Aren't nuclear reactors kind of heavy

Yes. Though most of it is shielding which doesn't need to be installed near the rear of the vessel. Turbines are surprisingly compact and are not as heavy as one might assume. Same with the nuclear core.

and full of moving parts that break?

No. Or more precisely, "not exactly". Nuclear reactors generate heat. How you translate that heat into electricity is where your moving parts tend to come into play. Using a working fluid like helium combined with a Brayton cycle turbine is efficient, well tested, and highly reliable.

Some serious engineering is going to have to go into making the power source light enough

NASA already did the work for a probe called Prometheus. While Prometheus was scrapped to make budget room for the Moon return plan, the reactor designs should still be on file. There might need to be more development done to scale the reactor upwards, but it's not brand new engineering.

I have a problem where my X-server grows to over a gig of RAM even when all windows have been closed.

You do realize that X memmaps the video memory into its memory space? This gives it some rather crazy numbers for RAM usage even when it's running thin. Especially with modern cards having 512MB of video mem or more.

Otherwise I do not disagree with your general statements. :-)

No kidding. But this engine is able to take large amounts of energy and impart it into very little matter. 200 kW per engine for 5 newtons of thrust to be specific. At an Isp ranging from 3,000 to 30,000 seconds, it needs very little reaction mass to operate. But to sustain that much power, it needs a nuclear generator. And if you're going to pay for the weight to generate 200 kW, you might as well pay to generate a megawatt or more. You'll need to carry more nuclear fuel and reaction mass, but the high efficiency means that the reaction mass will be a relatively small mass penalty. The nuclear fuel is already pellet sized, so that's the least of your concerns.

Follow? Good.

I'm trying really hard to come up with an appropriate response to that. Unfortunately, you have left me speechless and dumbfounded. Maybe some day I'll even understand what the point was you were trying to make.

Thanks for the excellent post! But there's one minor detail I'd like to inject here:

If you're going to pay the weight cost for a nuclear reactor (which is the only technology that can feasibly produce that much energy for sustained periods), you might as well attach more engines to it. This will give you far more thrust for the journey as well as spread the weight cost of the reactor across many engines. And since much of the cost of the reactor is fixed (e.g. You've got to shield the thing from the crew) installing a higher output version should not add significant weight to the vessel.

I'll grant you that you'll need more fuel for more engines, but part of the point here is that these engines are extremely fuel efficient. So the additional fuel cost should not be unreasonably for the additional engines.

The problem with speeding up is that you eventually have to slow down, and slowing down takes plenty of energy and time too.

When you have constant thrust, this is an easy to solve problem. You speed up until you reach the halfway point. Then you turn the ship around and begin thrusting the opposite direction for the second half of the journey. Assuming sufficient constant thrust, you'll still get to your destination faster than the yahoos attempting a low-energy transfer.

As a bonus, thrusting forward and thrusting backwards are exactly the same from a relativity perspective. Which means that you'll get artificial gravity for the entire journey.

Also, when building up to this insanely fast speed, what are they planning to do if some random debris gets in their path?

At such a small fraction of c, there's no difference between a fast ship or a slow ship. Meteorites could be moving toward you at high speeds no matter what your speed is relative to Mars and Earth. The velocity imparted on the spacecraft only becomes a concern when the speed imparted on the craft is enough to move interstellar distances. At those speeds (relative to stars), the various materials floating around are going to be much slower than the craft because they're aligned to the gravitational forces of the surrounding stars and galaxies while you're moving against that flow like a bat outta hell. ;-)

Gas Core Nuclear Thermal Rockets are still science fiction. No one has yet built the necessary components, and there is a great deal of argument over whether or not "nuclear light bulbs" are even possible.

I'd love to see a 3,000 - 5,000 second NTR engine as well, but it would still be better suited for liftoff. For interplanetary travel, you simply can't beat the efficiency numbers of VASMIR. They start at the theoretical limits of NTRs!

these numbers are from the 60s

I don't have the reference in front of me, but I seem to recall that solid core NTRs were brought as high as 1200 seconds. On paper, anyway. No one has built them since the 80's timberwind project.

You're partially correct. But only partially. While you generally need to wait for proper alignment to make your journey, the length of the journey is still dependent on how fast you go. Chemical rockets are so slow that we need to begin the orbital transfer ~260 days before the expected orbital intersection with Mars. With more acceleration, the ship could leave later and still make the rendezvous.

Ok, that's horribly simplified. But I simply don't have the time to look up and explain the myriad of orbital transfers available. Suffice it to say, a little bit of extra speed won't help much at all. A lot of extra speed will open up many more options.

Or in other words, how fast you get somewhere depends on how much energy you want to waste to accomplish that goal.

Try looking at the specific impulse on those. ~800-1000 seconds. Now compare to 3,000-30,000 seconds. Which one is more efficient with its fuel?

