Hacker News Reader

Comment by jsprogrammer | Hacker News Reader

An electromagnetic launcher seems to be the way to go. Why is everyone spending time on rockets?

Using a mass driver to launch, if youkeep the G forces down to the levels NASA wants I think you need something like 100 miles of track?

You can build them much shorter, but they could never be used to launch living cargo. My guess would be that we'll see more investment in them once we have a big rocket (BFR/New Armstrong) capable of launching larger habitation units into higher orbits. Once you have more than 10 humans living up there you start having to think about how to supply them, and then a rail gun looks pretty good, but in the meantime we need to be able to lift large (dimension) things as well as humans, and rockets are actually a pretty good choice for that.

Don't rocket-based vehicles end up burning a significant amount of fuel just to lift the weight of other fuel? What if there was a hybrid approach of using a mass driver to launch fuel containers into orbit, and then the rocket vehicle carrying people (and other more delicate cargo) could rendezvous with and pick up the previously launched fuel containers?

Would certainly make for some interesting failure scenarios to consider. Such as, what happens if the engines fail to start? Presuming they are intended to start after the vehicle is some distance down the rail.

Edit: oops, misread your comment.. sorry for the non-sequitur.

How about launching the personnel carriers this way?

https://en.wikipedia.org/wiki/Laser_propulsion#Laser_Thermal...

Then send cargo up that's hardened for 30g in a vehicle launched from a 20 mile long mass driver that's tilted 45 degrees. We should be able to build such a structure by starting several miles underground, then building up the side of a mountain. 30g over 20 miles yields a final speed of about 9735 miles per hour. With a boost like that, the vehicle wouldn't need a 1st stage and could probably get to orbit without staging.

If it's tilted at 45 degrees, how high will it be after twenty miles????

Make the bottom of the launcher a tube connected to some really deep part of the ocean and use massive floats to help support the weight of the structure then you have less structure to support in open air. You load the pods from sea level, then drop them down the rail-mortar and let computers and robotics handle the rest

At 45 degrees...it would be 14 and half miles high. If I did my math right triangled

I am sure you are correct, but I was reading the Saturn V wiki yesterday, and while it has a huge fuel load, that fuel load is completely consumed just to get the rocket to the point where the first stage can drop. It burns 400 tons of fuel per second!

Seems plausible, although I suppose it depends on what you want to do once you're in space and where you want to go.

If you could get a mass driver system to work very reliably (more reliably than chemical rockets), I could see it being useful to deliver the critical assemblies of RTGs into orbit. Plutonium is pretty resistant to high acceleration. One of the main objections to putting a lot of RTGs in space (or even better, actual nuclear reactors) is what happens during a launch failure. A system that doesn't involve any fuel, and thus won't blow up and scatter the contents around, or require tons of containment, might be useful.

Nuclear reactors have a pretty nice energy-density profile, particularly if what you want is long-duration energy for interplanetary missions.

Still hard to get around the capex problem though.

This is exactly the problem that SpaceX (And Blue Origin eventually) are trying to solve with re-usability. Think of a SpaceX launch, the "big" part of the rocket is thrown away just minutes into the flight, once it's fuel is expended. There's really no way to shortcut that, you couldn't refuel that part in-flight. Refueling that part after it lands again is much less expensive than building a new one entirely.

Meanwhile the payload on the top is now "in orbit" but operating in a vastly different environment than the larger component, with different engineering constraints, quite likely using different types of fuel.

If you could couple the initial launch vehicle being "cheap and easy" with another system for moving around to different orbits or getting into even higher orbits (https://en.wikipedia.org/wiki/Skyhook_%28structure%29 or a "space tug" type orbital vehicle) now you have access to high-orbit/cislunar space, and you can start assembling your larger inter-planetary vehicles or space stations there. (Perhaps using a mass driver to move heavy but necessary things like water and other raw materials to the site at this point, but that greatly depends on what the cost-per-pound of launch can get down to, and what resources you can scrape up out of the solar system)

It comes down to the numbers. Falcon 9's first stage accelerates at about 12.5m/s^2. That's about 1.25g's. It does that for 161 seconds, and first stage separation happens at an altitude of 100km, at Mach 6. That'd be A LOT of track to build straight up. The troposphere ends at 20km. People have trouble breathing before you get to there.

