Has space plane performance changed in the recent past

[davidpsummers]

I recently used a space plane which I have used quite a number of time, but not too recently.  (I think maybe since before I last upgraded to the new version of the program).  The amount of fuel left when I hit a 75 km orbit was lower than when I had used it in the past.  Now either I've forgotten something about flying them, or something changed about it.

Has anything change that would affect space plane performance in the last version upgrade or two?

[Vanamonde]

Drag was reworked fairly recently. Might have something to do with your issue, depending on the shape of your craft. 

[AeroGav]

I joined the game when 1.04 was the live build,  in my opinion it's never been easier to make a good one.    All engines now drain fuel evenly from all tanks, so you don't get fuel imbalance/fuel starvation for using engines other than the RAPIER.  No more draggy fuel ducts.

Autostrut means no more wobbling. No more draggy struts.

It was almost impossible to heatsplode in 1.04,  1.05 got ridiculously punishing, the current build is somewhere in between.   Part placement is now very important, inline cockpits mounted a few modules back from the front do really well, pointy ones not so good.

Adapters have had their drag values balanced.    There is still an issue with mk2 fuselages and mk0 liquid fuel tanks in my opinion, and 2.5m are OP.

[davidpsummers]

The craft still makes orbit and returns just fine.  It's just that the margin to reach my station once its up there has become tight.

[Red Iron Crown]

To expand on Vanamonde's post about drag changing: Blunt objects got draggier and pointy objects got less draggy, whether this helps or hurts performance of a particular craft is very much design-dependent.

[Cunjo Carl]

To expand slightly further on Red Iron Crown's post about drag changing, the effect is only really noticeable past ~400m/s, but then it becomes really noticeable. When I checked, lifting and control surfaces appeared to be unchanged.

[davidpsummers]

its "pretty pointy".  See below.  It is probably the parachutes.  The problem is that I don't find spending hours practicing landing "fun".  I can land some, maybe I'll just use quick save to do it again if I crash on the grounds that Jeb is a better pilot (and has a proper joystick).

[AeroGav]

I suspect it's because you're using a mk2 fuselage, they are very draggy, probably always were but it's especially noticeable now.  I noticed a big drop in the performance of my mk2s at the 1.1 update.     Looks like you already have some incidence built into the wings.  I try to minimise the amount of mk2 sections, using them for cargo and kerbals only, not fuel, since better options exist.

As for landing, I always savescum a few hundred metres up because you don't want to loose a three hour mission to a freak bit of bad luck.  More wing area would probably make that land slower however.   Or are the weak mk1 to mk2 joints breaking on touchdown leading to self disassembly?

If it was my ship, I'd probably make the following changes -

• swap to mk2 inline cockpit - better heat tolerance
• cones on the back of engines for lower drag
• swap two of the rapiers for nukes
• eliminate the mk2 fuel tanks
• more wing area

 

[davidpsummers]

23 hours ago, AeroGav said:

I suspect it's because you're using a mk2 fuselage, they are very draggy, probably always were but it's especially noticeable now.  I noticed a big drop in the performance of my mk2s at the 1.1 update.     Looks like you already have some incidence built into the wings.  I try to minimise the amount of mk2 sections, using them for cargo and kerbals only, not fuel, since better options exist.

As for landing, I always savescum a few hundred metres up because you don't want to loose a three hour mission to a freak bit of bad luck.  More wing area would probably make that land slower however.   Or are the weak mk1 to mk2 joints breaking on touchdown leading to self disassembly?

If it was my ship, I'd probably make the following changes -

• swap to mk2 inline cockpit - better heat tolerance
• cones on the back of engines for lower drag
• swap two of the rapiers for nukes
• eliminate the mk2 fuel tanks
• more wing area

 

How do you put cones on the back of engines, doesn't the thrust need to come out there?

Nukes are awfully heavy, you think I'll win after the push to space?  Not sure about how well it will pick up speed on only three engines in the air, especially with the extra weight.

[AeroGav]

1 minute ago, davidpsummers said:

How do you put cones on the back of engines, doesn't the thrust need to come out there?

 

RE: Cones 

There is a collider that checks the path out the back of the nozzle for obstructions out to a certain distance.  If you attach a 1.25m cone to the back of your engine, then use the offset tool to move the cone forwards, the "obstruction collider" never meets this cone.    It sounds like a cheat but all it does it lower the drag of  engines with attachment nodes to the same as that of engines which don't have them i.e jet engines.  It's because of the way the aero model treats open nodes as flat plate drag.

video here of me "coning" my engines -

Spoiler

 

.

.

 

.

Quote

Nukes are awfully heavy, you think I'll win after the push to space?

If you want to just bring the heaviest possible load to 75km orbit then stick with chemical propulsion, but if you want to reach another celestial body then definitely go nuke.  

