Atmosphere still too soupy to anyone else?

[adsii1970]

6 minutes ago, Alshain said:

I don't recall off the top of my head.  It's been a while.  It had to do with reentry though.

I think it was a Scott Manley video...

[OhioBob]

To me, this is "the atmosphere", and it hasn't changed since the introduction of v1.0.

Spoiler

		Atmosphere
		{
			enabled = True
			oxygen = True
			altitude = 70000
			atmosphereMolarMass = 0.0289644002914429
			staticPressureASL = 101.324996948242
			temperatureSeaLevel = 288
			pressureCurve
			{
				key = 0 101.325 0 -0.01501631
				key = 1241.025 84.02916 -0.01289846 -0.01289826
				key = 2439.593 69.68138 -0.01107876 -0.01107859
				key = 3597.11 57.78001 -0.009515483 -0.009515338
				key = 4714.942 47.90862 -0.00817254 -0.008172415
				key = 5794.409 39.72148 -0.00701892 -0.007018813
				key = 6836.791 32.93169 -0.006027969 -0.006027877
				key = 7843.328 27.30109 -0.005176778 -0.0051767
				key = 8815.22 22.63206 -0.004445662 -0.004445578
				key = 10786.42 15.3684 -0.003016528 -0.00301646
				key = 12101.4 11.87313 -0.002329273 -0.00232922
				key = 13417.05 9.172798 -0.001798594 -0.001798554
				key = 16678.47 4.842261 -0.0009448537 -0.0009448319
				key = 21143.1 2.050097 -0.0003894095 -0.0003894005
				key = 26977.92 0.6905929 -0.0001252565 -0.0001252534
				key = 33593.82 0.2201734 -3.626878E-05 -3.626788E-05
				key = 42081.87 0.05768469 -9.063159E-06 -9.062975E-06
				key = 49312.13 0.01753794 -3.029397E-06 -3.029335E-06
				key = 56669.95 0.004591824 -8.827175E-07 -8.826996E-07
				key = 62300.84 0.001497072 -3.077091E-07 -3.077031E-07
				key = 70000 0 0 0
			}
			temperatureCurve
			{
				key = 0 288.15 0 -0.008125
				key = 8815.22 216.65 -0.008096968 0
				key = 16050.39 216.65 0 0.001242164
				key = 25729.23 228.65 0.001237475 0.003464929
				key = 37879.44 270.65 0.00344855 0
				key = 41129.24 270.65 0 -0.003444189
				key = 57440.13 214.65 -0.003422425 -0.002444589
				key = 68797.88 186.946 -0.002433851 0
				key = 70000 186.946 0 0
			}
			temperatureSunMultCurve
			{
				key = 0 1 0 0
				key = 8815.22 0.3 -5.91316E-05 -5.91316E-05
				key = 16050.39 0 0 0
				key = 25729.23 0 0 0
				key = 37879.44 0.2 0 0
				key = 57440.13 0.2 0 0
				key = 63902.72 1 0.0001012837 0.0001012837
				key = 70000 1.2 0 0
			}
			temperatureLatitudeBiasCurve
			{
				key = 0 17 0 -0.3316494
				key = 10 12 -0.65 -0.65
				key = 18 6.36371 -0.4502313 -0.4502313
				key = 30 0 -1.3 -1.3
				key = 35 -10 -1.65 -1.65
				key = 45 -23 -1.05 -1.05
				key = 55 -31 -0.6 -0.6
				key = 70 -37 -0.6689383 -0.6689383
				key = 90 -50 -0.02418368 0
			}
			temperatureLatitudeSunMultCurve
			{
				key = 0 9 0 0.1554984
				key = 40 14.2 0.08154097 0.08154097
				key = 55 14.9 -0.006055089 -0.006055089
				key = 68 12.16518 -0.2710912 -0.2710912
				key = 76 8.582909 -0.6021729 -0.6021729
				key = 90 5 0 0
			}
			temperatureAxialSunMultCurve
			{
				key = 0 0 0 0
			}
		}

 

 

[Alshain]

16 minutes ago, adsii1970 said:

I think it was a Scott Manley video...

