Solar power is not a method of propulsion, it's a method of collecting energy. Perhaps you mean ion engines paired with solar panels? If so, the difference in thrust is massive, as in several orders of magnitude - to the point where getting anything larger than a satellite moving with electric-based propulsion would take many orbits of acceleration. You also run into issues of not being able to collect power when in shadow, needing power storage banks (batteries are very heavy), and diminishing power output as you move away from the sun (ex. at Mars, solar energy density is 40% less than at Earth).

You have to have some pretty big engines before the mass of the reactor is a small enough portion of the mass that the Thrust to Weight ratio of a nuclear thermal engine + tank + reaction mass is better than that of a chemical engine + tank + fuel. The only place that kind of thrust is needed is getting off of Earth, and a nuclear engine doing that is a non-starter.

And inside the orbit of Jupiter, solar + ion engines are way more efficient.

The Ars commenters are very well informed, they go back and forth on this a lot. I highly recommend reading the comments there.

The implied goal is to send humans to places in the solar system. That means life support, food, water, etc. So a lot of mass anyway. Worse, the longer the trip the more mass you need to bring onboard. Nuclear would allow for the shortest trip durations while ion engines would be the longest. With ion engines you have the issues of micro-gravity health risks on the crew, psychological risks and the need for a lot more supplies.

For super long trips, efficiency is king. Coupled with the ability to ability to burn continuously, solar-electric and nuclear-electric are way more efficient than nuclear-thermal. Ion engines may have low thrust -- just use a lot of them.

Nuclear-thermal's Achilles heel is that you can't use it to boost off Earth. So you need a high thrust chemical engine to get it into space. It's hard to make the numbers work -- nuclear-thermal may be more efficient than that chemical engine, but it's dead mass while you're boosting it out of the Earth's gravity well. If instead you can use the same chemical engine to get to Mars that you used to boost yourself out of Earth, you don't have that dead mass. If you want to get to Mars fast, just boost some extra fuel on an extra ship.

Nuclear-thermal looks nice on paper; it's probably the highest efficiency high thrust engine that's achievable with today's materials. But it's hard to find a good use case for it. A manned Earth->Mars run used to be it; but SpaceX's plan for in-space-refueling shortens the trip time just as much, if not more.

I think you’re overestimating the minimum size of a nuclear device, they can be small as a larger watermelon.

That's still several orders of magnitude more massive than the chemical and ion thrusters used on small satellites.

But don't take my word for it, there are 12 pages of intelligent comments on the Ars article, comments mostly of much higher quality than any comment here, including and especially mine.

And those comments are fairly disparaging. Unless you're going past Jupiter, in which case Nuclear-Electric-Propulstion is your best bet. And nuclear-electric is very different than nuclear-thermal.

If you can get 1g of acceleration, the trip to mars is a couple of days, I don't think you could ever get that with solar.

Edit: reaction mass becomes the problem, say to move a ship 10^4kg, 2x10^9 m (the distance to mars) at 1g, I adapted an answer from [1]

So say the craft has a mass of 10^4 kg. To accelerate at 1 G you need F = ma = (10^4 kg)(10 m/s^2) = 10^5 N. To get a force of 10^5 N over 2x10^9 m (distance to mars), you would need (10^5 N)(2x10^9 m) = 2x10^14 J, lets say 10^14J. Antimatter is the most energy dense material we know. To get that from antimatter you would need m = E/c^2 = 10^14 J/10^17 m^2/s^2 = 1gram. So you'd only need 1gram of antimatter - is that right?

CERN has made around 1 nanogram so far, 1gram would cost around $25 billion in a 2006 estimate, its down from $62 trillion in a 1999 estimate [2], so it may be feasible one day...

Edit2: The moon is 4x10^8m, so you'd need 0.1gram of antimatter, you'd be there in an hour or so.

[1] https://forum.nasaspaceflight.com/index.php?topic=34996.0

[2] https://en.wikipedia.org/wiki/Antimatter

Only if really all the mass of the antimatter is usable energy for propulsion. A large chunk this energy would probably be in the form of gamma ray photons: they are not usable for thrust and irradiate the materials of the spaceship, another large chunk would be heat. And I assume you would also need quite a bit of energy for whatever vessel is holding the antimatter in place (magnetic fields maybe?), to avoid that the entirety of the 10^14J is produced instantly by annihilation with normal matter.

