Concerning drag

[Shaun]

I've been making some gliders in KSP. The smallest one is 3 tonnes empty (yeah, heavy I know), a lot of lift area, and has a Juno engine with 30 fuel to get it up to altitude (I don't detach it as it only saves 282kg). I've tuned all surfaces so the aircraft will climb, level off at around 12,500m without using SAS or manual control. When the 30 fuel I have runs out, I can turn back and pass KSP with 5,000m+ to go, as my glider glides at about 30-40m/s IAS with a vertical speed of -10m/s initially, slowing down to -2.5m/s in denser air. Even with this incredible performance (for KSP), 2.5m/s sink rate is 500ft/m, and I only get a L/D ratio of between 8 to 10 (less than the Gimli Glider!). Real sailplanes get ratios between 30 to 60 and descend at about 0.5-1.0m/s. Because my glider descends at a stable speed, it has lead me to believe that KSP's drag is too high, and if I reduced the drag by a certain factor, my L/D ratio would climb and my descent rate would slow down while maintaining speed.

As well as this, I've noticed that turbofan aircraft often 'hit a wall'. I am aware that their thrust diminishes with speed and altitude, but they seem to accelerate very fast on take off, then find it difficult to creep up to speed. This seems to be why if I create an aircraft with a similar T/W ratio as an airliner, it won't get anywhere near cruising speed and will often not be able to accelerate to take-off speed. Thus, KSP aircraft tend to have crazy T/W ratios to get up to a somewhat realistic speed, but are able to take-off and climb extremely fast.

So my question is, is KSP's drag too high, or is there another underlying issue? If drag is the issue, I would be happy to develop a simple plugin to correct this. I am aware of FAR, but is rather complicated and currently not updated.

[Boris-Barboris]

19 minutes ago, Shaun said:

So my question is, is KSP's drag too high

yes

19 minutes ago, Shaun said:

I would be happy to develop a simple plugin to correct this

There is a global drag magnitude modifier somewhere in physics.cfg or smth, I used it along with some MM patches to scale up parachute drag cubes accordingly in order to tune the stock aerosoup up to my taste.

[Shaun]

I have found them! But there's a few... hmm. Drag... Angular Drag... Drag cube multiplier...

[Jas0n]

Ah this make me think of the good old days of flying through soup.

[Boris-Barboris]

1 hour ago, Shaun said:

Drag... Angular Drag... Drag cube multiplier...

dragMultiplier = 4 // Global multiplier to drag

This one is the easiest to tweak. Should suffice for your tasks.

[Shaun]

I set it to 1. Seems to be more realistic. Things can speed up now. The only issue is it's gonna completely break SSTOs and rockets. Might set it back to 8 and play with the x,y multipliers so subsonic drag is lessened, but the supersonic drag will remain the same.

[Sampak]

Related to OP somewhat, I also constantly run into a 'wall' speed wise with my various jet setups BUT they're random as hell as far as my A.D.D. self can tell. As in, if I take off and let things go, the Afterburner Jet you get (the 3rd one with wet/dry mode you get I believe) will often break neck speed up and get to 750-900 m/s easily within a minute of take off, and given T/W ratios, I can go vertical for example, and it will burn the skin off no problem.

Other times, I notice if I don't do certain things (I can't remember what it is, sorry) or do certain things, I can roll out, take off, climb at 15-25* and the jet will surpass the 'speed wall' and get up to the usual mach 2-3 (750-900 m/s), with almost no effort on my part, no shallow dives for speed as far as I recall. OTHER times, no matter what I do, no matter how many shallow dives I take for speeding up, regardless of flying at 1000m or 6000 or 1000m, the jet will not 'turbocharge' itself with the increased speeds it gets to, usually 300-350m/s sometimes less, and will simply refuse to accelerate as I would expect, short of nose diving into the ground to get beyond 350-400 m/s. And even if it does say, get to 400 m/s, in those 'off' cases I'm talking about, the same jet that went supersonic and beyond in an earlier flight, this time might quickly lose thrust/speed if I nose up even a tad. It's like some sort of vicious cycle I can't break.