No, I recall Gene or Werner saying atmosphere was what holds you up and what slows you down (paraphrased of course, there is no way I could provide a direct quote from that long ago).  You know those tutorials, they intentionally dumb it down for the sake of explanation, but that's the terminology people are going to use on the thread if it is in the game.

Edited September 4, 2017 by Alshain

[OhioBob]

5 hours ago, Alshain said:

No, I recall Gene or Werner saying atmosphere was what holds you up and what slows you down (paraphrased of course, there is no way I could provide a direct quote from that long ago).  You know those tutorials, they intentionally dumb it down for the sake of explanation, but that's the terminology people are going to use on the thread if it is in the game.

I just went through several of the tutorials, as well as the KSPedia.  The only thing I saw that sounds like what you're talking about is one reference in the Orbital tutorial in which Gene says "the atmosphere will start slowing you down."  I saw far more usage of terms like drag, air resistance, and aerodynamic forces.  I think it is a real stretch to say that because of that one obscure reference that is perfectly okay to substitute the word atmosphere for terms like drag and air resistance.

Why do we slow down when we pass through an atmospheres?  Because drag forces are being generate that oppose the vehicle's motion.  And what determines the magnitude of the drag forces?  Velocity, air density, drag coefficient, and area normal to the air flow.

If somebody thinks that drag in KSP is too high and would like to do something about it, then they need to know the source of the problem.  In this case the problem is not that the atmosphere is too thick, and it's not helpful to say that it is.  If somebody wants to get to the root of the problem, they need to understand that it is either because the drag coefficients or the drag multipliers are too high.

I'm not trying to be difficult here, I'm just trying to clarify what is really happening.  I thought I was being helpful.
 

Edited September 5, 2017 by OhioBob

[OHara]

On 9/2/2017 at 7:56 PM, ZooNamedGames said:

my Saturn Shuttle [...] one of my biggest problems is that I lose so much speed during the final moments of the landing

I think that earlier I had understood you backwards.  I thought you meant you float down the length of the runway while slowing down to the strangely low stall speed of 30m/s (i.e., too much lift at low speed).  Did you mean instead that drag decelerates you more quickly than you expected (too much drag) ?

I found your SSM-X3 and tried it.  It flew final approach nicely* at 90m/s on a 15°--20° glide-slope.  The flare slowed me to 70m/s, then it slowed down very quickly and I couldn't float very far at all.   That seems to be how a shuttle should behave, slowing quickly upon the steep nose-up flare.

I do find it strange that 6-tonne jets in KSP lift off at 40m/s, 80kts, but I have to admit that the various numerically-unrealistic parts of game are reasonably balanced.

Edited September 5, 2017 by OHara
*after I disabled roll on the rudder and inner ailerons, and disabled pitch on the outer ailerons

[ZooNamedGames]

9 hours ago, OHara said:

I think that earlier I had understood you backwards.  I thought you meant you float down the length of the runway while slowing down to the strangely low stall speed of 30m/s (i.e., too much lift at low speed).  Did you mean instead that drag decelerates you more quickly than you expected (too much drag) ?

I found your SSM-X3 and tried it.  It flew final approach nicely* at 90m/s on a 15°--20° glide-slope.  The flare slowed me to 70m/s, then it slowed down very quickly and I couldn't float very far at all.   That seems to be how a shuttle should behave, slowing quickly upon the steep nose-up flare.

I do find it strange that 6-tonne jets in KSP lift off at 40m/s, 80kts, but I have to admit that the various numerically-unrealistic parts of game are reasonably balanced.

I've advanced to the eighth iteration of the vehicle now. I've had to disable the flaps in a similar manner to what you did for launch and reenable them for landing. Quite a pain without AG toggles. The monopropellant tank inside the command pod has been locked to ensure that the CoM is forward during the return. Airbrakes removed since frankly the space shuttle didn't have them so neither do I. Clipped tail fins to create an airbrake. Minor angle adjustment to the wings. New RCS layout. Made the SAS disablable as they could make landing maneuvers too significant and cause stalls.

