Planes always fishtail

[Gaarst]

Just now, pincushionman said:

While this is true, it doesn't mean that a user should rule out quirks in his own design being a problem, or that tweaks to it couldn't minimize the effects of the bug.

Buggy or not, the behavior wheels has explicitly changed, so it is reasonsble to assume that some designs will not work over the transition, while others port just fine.

The posters above, who make plane with no issues, have offered to inspect  the design of a buggy plane. Their findings could be "It works fine for me," which means it's all bug and an interaction between Unity and the particular system (worst case), "It breaks a little, but if you do this it goes away," or "this design is terrible, do this instead." Until we see, we don't know which.

This comment was not directed at the people proposing the design to be a cause, but at the ones saying the design is the cause. Of course a badly designed plane will not function properly, but from what I read in this thread, besides the original issue, several people have observed abnormal behaviours with sensible designs, though some seem to still deny the problem as lying in the game itself.

Now, you're right keeping: all the possibilities open to solve a problem is a good thing; but IMO we have gathered enough information about the wheels bugs to rule out the design as a general cause. For this particular issue the OP is describing, design may be a cause, though I don't think it is, but people are free to investigate whichever cause they think is causing the issue.

[bewing]

4 hours ago, Gaarst said:

The issue is not with crafts. A craft that works fine for you might be uncontrollable for me, and the other way around.

Complete BS. There is a class of machine-specific issues that involve RAM limitations, disk limitations, specific graphics cards, and OS interactions. Basic math, which includes physics simulations, is precisely identical across all machines.

[Rocket In My Pocket]

3 hours ago, pincushionman said:

And for all of you insisting you can fly drive with no real problems, can you link to one of yours

Here's one, not even a terribly good design, uses very basic parts including the basic landing gear which is the buggiest, they are also mounted to the wings which tends to cause problems. No wheel tweakables have been changed. However this takes off and lands fine for me besides some very light drifting that is easy to compensate for. I've landed it going 150+ m/s horizontally. As I stated before the only way I could get this or any other plane to fishtail was to hold the nose down forcibly while rocketing full speed down the runway. Give it a try I suppose? Landing flaps are on the RCS action group. (The "R" key)

Dropbox Link: https://www.dropbox.com/s/lneuz6bexbffge3/ExamplePlane1.craft?dl=0

If anyone would like to upload a plane they feel is "impossible to take off" I would most certainly be willing to try my hand at it and report back on my findings.

Edited July 16, 2016 by Rocket In My Pocket

[Gaarst]

11 minutes ago, bewing said:

Complete BS. There is a class of machine-specific issues that involve RAM limitations, disk limitations, specific graphics cards, and OS interactions. Basic math, which includes physics simulations, is precisely identical across all machines.

Have you even read my post before calling BS?

You're basically saying that a significant part of the community (including the guys at Squad that acknowledge the issue) are noobs that can't design a plane. Though maybe you're right, maybe something went wrong in my brain after 1.1 and now I'm just too stupid to make a plane, who knows? 

Some people experience the issue. You don't, great for you, but please stop denying the issue exists when even Squad confirms it's there.

[bewing]

9 minutes ago, Gaarst said:

Have you even read my post before calling BS?

You're basically saying that a significant part of the community (including the guys at Squad that acknowledge the issue) are noobs that can't design a plane. Though maybe you're right, maybe something went wrong in my brain after 1.1 and now I'm just too stupid to make a plane, who knows? 

Some people experience the issue. You don't, great for you, but please stop denying the issue exists when even Squad confirms it's there.

I read it. Of course there are bugs with wheels, but "machine-specific fishtailing" is not one of them. Squad is acknowledging all the other ones.

[Gaarst]

4 minutes ago, bewing said:

I read it. Of course there are bugs with wheels, but "machine-specific fishtailing" is not one of them. Squad is acknowledging all the other ones.

The reported bugs include situations where wheels take abnormal grip values (at least seem to) causing loss of control. IMO this is enough to explain the OP's fishtailing, especially since he states that he sees the issue with stock crafts which rules out the "design" cause.

