Are shielded docking port at the front a bad idea since 1.2.1?

[juanml82]

I have this spaceplane built in the previous versions, which doubles as a reasonable lander

And it used to make it to orbit with plenty of dV to spare. I've just loaded it in 1.2.1 though and the rapiers can't keep up the thrust when they are breaking the sound barrier: They'll hold up to about match 1, but between match 1 and match 1.1 they will consistently keep loosing thrust, even if I level out or even descend slowly, which keeps me stuck below match 1.2. I've tried placing precoolers in front of the engines, but they don't change performance.

I've replaced the front of the spaceplane with the MK2 (not inline) cockpit and while there is a thrust loss, it isn't nearly as pronounced and I can more easily break the sound barrier, so I guess the issue is related to the changes in drag. Am I right? (And it's a pity, because that's a nice position to put the docking port).

I also installed Interstellar (and removed the rapiers MM patch as it caused them to overheat) but I don't think Interstellar should be the curprit

[AeroGav]

21 minutes ago, juanml82 said:

I have this spaceplane built in the previous versions, which doubles as a reasonable lander

: They'll hold up to about match 1, but between match 1 and match 1.1 they will consistently keep loosing thrust, even if I level out or even descend slowly, which keeps me stuck below match 1.2. I've tried placing precoolers in front of the engines, but they don't change performance.

I've replaced the front of the spaceplane with the MK2 (not inline) cockpit and while there is a thrust loss, it isn't nearly as pronounced and I can more easily break the sound barrier, so I guess the issue is related to the changes in drag. Am I right? (And it's a pity, because that's a nice position to put the docking port).

Terminology my dear boy, tsk tsk.

Your engines aren't loosing thrust,  but they aren't gaining it as fast as drag is rising, so you cannot get past the transonic region.  

Yes,  I've discovered this port is really draggy as well.

Use an Inline Clamp-o-tron for docking.    As a heat resistant nose cap,  use the NCS adapter and Fly by wire hub instead, 2400k rating, they are also long and pointy which takes your cockpit further away from the hot stuff.

Another thing you can try, if a plane is having trouble with the sound barrier, is not try and break it right down on the deck.   Try climb up to thinner air then use a burst from the nukes and set prograde to try get sonic.  I'd say go for 10km but your ship looks really heavy and only has small wings,  so better lower down.     If your subsonic climb is forcing you to raise the nose 10 degrees off the prograde in order to get enough lift to fly, then that is a sign you've gone too high subsonic,  the large-off prograde angle is getting to be as bad as the sound barrier.

[juanml82]

16 minutes ago, AeroGav said:

Terminology my dear boy, tsk tsk.

Your engines aren't loosing thrust,  but they aren't gaining it as fast as drag is rising, so you cannot get past the transonic region.  

Yes,  I've discovered this port is really draggy as well.

Use an Inline Clamp-o-tron for docking.    As a heat resistant nose cap,  use the NCS adapter and Fly by wire hub instead, 2400k rating, they are also long and pointy which takes your cockpit further away from the hot stuff.

Another thing you can try, if a plane is having trouble with the sound barrier, is not try and break it right down on the deck.   Try climb up to thinner air then use a burst from the nukes and set prograde to try get sonic.  I'd say go for 10km but your ship looks really heavy and only has small wings,  so better lower down.     If your subsonic climb is forcing you to raise the nose 10 degrees off the prograde in order to get enough lift to fly, then that is a sign you've gone too high subsonic,  the large-off prograde angle is getting to be as bad as the sound barrier.

Actually, if I look at the right click menu, yes, they are loosing thrust.

Will try the NCS and Fly by the Wire

About the wings, since they add drag as well, shouldn't I be trying to use as little wings as I can get by with? Is there a rule of a thumb or way to check how much wing a specific plane needs to get out of the atmosphere ASAP?

[MiniMatt]

Below is a neat thread on drag experiments in 1.2:

The suggestion from that thread (particularly the appendix section) is that the shielded docking port is better than leaving the node completely bare, but considerably more draggy than an aero nose cone or shock intake.

Further, though this potentially runs contrary to some stated dev clarifications, the experiments listed suggest that detached shockwave effects and resultant radial occlusion are not modeled, at least as they pertain to transonic drag.

All that said, the shielded docking port remains a useful part, and likely viable if you can engineer around it's shortcomings.

