Ion engines - can we really use them?

[Jean Deaux]

I've been a long time fan of the ion engine so I've been thrilled at the chance to use/play with them. Yah, they're slow and use little fuel, but they do offer an option for the long distance drive. But, I'm finding it a real challenge to use them in the game simply becaues of the TIME. I was attempting a short little run from Kerbal to the moon to act as a tug there and I'm supposed to throttle up and maintain a heading for 3+ hours? Really? I've got to sit at the keyboard and do nothing but watch my heading for 3 hours? No way I'm sitting at the keyboard doing that for that period of time. What about those LONG space flights, 4 days of holding? Surely there is a game sensible way to make use of these engines that's practicle to we the gamer behind the keyboard? Am I missing a trick?

[hoojiwana]

You're not missing anything. Ion engines are only really effective on small probes because of their very low thrust, although they will still take a very long time to do any lengthy maneuvers. You can use physics warp to cut down the time, but even then you're still sitting there doing nothing for ages.

One ion engine on one XenonGas tank in stock KSP on a probe with the solar panels to run the Ion has roughly ~4k dV, and it'll take 96 minutes to deplete all of it. For comparison, it takes ~4.5k

dV to get into orbit, and that takes only a few minutes, whilst fighting against drag and gravity to do it. I made some alternative electric engines that can complete a burn in 6 minutes, although they will eat through your Xenon four times faster.

[ThatBum]

You could fractionate your orbital maneuvers by doing periapsis "kicks." It maximizes efficiency due to the Oberth effect.

Illustrations:

(Images taken from this thread on nuclear engines.)

[Jean Deaux]

Yah, doing the staged elliptical burn is something you pretty much have to do simply because you don't have the time/thrust to do it all in one maneuver. Still, I don't see any practical way of being able to use these engines in game as-is simply because of the amount of user time required. Some type of automation is really needed here, else I think these engines are just going to collect dust. I hope this engines sees some kind of love in future releases.

[Kosmo-not]

I recently used the ion drive to complete a tour of the solar system. I made a nice video of the sights I saw: http://forum.kerbalspaceprogram.com/showthread.php/26990-The-Journey.

That has been the only practical application of the ion drive I have used.

[Francesco]

wow!

now it would be even more impressive to see the raw footage of that mission

[SpaceK531]

The ion engines are not the engines you want to use for a tug, unless the payload is ~1 ton. If you want low-powered space tug engines, use the LV-909

[ThatBum]

I tend to put them on probes in the stage that they're already in orbit with. They're good for orbital adjustments, like when trying to get a precise orbit for mapping or something.

[SciMan]

Good point, and you can kinda use them as a super-efficient translational RCS thruster by setting up some action groups to activate only one direction at a time (ie. the "activate prograde ion engines" group would also turn off the up, down, left, right, and retrograde ion engines) Sure it's heavy on the action groups, and it would take some time. However, I think it would be a good solution for docking truly massive modules (50-100 ton per module at start of docking) to each other without needing 10 tons of the module for just RCS thrusters+tanks!

[M5000]

Good point, and you can kinda use them as a super-efficient translational RCS thruster by setting up some action groups to activate only one direction at a time (ie. the "activate prograde ion engines" group would also turn off the up, down, left, right, and retrograde ion engines) Sure it's heavy on the action groups, and it would take some time. However, I think it would be a good solution for docking truly massive modules (50-100 ton per module at start of docking) to each other without needing 10 tons of the module for just RCS thrusters+tanks!

As I love the idea of non-RCS control systems, I don't think the benefit gained would be all that great. The individual ion engines combined with the way you have to mount them, Xenon tanks, and solar/RTG power systems to power them would be probably very close to the mass of the RCS's small near-massless thrusters, and the single-part tanks. Also, part count. But it's an interesting concept.

[numerobis]

MechJeb is very handy for long burns, but yes, ion engines have long burn times. 3h I've never seen, but I've burned tens of minutes.

[CalculusWarrior]

4x physical acceleration is very helpful, as most of the time you're dealing with a small probe that is structurally sound. Though ion engines are extraordinarily useful IRL, as demonstrated by this current NASA mission. Simply accelerate to 4x, then go read a book or something while you're waiting for the burn to finish.

[Tristavius]

Like many here my only main use I've found are as a final-stage engine for orbital adjustments on probes under say 2 tons.

I did also once mount one on the back of a large manned rover and it was actually kind of useful... it would slowly (about .1m/s per 2-3 secs) increase my speed allowing a much higher top speed than previously.

Finally I tried making a manned ion-hovercraft for Minmus a few days back. It ended up looking like the Nebakanezer from the Matrix. It actually just about worked, it seemed almost stable and had a very balanced TWR of about 1.02, just enough to get off the ground. The only problem was electricity - there was no WAY I was going to be able to power them all for more than about 5-10 seconds, so until I find a solution for that, no Ion hovercraft

[The Ideal Gas Lawyer]

Like many here my only main use I've found are as a final-stage engine for orbital adjustments on probes under say 2 tons.

I did also once mount one on the back of a large manned rover and it was actually kind of useful... it would slowly (about .1m/s per 2-3 secs) increase my speed allowing a much higher top speed than previously.

Finally I tried making a manned ion-hovercraft for Minmus a few days back. It ended up looking like the Nebakanezer from the Matrix. It actually just about worked, it seemed almost stable and had a very balanced TWR of about 1.02, just enough to get off the ground. The only problem was electricity - there was no WAY I was going to be able to power them all for more than about 5-10 seconds, so until I find a solution for that, no Ion hovercraft

Have you tried the H.O.M.E. Nuclear reactor? There are also mods with XXL solar panels and liquid fuel powered generators

[Carl Sagan The S Stallion]

I would recommend that you put a probe into prbit using your favorite launch vehicle (whether it is a shuttle or rocket). Then, use nuclear engines to get you as far as possible, then switch to Ion power. Use the Ion engine for tweaking your orbit on small probes.

