Best air intake for Laythe-bound SSTO?

[Aegolius13]

Hi folks,

I just completed a sandbox-mode dry run to Laythe with my six-kerbal, full-science SSTO, the Sojourner II:

The dry run went pretty well - I got out of Laythe orbit with about 1100 delta-v left in the tank, which should work for my eventual plan to dock in Laythe orbit with a fuel pod.  It would be nice if I could get back Kerbin without refueling but I have a feeling that's not going to happen, even if I improve my maneuvering a bit.  

In any event, I am always looking to optimize the plane for my career-mode run, and one area I'm unsure about is intakes. One preliminary note - due to my ruthless culling of extra jet engines to save weight, this thing can barely get out of its own way in-atmosphere.  It has to launch half-empty on Kerbin (I refuel in orbit), then has to do a slow climb and then dive at around 7k meters to get velocity into the Rapier sweet spot.  Laythe performance was a little better, but I still had to turn on my nukes for about 5 seconds on takeoff to make sure I didn't hit the ground right away. 

I don't care much about Kerbin efficiency as long as the darn thing makes it to orbit, since I will be refuling in LKO anyway. I do care about saving fuel on Laythe, though. But since so much delta-v is going to be expended in space, weight may be more important than atmosphere performance.    

As you can see, I've currently got 2 Ram Air Intakes on each of my Mk 1 nacelles.  They generated way more than enough intake air throughout the journey, and are fairly light, but I hear their drag is not the best.  Other options:

-The shock cones.  These seem to be the go-to choice in most planes I see.  I do not need the extra air at all, but I do hear they have better drag than the Ram intakes.  I did a test climb with shock cones and could see maybe a minor speed improvement, but it was overshadowed by inaccuracy in my climb profile.  The extra weight shaved off about 20 of my plane's 5,800 vacuum delta-v versus the Ram intakes, which does not seem like a big price if there is a decent efficiency improvement on Laythe.  

-The precooler, which also seems pretty popular.  I like the zero-drag factor, but between the weight of the part itself and the fact I'd have to put something else in front of my nacelles (maybe NCS adapter and tiny-size nosecone), it will likely end up heavier than the shock cones.

 It will also add bulk, being the size of a Mk LF tank but with much less fuel capacity.

-The radial adjustable ramp intakes.   These are very light, which I like, but they add a new source of drag and, again would require me to put a nose cone in front of my nacelles.  Air supply seemed adequate (maximum draw with 2 of these guys was about 50% according to KER), except for a weird thing where my Rapiers flamed out 1 second after initial engine startup on the runway.  This was remedied by starting at half throttle for the first few seconds, then maxing out once my plane was moving.    

None of the other intakes seem like good options.  The subsonic ones would flame out, and the divertless intakes seem like worse alternatives to the precooler.  

I am tempted to just go with the shock cones, but was wondering if anyone else has any thoughts on what works best for a Laythe mission.  At the end of the day I doubt there's a big difference, but half the fun of KSP is agonizing over the best part choice, right?

Any other thoughts on ways to optimize the plane are welcome as well.  I am tempted to remove the drone core, clamp-o-tron and/or RCS, but usually end up regretting those decisions.  I'm only keeping 15 monopropellant.  May also reduce the oxidizer capacity as I don't need it all to escape Kerbin, and didn't need any on Laythe.

Final note - part cost is not an issue.  

Thanks!
 

[bewing]

Are you willing to exploit a bug? It has been brought to my attention that when you attach a radial air intake to an MK2 bicoupler, the combination suddenly generates massive body lift. Very much overcoming any drag losses.

I can attest to the fact that shock cones have lower drag than the others from experiment. And the NCS adapters with the small nosecones really are a pretty good deal.

[Aegolius13]

Nah, probably wouldn't do an exploit (though just running stock aero is kind an exploit).  Interesting info, though.  

[Val]

Like @bewing suggests. A single shock cone and some NCS adapters works great.

For inspiration, here's one I made back in KSP 1.0.5.

The craft weighed ~53 t at take-off from Kerbin and had about 2400 m/s dV in LKO. Could go directly to Duna or Laythe before needing to start up the ISRU.