First, some problems.
Your BFG isn't Big enough.
Wouldn’t this mean that this gun could destroy any planet in the galaxy?
Depends on what you mean by "destroy".
A 1,000 ton projectile at 0.9c is about 1e23 J. This is a lot, it's about 1/4 the energy of the meteor that killed the dinosaurs, but the impact will not "release more energy than a matter/antimatter reaction with the same mass". It's roughly the same.
Getting smacked by a dinosaur killer is bad, but it won't destroy the planet. A civilization advanced enough to be worth expending 1e23 J on can recover and fire back. We'll get to that.
You can, of course, pump this thing up with as much energy as you like. At 0.999c it's 2e24 J, but it still isn't a planet cracker. More like a major ecological disaster.
Sorry about your atmosphere.
firing a 1,000 ton “bullet” out of the atmosphere at relativistic velocity and into space
This 1000 ton bullet has to plow through your atmosphere adding drag, reducing it's final velocity, and greatly increasing the amount of energy you need to fire it.
Worse, it will compress and heat the atmosphere creating an enormous fireball at the launch site doing a lot of damage to your own planet.
1e23J is roughly a week's worth of sunlight for Earth. One shot will disrupt weather patterns. Fire a few of these and you'll cook your planet.
The first victim of the BFG will be the ones who fire it. This needs to go on a moon with no atmosphere, or in space.
How do you get it that fast?
Assuming you have the energy, and we're talking "Type II Civilization" levels of energy, how do you apply it to the projectile without obliterating it? This is the basic problem of a rail gun, how do you accelerate an object to relativistic speeds before it leaves the "barrel"?
You could make the barrel longer and longer and longer, but at 0.9c this thing will go the diameter of the Earth in 47 ms. Not a lot of time to apply energy.
Better to use a ring accelerator. Basically a giant cyclotron particle accelerator. The projectile spins around the ring, going faster and faster, held in the ring by powerful magnets. When it's reached its final velocity, the magnets release and it goes flying off.
This would take even more energy to hold the projectile in the ring while its accelerating. The bigger the ring, the less energy necessary. So let's build a big ring.
Everyone Will Fire Back
The problem with a purely kinetic is whoever gets hit can trace it right back to you. And they're going to be pissed. And they'll throw whatever they have left at you. As before, this is Type II Civilization levels of energy. Presumably your enemies are also Type II Civilizations to be worthy of such an investment which means they have more than one planet, and ways at striking back at you.
Even if they don't, everyone else is likely to be pissed and come to destroy this threat to the galaxy.
The Big Falcon Ring
This can't be in an atmosphere. It has to be huge. It needs enormous amounts of energy. It needs time between shots to gather energy and accelerate the projectile.
Pick a large moon with no atmosphere and build the ring on the surface (turns out this isn't going to work, see below). As others have pointed out, you'll need several rings to be able to fire at all points in the sky. Putting it in orbit doesn't help. Mine the moon itself for material. Put solar arrays in orbit to power it all, possibly even a Dyson Swarm around the star.
The amount of energy needed to keep the projectile in the ring path is related to its centripetal force. The formula for relativistic centripetal force is $$F = y m v^2/r$$.
y is The Lorentz factor, $1/\sqrt{1 - v^2/c^2}$. This accounts for relativistic velocities.
- v = velocity, ultimately 0.9c
- m = mass of the projectile, 1000 kg
- y = the Lorentz factor, ~2.3 at 0.9c
- r = radius of the ring
The force necessary to keep the projectile on a circular path is inversely proportional to the size of the ring. Double the size of the ring, halve the force necessary to keep the projectile in a circular path.
Starting with the radius of Earth's moon, 1.7e6 m, we get $F = 1000 kg * (0.9c)^2 / 1.7e6m$ or about 1e14 N. This is a lot. A Saturn V rocket puts out about 1e7 N. However, we can scale this down by scaling up the ring. Put it at Earth's orbit, 1 AU, and we're down to 1e9 N. Assuming the force is exerted over 1 m^2, this is 1 GPa inside the yield strength of steel.
Stronger and more exotic materials, like graphene with a tensile strength of about 100 GPa, would allow a 0.01 AU ring or a "mere" 1.5e6 km. Sci-fi materials would make it even smaller.
Put it in a system that isn't terribly important to you, because that's where everyone will trace your shot back to, and that's where everyone is going to direct their ire. The ring need not be orbiting the Sun, it could be at a stable Lagrange point.
I can't underscore enough just how large a project this is, even for a Type II Civilization.
The BFG needs Big Falcon Power.
We can calculate the forces involved for the cyclotron gun at maximum velocity. The force required to keep a non-relativistic projectile in the ring is F = mv^2/r. For your 1000 tonne projectile at 0.9c on something like the Earth's moon it's about 5e16 N: a lot. To put this in perspective, 5e16 N is the equivalent of lifting everything humanity has ever made 1 meter up in 1 second. And that's before relativistic considerations.
Putting aside the question of how you'll build a moon sided cyclotron that can withstand 5e16 N of force, something very, very large will have to power this. Not only is this a major investment by your civilization, but it's vulnerable to attack, and vulnerable while its being built. You'd need to pretend its some sort of civilian project.
But once it's fired your cover is blown and it is vulnerable. Like a nuclear weapon its role is a balance of terror. Once you use it its value to protect yourself is lost.
