vector addition means you save energy, but not one to one. EX: if you add 1 going north and 1 going east then your final vector is 1.414 going north east. So, it might only be worth 1% or 5% of the final velocity.

Orbital ATK has been building and launching the Pegasus XL (https://www.orbitalatk.com/flight-systems/space-launch-vehic...) for a bit now. It is an extremely cheap platform that is launched from an airplane.

What's neat is the User Manual is available online: https://www.orbitalatk.com/flight-systems/space-launch-vehic... [pdf]

Extremely cheap? Initially it was $6mm, now it costs almost as much as a Falcon 9 launch.

Part of that is due to its extremely low launch rate -- this was the first launch since 2013 -- but still.

if Orbital launched as much as SpaceX, the Pegasus cost before margin could be 20% of what it is today.

With Stratolaunch buying a bunch of Pegasus XL rockets, we'll know in a couple of years.

How is it extremely cheap? It can almost not compete now and it has very little option of re-usability. Everybody else is pushing for reusable systems, SpaceX is almost there.

If they could get a extremely high lunch rate the might be able to do it at a good price, but that will be hard.

I worked on the software for the star trackers in these. Glad to hear they have already made contact. :-)

I can't get my head around how this thing works. Apart from the direct GPS signal it's able to also receive the part of it that was reflected by the ocean? How does the reflected signal tell anything about the wind speed?

And btw what are star trackers used for? As a fallback for GPS?

IIRC, star trackers are used as both a position and an attitude reference. GPS gives you only position (though more sophisticated systems can be used to determine attitude).

While a GPS outage is a minor inconvenience for most of us, it could be pretty catastrophic for spacecraft if they didn't have another position reference.

EDIT: See below - I was off; they're just attitude, not position

Star trackers are usually just used for attitude, not position. I've heard of research to use them in conjunction with a camera pointed at Earth to resolve coastlines to get position, but it's very theoretical at this point AFAIK.

For those of us just watching, what is attitude? At first I though spellcheck might have corrected it from altitude, but 3 times among 2 differently people I see attitude.

It's orientation, AFAIK, what direction it's facing.

Yep. Think of it like right ascension, declination, and roll from astronomy. Given attitude, your position, and the time, you can know where you are pointing on the Earth (lat/lon).

Altitude is how high your body is.

Attitude is which direction your face is pointing. Imagine a satellite meant to photograph the earth at the right altitude but the wrong attitude could just be taking nice shots of outer space.

What's the difficulty level of writing star tracker software for consumer grade CCD devices? I've always had writing and open sourcing a star tracker on my bucket list, similar to how this person did the same for a from-scratch GPS receiver [1]. Same for determining your position in space using pulsars as beacons [2].

[1] http://www.aholme.co.uk/GPS/Main.htm

[2] https://ntrs.nasa.gov/archive/nasa/casi.ntrs.nasa.gov/201500...

Interesting, thanks. I wonder at what altitude GPS becomes ineffective? What then? We'll likely need some positioning satellites further out at some point? maybe at L4, L5 Lagrangian points?

The earth is bathed in radiation from GPS transmitters and these clever satellites are using doppler radar techniques to see how those signals are being reflected from ocean waves. From the doppler shift you can infer wind fields. Consumer GPS receivers only compute position from these signals, but there is a lot of interesting science you can do from the raw signals with a hacked GPS receiver.

Whoa, doppler shift from this mess from ocean waves, that's amazing. Looks like you registered just to reply, thanks a lot.

Can you tell us some more? What the software is, what language, what hardware they're based on, etc.?

It's a LEON CPU (SPARC architecture) with a FPGA for hardware acceleration. Languages are C, MATLAB, and VHDL. The function is to take a picture of stars and return where you are pointed in inertial space with lots of accuracy.

Very interesting, thanks.

Just an FYI, NASA live streams mosts of these launches here: https://www.nasa.gov/multimedia/nasatv/#public

The Pegasus launch system has been around since 1990. First private orbital launch system that worked.

A little off topic.. I've always wondered why NASA lunched rockets from Cape Canaveral (only 3 meters above sea level).

Granted, it's a little tricky to transport rockets, but surely it would have made more sense to move the operations to a mountain near the equator?

How much fuel would be saved by lunching 3000 meters above sea level instead?

Not much at all. Low earth orbit is ~200km so the mountain is only helping by 1.5% of the needed altitude.

Altitude is not even the most difficult part of getting to orbit. Getting to the necessary horizontal speed to stay in space takes a hell of a lot more fuel than going straight up.