However while passing through 1800m/s to 600 (usually flown around 10-20km); the vehicle strangely becomes unstable and no matter what I do it refuses to stay locked to prograde but rather performing an outward spin. The best solution to this was to ensure there be weight during re-entry and locking the Mk3's monopropellant however the issue still appears however SAS is occasionally toggled and help with the roll but the vehicle still seems resistant to point towards prograde. As it performs these outwards rolls consistently the speed is bled too quickly and without any means to control it. Making a safe gradual return, turn into a splatting east of the KSC mountain range. Granted, once I pass under 6km or so, stability returns. Giving the vain thought I might be able to recover it only to land far short or try to overcorrect again, leading to a stall, before crashing into the ground.

[OHara]

At high altitude drag seems to be relatively more important, so the aerodynamic center moves forward despite the lifting surfaces in the back.  I also noticed the shuttle tends to tumble around 20km altitude.

Once you are stable again, SAS off, control surfaces limited to 30%, trimmed for nose about 4° above prograde, leftover monopropellant can stretch the glide somewhat, and the game is to adjust your path without turning so far as to be a draggy brick.  

After playing around with it, I don't think the atmosphere is too soupy.  This shuttle might be easier with lower global scaling of low-speed drag coefficients, but then other aircraft will float even longer.

[OhioBob]

36 minutes ago, OHara said:

After playing around with it, I don't think the atmosphere is too soupy.  This shuttle might be easier with lower global scaling of low-speed drag coefficients, but then other aircraft will float even longer.

Also beware that if changes were made to the atmosphere (i.e. air density) to reduce drag, lift would also be reduced.  To reduce drag without reducing lift requires selective targeting of the drag coefficients.

[Gaarst]

After playing around with FAR for a little while for the first time and comparing it to the stock atmosphere, I now officially hate the stock atmosphere (or drag coefficients, whatever you like) and will never build a plane on it ever again.

I realised flying an airliner designed for efficient long range flight with a L/D ratio lower than the Space Shuttle is stupid. Not being able to get it to cruising altitude is even more stupid. Doing both these things while having more than twice the thurst of an equivalent real plane is what made me stop trying.

My days of brute-forcing my way through the atmosphere are over. Do yourself a favour and get FAR too, plus it makes designing planes a lot more interesting than stock.

Edited September 11, 2017 by Gaarst

[MR L A]

On 03/09/2017 at 3:56 AM, ZooNamedGames said:

I have spent time since .25 refining my Saturn Shuttle (and still continue to) but one of my biggest problems is that I lose so much speed during the final moments of the landing that I can almost stall moments before touching down with less than 20m/s remaining. Which I find not only problematic (as it's stall speed is far above that so stall to that extent is catastrophic or possibly even fatal). In reality the Space Shuttle landed at 214 miles per hour (344.4 kmh) or ~95.6m/s (on the low end). Whereas I end up landing at highway speeds of 44 miles per hour (71.9kmh).

Understanding this takes place on a unrealistically small world with vastly simplified physics but I still have to wonder-

Am I the only one who thinks the atmosphere is still too thick?

Or are there players out there who find the game perfectly suitable, making me into a raving madman? Or do others agree?

_{^{Ps- I am aware there are mods to solve my problem however I am a vanilla kinda guy and more importantly feel that it's a lingering stock problem that should be resolved... again.}}

Discuss

I had a VERY similar problem with a mk3 shuttle I was using... the solution was to land extremely fast with an extremely low vertical speed. Which takes a huge amount of piloting skills and using a LOT of runway. Pretty much how the real shuttle landed as you described. You just need a steeper re-entry profile or use jets so you aren't actually gliding

[AeroGav]

The problem isn't the atmosphere it's the stock flight model.    There is too much lift at low speed and not enough at high speed.    There is too much drag at low speeds but not enough at high.     Lift drag ratios in KSP are very poor compared with real world examples.   Jet airliners get about 20:1, in KSP you'd not get much more than 5:1 at similar mach number.   This is made up for by the ridiculously OP jet engines.     However in the real world orbital velocity is >mach 20 rather than mach 7 and by mach 17 or so I think the Space Shuttle orbiter would be seeing L/D numbers not much better than unity.   Hard to optimise an airframe for such a vast speed range.