Pilots try to keep plane flight take-offs below 1g. You could possibly increase the g load but not everyone is conditioned like a fighter pilot. Even supposing you could get up to speed faster, you'd have a lot more drag in the lower atmosphere. You can't really get up to any significant speed until you exit the troposphere.

And that means building at least 20km, straight up.

Because:

(1) Rockets are proven technology, nobody has yet been able to put anything into LEO using a railgun

(2) Rockets have significant freedom when it comes to choosing the final orbit to deploy to

(3) Rockets are relatively cheap on the 'capex' front, a 200 km long electromagnetic launcher up the side of a mountain will be extremely expensive and may not even work...

Of course if you're so sure about this then maybe start a company to create an electromagnetic launcher, it should be easy to raise money assuming you are right.

> Of course if you're so sure about this then maybe start a company to create an electromagnetic launcher, it should be easy to raise money assuming you are right.

Could you explain how it would be easy? In my mind it seems hard to raise several billion dollars, no matter how you put it...

It's a joke: if it really was clear that a railgun was "the way to go," companies would be doing it.

It's apparently not hard to raise several billion dollars if you're trying to sell pet food online, create an app that dispatches taxis, or post extremely short messages on a web site.

>It's apparently not hard to raise several billion dollars if you're trying to sell something that makes hundreds of billions of dollars

Sorry, when did sell online pet food retailing, taxi dispatching apps, or instant messaging apps, come close to making anywhere near "hundreds of billions of dollars"

https://en.wikipedia.org/wiki/Going_concern

Efficient market hypothesis.

I haven't estimated building costs yet. Several billion could be in the ballpark though.

Serious issue with rail guns is the acceleration. Might be suitable for cargo lift, but accelerate a human quickly enough and it squishes rather unpleasantly.

In theory you need to apply 7.8 km/s of deltaV to an object to put it into orbit. In practice, you also need to use more to overcome to factors. There's gravity drag, the action of gravity on your rocket while it's speeding up to orbital velocity. There's also atmospheric drag, the resistance of the atmosphere to your rocket when it's traveling up to space. The faster you accelerate the less gravity drag but the more atmospheric drag you experience since you spend less time accelerating but are traveling faster lower in the atmosphere, and drag is proportional to the square of the velocity.

As you can see, using an electromagnetic cannon where you are at full speed leaving the barrel of the gun, you will have zero gravity drag but an immense amount of atmospheric drag. This means both that you will have to apply much more deltaV and that the heating caused by the passage of your capsule through the atmosphere will be enough to destroy any known material.

Finally, if you launch at full velocity from the surface of the Earth then the rules of orbital mechanics mean that, if you are going into an elliptical orbit, your orbit will intersect the launch point eventually and your capsule will lithobrake on the ground, also destroying it.

Neither of these problems apply to launching from the Moon back to Earth, for the record. The moon has no atmosphere to apply drag and heating. And if you're leaving the Moon's orbit for Earth you're on a hyperbolic orbit rather than an elliptical one and won't be coming back around in any case. So for getting off of airless Moons into orbit around their primary electromagnetic launchers are an excellent idea.

It's not impossible. After all, a "space gun" reached 117 miles in altitude in 1967.[1] A small satellite launcher is probably possible, especially with a rocket upper stage.

In recent years, military work on electromagnetic launchers and weapons has increased substantially. The US Navy has a railgun in test, and plans to mount it on a ship (probably the USS Zumwalt) next year. The new generation of carriers use electromagnetic catapults. Many of the components thus already exist.

The acceleration in a gun-type launch would be huge. All the acceleration has to take place in the length of the gun or track. Hardware, yes; humans, no. Think in terms of bulk launching of small satellites for mobile/Internet access.

[1] http://www.atlasobscura.com/places/project-harp-space-gun-ba...

An electromagnetic launcher isn't the way to go not from earth, especially if you want to put people into orbit.

Forget about the energy required, forget about the electromagnetic field and what it can do the payload even non-biological one.

