high school history

Recently, someone asked me about a startup called Longshot Space. They're trying to make big guns that're useful for space launch. Here's a video on them from Scott Manley. Many startups exist, and normally I wouldn't write a post about something like this, but it reminded me of my younger self.

Yes, back when I was a freshman in high school, I liked some flawed ideas, including:

• non-rocket space launch systems, like launch loops
• concentrated photovoltaic solar
• text compression as a route to AI...with low compressor size as a metric

So I had email conversations about stuff like that with various people at universities instead of doing homework.

Those seem pretty cringe now, right? What I didn't realize at the time was:

• If rockets are expensive, then the solution is to just make rockets more efficiently. Sure, governments already spent a trillion dollars on rocket development, but that doesn't matter for smart people with internet access.
• Similarly, if solar panels are too expensive, then just make solar panels more efficiently. Notably, diamond-studded wire saws for slicing thinner wafers was an obvious idea but it took a long time to go from invention to implementation.
• Neural networks with certain activation functions and L2 norm regularization don't overfit; see this post.

Now, past me was well ahead of Longshot Space conceptually, despite the tech progress since then, but I still have some sympathy.

Longshot's plan

The speed of gun projectiles is limited by the speed of sound of the propellant gas. Light gas guns can get high speeds by using hydrogen, which has a higher speed of sound.

It looks like Longshot Space has a bunch of gas tanks containing H2/O2, connected with burst discs to a gun barrel. The projectile has a long tail that gas pushes on sideways. The gas in a tank is ignited, the disc ruptures, and the hot gas pushes on the projectile tail.

Well...at least they're thinking about some of the problems...? Their design could well be cheaper than a comparable railgun...?

ram accelerators

Railguns and coilguns have a bunch of capacitors + wires + switches. This fancy gun thing has a bunch of tanks + igniters + rupture discs + gas fill lines. You know what's cheaper than those systems? Just the barrel part. Just a big metal tube.

Instead of the gun doing something as the projectile passes, we can simply have a ramjet projectile that adds fuel to the gas. Ramjets work in open air, but since the projectile is in a tube, it can compress gas against the tube walls. Now we don't need a path for gas thru the middle of the projectile. We can also use high-pressure gas in the tube to get faster projectile acceleration.

Of course, ramjets work best around Mach 3.5 and we have a wide range of speeds. Ramjets can compensate for speed changes with variable geometry, but since we have a controlled environment, we can instead change the speed of sound to maintain similar Mach numbers. For example, by going from (CO2 + O2) to (N2 + O2) to CH4 to H2, we can vary speed of sound from 0.9x to 4x that of air. If we use cryogenic air at the start, the initial speed of sound could be much lower, making ramjet startup at 400 m/s feasible.

Since we control the gas composition, we can also just have premixed fuel + oxidizer. And now we have a ram accelerator. That was demonstrated at the University of Washington in the 1980s, and has gotten up to 2.7 km/s.

As I recently mentioned, the main problem with scramjets is fuel mixing speed, so with premixed gas, scramjets are much easier. Thus, 8 km/s is considered feasible for ram accelerators.

that's still expensive

So, we just build a big metal tube. No fancy tanks or valves. No wires and power electronics like an electromagnetic gun would need. Is that cost-effective?

I suppose that depends on what you're comparing it to. Is it cost-effective compared to a railgun, or a big centrifuge in a vacuum chamber? Sure, I guess. Compared to good rockets? No.

"Just" build a big metal tube that's very long and very straight, huh?

Oil pipelines are expensive, and they don't have to be straight. This would be much more expensive than an oil pipeline. If you could build a big metal tube that's long and straight cheaply, then I think a better use for it than a space gun would be hydrogen tube transport.

There are also some obvious issues with a gun pointing sideways not being ideal for going to space, and drag at Mach 20 at ground level being rather high. But this has all been discussed before.

cost reduction

For compressed air energy storage using air at ~70 bar, it's well-known that underground caverns are cheaper than metal tanks on the surface. Perhaps an underground tunnel made with a TBM would be cheaper than a tube on the ground, especially considering land costs? But of course, tunneling thru rock is still expensive. Perhaps something else would be cheaper to make a tunnel thru?

Back in 1997, Andrew Nowicki made 2 proposals for mitigating the construction cost of guns for space launch:

• Make a tunnel through the antarctic ice sheet to act as the barrel of a gun for space launch.
• Instead of making a solid metal tube for a ram accelerator, make a disposable balloon from eg mylar. Fill it with hydrogen and possibly a little oxygen, and run a ramjet projectile through it.

I'm not convinced those designs are particularly practical, but at least they were interesting ideas, which is more than I can say for Longshot Space or SpinLaunch. Interesting ideas tend to be in areas of conceptual space with rich possibilities. Ramjets, guns, pipes, tunneling, and balloons are all useful. Even if using guns for space launch is impractical in general, we can say, for example:

Natural gas pipelines use high pressure, making them similar to pressure vessels. If underground chambers are cheaper than pressure vessels for CAES, then perhaps making underground pipelines for natural gas transport is practical?

a minor point about high-velocity guns

Rockets are often considered inefficient, but they're actually quite efficient at expelling gas at ~3 km/s. The inefficiency generally comes from mismatch between rocket speed and exhaust speed. So, any reasonable launch system should probably involve rockets, and any launch assist system should probably involve launching a rocket at under 1500 m/s.

— me

If it's not worth launching a rocket from a low-speed gun, and it isn't, why would it be worth launching it from a higher-speed gun? Yes, rockets need to be adapted for firing from guns, but for a given reduction in rocket mass per gun cost, a lower-velocity gun can be wider and have lower acceleration.