How Fast Can You Fire Again?
This thing needs enormous amounts of energy. And it needs time to accelerate the projectile. And the apparatus might be damaged in the firing. How often can you fire it?
This is largely up to you. You can tweak the numbers for your story. If it's once a year, then they can get maybe 4 or 5 shots off at a close neighbor before anyone realizes what's happened. Maybe it's longer. Maybe its shorter.
Narratively it gives time for the attacked civilization to react before another shot can be fired. Even better if they know they're doomed. The first shot has already landed, maybe somewhere relatively unoccupied to limit the immediate damage. They know more are already on the way and they can't stop them. But they can try to destroy BFG before it harms anyone else, and launch themselves at it knowing by the time they get there their planet will probably be wrecked.
Time On Target
If your civilization wants to be really smart, it conducts a Time On Target bombardment. It fires at its furthest targets first, then closer and closer ones. The end result is every target is hit simultaneously. Nobody can see it coming. Nobody gets any warning.
How Do You Know Who You're Firing At?
Is there some way that a planet could defend against this galactic scale weapon of mass destruction?
"Galactic scale" is a bit of a problem for your targeting. We're talking 100,000+ light years end-to-end. Being able to fire a relativistic kinetic projectile and hit a moving target 100,000 ly away is crazy complex and requires information you literally cannot have.
At the time of firing you'll be seeing your target as it was 100,000 years ago. While you can do rough calculations to determine its motion, you can't do this with sufficient accuracy to hit a planet both because of the crazy complicated and chaotic math involved, and because you cannot have sufficient detail at that range. You probably can't even see the planet.
There's also the question of how you got into a fight with someone it takes 200,000 years to communicate with (100,000 years out, 100,000 years back). When your projectile falls, who will even be on that planet? Will it even be the same species? It's like firing at us for something the Neanderthals did.
Either you're doing some extremely grand space opera, or you should scale it back. A lot. 20 light years offers you about 100 or so systems to play with and time scales that are inside a lifetime.
Do we even need a BFG?
What if we put a motor on this 1000 kg mass? What would that take?
Running the numbers we can use the rocket equation to find an ideal solution of how much reaction mass we'd need for a theoretically near perfect sci-fi space motor.
$$m0 = m1*e^{dv/ve}$$
- m0 = starting mass
- m1 = final mass, 1000kg
- dv = total change in velocity, 0.9c
- ve = exhaust velocity
The tech details don't matter because ultimately space motors throw mass out the back as fast as possible and rely on Newton's second law to be thrusted forward. It all depends on how much (reaction mass) and how fast (exhaust velocity) we throw it out the back. The higher the exhaust velocity the less mass we need. Less mass means a lighter spacecraft which requires less thrust for the same acceleration and the Tyranny Of The Rocket Equation works in our favor.
The exhaust velocity of a very, very efficient ion thruster is 210km/s. So that's $1000kg * e^{0.9c/210km/s}$ or $1000kg * e^{1286}$. $e^{1286}$ is so large even Wolfram Alpha won't give me an answer. So much for known tech.
If ve = 0.1c that's $1000kg * e^9$, 8e6 kg or 8000 tonnes of reaction mass. Not infeasible! Probably about as much mass as your average sci-fi space cruiser. Let's go faster!
Let's say we can throw mass out the back of this engine at ve = 0.9c! $1000kg * e$ is a mere 1700 kg of reaction mass. Great! We're in business... maybe.
What about the energies involved in throwing all this mass at relativistic velocities? How much mass are we throwing out the back and how much energy does it take? The mass is easy enough to calculate since we're at constant acceleration, $reaction mass / time$. How long do we need to be accelerating? Depends on how far the target it. The worst case is a nearby star system at about 4.5 ly. It's accelerating constantly to 0.9c, so the average projectile velocity is 0.45c. It'll be accelerating (hopefully) tiny fractions of this mass for roughly 10 years.
If ve = 0.1c, that's $8e6 kg / 10 years$ or 25g/s. This is a lot. The kinetic energy of 25g at 0.1c is about 1.1e13 J. This is a lot of energy. This means our engine must produce 11 TW for 10 years: 3.2e21J. Assuming the most mass-efficient generator possible, a matter/anti-matter reaction, and using $e = mc^2$ and $m = e/c^2$ that would require 35,600 kg adding significantly to our projectile's mass and throwing off the rocket equation. ve = 0.1c won't cut it.
If ve = the ludicrous 0.9c that's only $1700 kg / 10 years$ or 5.4mg/second. At 0.9 it has a kinetic energy of 6.3e11 J. That's still a lot of energy requiring 0.6 TW over 10 years: 2e20J or the matter/anti-matter reaction of over 2000 kg again throwing off our rocket equation, but not unrecoverably. I'm not sure on the math, but I'd estimate we'd wind up with something like 5000 kg of reaction mass and 5000 kg of matter/anti-matter.
What about beaming it power from a laser? If you can focus a 1 TW laser on a moving target 5 light years away and sustain it for 10 years, why are you messing around with throwing rocks? Just cook them.
As assuming we can come up with some lightweight matter/anti-matter power source, and a lightweight engine that can fire milligrams of matter at 0.9c, this can be done inside known physics... but not known engineering.