An electromagnetic launcher is effectively a canon, calculate the acceleration needed to achieve orbital velocity not to mention escape velocity and you'll get to a point where you can't have anything surviving the launch. Not to mention that achieving Max-Q effectively on the launch pad is both the least effective thing you can do as well as the most dangerous one.

Acceleration over 200km will get you to orbit. The electricity required for a single launch is measured in US pennies.

An aerodynamic vehicle could be extremely safe; in the event of a failure, you could just glide back to the ground. The possibility of violent explosion would be practically non-existent.

Have a source on the electricity costing pennies?

It's not going to cost pennies, it's an easy calculation treat energy efficiency as 100% efficient and calculate how much watts you would need to put say a 10,000KG payload in orbit and then calculate how much it would cost you in terms of electricity.

F (d)x/(d)t = F * v = m * a * v = joules per second = watts

Please see my response to kragen, where I derive the cost.

The electricity and cost of the rail gun isn't measured in pennies.

An Aerodynamic vehicle could be safe, but it would also mean it would generate considerable drag on the way up requiring more power, it also would be safe only after almost reaching orbit because it would be hypersonic out of the launch pad, if you can design a hypersonic glider NASA would like hear about it.

Building a launch pad over 200KM is also not a simple feat, if you ask why people are building rockets it's because we have no clue how to build EM and by all accounts it's not sustainable for earth.

EM launchers for the moon and even mars as well as large asteroid bases are considerably more sensible.

> The electricity and cost of the rail gun isn't measured in pennies.

I think you should back this up with real numbers.

But where is the 200km of track? You also have to deal with the atmosphere. If the track is on the ground or low in the atmosphere, then when you come out of the "cannon" you would be torn apart. So that means the track (or at least the exit) needs to be above the atmosphere, which is quite a challenge in itself.

You could use a 200km evacuated tube. There is of course the issue of maintaining the vacuum as the vehicle exits. This could be managed using a pair of diaphragm shutters. Then only the space between the shutters would need to be reevacuated after a launch. Shutter failure would be disastrous.

According to this document from the USAF, the estimated energy needed to launch into orbit is 67GJ. Converting that to the US measurement of electricity is 18,611 kWh. According to the EIA, the US average for electricity, specifically for the transportation sector, was 10.22 pennies per kWh.

So yes, a lot of pennies...

edit: The same document mentions that 20 degree launch trajectory gives LEO == HEO delta-v. So that 200km track also needs to, naively, be 72km high at the end to have constant acceleration through that 200km.

[0] - http://www.star-tech-inc.com/papers/lcls/low-cost_launch_2.p...

[1] - https://www.eia.gov/electricity/monthly/epm_table_grapher.cf...

That's 67GJ per KG? because otherwise either I'm very very wrong, or they are very very optimistic.

Looks like that's for a 300kg total weight projectile @ 20% energy efficiency. Details can be found on page 317-318.

>An aerodynamic vehicle could be extremely safe; in the event of a failure, you could just glide back to the ground. The possibility of violent explosion would be practically non-existent.

That is a ridiculous statement, considering the deaths that have happened during reentry

https://en.wikipedia.org/wiki/Space_Shuttle_Columbia_disaste...

https://en.wikipedia.org/wiki/Soyuz_11

https://en.wikipedia.org/wiki/Vladimir_Komarov

Don't forget that the vehicle would be leaving the launcher at greater than reentry speeds, so it would be even worse than a normal re-entry.

Not all vehicles are safe. The shuttle was >100,000kg incoming. Passenger vehicles to orbit could be on the order of 1,000kg, which would be much easier to make safe for gliding. Emergency parachutes could even be feasible.

LEO is -29.8 MJ/kg - -62.6 MJ/kg = 32.8 MJ/kg (https://en.wikipedia.org/wiki/Orbital_speed#Tangential_veloc...) so one tonne is 32.8 GJ. At US$0.04/kWh (a common benchmark wholesale price for electrical energy in the US, usually stated as US$40/MWh) that works out to US$364, which is thirty-six thousand pennies.

What are you launching, a sparrow?

Also, doing it over only 200 km will require at least 16.7 gees. That's not survivable for people. (Or probably sparrows either.)

