Exactly how bad is the aerodynamic model in KSP?

[RidingTheFlow]

Good video, my major gripe with KSP "placeholder" model always was that it completely counter-intuitive when it comes to anything which involves "balancing", and its nigh impossible to make your craft act as you want in the atmosphere.

Basically, the most complex thing that it does right, is "faster craft goes in atmosphere, more friction force it gets". If you try to look at anything more detailed, its wrong.

And issue is, all stock parts are balanced in mind for this broken model, so game experience changes too much, if you, say, use FAR.

[RidingTheFlow]

It does on your craft because you made it do something. By default, it doesn't do anything.

Umm.... Yes? And? "Doing something" is the whole point of the game?

Anyway I don't exactly understand how it supposed to work automatically, considering multitude of possible abort modes. Even auto-staging gets it wrong very often.

[Hodo]

It does on your craft because you made it do something. By default, it doesn't do anything.

And I was talking metaphorically... Do you think Bob can actually hit the abort button??

[Tex]

Also, your suggestion for problem solving might be THE WORST suggestion ever. People can have whatever opinion they want.

Absolutely, I agree that anyone can have an opinion. As for my suggestion, it isn't really a suggestion; it's the only way around his problem. Mods fix the aerodynamics while we all wait for SQUAD to do it after they finish building up the game we all lovingly support. I apologize if you took my post the wrong way, but you agree that KSP is wonderful so far, isn't it? I do, too. Also, where did the "spokesperson" thing come from? Anybody would have said the same thing if they thought like I did.

[WafflesToo]

I don't see how it flipping front ways first is a problem, considering your COM was pretty close to the top. You essentially threw a shuttlecock from orbit, it is always going to go that way since COM is a lot further from your wanted forward facing surface.

That's the problem, it ISN'T flipping frontways into its direction of travel (watch the navball). It is actually unstable in an atmosphere when flying straight into its prograde marker. Why does it only become stable when flying side-on to the prograde marker? Took me a bit but I figured it out eventually. (hint; there's a parachute mounted on the left side of the equipment bay opposite the goo container, there is one statistic about parachutes that's different from almost every other component in KSP).

The model isn't "gooey", "syrupy", or even really "quirky", it's flawed on a very fundamental level.

[Lanzfeld]

Does on my craft... I have my cockpit/capsule attached to a decoupler and a few radial sepatrons all cued to the Abort button so if things go bad, I jettison the crew and it blows clear of the craft, then I deploy the parachutes and land safeish to the ground.

Sweet idea! I am going to set that up as well.

Learn something new everyday.

[weezl]

You'all are missing the point. The probe exploding isn't the point, the probe exploding is comedy. -_^

The point is that despite being shaped and weighted like a lawn-dart this thing is only aerodynamically stable flying sideways into the jetstream because of the wonky way KSP handles aerodynamics (KSP sees everything as spheres). Squad is on record multiple times having said that the aerodynamic model is a temporary placeholder and that it will be updated. That was two years ago and here it is still holding that place so why is it so out of line to try and put it back on their radar?

(and yes, I know about heatshields. There are two stacked on the leading edge of the probe, and yes it survives just fine when flown manually to maintain heading, but that wasn't the point either, and yes, it flies with perfect stability in FAR, but guess what, also not the point). I don't understand why everyone gets so defensive whenever someone points out this walrus sitting on the coffee table.

I don't want a walrus on my coffee table...

If I might be so bold.. please take your own advice.. many posters here have pointed out that this game is in development and that the aerodynamics are a place holder, you even say so yourself so I know you know this. Your reference to timeframes is pointless.. game development takes time.. especially for something like KSP that has so much to model to even look 'kind of' like reality. They will get to it. Please show patience. THAT is the point all the other posters are making. We all get you want it now. I'm pretty sure if you asked the devs they would want to program everything instantaneously too.. its in the nature of many folks to want to do their best. Unfortunately, making working code is time intensive. You are free to ask.. just don't expect everyone to agree.. we all want the best working game possible.. and many of us trust Squad with having good judgement.. and know the time involved in making something that works.

Its not 'being defensive' to point these things out to you.

