Plane veers to the left. Stumped.

[foamyesque]

@jsisidore: You might want to take a look at this. I kept your cargo layout, crew count, and engine mass, and basically redid the plane with more fuel, more rapiers, and fewer nukes.

 

(I haven't flown a proper mission, Duna not being in the proper window for a transfer, and instead used alt-F12 to put me in a very elliptical orbit after a simulated ejection burn.)

I decided to change up the usual nuke positioning and instead have them vertical in cargo bays aligned on the CoM of the ship. My reasoning on this is that for two major burns -- orbit insertion and ejection to Duna -- high TWR is very useful and so you might as well do those on the RAPIERs. That means the nuclear engines don't need to have their thrust axis along the ship's low-drag axis, since they're just being used for orbit work. By pointing them straight "down" relative to takeoff, they can also be used as landing jets on Duna, supporting almost all of the plane's weight, if necessary, and allowing much slower, softer landings (as well as sharper turns). It also makes it miles easier to reduce CoM shift.

There's a Clampatron Jr on the floor of the rover's cargo bay to act as a control point when using the nuclear engines.

The RAPIERs I pulled further up the ship, again to reduce CoM shift.

The wing carries a 5 degree AoI and is oversized for the ship on Kerbin, because I wanted something that could provide significant lift on Duna. This also applies to the aerodynamic control surfaces, which are generally larger than I would use for a strictly Kerbin spaceplane.

There's a set of vernor RCS thrusters, which aren't strictly necessary but help significantly with holding prograde during orbital burns (the RAPIER thrust axis is slightly misaligned with the CoM, due to some positioning restrictions on where I can put the engines) and with control during Duna entry.

I cut the parachutes as I didn't see a practical use for them; the gear brakes are more than adequate to stop the plane.

Action groups:

1: Toggle airbreathing/rocket mode and RAPIER gimbals on/off.

2: Toggle RAPIERs on/off.

3: Open/close nuke cargo bay doors.

4. Toggle nukes on/off.

5. Open/close cargo ramp and rover cargo bay doors.

 

Dropbox link: https://www.dropbox.com/s/zhfevvas4zhdct8/DragOn 003.craft?dl=0

 

(EDIT: @AeroGav, you might be interested too)

 

Edited December 19, 2016 by foamyesque

[jsisidore]

6 hours ago, foamyesque said:

@jsisidore: You might want to take a look at this. I kept your cargo layout, crew count, and engine mass, and basically redid the plane with more fuel, more rapiers, and fewer nukes.

Hmm, now I'm confused. I don't see any nukes, where are they? All I see is 8 rapiers. Also why delta-v is only 3700m/s ? Originally I had like 12k and still it wasn't enough. Also, your twr is only 0.1 isn't that like too low?

[jsisidore]

I gave it a try and it flies great, maybe the front wings are a bit too responsive when flying prograde and trying to maneuver, but then you can switch off assist for that, so it does not really matter.

I assume you have to use rapiers in a closed cycle... or was it open cycle? I don't remember. Thing is you have not enough oxidizer later on if you reach 20km or so on liquid fuel alone.

[foamyesque]

6 hours ago, jsisidore said:

Hmm, now I'm confused. I don't see any nukes, where are they? All I see is 8 rapiers. Also why delta-v is only 3700m/s ? Originally I had like 12k and still it wasn't enough. Also, your twr is only 0.1 isn't that like too low?

The nukes are hidden in cargo bays in the nacelles. They point down, instead of back. The TWR and deltaV numbers are lies, because I was testing the balance of the nukes, so they're exclusively the first stage.

I fly the RAPIERs on open-cycle until ~22km and ~1550m/s (~5 degree nose pitch until ~9km, then nose-to-horizon or below until ~1450m/s @ 15km, then just let it climb). Once the RAPIERs cease to provide accel, I kick them to closed-cycle (action group 1) and lock prograde until I have an apoapsis in the area of 107km. Friction will drag that down to about a 100x0km orbit. Once I'm in space, I turn off the RAPIERs (action group 2), open the cargo bay containing the nukes (action group 3), turn the nukes on (action group 4) and set the control point as the clampatron Jr underneath the rover. Then I circularize on the nuclear engines and plan the ejection burn. For that ejection burn, I switch back to the RAPIERs and the cockpit control point, because the higher TWR makes it much more comfortable. Disengage the RAPIERs at around 200 units spare of oxidizer to keep a reserve for RCS and, if necessary, finish off on the nuclear engines. Use the nukes for any further manouvering in space (plane changes, trimming Duna periapsis, etc); there's lots of spare fuel for that sort of thing.

 

Edited December 19, 2016 by foamyesque

[jsisidore]

Didn't know KER is not to be trusted with.

1. Quite clever. So nukes hidden in cargo bays (i thought they were fuel tanks) are hidden from drag... You got so many rcs ports you need them for Duna?

2. At what altitude you switch to nukes when in space? It will take a while for them to circularize orbit. Why 107km?

3. In closed cycle are you dead on prograde or close to it?

4. Do you burn away from Kerbin's orbit through the Mun? I think it wastes less delta-v that way but I'm not sure.

5. When in Duna's atmosphere how do you land, with rcs alone plus nukes?

 

[foamyesque]

25 minutes ago, jsisidore said:

Didn't know KER is not to be trusted with.

My version is out of date; I believe the newest ones are correct, but the staging is set up differently than actual flight, so even those will get it wrong b/c they can't anticipate toggling back and forth or the airbreather/closed-cycle switch and so on.

I chose 107km because, coming out of the atmosphere, you want some climb so that you're out of the drag a little quicker. 105-110km as a target apoapsis does that, and the drag smooths it out to around a 100km apoapsis, with a zero (or slightly negative) periapsis. The nuclear engines can circularize that to 100x100 quite quickly, under half a minute or so of burn.

Yes, when I'm running the RAPIERs in closed cycle I'm generally locked prograde (I use stability-assist for open-cycle). If my climb rate drops under 100m/s or so I'll generally move it back to stability assist until it recovers. I don't bother with a Mun assist; it's a lot of fiddling about for a completely unnecessary savings. Much simpler to just burn for Duna directly, without other bodies (like IKE, damnit) getting in the way.

As far as landing on Duna goes, I fly a conventional plane landing, but once I'm flared for touchdown (~15 degrees nose-up) I use the nuclear engines as both brake and support. You could even land straight down, I suppose, with the assistance of the RCS units, but I had a horizontal touchdown speeds of ~50m/s. The nukes also are very useful if you're coming in on a higher-altitude landing; I was aimed at one of Duna's lowest points in that screenshot, for maximum aerodynamic benefit.

A word of caution: on Duna, you're always going faster than you think. Keep a real sharp eye your speed indicators, because reflexes from flying on Kerbin will be out-of-whack.

Edited December 19, 2016 by foamyesque

[jsisidore]

You burn directly to Duna from Kerbin's orbit, without escaping it first?

[foamyesque]

1 hour ago, jsisidore said:

You burn directly to Duna from Kerbin's orbit, without escaping it first?

 

Yes, of course. Ejection burns to anywhere should always be done in LKO (assuming you're not doing things like orbital refuelling). That gets you the maximum benefit from Oberth. Doing a Kerbin escape burn, waiting until you clear the SoI, and then finishing it into a Duna transfer orbit is much less efficient, particularly if you exited Kerbin at the wrong angle.

Edited December 20, 2016 by foamyesque