Calculations in units(1) format in case I got something wrong (should the delta-specific-energy really be 32.8 MJ/kg rather than, say, 24 to 30?):

    (-29.8 - -62.6) MJ/kg * 1 tonne * US$ 0.04 / kWh
    (-29.8 - -62.6) MJ/kg / 200 km / gravity

    m=10,000kg
    a=90m/s^2
    vf=7500m/s
    vi=0

    F=ma
    W=Fd
    P=W(t^-1)

    t=(vf-vi)/a
    d=(vf-vi)^2/(2a)

    t=7500/90
     ~83.33s

    d=7500^2/180
     =312,500m

    F=10000(90)
     =900,000N

    W=900000(312,500)
     =281,250,000,000J

    P=90(281250000000)/7500
     =3,375,000,000W

     3,375,000kW * ((7500 / 90) / 3600 s/hr) = 78,125 kWh
     78,125 kWh * $0.04/kWh = $3,125 / launch

     10,000 kg/launch / 3,125 $/launch = 3.2 kg/$ = $0.3125/kg

Looks very close to your numbers. I maintain my claim.

The electricity required for a single launch is measured in US pennies.

78,125 kWh x $0.04/kWh = $3,125 / launch

I maintain my claim.

The two claims above are incompatible.

$ is a unit of pennies

    $=100p

The calculation above was for a 10Mg payload. Payloads up to 3.2kg would cost under 100p under the same assumptions.

Your calculation is very very incorrect, not the math, but the concept you are missing too many variables.

The US navy actually is building railguns, their efficiency is very very low due to the resistance from inductance it seems that if we go the the equations for railguns your 200KM EM gun cannot be built.

Overall the US is designing a 64MJ railgun, this gun can't put anything into orbit, it will have a range of about 20 miles, the ship that is going to be equipped with it is going to have a 78MW power plant and while it can power a single rail gun it will not be able to power multiple ones. By the US Navy's own calculations it would require 28MW to launch a projectile at 32MJ which which means yeah.... these figures are all off by orders of magnitude.

It seems there is much more to railguns than classical mechanics.

I believe the barreled design to be fatally flawed. An earthquake or other disruption during launch could be catastrophic to the vehicle. Much better for the vehicle to ride next to the track; in failure scenarios, the vehicle can simply detach and glide back to earth.

Yes, it means you have to go through the atmosphere, but it doesn't take far to clear. That will effect the calculations some, but not much. What will really effect the calculations though, is the cost of electricity. Generation costs will surely drop below $0.04.

Further, transmission loss can be almost entirely mitigated by generating and supplying the required power on-track.

Launches in favorable conditions (moon and/or planet alignments) would probably make some launches even cheaper. Of course, if the launcher were operating continuously, the savings would be used up during unfavorable conditions.

You're suggesting that we should magnetically levitate the launch vehicle in free air above a track running along the ground? That seems like it makes the problem a lot harder — instead of having to push it through just the air between here and space, which is about ten tonnes per cross-sectional square meter, you're pushing it through another two or three hundred kilometers of air, which is about another 200 tonnes of air per square meter. You know that shroud of plasma surrounding a re-entering spacecraft? That's the power required to push an orbital-speed object through air — but in that case without even maintaining velocity, let alone rapidly accelerating, and in that case it's the rarefied air of the stratosphere. You're proposing to do that for the majority of the track. That seems like a bad idea.

Yes, generation costs will likely drop significantly below US$0.04/kWh eventually. But that's a Kardashev-Type-1 kind of event. Generating the power on-track may not turn out to be less expensive than long-distance transmission, because it depends on things like sunlight availability. Of the few suitable sites, most are pretty cloudy on one side.

No moon or planet alignments significantly reduce the energy barrier to get to orbit.

The track would not run along the ground. You'd want to get height as quickly as possible to minimize atmosphere and track length.

Originating somewhere around Mojave and launching towards Las Vegas could work.

Alternatively, you may want to launch over the ocean for safety reasons, but it seems like you may be subject to more weather concerns.

You don't want to run the track along the ground? Where then, on a dirigible at the mesopause as suggested in http://postbiota.org/pipermail/tt/2011-August/009613.html?