[Jouni]

It is actually unstable in an atmosphere when flying straight into its prograde marker. Why does it only become stable when flying side-on to the prograde marker? Took me a bit but I figured it out eventually. (hint; there's a parachute mounted on the left side of the equipment bay opposite the goo container, there is one statistic about parachutes that's different from almost every other component in KSP).

That's actually one of the few nice features of stock aerodynamics. Unless you use parachutes in a weird way, the lander will automatically reorient itself to an attitude that minimizes the chance of breaking apart on parachute deployment.

[MR4Y]

I apologize if you took my post the wrong way, but you agree that KSP is wonderful so far, isn't it? I do, too.

Not really. KSP has a cloudy development model that JUMPS between Alpha and Beta and the total lack of communication between the developers and the community, along with the need of oranges to repass info to the community is abismal at best.

Also, where did the "spokesperson" thing come from? Anybody would have said the same thing if they thought like I did.

Exactly. If 50 people say a foolish thing, it still is a foolish thing. Just because you like the game, that doesn't give you a license to go all out and defend it on a personal level, as if you were part of SQUAD or part of the development team. And even, if you were, that would be an unethical and unprofessional attitue at best.

[[email protected]]

*cough* FAR *cough cough*

Seriously. In a game this easily modded (clicks download for FAR, drags FAR folder into gamedata folder, launches KSP) with such a small dev team, I'd consider it a pretty low (as in not) a priority. I'm confused everytime I see a topic requesting features easily added or corrected with mods as part of the stock game....why? its just so easy to stop complaining and mod the game!

I'd rather see squad focusing on adding new features and getting scope-complete as they say.

[Dodgey]

I'm honestly confused to what it is you are complaining about? Are Squad not developing the free updates for you fast enough? In which case see here.

[RocketPilot573]

If you take a fighter jet into a full power dive it will... slow down...

In the .24 Update Update 2, Harvester talked about how the expected quality of new features is a bit higher now because the game is closer to being finished (or something like that). Meanwhile, the aerodynamics system has been around since the FIRST update and has never been even been touched at all since then. It was fine before, but now it's just ridiculous.

[WafflesToo]

The italians have a saying, "Everything temporary is permament".

Two years ago it was easier to ignore it, things were in a very unfinished state and we knew that they had their hands full just debugging the parts that were there.

It's starting to worry me though as we keep moving forward because aerodynamics is a fundemental part of the core mechanics; I say that because everything new they add needs to interact with it. The more that gets added without fixing it the harder it will be to fix. It's like they skipped right over the "finalize the engine" step and went straight for the "add more bling" step with the very flawed assumption that "well go back to it later".

I don't think it's unreasonable to ask that the game we paid in advance for to be released in a completed, non-broken state.

Edited June 12, 2014 by WafflesToo
clarification

[Hodo]

Sweet idea! I am going to set that up as well.

Learn something new everyday.

Like those... granted I use a few mods but it can be done without mods.

You can see them pretty well behind the cockpit in this picture.

[RidingTheFlow]

I still have hope for one of next big updates bringing both re-vamped spaceplane parts, and improved aerodynamic, this will make perfect sense from development & release standpoint.

[WafflesToo]

I still have hope for one of next big updates bringing both re-vamped spaceplane parts, and improved aerodynamic, this will make perfect sense from development & release standpoint.

The next big update will all be career mode stuff (specifically money and contracts), but I do hope that they make fixing the aerodynamics a priority before they add many more parts to the game. Like I just said, the more stuff they add the harder it will be to fix.

[Kasuha]

If I'm not mistaken, the next major feature to be implemented after Career is Multiplayer.

Drag may not work exactly how it does work on Earth, but honestly if you tried with Earth atmosphere what OP tried in that video, your return module would probably not survive. Real crafts need to pay proper attention to being oriented the right way before they start entering the atmosphere or bad things may happen. Reentry is very little about friction, most of it is high velocity impacts of individual gas molecules to your module - and as they impact it, they transfer the energy of the velocity difference into heat. Not enough energy to turn your ship, but enough energy to destroy it if you leave a weak point exposed.

And there's more about KSP aerodynamic model than its drag everybody seems to be complaining about. What I have major problem with is actually the Lift model. Wing parts are just ridiculous, and control surfaces double so.