I've built quite a lot of SSTOs that don't use oxidizer at all,  it's not as scary as it sounds.  When i first played KSP, I used to yank the stick back when going to rocket mode, trying to climb vertically over 70km, since that's what rockets do isn't it?   On a nuke design, you just maintain your best lift:drag ratio AoA and cruise upward,  thrust > drag and lift > weight  inevitably gets you to space in the end.

Recent example on a low tech design (Panther and Nerv, no shock cones, no Big S wings or strakes) >1900dV in orbit.   As regards to the efficiency of that launch profile,  the Panther flames out at 22km and 848 m/s,  at this point Kerbal Engineer says we got 3809dV.   At the end of the video we've gained 1235 m/s velocity and another 52km altitude , for  a dV cost of 1832 m/s.

 

Quote

 Not sure about how well it will pick up speed on only three engines in the air, especially with the extra weight.

It's all about drag ! Coned engines and shortening the mk2 fuselage will help.   As will extra wing area because you can fly higher while staying near prograde.

You can see that at high speed the RAPIER produces more than eight and a half times it's static thrust .      In fact for the design you posted, i'm tempted to try a single RAPIER with two Panthers and two Nukes.      The hard bit is getting through the sound barrier.  After that the crazy thrust buildup from the RAPIER solves all problems,  including hair loss and E.D.      Climb at <240 m/s until the wings are struggling to get lift, then go into a shallow dive, light the afterburners and toggle the nukes on till > 440.   Resume climb, level off at 22km for speedrun,  then pitch to optimum lift/drag restart nukes and go for orbit.

 

[AeroGav]

@davidpsummers   OK, just finished making a 5 engine mk2 - 2 panthers, 1 rapier, 2 nukes.  It reaches Minmus easily, the drawback is it's fairly slow getting up to 22km/1400m/s - about 15 minutes.     A high thrust, pure chemical SSTO that only has enough fuel for low orbit  could probably save 10 minutes getting to that point.    

But, to accomplish the same mission , you'd either need to

1. rendezvous and dock with an interplanetary shuttle,  
2. make the SSTO carry an upper stage in its cargo bay to fly the Minmus leg.    Which means re-entering two vehicles, the SSTO and the upper stage, and bringing them both down close to KSC.

https://kerbalx.com/AeroGav/Penta-Shuttle

The other drawback, well, it's a bit ugly.  

Edited March 13, 2017 by AeroGav

[davidpsummers]

Thanks.  Something to think about.

[Spricigo]

9 hours ago, AeroGav said:

RE: Cones 

There is a collider that checks the path out the back of the nozzle for obstructions out to a certain distance.  If you attach a 1.25m cone to the back of your engine, then use the offset tool to move the cone forwards, the "obstruction collider" never meets this cone.    It sounds like a cheat but all it does it lower the drag of  engines with attachment nodes to the same as that of engines which don't have them i.e jet engines.  It's because of the way the aero model treats open nodes as flat plate drag.

Well, it may sound like a cheat, we can even consider it a cheat but in a single player game it really don't matter. Personally I'd consider the trick equivalent to a modded rapier, which is fair game for most people. (incidentally I modded the rapier to behave more like a scramjet {no thrust at subsonic speed, works at bit higher speed and no rocket mode} that I, in all my ignorance, consider balanced . I'm a shameless cheater)

Quote

  A high thrust, pure chemical SSTO that only has enough fuel for low orbit  could probably save 10 minutes getting to that point.    

But, to accomplish the same mission , you'd either need to

1. rendezvous and dock with an interplanetary shuttle,  
2. make the SSTO carry an upper stage in its cargo bay to fly the Minmus leg.    Which means re-entering two vehicles, the SSTO and the upper stage, and bringing them both down close to KSC.

Some people will even argue that its not efficient to bring a spaceplane further than LKO. But yes, sometimes is not about efficiency (at least not cost/launch efficiency)

Anyway: I don’t think its relevant to compare a chemical spaceplane that only get to orbit(orange) and a nuclear SSTMinmus (banana). If instead we compare the whole systems seem more a matter of personal choice to me. You can either:

1.Use jets and nukes, and thus taking longer to get to orbit

2.Reach the orbit with jets and chemical power, transfer to another vessel for the rest of the trip. Rendezvous and docking takes time.

3.Reach the orbit with jets and chemical power, use a upper stage for the rest of trip. This spaceplane will be bigger(heavier, more expensive) than other options.

For 2 and 3 you are also re-entering two vehicles or instead taking the extra step to circularize, rendezvous and docking in LKO.

Then we have the whole matters of how far are you going, how big is your payload , how often you are doing it and how much you enjoy doing it.