Edit:
Oh, and a "trick" with the wings which seems to remain valid in 1.2 is to dial in an angle of attack in their placement - ie. rather than pointing the whole plane 10 degrees above prograde to generate lift, rotate your wings (say one shift-click on the rotate widget) - this allows the plane to point prograde whilst still generating lift through angle of attack on the wings. This results in less drag from the body, but more drag from the wings, but the net result is generally less overall drag.

 

Edited November 6, 2016 by MiniMatt

[AeroGav]

33 minutes ago, juanml82 said:

 

About the wings, since they add drag as well, shouldn't I be trying to use as little wings as I can get by with? Is there a rule of a thumb or way to check how much wing a specific plane needs to get out of the atmosphere ASAP?

Parasitic drag just goes up the faster you go and the more stuff you have sticking out into the breeze.  So yes, for lowest parasitic drag, smallest possible wings.  In fact no wings at all would be best.

Induced drag (lift induced drag) comes from flying with the nose at an angle to your direction of travel.   The larger the wings the lower your induced drag, because your AoA wont need to be as high to get lift.  

So, let's assume for a moment speed and altitude are beyond your control, but you have a choice about how much wing to fit.  For lowest drag you want there to be enough wing for them to give you enough lift at about 5 degrees off the prograde, since this is the minimum drag point in KSP at the sort of speeds we're interested in.  

In practice speed and altitude are under your control, so you can compensate for the characteristics of your craft.   If you have something with a lot of wing ,  but also a lot of parasite drag (possibly , from the wings themselves),  you can climb more steeply to thinner air, where parasite drag is not so bad. Your larger wings will still get lift at the lower speeds and thinner air.   Or if you have tiny wings and start getting bad induced drag penalties, try flattening out your angle of climb,  so you stay in thicker air.      In both cases, this boils down to nothing more than keeping the nose about 5 degrees off the prograde for most of the flight,  and your trajectory will adjust itself to whatever suits your airframe best.

There are limits to this however.  If you have wings that are too big, flying at your best angle of attack will mean climbing into thinner air where engine power is less, before any decent ram air effect from supersonic flight kicks in to compensate.       And if your wings are too small you might gain altitude too slowly when flying at the most efficient angle, and have an overheating problem.

 

[AeroGav]

51 minutes ago, MiniMatt said:

 

Edit:
Oh, and a "trick" with the wings which seems to remain valid in 1.2 is to dial in an angle of attack in their placement - ie. rather than pointing the whole plane 10 degrees above prograde to generate lift, rotate your wings (say one shift-click on the rotate widget) - this allows the plane to point prograde whilst still generating lift through angle of attack on the wings. This results in less drag from the body, but more drag from the wings, but the net result is generally less overall drag.

 

I find this works best on aircraft with high wing loading - like the OP's.    The drawback is that you can't reduce the angle of the wings and body simultaneously to zero/very low values as a temporary measure to get through the high drag transonic region.  That is because when your body is on prograde, the wings will still be at +5 or +10,   if you angle so your wings are at 0 the body will now be at -5 / -10 .   But on a plane with small wings relative to its weight you probably dont have to shallow dive through the sound barrier anyway.

I'd like to be able to angle the wings on this beast actually, because i think it would help when landing on Duna

If I pitch up more i can get slower landing speeds, but i'll bust the tail off on touchdown.   Angled wings would be great for this offworld stuff.   But it's got too much wing area for this to work properly on kerbin.   It'd be trying to climb above 10km every time you go over 200m/s,   you'd not get through the sound barrier.

[Foxster]

One thing (amongst a few) that's not helping that craft is the very draggy back end. All the bi/tri/quad adapters are very draggy and you want to avoid them, especially on an aircraft. 

An alternative is something like this...

 

[TwinKerbal]

3 hours ago, Foxster said:

One thing (amongst a few) that's not helping that craft is the very draggy back end. All the bi/tri/quad adapters are very draggy and you want to avoid them, especially on an aircraft. 

An alternative is something like this...

 

All I can think of when I see that is me accidentally decoupling the engines mid-flight.

[AeroGav]

3 hours ago, Foxster said:

One thing (amongst a few) that's not helping that craft is the very draggy back end. All the bi/tri/quad adapters are very draggy and you want to avoid them, especially on an aircraft. 

An alternative is something like this...