Any ideas what I've run into? Must be a common issue for newbs I take it. FYI my T/W ratios surpass 1.0 and as I said I can go vertical and get to desired speeds, yet same plane, traditional rollout and I can't gain steam.

[natsirt721]

Jet aircraft performance is highly dependent on (primarily) velocity and air density in complex ways.  Flying at a variety of altitudes and velocities you're probably experiencing this and can't make sense of it.  Thrust from the jet engine, and lift and drag on the airframe are three of the four forces governing vehicle dynamics (weight is the other) and each is affected by the freestream conditions (temperature, pressure, velocity, and derivatively density and Mach number). In turn, these forces govern acceleration and velocity which affects how the freestream conditions change.  Under some conditions you can be unable to change your position and velocity like you would expect; e.g. shallows dives don't change your conditions enough to accelerate rapidly like you would expect.  

KSP doesn't exactly model real world interactions (modeling subsonic compressible aero is tricky enough, let along super- and hypersonic aero), but it's close enough to be passable and the interactions are still very complex.  Short of learning the governing equations and how KSP modifies them I'd say the best thing to do would be trial and error - what conditions are required to get to to other conditions?  If you are so inclined, there is a "show areo dialogue" tick in the F12 menu (sorry if you're on console :/) which will enable a button on most parts to show you a window with all the nitty gritty values.  

tl;dr aerodynamics is a complex topic, learn the trends and you should be able to get by.

[natsirt721]

On 8/19/2017 at 1:12 PM, Shaun said:

As well as this, I've noticed that turbofan aircraft often 'hit a wall'. I am aware that their thrust diminishes with speed and altitude, but they seem to accelerate very fast on take off, then find it difficult to creep up to speed. This seems to be why if I create an aircraft with a similar T/W ratio as an airliner, it won't get anywhere near cruising speed and will often not be able to accelerate to take-off speed. Thus, KSP aircraft tend to have crazy T/W ratios to get up to a somewhat realistic speed, but are able to take-off and climb extremely fast.

This could be because in KSP turbojet and turbofan engines take much longer to spool up that real engines.  Less thrust & acceleration earlier means lower velocity for a given distance (with the same TWR) in KSP than IRL. However, KSP engines usually reach full spool by the end of the takeoff run, and combined with the high TWR allow for a more rapid ascent. 

Also, KSP totally fails to model the sharp increase in drag in the transonic and hypersonic regions, which is probably why the model uses the increased drag multiplier.  With an accurate subsonic model, drag would be far too low at high Mach to be practical - the multiplier combined with the polar curve at low velocity isn't too disparate, but enough so that you noticed the improper L/D (props for that!).

[Sampak]

6 hours ago, natsirt721 said:

Jet aircraft performance is highly dependent on (primarily) velocity and air density in complex ways.  Flying at a variety of altitudes and velocities you're probably experiencing this and can't make sense of it.  Thrust from the jet engine, and lift and drag on the airframe are three of the four forces governing vehicle dynamics (weight is the other) and each is affected by the freestream conditions (temperature, pressure, velocity, and derivatively density and Mach number). In turn, these forces govern acceleration and velocity which affects how the freestream conditions change.  Under some conditions you can be unable to change your position and velocity like you would expect; e.g. shallows dives don't change your conditions enough to accelerate rapidly like you would expect.  

KSP doesn't exactly model real world interactions (modeling subsonic compressible aero is tricky enough, let along super- and hypersonic aero), but it's close enough to be passable and the interactions are still very complex.  Short of learning the governing equations and how KSP modifies them I'd say the best thing to do would be trial and error - what conditions are required to get to to other conditions?  If you are so inclined, there is a "show areo dialogue" tick in the F12 menu (sorry if you're on console :/) which will enable a button on most parts to show you a window with all the nitty gritty values.  

tl;dr aerodynamics is a complex topic, learn the trends and you should be able to get by.