So say a 4000m launchpad is feasible, that's 2% of the needed altitude. While gravity will be roughly the same (yeah it's higher altitute, but there's the mass from the mountain). There is much less atmosphere* (~50% less pressure & 1/3 less of it to get through).

Given that payload to fuel ratio is roughly 1:10, this does seem to add up to a significant saving.

*Air pressure numbers http://www.mide.com/pages/air-pressure-at-altitude-calculato...

you're missing the point. altitude and drag losses are nothing compared to horizontal velocity. you need to be going 7.8km/s. Per second! That takes a lot more energy than gaining any kind of altitude or overcoming atmospheric drag. It's not significant savings.

Obligatory xkcd: https://what-if.xkcd.com/58/

In short, gaining altitude, even 3000m, isn't that hard for a rocket. The problem is that they have to go fast.

The short answer is being as close as possible to the equator and able to launch over an unpopulated area (like an ocean) is more important than altitude.

There might be some benefit to launching off of a really tall mountain, since it would make your first stage engine more efficient, but not enough to justify the extra logistics.

Given how SpaceX has landed things on a drone ship, presumably they're considering launching from a rocket that's been lifted above the bulk of the atmosphere using a balloon of some sort.

You'd probably save a ton of fuel if you could launch from 40 or even 100K feet.

Eh, not really. It's not getting to altitude that's expensive, it's speed. Orbit is blisteringly fast.

It's expensive to build up speed while fighting through the lower atmosphere. Density drops pretty fast[1], so you don't need to fight it with the rocket if the launch vehicle has already paid the cost.

[1] https://upload.wikimedia.org/wikipedia/commons/d/de/Atmosphe...

Why aren't space ports then at higher altitudes but mostly at sea-level like cape canaveral? At the beginning the rocket flies comparably slow, so the lower atmosphere density is not that important, its the speed that is hard.

Because there aren't any places in the US where you've got 20kft mountains right next to the ocean, on the east coast. The idea is that you launch over the ocean, near the equator so that you're getting the first ~1000mph for "free" since you're launching with the earth's rotation instead of against it. Next to the ocean is for range safety. If it's over the ocean and it blows up you don't rain down parts on populated areas. If you tried launching from the top of a mountain in the Rockies and something went wrong it could theoretically rain down on DC which would definitely get your funding cut.

No need for the east coast. Hawaii's highest point is 13,800 ft above sea level and much closer to the equator both significant advantages. The real advantage to Cape Canaveral is it was in the middle of nowhere with lot's of open land.

It's a lot easier to get up to speed if you don't have to deal with cutting through a heavy chunk of atmosphere on the way up. Plus, rocket designs could vary, be less concerned about aerodynamics, if they were starting from a launch platform already in very thin air.

A wider, shorter rocket might work better in thin air, for example, but that would be impractical launching from sea-level as is tradition.

I have this dream that someday we'll build a SR-710 that's 10x the size and carry a BIG payload. And then putting stuff into orbit will basically be strapping your thing to a "big" upper stage (but small relative to a whole rocket stack), loading it into the cargo bay and then some fancy flying.

I know it'll never happen for a variety of reasons. But it's such a cool plane and when you think about cruising at mach 3 (2000mph+, 15% of escape velocity (http://www.sr-71.org/blackbird/sr-71/)) it seems like it could make the rocket equation for a plane-launched rocket much, much more forgiving.

One big reason it'll never happen: The square-cube law. https://en.wikipedia.org/wiki/Square-cube_law

Physics is such a killjoy.

Well 10x doesn't necessarily mean 10x longer, it might mean something with 10x the total wing area and thus 10x the lifting capacity. Which might only be say 3x the length and 3x the width.

With a wider, shorter rocket and lower pressure you could also use wider engine bells, for more efficient thrust.

When you're dealing with LEO velocities, the difference between the speed of the aircraft they launched with (significantly <M1) and a drifting balloon is relatively negligible considering the benefit of start at the edge of the atmosphere inexpensively, I would think. Definitely a new technique that would have some hard problems, but not fundamentally infeasible, I would think.

SpaceX was originally going to build a 4-engine Falcon for Stratolaunch, which is building a huge carrier aircraft, but later decided it was a distraction.

https://en.wikipedia.org/wiki/Stratolaunch_Systems https://en.wikipedia.org/wiki/Falcon_9_Air

Stratolaunch's latest strategy is to use Pegasus XL to launch from their plane, as many as three in a single flight. While they had flirted with a rocket that had one cryo stage, Pegasus XL and the other rocket that OrbitalATK might build for them are both all-solids.

