Ion Drives Are Doing My Head In

[NeoMorph]

So I have been trying all day to get my Ion drive powered sat to Duna and for the love of god I JUST CAN'T DO IT. I even tried letting MechJeb do it and even it ended up flying the craft to Jool lol.

Either my mods installs have messed up the nodes or there is something well wrong with the sphere of influence changes. It doesn't happen with normal rockets... it just seems to get it wrong with Ion engines as if it can't handle the low thrust and adds 1 + 1 and comes out with 5.346343634346346364643563464366234.

I think tomorrow I am going to strip out all my mods and do a vanilla launch to see if it's a bug. Because Mechjeb showed the Delta V needed as climbing and climbing and climbing... which is what I would expect for the low power but I had FIVE Ion engines running and the I used to be able to fly Ion ships easy.

Has anyone else noticed anything weird regarding Ion drives? I tried doing a search on here but the search engine sucks and picks up things such as "description" and so nothing about Ion engines seems to show up.

[Arsonik]

Did you have the probe on upside down? Was it plugged in? You made sure you pressed the power button before calling tech support, right?

[cantab]

Compare the length of the burn to the period of the orbit you are in. If the burn is too long compared to the orbit you're going to have all sorts of problems. The solution is to first get into a higher parking orbit, either from the lifter or on the ions themselves.

Also check you have good solar power. If you're heading to Duna your ejection burn would normally be on the night side of Kerbin, a problem for ions. A higher parking orbit will help, because you're in Kerbin's shadow for less time. Alternatively launch into a retrograde orbit, then you can make your ejection burn on the day side.

[xebx]

I've been using a lot of Ion engines recently, including tweakscaled ones, and didn't encounter any major issue.

I had minor issues, like deltaV reading by KER, maybe because staging was a little complicated and i used a lot of octagonal strut to connect engines and tanks.

All engines were running at full power and ship had enough acceleration ? (>1m/s2 and a first kick using a LFO stage is welcome if your ship is heavyloaded during ejection burn from kerbin soi)

+1 for a higher parking orbit, and you can use fuelcell array (1 per 2 engines) when the ship is in the shadow

Did you retry after quit/reload ?

Did you retry by manually following the waypoint created by MJ (not execute) ?

Did you install recently another mod ?

Edited October 3, 2015 by xebx

[nalfz]

adds 1 + 1 and comes out with 5.346343634346346364643563464366234.

That's what you get for using one of these

This might help you, though it does lower the thrust to a more reasonable level for ion thrusters. You could probably just find the patch for that in the mod and edit it back to the stock thrust if you want (don't delete it obviously, because it adds a module to the engine that it needs to work with the plugin)

The thing with ion engines is that though they look like they would be far more efficient from the ISP figures, the low thrust means that the assumption with manoeuvres being an infinitesimal point in time breaks down even more, and you have to take totally different trajectories, as well as the fact that conventional chemical engines, having loads of thrust, can exhaust more propellant deeper in a gravity well and take more of an advantage of the oberth effect, so basically unless you really, really need to use an ion drive, it probably isn't going to do what you want

[cantab]

The thing with ion engines is that though they look like they would be far more efficient from the ISP figures, the low thrust means that the assumption with manoeuvres being an infinitesimal point in time breaks down even more, and you have to take totally different trajectories, as well as the fact that conventional chemical engines, having loads of thrust, can exhaust more propellant deeper in a gravity well and take more of an advantage of the oberth effect, so basically unless you really, really need to use an ion drive, it probably isn't going to do what you want

When you've got 10, 20, or more km/s in the tanks you don't tend to give two hoots about the Oberth effect Different trajectories is a fair concern in reality, but less so with stock KSP ions.

That said, Kerbin-Duna isn't really in ion engine territory delta-V wise, unless the OP wants to do a lot of orbital manoeuvring at Duna for whatever reason. Kerbin>Moho>Gilly>Bop>Kerbin, that's the kind of mission profile ions shine at (just watch for the weak solar power out by Jool).

Edited October 3, 2015 by cantab

[xebx]

@nalfz : I didn't have to use totally different trajectories during travel with my different ion ship, the ship followed the path defined by MJ, no problem. I needed small adjustments like with LFO engines.

Using fuelcells if needed you can start an ejection where you want and fly a few km above the ground of a celestial body without atmosphere (for example Eeloo or Moho), this way you get oberth effect.

In the end i had the very high deltaV and velocities that were promised, thanks to their high ISP.

This said, Ion are clearly for high deltaV mission, i don't use them for basic mission, they don't like high-med gravity like Eve (mostly) or Kerbin and you need more planning of energy requirements and burn timings.

They are highly efficient engines for the experienced navigator.

LFO are very good complementary, when you need high thrust (landing or liftoff), and can do most of the classic missions alone. (including a Jool-5 challenge)

Note : more recent and bigger Ion/plasma engines have good thrust and high ISP (and high electrical requirements)

Edited October 3, 2015 by xebx

[NeoMorph]

The reason I wanted the large amount of Delta-V is that I want to do a skip mission... go to Duna, scan, do Ike, and then see how it goes with skipping to another planet using a slingshot round Duna itself.

Like I said, I've not had problems with Ions before (used them a lot in .20) but now there are weird things happening. It's like the node is plotting for a normal engine and the estimated burn time is WAY off. I mean like hours off.

I really think it's a mod problem. I have way too many mods installed and need to prune it down a bit.