[MarvinKitFox]

Like many here my only main use I've found are as a final-stage engine for orbital adjustments on probes under say 2 tons.

(

All I use ion drives for are silly stunt rockets.

Either flying to Eeloo return trip using under 2 tons on the launchpad,

or for squeezing the last oomph out of a Kerbol-slingshot behemoth, trying to reach that elusive 100km/s interstellar speed on solar system escape velocity. (my best so far is 86km/s when passing Jool)

Wot, you mean for practical stuff?

Nuke is more than good enough, and just about the weakest I'm willing to work with.

[MarvinKitFox]

Finally I tried making a manned ion-hovercraft for Minmus a few days back. It ended up looking like the Nebakanezer from the Matrix. It actually just about worked, it seemed almost stable and had a very balanced TWR of about 1.02, just enough to get off the ground. The only problem was electricity - there was no WAY I was going to be able to power them all for more than about 5-10 seconds, so until I find a solution for that, no Ion hovercraft

A grid of cubic octagonals, with multi-layeres of single solar panels.

As the smallest solar panel is massless and physics exempt, it is also pants at noticing when it is shaded by one of the 43 other solar panels you have piled on top of it.

As long as you have any sun at all, you can have any arbitrary amount of solar power.

You must just be willing to use clipping (not even the alt-f12 type, normal VAB clipping is strong enough). The only real limit is your CPU, as in how many parts can your machine handle.

My one-man microJalopy Eeloo vehicle (2.6 tons on launchpad) was ion powered, and generated enough solar AT EELOO to saturate one Ion engine.

http://forum.kerbalspaceprogram.com/threads/77034-See-Eeloo-on-5-a-day-%28image-heavy%29

If I recall correctly, I had 22 solar panels on each of 8 'wings', total 176.

Combined weight of the struts, batteries and solar panels was, of course, under 1 microgram (measured to the nearest µg)

[juanml82]

I made this light utility rover/lander for use within Minmus SOI

It's so light that the trust on Minmus is in the 1.7 range, it can rest vertical on four LT-5 landing legs or rover around on the landing gear, provided I give it a kickstart with ions or monopropellant. It's useful if you want to send kerbins around and electricity isn't a problem. And it's sort of fun to land on a silent and slow engine.

But it is hard to make use of them. To make matters worse, they are, understandably, expensive in career mode. Maybe as a transfer engine for SSTO planes but the question remains about why are you taking an SSTO anywhere on ions.

IMHO, the game engine has to change for ions to be useful. They ought to work as real ion engines, ie, they are fired for months, which means they need to work in the background and the whole maneuvering node becomes useless to plan their trajectories.

[Rusty6899]

Good point, and you can kinda use them as a super-efficient translational RCS thruster by setting up some action groups to activate only one direction at a time (ie. the "activate prograde ion engines" group would also turn off the up, down, left, right, and retrograde ion engines) Sure it's heavy on the action groups, and it would take some time. However, I think it would be a good solution for docking truly massive modules (50-100 ton per module at start of docking) to each other without needing 10 tons of the module for just RCS thrusters+tanks!

An interesting idea, but it would be vastly more expensive than RCS + Monopropellant costs. I'd say you'd be far better with regular RCS.

[Dave Kerbin]

As I love the idea of non-RCS control systems, I don't think the benefit gained would be all that great. The individual ion engines combined with the way you have to mount them, Xenon tanks, and solar/RTG power systems to power them would be probably very close to the mass of the RCS's small near-massless thrusters, and the single-part tanks. Also, part count. But it's an interesting concept.

Given that the game has the Vernor Engine now there isn't a reason we couldn't have the same thing for xenon thrust too, with a radially attached ion engine that is linked to RCS controls like the vernor. It can be lighter and require less power, but produce much less thrust (0.1 kN) at a lower ISP (2000s), making it more useful for satellites and small probes.

[RocketPilot573]

I never find myself using them... I can usually get enough Delta-V from the liquid fuel rockets.

[juanml82]

Given that the game has the Vernor Engine now there isn't a reason we couldn't have the same thing for xenon thrust too, with a radially attached ion engine that is linked to RCS controls like the vernor. It can be lighter and require less power, but produce much less thrust (0.1 kN) at a lower ISP (2000s), making it more useful for satellites and small probes.

That's interesting. It should be a lot more expensive than monopropellant, though. Maybe a mod can do it.

[cantab]

Bit of an old thread here. In particular the previous discussion predated .23.5, which significantly buffed the ion engine to reduce the burn times.

Anyway, they're good for low-mass high delta-V applications, where the LV-N is just too heavy. They're also good for extreme delta-V applications, though the part count will really spike if the ship isn't small. The LV-N's might be good for 5, even 10 km/s delta-V in a stage, but the ion can go north of 20.

[Jean Deaux]

Yah, this is pretty old, but the basic gripe I had about using them was having to sit behind the keyboard for so long to use those engines. To provide some closure to this topic, I soon after discovered MechJeb to serve as an autopilot so that I wasn't tied to the computer for such lengthy burns.

[davidpsummers]

I've been using them. I stacked four of them and used them with a one-manned capsule. The way I deal with time is a set up a "burn" and see how long it takes, then I just use Kerbal Alarm clock to pause the game 5 min before the burn is over and go and take a shower, eat dinner, go to bed, etc.

But it does seem that, late in the tech tree, we get some cool engines that can only be used for small craft. But the thing is that at the point one is good at sending small things almost anywhere and what one needs, for example, send a manned mission to Jool.