Bouyant structures could be used, as well as wings, to suspend the track.

You can't run along the ground because you don't want to go too fast through atmosphere.

jsprogrammer has the merit of having actually done some calculation, which you evidently have not — although if you think resistance is a thing that inductance can have, maybe you are not in a position to do any calculations. Either way, you should rethink your attitude.

Naval railguns don't have the luxury of accelerating over hundreds of kilometers; they are optimized for muzzle velocity, not efficiency. The things I've seen videos of launch projectiles at about Mach 7.5 (2500 m/s) over about 7 meters. That implies an average acceleration of at least 45000 gees, which means that you're going to have to accept significant inefficiencies that you can avoid in a design that accelerates three thousand times more gently.

In practice, both rotary and linear electric motors typically have efficiencies of over 80%, often over 95%. The proposal in question is a 200-km-long linear electric motor running up the side of a mountain. It's clearly feasible, but it won't cost pennies per launch.

You do need to partially evacuate the launch tube to shove your launch vehicle through hundreds of kilometers of it.

It is not plausible that a naval railgun uses only 28 megawatts. Traveling 7 meters at an average of Mach 3.75 takes 5.6 ms; if the total energy output is 64 MJ, that's an average of 11.3 gigawatts†, which is a lot more than 28 megawatts or for that matter 78 megawatts. So what you do is you charge a big low-ESR capacitor bank (at less than 78 megawatts) before the shot, then discharge it during the shot (at tens of gigawatts). If you're doing that, though, you have no limit on how many railguns you can run from your 78 MW power plant, just a limit on how many total shots per second you can fire among all of them. Your entire paragraph on the topic is incoherent nonsense.

For comparison, a .22 LR 30-grain (1.94 g) copper-plated hollowpoint bullet traveling at 500 m/s out of a 510 mm AR-15 barrel only has at most 2 ms to accelerate to its final 240 J energy and therefore requires over 120 kW of power.

Classical mechanics are perfectly adequate for all of this. No relativistic or quantum effects are relevant. Your suggestion otherwise is absurd.

Calculations in units(1) format for those who want to check them:

    (mach 7.5)^2/2 / 7 meters / gravity
    7 meters / (mach 7.5/2)
    64 MJ / (7 meters / (mach 7.5/2))
    30 grains
    510 mm / (.5 500 m/s)
    30 grains (500 m/s)^2/2
    30 grains (500 m/s)^2/2 / (510 mm / (.5 500 m/s))

† With constant acceleration the power output ramps up linearly and ends up at twice the average. With constant power the acceleration ramps down instead, which means that you have to start out at even higher accelerations to get the same average acceleration, and your total time in the barrel is shorter, so your average power is higher, although I don't feel like doing the simple calculus to quantify this at the moment. In either case you have at least a point where the power is a few times higher than this average.

Your calculations are correct, with the caveat that that's the average power assuming constant acceleration, not the maximum power. But you can't launch a one-kilogram thing and get it through Earth's atmosphere; you need an aerodynamic fairing and ablative tiles and whatnot, not to mention whatever kind of inductive apparatus you're using to grab hold of it in the first place. Thus my remark about the sparrow!

Until your objects get long and thin enough that drag matters, Newton's impact depth approximation applies to the atmosphere: if you're going straight up, you have roughly 10 g of air per cross-sectional mm² that you will run into on the way, assuming you can keep the hypersonic aerodynamics sufficiently under control to keep from just totally tumbling end over end, which is harder than it sounds. At the shallow angles available running up mountainsides, the situation is several times worse.

It's unfortunate that people incapable of doing calculations themselves are downvoting you. Calculations like these, plus experiments to validate them, are how we got rockets in the first place.

> That's not survivable for people. (Or probably sparrows either.)

A peregrine falcon can take 20Gs [1], at least over short periods. Maybe we need to start an astrobird training program.

[1]: https://www.quora.com/How-much-G-force-can-a-bird-take

> That's not survivable for people.