Edit: Hmmmm.... The Kerbin hours in a day changed between 0.20 and 1.0.4... oh and the Ions themselves changed IIRC. Perhaps I'm just planning it wrong now. It's just that the burn starts fine and the time is clicking down normally but then towards the middle of the burn it's like the ship is turning away from where it is supposed to be going and the time to complete the burn starts ticking up again. For example it gave me an 18 minute burn... I locked it on the node, started the burn and set the Kerbal Alarm Clock for 16 minutes. I check at about 8 minutes and it's around half done... Come back 8 mins later and the time to complete is now 20 minutes lol. Something sure is going bonkers. It's still locked onto the node point, still at full thrust... still plenty of energy.

Edit2: So I used a standard kick booster and everything worked fine. Went back to the Ion and it went GAGAGAGAGA and got lost half way through the burn again. The normal rocket booster was on just over a 1000 dV. For the Ion it wants 5000+ for some reason.

Edited October 3, 2015 by NeoMorph

[richfiles]

18 minute burn.... there's your problem!

You have to split the burn. NASA does this with their Dawn probe too. What's happening, is you are getting too far away from your Periapse and still burning. Eventually, you'll swing around the other side of the planet and be burning the other direction (on the wrong side of Kerbin, that maneuver node is making you lose velocity).

You MUST split your ion drive burns. If it's an 18 minute burn, start burning 3 minutes befor periapse, and shut down 3 minutes after. Orbit once, with engines off, and repeat it a total of 3 times. The last time should have you close to your desired escape trajectory. You may find that the second orbit leaves you waiting days to come back around to periapse. You might even get close to escape trajectory. Try to save the actual escape for the last burn, and split it in half at periapse, half before, and half after. If you are taking more than 1/3 of an orbit for thrusting

- - - Updated - - -

Then you either need to split it more, or plan more thrusters.

[xebx]

18 minute burn.... there's your problem!

You have to split the burn. NASA does this with their Dawn probe too. What's happening, is you are getting too far away from your Periapse and still burning. Eventually, you'll swing around the other side of the planet and be burning the other direction (on the wrong side of Kerbin, that maneuver node is making you lose velocity).

You MUST split your ion drive burns. If it's an 18 minute burn, start burning 3 minutes befor periapse, and shut down 3 minutes after. Orbit once, with engines off, and repeat it a total of 3 times. The last time should have you close to your desired escape trajectory. You may find that the second orbit leaves you waiting days to come back around to periapse. You might even get close to escape trajectory. Try to save the actual escape for the last burn, and split it in half at periapse, half before, and half after. If you are taking more than 1/3 of an orbit for thrusting

- - - Updated - - -

Then you either need to split it more, or plan more thrusters.

Start from a higher parking orbit : slower orbit velocity will give your ship the time to do 50% of the burn before Node and 50% after, add engines to reduce total burning time (it'll reduce deltaV but it'll remain high), and use LFO when around a high/med gravity planet (eve/kerbin) or split the ion drives burns.

And when using MJ execute mode, abort the last 5%, so you can adjust the final encounter without losing deltaV.

Edited October 5, 2015 by xebx

[magnemoe]

When you've got 10, 20, or more km/s in the tanks you don't tend to give two hoots about the Oberth effect Different trajectories is a fair concern in reality, but less so with stock KSP ions.

That said, Kerbin-Duna isn't really in ion engine territory delta-V wise, unless the OP wants to do a lot of orbital manoeuvring at Duna for whatever reason. Kerbin>Moho>Gilly>Bop>Kerbin, that's the kind of mission profile ions shine at (just watch for the weak solar power out by Jool).

Ions are nice for farming science contracts, has never done around Duna, this was done by probe or ship opperating out of Ike base.

[NeoMorph]

Well I finally found out what the problem was. It was a bug with the auto-start of my nuclear reactor.

Here is the sequence of events...

1. My interplanetary probe was kicked out of Kerbin orbit by a booster I had re-used that had a few tiny solar panesl that I used to keep the boosters computer running so that I could de-orbit it after use (I only burn 95% of the tank's fuel so I have enough to de-orbit it).

2. I use Smart Parts which let me trigger action groups at specific altitudes (amongst other things). One of the action groups that got triggered when leaving the planet was the Nuke Reactor startup. At that time the booster that was kicking the main package out of Kerbin was still connected. Unfortunately, some bug meant the Smart Part had turned into a Dumb Part because it didn't start up the Nuke. It had been working fine previously when testing around the Mun so I thought it was fine.

3. The only way to check that the pile had started was to see if the charge level was sticking at max even when the package was being rotated by the gyros. And here was the reason I got mislead... The solar panels... the damned tiny solar panels was supplying the charge to top up the electrics. So when I THOUGHT the Nuke had been turned on, it hadn't.

4. So then I detached the booster and turned on the Ion engines and they all ran fine... but they were on battery power. Oops. So the batteries drained half way through the burn and then... the darn delta-V needed kept increasing.

I always have problems with the electrics on my rockets... It's never running out of fuel or not having enough deflection from the aerofoil parts. I just have blind spots with running out of electric heh. That's why I shoved a nuke onboard in the first place. The fix in this case was to remove the smart part and put a new one on and rebuild the links. Worked perfect after that. My Scansat satellite is now around Duna and scanning away like normal.

Morale of the story? Never assume parts work because they did earlier.

Oh and the MAIN reason for the probe build?... I want to see if I can use one sat to do the whole of the Jove system. It uses Aerobraking to save dV on entry then has around 10k dV to tootle around the system if needed. Multiple sensors need different heights so it needs to do a lot of altitude changes. Whether the plan will work or not is to be seen.

Edited October 4, 2015 by NeoMorph