 

6 hours ago, natsirt721 said:

This could be because in KSP turbojet and turbofan engines take much longer to spool up that real engines.  Less thrust & acceleration earlier means lower velocity for a given distance (with the same TWR) in KSP than IRL. However, KSP engines usually reach full spool by the end of the takeoff run, and combined with the high TWR allow for a more rapid ascent. 

Also, KSP totally fails to model the sharp increase in drag in the transonic and hypersonic regions, which is probably why the model uses the increased drag multiplier.  With an accurate subsonic model, drag would be far too low at high Mach to be practical - the multiplier combined with the polar curve at low velocity isn't too disparate, but enough so that you noticed the improper L/D (props for that!).

 

Thanks for your replies. I do understand to a degree that the aerodynamics in KSP are game-i-fied due to many reasons like many of the space related physics (N body etc.). With that said I have been long aware of the F12 aero drag display. To my surprise I didnt pay attention to it before but after reading another post I looked again, and the troubled aircrafts I test all seemed to have at least 1 piece of fuselage (usually the Mk2 lifting body components) that have a 20-30 yard long drag line, usually just the one bad one at around 300+ m/s. These are the same craft that have trouble gaining any kind of speed above 0.9xx Mach. So, I oberved another craft that was more a simplistic joke build, with mk1 cylinder shaped components. Sure enough, the worst drag lines were barely a couple odd yards long at hypersonic (super? as in well above 500+ m/s) so it was able to pierce through the air. Keep in mind I'm a minimalist in most of my builds so I'm using 1 or 2 Panther jets in Wet mode (that one I described again) that have PLENTY of t/w on paper. The ones that seem to speed up without issue are dissappointingly NOT the lifting body Mk2 (they look sooo much better/scale to me) but the mk1 builds.

And yes, I've used the slanted scoop as well as this beginner circular intake. No diff as far as any change in success rate....

And no, I'm not going to use the EZ mode Rapiers because, 1. I dont have them unlocked yet, 2. Why not use Panthers if they work for others and me on occasion?

So long story longer, are Mk2 fuselage just gimped in the game at the moment and just need a hideous amount of power behind them to work (i.e. ugly big quad jets etc) and push through their disgusting drag. OR, am I making a mistake somewhere? I'll try to update this post with some pictures of the afflicted aircraft for reference so you might point out some errors on my part (hopefully?). I try to go for somewhat realistic looks and only now and then clip parts to make something look better.

LOL ok as soon as I fired a "bad" design up and took to the air, it again, defied the norm and kept speeding up. It's now been steadily climbing in alt and speed.  WTH... last night I literally spent 2-3 hours trying to figure out why it works and then doesn't. LOL. Same plane, different result.

The "usually bad one" that seems to have worked just now but look at those drag lines!

The cheap one that achieves almost 1000 km AP if I keep pointing up without trying lol  << what drag?

Edit: ok tested the plane again TWICE, now it won't get past 370 m/s, seems to be the same shallow climb or even leveling off at ~ 1000-1500m ASL. These next 2 times the drag line was as long as before (or seemed longer, again the same mk2 part) and I couldnt do anything to get up past sub sonic speeds. I give up!

Edited August 23, 2017 by Sampak
Tested the plane again

[MR L A]

On 19/08/2017 at 6:35 PM, Boris-Barboris said:

On 19/08/2017 at 6:12 PM, Shaun said:

So my question is, is KSP's drag too high

yes

I'd answer this yes and no tbh. I haven't read all the comments so pardon me if I'm repeating someone else, but KSP drag is too high for your use case yes, but the answer is no in terms of drag being deliberately high as a game balancing measure. I felt like I had to state this, not trying to be a doorKnob  

[SchwinnTropius]

On 8/22/2017 at 5:42 PM, Sampak said:

are Mk2 fuselage just gimped in the game at the moment and just need a hideous amount of power behind them to work

Yes. Mk2 fuselages look the best but have insane drag. You will most definitely need lots of power to push them to the speeds you need.