Got me wondering. https://en.wikipedia.org/wiki/G-force#Human_tolerance reports an early experiment of a human taking 10g for 1 minute. That'd add up to 5.8 km/s, which is still under delta-v to LEO (9.4 km/s or more including air drag according to https://en.wikipedia.org/wiki/Low_Earth_orbit). But it's close, and the Wikipedia page doesn't say it's an upper bound. OTOH at 10g to 9.8km/s (higher delta-v to be conservative and for easier math) the track would need to be more like 490km long (and it'd take ~100sec). And maybe the deceleration on hitting the atmosphere would be worse, I don't know. It sounds more plausible for cargo.

Doesn't have to be all electromagnetic. First or second stage could be. And it would be used for cargo perhaps, not people.

Yet it doesn't seem any private enterprises are trying to do it realistically. So you're right, calculations are probably telling them it doesn't work.

Laser propulsion is thought to be a better way for launching from the Earth, and has been tried with small models. You build a powerful laser ground laser system, and through some method I'm not clear on, heat air in pulses? in the bell of the rocket chamber. Once that gets too thin, the top layer of the chamber is ablated to provide the final thrust needed.

Supposedly can be gentle enough for humans, see some of Jerry Pournelle's fiction and non-fiction for examples of this.

Or, if you're in a hurry, there's always Orion (see Niven and Pournelle's Footfall): https://en.wikipedia.org/wiki/Project_Orion_(nuclear_propuls...

Laser propulsion could work. I'll need to learn more about the physics of it.

Because it's easier to build a reusable rocket than a reusable railgun, especially a reusable railgun that's hundreds of kilometers long.

Easier solutions aren't necessarily good. Rockets are extremely dangerous and expensive.

The ~200km required is daunting, but imminently possible.

Many things are possible. But military experiments on a scale four orders of magnitude smaller have revealed many previously unsuspected engineering difficulties with putting the technology into production. Solving those problems at the 20m scale would seem a necessary prerequisite to successfully building a space launcher at the 200km scale.

Pointing your rail gun up is best for getting out of the atmosphere ASAP, but how do you build a gun that goes, say, 20 km high into the air?

A mostly (the curvature of the earth will make it point up) horizontal gun will have the problem that the projectile will leave the gun at escape velocity, at around 1 atmosphere of air pressure (either after accelerating through 1 atmosphere of air pressure, which takes lots of extra energy, or leaving a vacuum, which makes exiting the gun barrel quite a bang)

Also, your barrel will have to extremely straight.

I don't think either is 'imminently' possible.

Pointing your rail gun up is best for getting out of the atmosphere ASAP, but how do you build a gun that goes, say, 20 km high into the air?

It's physically possible for us to build a 100km tall tower with current technology. (Columns formed from many highly pressurized tanks made from boron.) It's not really economically feasible, however.

I wonder how well such a tower would withstand the forces involved in also launching projectiles at escape velocity.

As of around the year 2000, I remember a figure of 11 miles as the height limit for a fairly conventionally built skyscraper. If we used lightweight truss construction, I suspect we could build a load bearing tower. If we used aerospace grade materials, we could do even better. (Yes, it would be quite expensive.)

Quite expensive like "a significant portion of the annual gross global product"

Most of the load could likely be mitigated by wings and buoyant structures.

Most of the acceleration track floats between San Cristobal (Galapagos Islands) and Ecuador. Near the end, it ascends Mount Chimborazo to 6000m and punches through a plasma window separating only 0.46 atm external air pressure from the low vacuum inside the accelerator.

At our current level of technology, I estimate this would cost at least $200 billion in capital expenses.

A neutral-buoyancy under-ocean train between two cities otherwise un-connectable by terrestrial transit would prove some of the necessary technologies. Liverpool to Belfast would probably work. Miami to Havana to Cancun would work, if not for politics.

Relevant search terms: "vactrain", "StarTram", "transatlantic tunnel"

You'd be better off putting that money into space elevator research.

I was just explaining how possible it is. I am skeptical that anything like it will ever actually be built on Earth.

Other non-rocket launch technologies do seem more promising. If you could build a mass driver launcher between Galapagos and Ecuador, you could likely build a launch loop in the same spot for less capital, less operating cost, and higher capacity.

Reply on news.ycombinator.com