 

I did some testing in 1.05 with hack gravity and a wingless fuselage, autopilot set to prograde at 300m altitude and an airspeed of 250m/s.   The 1.25m rocket bi , tri and four-way adapters were all terrible.  The 2.5m to 1.25m two and four way were nearly as bad, but the triple was an exception for some reason. Also the mk2 bicoupler and mk2 to mk1 adapters were nice and slick. The space shuttle style mk3 engine mount was a horror show

 

The space shuttle engine mount has now been fixed and is low drag, provided you occupy all 3 of its 1.25m nodes and the central 2.5m node which everyone forgets about.  I just repeated the test of the 2.5m quad mount and it now has a drag of about 6, which is similar to the 4 slanted nose cones you'd need to attach those engines without it.   The 1.25m quad is still horrible.

The issue is going to be when he decouples.  If he's already out of the atmosphere by that point, fine, otherwise bear in mind he's now got an open 2.5m node exposing flat plate drag to the slipstream.   To the stock aerodynamics, having a flat plate tail is the same as having a flat plate nose.  And to be honest, real airplanes all have pointy tails as well, so it must matter.

So yeah, if the OP is bothered about drag after the decoupling event,  then mounting radially with "small hardpoint" decouplers is a much better idea. 

Also, the rear attach nodes on those rapiers (and his nukes) will be causing a lot of drag.  To avoid this, attach cones to the back of those engines and offset them into the engine so they can't be seen.

I don't think this is cheating, because jet engines without attach nodes (panther, whiplash etc) have a drag value almost as low as a cone, so I think by doing this you're just correcting for something the developer didn't intend.  

Clipping engines inside engines or fuel tanks inside fuel tanks  - now that's another thing entirely.

important - if you're going to enable these aero data values, make sure you turn off "part highlighting fx" in the graphics settings off the main menu, or you'll experience random crashes to desktop.

 

[Raideur Ng]

One very good alternative to the shielded docking port is a fairing! If you require a docking port, you can simply place that on the fairing base. Make the fairing as pointy as possible and you now have a nosecone FAR more heat resistant than any aero part, very minimal drag, and will function as a docking port. The only time you generally need extremely optimized drag is during initial ascent anyways. Though, if you jettison it, its TECHNICALLY not an SSTO anymore.

[juanml82]

5 hours ago, Foxster said:

One thing (amongst a few) that's not helping that craft is the very draggy back end. All the bi/tri/quad adapters are very draggy and you want to avoid them, especially on an aircraft. 

An alternative is something like this...

 

It's good to know, but the catch is that the mk2 to 2.5m adapter in that design also doubles as landing legs in low gravity bodies such as Minmus.

Edited November 7, 2016 by juanml82

[AeroGav]

18 hours ago, juanml82 said:

I have this spaceplane built in the previous versions, which doubles as a reasonable lander

 

One other thing (Peter Faulkner impression)

You've got over 8 tons right at the back in those Rapiers.  The centre of gravity will be much further aft with them present, than with them gone.  So to ensure stability in all conditions, you're going to have to tune the Centre of lift way aft.    After separation you'll be very nose heavy,  the elevons will be making a lot of drag, and creating a lot of downforce pushing the tail down.

Worst case, this can end up looking like the space shuttle, with the Elevons doing their best to sabotage the efforts of the wings (yellow lines are the negative lift from the elevons)

You can work around this by putting a wing segment on the module to be decoupled, that way when it goes, the CoL moves forward just as much as the CoG.  But I suspect you're not planning do any performance critical atmospheric flight after releasing the engines anyway.

Btw,  if you want a wider footprint for tail sitting landings, this might be a lower drag way of doing it than by flaring out to a 2.5m fuselage, though tbh 2.5m are nowhere near as bad as mk3

There is a long mk2 to 1.25m adapter in the middle.  Two of the rapiers decouple.  the pic on the right shows with fins added to the decoupling nacelles, to adjust the CoL

[juanml82]

I will try that last pics you've put. But as for the CoM, I'm not that concerned: the rapiers gets decoupled in near vacuum (around 40km if memory serves me right) so at that point aerodynamics aren't much of a concern.

As for the reentry, even if I loose control below 25km, the ship has parachutes and it's meant to land vertically anyway. Basically, once the rapiers detach, it's pretty much a rocket instead of a plane as far as I'm concerned. And, by detaching those 8 tons of engines, I get a lot more dV in orbit than I would if I went the full SSTO route.