[RidingTheFlow]

I don't mind exact math behind drag (e.g. how exactly it slows ship down for particular mass or will it burn down or not). My main issue is that it creates very strange and counter-intuitive rotational forces (e.g. in most cases crafts orient themselves in opposite direction you'd expect them to, and it nigh impossible to make them behave the way you want, at least unless you use non-stock parts like B9 airbrakes).

[cantab]

My main issue is that it creates very strange and counter-intuitive rotational forces (e.g. in most cases crafts orient themselves in opposite direction you'd expect them to, and it nigh impossible to make them behave the way you want, at least unless you use non-stock parts like B9 airbrakes).

When you know how it works it's easy enough. (Just put your higher drag coefficient parts, for example stowed parachutes, at the rear.) The stock aerodynamics may be unrealistic, but they're at least fairly simple and easy to understand, and they haven't stopped players building some pretty neat aircraft. And the air's better than the water, at least.

[RidingTheFlow]

Its not that simple. One time I think I've got how it works, next I am pulling my hair out trying to understand why my craft flipping around or turns to the side and starts rapidly spinning... Or why my plane is more stable flying tail-first and its impossible to point nose forward once it starts doing that. Probably one of the issue is that "drag values" don't really follow the parts "look and feel", and also there are some values which don't even show in editor I believe.

Edited June 12, 2014 by RidingTheFlow

[Moss]

I'm sure they'll update the aerodynamics before multiplayer at the very least.

[RocketPilot573]

Its not that simple. One time I think I've got how it works, next I am pulling my hair out trying to understand why my craft flipping around or turns to the side and starts rapidly spinning... Or why my plane is more stable flying tail-first and its impossible to point nose forward once it starts doing that.

Exactly, some say that FAR is hard, but at least you can just do some research on Google and find out how planes fly. But with the current aerodynamics, it's anybody's guess. No offense to anyone, but the tutorials on the forums are no help at all, as this system is just to unpredictable (the tutorials really only work with very simple crafts).

[RidingTheFlow]

Yup. Its actually not simpler than FAR, because it has low of own quirks - and you have to learn all these quirks which completely disjointed with what you expect from real world (even in crude approximation). This makes for really unproductive experience - and one of most respectable things of KSP I thought was how educational it is, in everything except this.

FAR is not that hard actually, because, well, it follows real world. And I've got experience with it, suddenly I've noticed that I can get same payload to orbit with launcher twice as small (fairings and all of that) - and this pretty much destroys balance, because I know that all crafts I design for FAR will be hopefully unusable with stock...

Edited June 12, 2014 by RidingTheFlow

[Kasuha]

Its not that simple. One time I think I've got how it works, next I am pulling my hair out trying to understand why my craft flipping around or turns to the side and starts rapidly spinning... Or why my plane is more stable flying tail-first and its impossible to point nose forward once it starts doing that.

It is very simple.

All that matters is part's drag coefficient. Size or shape of the part plays no role, mass plays very little role.

All parts with the same drag coefficient (0.2 is standard used by most parts) are undergoing exactly the same deceleration. They don't turn the ship in any direction.

Parts with smaller drag coefficient (nosecones, Kerbals, inline docking port, command pod) undergo smaller deceleration, therefore they're pulled to the front.

Parts with larger drag coefficient (SRBs, chutes, cupola module) undergo greater deceleration, therefore they're pulled to the back.

All you need is to free your mind of the prejudice how you think it should work and open it to how it actually works.

[RocketPilot573]

It is very simple.

All that matters is part's drag coefficient.

All parts with the same drag coefficient (0.2 is standard used by most parts) are undergoing exactly the same deceleration. They don't turn the ship in any direction. Mass, size, or shape of the part plays no role.

Parts with smaller drag coefficient (nosecones, Kerbals, inline docking port, command pod) undergo smaller deceleration, therefore they're pulled to the front.

Parts with larger drag coefficient (SRBs, chutes, cupola module) undergo greater deceleration, therefore they're pulled to the back.

All you need is to free your mind of the prejudice how you think it should work and open it to how it actually works.

Err, except that the very few parts that do have a different coefficient have very little mass, so they make no difference.