Nuclear thermal isn't chemical propulsion. It heads a propellant to a high temperature using an onboard reactor, and then expels it.

The advantages over solar power is thrust (which, within the solar system at least, is important if you want to get somewhere fast), and the fact that it works further out from the sun, which means you could use it to get mass to the outer planets (or mars) relatively quickly. Solar propulsion, yes, is limitless, but and can reach much faster speeds, but it can take so long to reach them, that by the time you're going faster than the NTR, the NTR has already arrived at its destination.

The advantage over chemical propulsion is efficiency.

You can also have solar thermal propulsion where you reflect large amounts of sunlight onto some hydrogen to heat it up before using it as propellant. The efficiency is roughly the same since in both cases you're basically limited by the melting point of your engine. Solar thermal designs tend to have lower thrust than nuclear thermal but much more thrust than solar electric.

There are also things like solar electric water plasma, check out Momentus. Might be a good 'middle' for thrust.

>I’d like to see them explain the benefits of this over solar power.

One big problem with solar energy is that it is a diffuse power source (i.e. energy per unit area is low), and gets exponentially more diffuse the further from the sun you go.

The second is that though solar energy is limitless, the components that need to use chemistry to turn solar energy into electricity, are not.

I assume you're talking about solar-powered ion engines, as solar panels cannot on their own move a spacecraft appreciably.

The big difference there is thrust. Ion engines are efficient but have very low thrust, so it can take months to change your orbit. Chemical and nuclear propulsion is high thrust, so we're talking minutes instead. Nuclear has higher propellant (i.e. mass) efficiency than chemical as well.

No, solar thermal can generate similar thrust to nuclear, using less mass. Nuclear only really makes sense past Jupiter.

The average solar irradiance on Mercury is over 9000W, on Earth about 1400W, on Mars 600W, on Saturn 15W, Pluto 0.87W, per square meter(or per 10ft^2, or per 14 medium sized Pizza). Farther you’re out the sunlight must disperse to a broader area thus gets weaker. Inverse square law you know.

Which means the solar array must be twice as large for same output on Mars alone, and solar propulsion is probably completely infeasible due to weight and volume constraints, if not structural as well, on Saturn orbit and beyond.

For missions further out or for Mars missions that cannot accommodate twice larger wings, we must “carry our own Sun” for power.

Which isn’t wrong at all, after all the Sun is a natural fusion reactor so solar panels aren’t like it’s free of stigmatic nuclear technology.

Oh and also nuclear propulsion aren’t considered chemical by the way...

Combination of both high thrust-to-weight ration and a high exhaust velocity. The first ensures you can accelerate quickly when not loaded with too much fuel, the second ensures you don't need to bring that much fuel.

Solar by itself doesn't provide reaction mass (outside of solar sails, which are very low thrust). You would probably be combining it with ion thursters, which have really high exhaust velocities, but low thrust to weight ratios.

Well, now that I think about it, our chemical propulsion requires an oxidizer, and judging by how this works, this doesn't need one. So by using this method they cut down on the oxidizer mass, which is something fairly significant. Not sure if it'll beat an ion thruster powered by the sun when it comes to convenience, though.

It's wrong to think of ditching oxidizer and keeping fuel. They both contribute as components of the rocket's energy source and its working mass. If you get rid of the chemical reaction, the "fuel" is not fuel any more, and you should replace your concept of it as such with a narrower concept of "working mass."

The questions here include:

- whether you can get more efficient propulsion overall when the heat comes from a reactor instead of from a reaction in the fuel

- whether switching from a fuel/oxidizer energy-source/mass combo to another working-mass-only substance might net you other engineering advantages (storage concerns, perhaps? though if you're switching to hydrogen like one often considers for nuclear rocket working mass, you will have plenty of storage problems)

- what sort of engineering disadvantages the nuclear reactor component brings (added mass, heat dispersal issues, etc)

Solar does not work very well beyond Mars/Asteroid belt distances.

But better than nuclear anywhere within Jupiter’s orbit.

The article says it is about getting to "close" destinations, like Mars, as fast as possible. I am not sure how solar would achieve that over short distances.