My KSP shuttle.        Glide ratio similar to real one (it looks very flat here but that's because we're loosing speed the whole time).   Landing speed ... not so much.  More like a Cessna.   With a STOL kit and Tundra tyres.

Oh and let's not mention my latest SSTO shall we..  not quite realistic landing performance.

Spoiler

 

Yes it had an interplanetary fuel load aboard at the time ....  the next video shows it going to orbit from the roof.

 

In Ferram,  there is more lift at very high speed.    At mach 4.5 we're almost doing level flight at 41km !   Made for a pretty gentle re-entry

 

But even this craft , with absolutely ginormous wings,  has a more realistic landing speed 

Subsonic L/D is around 10:1,  low supersonic is better than you'd see without Ferram (never got more than 4.3:1 supersonic in stock aero).   However, it declines fairly rapidly with increasing mach number, dropping below 2.7 at mach 4, whereas with stock aero i get about the same L.D ratios all the way from mach 1.5 to orbit

 

 

Edited September 11, 2017 by AeroGav

[Mr. Peabody]

KSP does not need to be perfectly realistic. The atmosphere is  too thick for my liking.  Atmospheric drag is already very high. A worst case scenario would be ending up with an atmosphere like Eve's.

[luizopiloto]

I've made my SSTO glide 800Km at 30Km alt after reentry from 300Km orbit...  It sustained mach6 for 10min until I had to do S-Turns to kill speed and land...
It all depends of the design...

[guesswho2778]

 

1 hour ago, The Dunatian said:

KSP does not need to be perfectly realistic. The atmosphere is  too thick for my liking.  Atmospheric drag is already very high. A worst case scenario would be ending up with an atmosphere like Eve's.

or maybe kerbins atmo is just thicker than ours

Edited September 13, 2017 by guesswho2778

[ZooNamedGames]

1 minute ago, guesswho2778 said:

or maybe kerbins  atmo is just thicker than ours

Considering the point of Kerbin is to be an Earth Analogue for new players, that renders the point moot. If it's ok to make the atmosphere unnecessarily thick, then why not make Kerbin have the some atmospheric density as Eve and force players to work for orbit.

[NSEP]

Can you edit how soupy you want your atmospheres in the configs? Can you do that with FAR? 

I like my atmospheres thick, it has less realism but its still a little more 'fun'. I completely understand why it is for some not enjoyable though.

[Azimech]

On 9/11/2017 at 3:03 PM, Gaarst said:

After playing around with FAR for a little while for the first time and comparing it to the stock atmosphere, I now officially hate the stock atmosphere (or drag coefficients, whatever you like) and will never build a plane on it ever again.

I realised flying an airliner designed for efficient long range flight with a L/D ratio lower than the Space Shuttle is stupid. Not being able to get it to cruising altitude is even more stupid. Doing both these things while having more than twice the thurst of an equivalent real plane is what made me stop trying.

My days of brute-forcing my way through the atmosphere are over. Do yourself a favour and get FAR too, plus it makes designing planes a lot more interesting than stock.

Or, just go to the debug menu -> Physics -> Drag -> Global Drag Modifier, and set the slider to your liking.

Some of us do it for stock turboprops, we set it to minimal. This way the planes approach the performance of KSP 0.90 turboprops, and close to FAR.

So yes, I agree. drag values in the lower atmosphere are way too high. But I wonder if Squad would ever choose to change them again. At least we now have the power to modify stuff ourselves.

Edit: @NSEP you'll like this.

Edited September 13, 2017 by Azimech

[Gaarst]

16 hours ago, Azimech said:

Or, just go to the debug menu -> Physics -> Drag -> Global Drag Modifier, and set the slider to your liking.

Some of us do it for stock turboprops, we set it to minimal. This way the planes approach the performance of KSP 0.90 turboprops, and close to FAR.

So yes, I agree. drag values in the lower atmosphere are way too high. But I wonder if Squad would ever choose to change them again. At least we now have the power to modify stuff ourselves.

