While building some more detailed spacecraft, I got to look at the fusion torch, as described previously for both interface and space propulsion, again.

And it turns out there’s a big issue here.

Energy

Four tons of water-fusion torch, as defined previously, have a thrust of 120t (TWR of 20), and an ISP of 15,000s. Let’s compute some statistics for that:

The energy output is simply the ISP multiplied by the total thrust. Including unit conversion (thrust is expressed in t at Earth gravity, i.e. 10,000N, ISP is in seconds), this gives an energy output of 90GW.

Oh, wait.

For reference, the engine exhausts less than a ton per minute, for a final energy density of 5.5GW/kg of propellant. This, I should add, heats the fuel to about 6 million Kelvin, roughly the temperature of the sun’s corona.

Remember the shuttle from the interface post? It turns out that this lands - possibly VTOL-stlye - on twin columns of fire hot enough to not just melt things like tungsten, but to boil it. It might not be enough to initiate fusion in the air molecules, though, so there’s that.

Clearly, that’s not going to work.

Alternatives

Now, let’s think back to the requirements that drove me to pick the fusion torch. First of all, efficient space propulsion. For this, high ISP clearly is important. Second, relatively easy takeoff from Earth-like planets. For that one, sufficiently high thrust and sufficiently ISP is necessary, but we can compensate for the latter by using RAM rockets.

But we’ve already seen a possible alternative which lets us keep an awesome space ISP: We don’t need high thrust in space, and we don’t need that high of an ISP during liftoff.

So, let’s introduce gears. Gears, or different operation modi, are something that actually exists. The VASIMIR drive, for example, has a high-efficiency and a high-thrust (well, relatively) mode, for example. In the simplest case, this can be done by increasing the mass flow into the engine, trading off efficiency for effectiveness.

In our case, let’s say the high-efficiency gear does exist, and does have an ISP of 15,000s, but does has a thrust of only a tenth of what we originally envisioned: TWR of 2. That’s less than most jet engines, and far below rocket engines.

But, if you really need the power, you can pump in additional water propellant: Reducing the ISP to 1000s, you get our original TWR of 20.

High-Efficiency Gear

If we plug in the high-efficiency numbers, we’re getting an energy output of only 9 GW for the 4-ton drive (“only” might have some different meaning here…). It still has a temperature of about 6 million Kelvin.

Does that mean the most effective weapon is the exhaust? Depends on how focused this exhaust is. And this is something that can be computed. If we assume it behaves like a gas, the root-mean-square speed of molecules is

where is said root-mean-square speed, R is a constant (about ), T is the temperature in Kelvin, and is the molar mass of the gas. Now, the root-mean-square speed isn’t quite the expansion speed - but I’m going to use it as that because I’m only interested in orders of magnitude.

For a temperature of 6 million Kelvin and water, this turns out to be about 90 km/s, which is about 60% of the exhaust speed of 150km/s. Assuming for a moment that that speed is only used for expansion (i.e. moving away from the thrust axis), and the exhaust speed is only used for moving the exhaust along the thrust axis, we get a cone with an opening angle of about 60°.

We can now compute the temperature at a certain distance (assuming the expansion speed stays constant), simply by reducing energy relative to the area of the cone’s base at a certain distance.

As an example: Starting at an assumed nozzle diameter of a metre and a temperature of 6 million Kelvin, the effective temperature is going to be about 375,000 Kelvin at five metres distance, and is reduced to 400 Kelvin at about two hundred metres. That’s dangerous near the ground, but quite safe in space.

Now are those numbers good? Hell no. They’re probably quite wrong. But they’re (a) an acceptable approximation, and (b) I don’t want to spend the time implementing a simulation of that.

High-Thrust Gear

The high-thrust numbers (TWR of 20, ISP of 1000s) gives an exhaust energy of 6 GW per four tons, and produces an effective exhaust temperature of about 25,000K. This should be survivable out to about 50 metres.

If operating in atmosphere, and using a RAM rocket, we can also add more air to the mixture. At a mixture ratio of 10:1 (10kg of cool air for every kilo of exhaust-air), temperature should decrease to about 2500K (about double of a jet engine), which makes it possible to stand closer without melting.

Medium Gear

If a torch drive can pump a hundred and fifty times as much mass into the engine, it surely can also pump in twenty times the normal mass. Accordingly, there’s yet another possible gear: Thrust is 4x as high, but ISP is only 1/5th. Temperature is dangerous out to about 50 metres.

Modes in Practice

The “new” fusion torch can be switched between two modes: By default, it uses a small amount of water and accelerates this to high speeds using a fusion reactor. In any emergency, it can pump much more propellant into the reactor, costing efficiency but greatly increasing thrust. In-atmosphere flight uses the latter, but accelerates air instead of water.

In THS’ spaceship system, one space masses masses 4t and can either produce 12t thrust at 15000s ISP (that’s 3t of propellant per hour per space) or 120t of thrust at 1000s ISP (432t per space per hour(!)). The latter isn’t really sustainable over longer times except for atmosphere; pay 15x dV for high-thrust mode. For a medium-mode, it gives 48t of thrust and an ISP of 3000 (57.5t per hour per space).

It is also available in compact (1.5x more thrust and weight per space), RAM (1/2.4x thrust per space), or vectored thrust (enabling VTOL with sufficient thrust; 1/1.5x thrust per space).

During space combat, each km/s of dV gives you 100 Burn Points. Maximum burn points per turn are, as usual, 100x sAccel.

As an example: Take an AKV with a total of 2.5 compact spaces of fusion torch engine. It has a dry mass of 136t, with 45t of reaction mass. This gives 45t of thrust in high-efficiency, 180t in medium, and 450t in high-thrust mode, for a minimum acceleration of 0.25g, 1g, and 2.5g respectively. dV is 19.25km/s, 3.25km/s, or 1.28km/s.

During combat, this translates to 1925 burn points. In high-efficiency mode, it can spend at most 25 points per turn (+6). In high-thrust mode, effective available burn points are 128, and it could spend 250 per turn (limited to 100 and therefore +10). In medium-thrust mode, it has 325 burn points, and can spend at most 100 per turn (+10).

Summary

While I’m not quite happy with the relative energy density (2.25GW/ton, compared to fusion reactors at 2MW/ton) and that you don’t need a drive core, it’s fine for now. I’d also note that this does not quite fall into Rick Robinson’s definition of a torch drive, as cited on Atomic Rockets: “a spacecraft with a Torch Drive and a specific power of one megawatt per kilogram or larger.”