I've heard of this concept lately, not sure what to think tbh. http://www.zero2infinity.space/

That'd require a rather huge balloon. Rockets are quite heavy. Pegasus XL has a mass of over 18 Mg, for payload mass of just over 400 kg.

Not sure how this relates to SpaceX' drone ship landings. Falcon 9 couldn't launch from the air, as cryogenic fuels mean you have to launch rather soon after fueling. Furthermore fuel is cheap (compared to the rocket), and their mode of operation seems to work rather well (despite CRS-7 and Amos-6) with normal launch and propulsive landing.

SpaceX also has made it clear that a lot of the things they're doing, trying, and learning are designed to work in other places as well. While you can land with parachutes on Earth, you can't do so on Mars. Same holds for balloons. They just won't do it.

The thing about balloons is they're relatively cheap, and the stuff you put in them isn't expensive either. If you want to flirt with danger, hydrogen filled is one way to do it.

The drone part is mostly to do with automation, as a terrestrial platform allows a lot of hands-on work to prep the rocket for launch. A high-altitude platform would have to be automated.

Plus the way rocket physics work is the less fuel you need to punch through the atmosphere at speed, the less fuel you need overall, which reduces the weight of the rocket exponentially, as the fuel itself requires more fuel to get lifted.

Anything they can do to get to a higher altitude without paying a heavy price is worth considering. Turning a theory into a viable launch strategy is not easy, I understand, but they've shown a remarkable ability to innovate.

There is a reason basically nobody ever uses a balloon, it just does not make sense. It does get you far to little for not enough. You don't save that much energy and you have way harder time starting.

You can't do deep cryo fuels and you can't do static fires in lunch conditions.

Couldn't we build a launch pad above the clouds, at 100k? I know building a space elevator is rough/difficult but if it's a lot easier/cheaper to launch from that high up, why not a permanent structure as high as you can get?

Like most ideas, you can read about it in science fiction! Not so practical, though.

https://www.dezeen.com/2013/09/24/science-fiction-author-pro...

Hint: if altitude were such a problem, why don't rockets go straight up to 80km or so, and only then thrust sideways?

Rockets clear the most dense portion of the atmosphere very quickly. Building up speed is what takes time (and fuel).

Of course, the exponential nature of the rocket equation means that you'd get to save a lot of fuel. You'd still have to lift a lot of rocket though. With a balloon so big that would put the ships from the original Independence Day movie to shame.

Maybe it would be worth it on Venus.

This is probably a tremendously dumb question, but what are the major impediments to literally sticking a big freaking balloon to the top of a rocket and floating it up much of the way?

It wouldn't save you nearly as much fuel as you'd think. The majority of the delta-V (fuel) to get something into orbit is for pushing it sideways, not pushing it up - you need about 8 km/s of sideways speed to get into even a low orbit, and that's where the the fuel is going. When you see balloons lifting cameras (or people) out of the atmosphere, they're going to space, but not to orbit - they fall straight back to Earth as soon as the rope is cut, even though they might be at the same altitude as the ISS which does not fall.

xkcd had a helpful explanation about the vertical/horizontal mistake that is often made when thinking about space travel: https://what-if.xkcd.com/58/

Why bother with the rocket? It sounds mad, but you can actually just send the balloon. to orbit... http://www.jpaerospace.com/atohandout.pdf

Well, with some help from "hybrid electric/chemical propulsion". Also, hydrogen. Yikes.

Why do they need the space station part?

Because the orbital ship is structurally unable to cope with higher atmospheric density so must be assembled and maintained at altitude and never go below the height of the station. There's a few challenges there...

It's a neat idea but I'm not sure the materials science is there. Right now out best balloon material leaks a fair amount of helium over time. Then there's trying to make a ship that's large enough to carry enough helium to float itself up there then rigid enough to survive the extreme velocities to fly up. Seems like there's a huge gap there that might just not be possible to cross.

Well, kudos to NASA but also kudos to ISRO for taking the initiative in making satellite launches more cost effective.

Pegasus XL is a private launcher (NASA is the customer), and it's not cost-effective. It costs nearly as much as a Falcon 9 for a much smaller payload.

I'm not sure where the India reference came from? The EU and Japan both operate smaller launchers. A couple of startups are getting close to their first launch, too.

I see. I made the wrong assumption then that sending 8 satellites from a rocket that gets some help from a plane (requiring less propellant/size) was more cost-effective. Thanks for clarifying.

Reply on news.ycombinator.com