Edit: @NSEP you'll like this.

It's probably what I would have done if I ended up not liking FAR, but besides the better atmo, I enjoy the complexity that it adds to building planes.

I personally enjoy complexity in the gameplay itself (e.g. RSS) rather than building complex stuff with no other motivation than trying to melt my PC with hundreds of parts. So I prefer building a simple plane working nicely in FAR rather than a gigantic SSTO in stock.

[Gordon Fecyk]

Even Realistic Atmospheres doesn't change Kerbin's atmosphere too much, though it makes some adjustments. Combine that with Ferram Aerospace and see how your planes behave.

I found that Advanced Jet Engines also messes with the stock jet engines in ways I didn't expect; varying thrust and specific impulse with more than just air density.

Here's a fascinating thing: One of Mark Thrimm's stock spaceplanes, his LTS Sparrow, originally had an extra Mk2 liquid fuel tank. With just FAR installed I could replace that extra tank with a short cargo bay. With FAR and AJE? I sometimes wish I put that tank back. Jets get seemingly nerfed with AJE, but what's happening is jets are running through NASA EngineSim code. Stock jets seem overpowered compared to realistic versions, possibly to overcome the stock drag model.

So it appears stock jets plus stock drag has similar performance to AJE jets plus FAR drag. I'll have to check against the other LTS craft I converted from stock to FAR, but I suspect it'll be similar results.

 

[StupidAndy]

meh, I don't care, its all the same, I have to figure out how to do planes good still, which is a problem with BDA and other aeroplane challenges

[Helmetman]

Part of what's going on might be the difference between your spaceshuttle and the real one. 
Yours probably uses a lot of fuel deducted from its total weight. Or is it actually a replica from the real Saturn shuttle considering all those parameters? You might if it's taken you since...

The spaceshuttle had about 240.000 total weight of which 38.000 lb was OMS propellant. Compared to other ksp build craft (yours?) that total fuel weight might be much lower in comparison.  So the real Spaceshuttle had a lot of empty landing weight on final touchdown and required more speed to land safely. Also the spaceshuttle had to approach faster to save headroom for required control inputs on touchdown. Things like sudden winds gusts which are common in Florida. You wouldn't want to land a unpowered glider with just enough accuracy to glideslope towards the threshold, maybe if yours is powered?

Haven't used excessive wing parts? Even for cosmetics for interior design or whatnot? Sad to say, but that may include extra wing lift into your model making it fly more easily.  It probably won't get you to 214 miles per hour but maybe the change shifts your consciousness from finding this bizarre to this makes sense mode. 

I do never really expect the atmosphere to be realistic, neither in FAR. People will always review aerodynamic quality to be obsolete. Whether that is in the stock or the FAR mode. Things get pretty accurate once you launch rockets that includes FAR. And yes planes fly and I had build a couple spaceplanes. 
Comparing flight sims and aerodynamic complexities in real life and you will see a wide gap, and KSP is not a flight SIM. FAR is preferably mixed with RSS due to the similar Earth atmosphere dimensions. Jet engines work better as they should. Aerodynamic losses are minimized as should compared to reality. But proper plane dynamics is not the best and ultimately not the greatest part of KSP. If it were to be stock it would require less Delta V to orbit for aerodynamic losses. And I wouldn't want that part to become easier if I'm honest. For that they would to have increase the size of Kerbin.

 

 

 

 

[ZooNamedGames]

On 9/16/2017 at 3:31 PM, Helmetman said:

Part of what's going on might be the difference between your spaceshuttle and the real one. 
Yours probably uses a lot of fuel deducted from its total weight. Or is it actually a replica from the real Saturn shuttle considering all those parameters? You might if it's taken you since...

The spaceshuttle had about 240.000 total weight of which 38.000 lb was OMS propellant. Compared to other ksp build craft (yours?) that total fuel weight might be much lower in comparison.  So the real Spaceshuttle had a lot of empty landing weight on final touchdown and required more speed to land safely. Also the spaceshuttle had to approach faster to save headroom for required control inputs on touchdown. Things like sudden winds gusts which are common in Florida. You wouldn't want to land a unpowered glider with just enough accuracy to glideslope towards the threshold, maybe if yours is powered?