NTRs are very, very cool. But they're very wasteful with the energy produced by the reactor. Potentially great for liftoff (if anyone ever building a modern variant without the graphite flaking problems), but nowhere near as useful for interplanetary travel as the VASMIR engines are promising.

So it seems like this will be a great way to power a spacecraft that's already in orbit

Correct. While it's theoretically possible to use engines like this as part of a liftoff stack (assuming enough engines, low enough weight per engine, and a high enough power budget), it's not really practical to consider such a concept at this time. For the short term at least, LEO access will remain the purview of chemical rockets.

And I quote (with emphasis added):

No matter how much of a cash cow your current product line is, you need to be investing in the R&D to compete in the next generation of products.

I never said the basic business had to change, only that a company's offerings had to change to remain competitive. IBM may still be in the office automation business, but they sure as hell aren't relying on punch cards and typewriters to pay the bills. Similarly, Hudson may be in a more modern form of their core retail business, but they're definitely not setting up forts and trading posts to acquire, process, and resell fur pelts.

The expected thrust is 5 newtons for 200 kW of power. However, they have only tested 30 kW of power.

For 200 kW per engine, we're thinking nukes.

For comparison, your car needs about 20 kW of power to maintain cruising speed on the interstate. 200 kW of power would be akin to running a 300 horsepower engine at its peak power output. With the way cars are designed, that doesn't happen much with the possible exception of expensive sports cars and pickups hauling a heavy load.

If we take the case of the sports car, we find that it's enough energy to slam you against your seat and hold you there while you do 0-60 in 3 seconds. (Hey look, ma! Artificial gravity!) In the case of a pickup pulling a heavy load, it's enough to accelerate reasonably while dragging a trailer full of spools of heavy steel cabling.

The difference between your car and the spaceship is that the spaceship will be powered by some sort of long-term fuel supply. e.g. A nuclear reactor. Which means that the spaceship will be able to continue accelerating for millions of miles while your car would have run out of gas after the first few hundred miles.

Since acceleration is cumulative, being able to continuously accelerate like that means that distances between planets become a lot smaller on one "tank of gas" as it were. Add more engines for greater thrust and redundancy, and you have a souped-up hot-rod of a ship that can take you interplanetary distances in record time.

Hmm... I'm sure someone is about to chide me for some horribly sloppy analogies, but look on the bright side. It's got cars in it! And hopefully it will make the energy budget a bit more understandable. ;-)

For those of you who are unclear on why the VASMIR system is so cool, allow me to give you a brief bit of background. Practically every propulsion method developed to date falls into one of two categories:

1. High thrust, low efficiency
2. Low thrust, high efficiency

Generally how it works is that the more power you get out of engines, the less energy you extract from the fuel. This is the case of chemical fuels like Liquid Hydrogen/Oxygen or Kerosine. These fuels provide the massive amounts of thrust necessary to get off the ground, but they burn through their fuel very quickly. Interestingly, LHOx is more efficient than Kerosine, but it's also harder to get as much raw thrust out of it. That's one of the reasons why Kerosine was the heavy lifter during the space race with the LHOx engines reserved for in-space stages.

On the other side of the coin, you have engines like Ion propulsion. These engines are able to inject incredible amounts of energy into tiny amounts of fuel, thus making them extremely fuel efficient. The only problem is that the amount of thrust is very low. Most of the ion engines that have operated to date produce thrust that matches the weight of a sheet of paper. Definitely not enough for liftoff, but perfect for extended missions in space where constant low thrust provides more velocity over time than the chemical engines which fire once, then coast the rest of the way.

The problem with both types of engines is that neither one gets spacecraft to their destination all that fast. Chemical rockets have the thrust to do it, but you couldn't feasibly build a chemical rocket with enough fuel to get you to another planet in a reasonable amount of time. A nuclear pulse propulsion craft could feasibly get fairly close, but it would just have more power in the intial thrust rather than providing a constant, high power thrust. (Obviously these have been discounted over the difficulties of building a large enough craft without using a nuclear ground launch. Nuclear ground launches are a no-no under current test-ban treaties.)

This is where VASMIR comes in. These engines are incredibly efficient. The specific impulse (measurement of efficiency) is between 3,000-30,000 seconds depending on the configuration and current thrust levels of the engine. This compares favorably with the ~450 seconds of shuttle engines and 3,000-10,000 seconds of Ion thrusters. Meanwhile, the thrust of Ion engines ranges from 90-3,000 mN while the thrust of VASMIR is expected to be ~5000 mN of thrust when tested at 200 kW of power.

What this means is that we may be able to build spacecraft where a trip from LEO to the moon is a daily affair and a trip from LEO to Mars takes only a few months (or less!) vs. the current flight time of nearly a year. The better these engines get (and the more we can put on a craft), the faster those flight times will get!