[Wanderfound]

13 hours ago, Spricigo said:

You can either:

1.Use jets and nukes, and thus taking longer to get to orbit

2.Reach the orbit with jets and chemical power, transfer to another vessel for the rest of the trip. Rendezvous and docking takes time.

3.Reach the orbit with jets and chemical power, use a upper stage for the rest of trip. This spaceplane will be bigger(heavier, more expensive) than other options.

For 2 and 3 you are also re-entering two vehicles or instead taking the extra step to circularize, rendezvous and docking in LKO

You appear to have left out one option: jets to 1,500m/s & 30,000m, just enough oxidising rocketry to quickly lift the apoapsis to 70km+, then circularise and head off to Minmus/Duna/wherever unrefuelled with a nuke on the same SSTO. You don't have to go pure LF and low-TWR to build a good long-haul spaceplane.

A RAPIER/nuke hybrid SSTO spaceplane can fairly easily be built to reach orbit with a few thousand m/s of ΔV in the tanks, enough for a Minmus return. Add in an orbital refuel, and you can do Duna or Laythe. Add an ISRU unit, and you've got an SSTA.

This sorta thingie:

You'd get a bit more range if you shaved the number of RAPIERs down to the bare minimum, but I've never seen a good reason to fly around in a truck when you can do it in a sportscar.

Edited March 14, 2017 by Wanderfound

[AeroGav]

1 hour ago, Wanderfound said:

You appear to have left out one option: jets to 1,500m/s & 30,000m, just enough oxidising rocketry to quickly lift the apoapsis to 70km+, then circularise and head off to Minmus/Duna/wherever unrefuelled with a nuke on the same SSTO. You don't have to go pure LF and low-TWR to build a good long-haul spaceplane.

A RAPIER/nuke hybrid SSTO spaceplane can fairly easily be built to reach orbit with a few thousand m/s of ΔV in the tanks, enough for a Minmus return. Add in an orbital refuel, and you can do Duna or Laythe. Add an ISRU unit, and you've got an SSTA.

This sorta thingie:

Whilst I do build LF only ships I acknowledge they are not as efficient as one that brings a little bit of oxidizer, for a 300-400m/s boost.  

BTW comparing a 2 rapier 1 nuke Minmus SSTO to a LF only  1 rapier 2 nuke version,  the 2 rapier ship will obviously get up to 1500m/s faster, but getting to 2800/2900 transfer maneuver velocity on one nuke is going to take a while.  Save time in one place, waste it somewhere else.

If you want the best of both worlds, don't be snobbish about 100% re-use and radially attach some Whiplash nacelles.  A pair under each wing gives a huge amount of extra thrust, but can be decoupled when they flame out so as not to add weight or drag.   Those engines only cost 2000 funds each.     It's perfectly ok to do that on SSTA designs too, because Laythe is so much easier to launch from you won't need as many engines as on Kerbin.

[fourfa]

Or get creative with undocking the engine nacelles in orbit, for recovery later... for instance the inline docking ports have the same drag open or closed and can be used in interesting ways

[Spricigo]

@Wanderfound actually I left many options out. Still the point its that if you need to get to orbit and then transfer to an interplanetary vehicle you can make the both the ascent and the interplanetary travel more efficient at the cost of efficiency at rendezvous/reentry while a SSTO that is also a interplanetary vessel is the other way around. The 'best option' its in certain measure a matter of taste.

[AeroGav]

1 hour ago, fourfa said:

Or get creative with undocking the engine nacelles in orbit, for recovery later... for instance the inline docking ports have the same drag open or closed and can be used in interesting ways

Oh yeah,  I forgot about that.    Hardly falls into the "time saving" category though, as does anything involving docking.   Might be good if you're trying something extreme like spaceplane to Moho.     On a similar note, i've been experimenting with fairings to cover up docking ports. Have the nerv engine and liquid fuel tanks and crew pod in the rear fuselage, attach the airframe and jets and intakes to the front.  Undock from the airplane bit in orbit.   Fairing will remove the drag penalty of the docking ports on the ascent, obviously you shed the fairing to undock and on redocking, there will be no fairing.   But on re-entry you're not so fussed about the n'th degree of drag reduction.

[fourfa]

That's a cool idea!  Another thing I've been messing around with (I have a project seaplane Laythe/Kerbin SSTO with removable wing and engine modules, to reconfigure the core airframe for vacuum landings): you can place Jr docking ports inside cargo or service bays, radially attached to a part that's node attached and shielded inside the bay, and offset them very carefully so that they docking surface is outside the bay but still shielded.  Probably can do the same with the regular size ports too, which could mate nicely with some Mk1 engine nacelles.

Tweakable Everything even allows you to build in the VAB/SPH using the open ports on the shielded versions