Haven't used excessive wing parts? Even for cosmetics for interior design or whatnot? Sad to say, but that may include extra wing lift into your model making it fly more easily.  It probably won't get you to 214 miles per hour but maybe the change shifts your consciousness from finding this bizarre to this makes sense mode. 

I do never really expect the atmosphere to be realistic, neither in FAR. People will always review aerodynamic quality to be obsolete. Whether that is in the stock or the FAR mode. Things get pretty accurate once you launch rockets that includes FAR. And yes planes fly and I had build a couple spaceplanes. 
Comparing flight sims and aerodynamic complexities in real life and you will see a wide gap, and KSP is not a flight SIM. FAR is preferably mixed with RSS due to the similar Earth atmosphere dimensions. Jet engines work better as they should. Aerodynamic losses are minimized as should compared to reality. But proper plane dynamics is not the best and ultimately not the greatest part of KSP. If it were to be stock it would require less Delta V to orbit for aerodynamic losses. And I wouldn't want that part to become easier if I'm honest. For that they would to have increase the size of Kerbin.

 

 

 

 

The craft file itself was posted a while back so you can compare and look to the details as you like.

That said, the only fuel remaining during descent was the RCS provided in the mk3 cockpit. The rest is exhausted.

There are only 4 wing parts. 5 control surfaces. 2 vertical stabilizers (clipped to form one; done so that when they are deployed they act like the Space Shuttle vertical stabilizer).

[regex]

Quote

Stock jets seem are very overpowered compared to realistic versions

FTFY

Quote

possibly to overcome the stock drag model.

Possibly, although it might also have something to do with how the engines worked pre-1.0 when they were some, what, 17 times more fuel efficient than IRL engines? Now they're like 2 or 4 times better, I think? While "nerfing" them to more realistic levels would have been much better for the game overall (as well as caused some very delicious tears in the community) it wouldn't have really kept with the spirit of the game as envisioned by Harv of having inventive and silly things still being able to fly. At least now you can't get to orbit on jet engines alone by "storing" intake air, god that was idiotic...

I think one of the most interesting things about discussions regarding the drag model or flight is that people don't seem to realize that we're not playing a traditional flight sim where the plane (my word for "the flying object") is one single piece and has defined flight characteristics. In KSP the plane's shape and characteristics are derived from the individual pieces that comprise it, and it's very tough to determine things like what is actually a wing and which way is up for the purposes of the airfoil from objects that can be assembled however the user so desires. IIRC even ferram4 is still struggling with these questions.

Pre-1.0 we had a placeholder drag system which was replaced by the newer drag cube model. I find the drag cube model flies more like FAR pre-1.0; obviously much simpler but the feel is close. That jives with what I actually know about the two systems. Pre-1.0 FAR had some arcane bits you had to learn like closing open attach nodes (like with an antenna) in order to reduce drag and using the most optimistic wing shape for the speed (I think?). Post-1.0 FAR uses a voxelization technique to create the craft but it still has to answer some questions from the individual parts rather than their shape (wings) and, IIRC, uses a single airfoil model. The new drag cube model doesn't handle occlusion or clipped parts very well (not sure if FAR did pre-1.0 either) and assumes the most optimistic wing shape for the speed so the user doesn't have to worry about area-ruling the craft and such.

Anyway, these sorts of things get more complicated as the simulation gets more complex. If you're talking about a single aircraft "piece" the flight characteristics can be individually well-defined. You could even have multiple pieces if you abstract well and work from an assumed airframe outline (wings in certain places, etc...). The problems become much more complex as the number of pieces grow; you have to make assumptions, your sim needs to maintain a certain speed, if "gameplay" is a concern you need to accept certain abstractions, etc... That complexity means no one will ever be fully satisfied with any drag model.

Bottom line, neither are perfect. If you want a more realistic drag model get FAR but the stock drag model is (IMO at least) perfectly adequate to the task of simulating flight on a completely unrealistic planet that makes absolutely zero sense (and it even flies okay).

Edited September 19, 2017 by regex