[bewing]

21 minutes ago, Gaarst said:

The reported bugs include situations where wheels take abnormal grip values (at least seem to) causing loss of control. IMO this is enough to explain the OP's fishtailing, especially since he states that he sees the issue with stock crafts which rules out the "design" cause.

Again, the abnormal grip value bugs (if they exist) are not going to be machine-specific -- they will either be mod-related, or they will be universal. And the stock craft stink, and need to be heavily tweaked to fly properly. During most of beta-testing, the stock craft couldn't even get themselves airborne without losing parts.

[foamyesque]

FIshtailing with gear are often caused by the rear wheels losing traction because of aerodynamic loading. If you're using trike layouts, design your plane with a bit of up pitch and put your rear gear behind both your CoM and CoP.

 

[Rocket In My Pocket]

16 hours ago, foamyesque said:

FIshtailing with gear are often caused by the rear wheels losing traction because of aerodynamic loading. If you're using trike layouts, design your plane with a bit of up pitch and put your rear gear behind both your CoM and CoP.

What he said^ Can we make this a loading screen tip?

Your nose should be "flying" long before the rest of your plane is. Imagine lift off as an extreme slow motion wheelie.

On a lot of real life smaller planes the front wheel isn't even steerable in any fashion. It's basically a "shopping cart wheel" that just rolls about wherever it pleases, and believe me; the first thing you do as you roll down the runway is get the pressure off that damn wheel.

[JPLRepo]

22 minutes ago, Rocket In My Pocket said:

What he said^ Can we make this a loading screen tip?

Your nose should be "flying" long before the rest of your plane is. Imagine lift off as an extreme slow motion wheelie.

On a lot of real life smaller planes the front wheel isn't even steerable in any fashion. It's basically a "shopping cart wheel" that just rolls about wherever it pleases, and believe me; the first thing you do as you roll down the runway is get the pressure off that damn wheel.

You can contribute your tip to the Community loading screen tool tips that my new mod supports. @Deimos Rast is putting it together (still a work in progress)

 

[foamyesque]

On 2016-07-16 at 8:11 AM, pincushionman said:

EDIT: And for all of you insisting you can fly drive whith no real problems, can you link to one of yours for him to try? That could be helpful too.

Sure. As it happens I have one handy from an argument with @Alshain a couple of days ago.

https://www.dropbox.com/s/ldl6f0jsv0hfmly/Shrike MK2 - Corrected.craft?dl=0

 

 

[Foxster]

Some of those here saying all their craft fishtail are re-using older craft. Some of those who don't seem to have the problems are creating new craft in the latest version of KSP. Is there perhaps an issue with older craft being upgraded?

Sticking friction control to max on the rear wheels is the workaround that has been working for me for a while now. Still not as stable as pre-1.1 though. 

Stable take-off and landing of a reasonably designed aircraft really should be stock behaviour. I can imagine the current behaviour being a nightmare for some nubs. 

[Diche Bach]

I played the game three years ago, but took a long break. I restarted recently. I have no memory of flying planes in KSP from before, though I have played quite a few other flight games.

In my ongoing career play, I got the first couple nodes of flight stuff researched (the crappy starting landing gear, and wonky-ass tail section). I built one: it fishtailed. I tweaked it: if fishtailed. I tweaked it more, for hours: it fishtailed . . . however, it deserves to be noted here that: my understanding of aerodynamics is pretty damn rudimentary, and worse I don't really know how squad has things like angle of attack modeled (seems a bit weird, but anyway) . . . part deux . . .

I restarted a new career mode with mostly all easy settings, max science and cash, etc. (and including stock models, which I don't normally do), booted it up, bought the first level of science nodes I could afford (using Kerbal Construction Time [KCT], so I'd have to wait a bit to upgrade my R&D center) and selected the Velocitize.

Took off with ZERO problem, even on a dirt airstrip. Two things should be noted here: this plane does not use the crappy starter landing gear (the stuff that looks like it belongs on a single engine propeller plane), and this plane was NOT designed by me.

I didn't change anything about it, and it flew just fine (even though I was only able to fly it in "Simulation" mode in KCT for the simple reason I had not unlocked the R&D nodes for a couple of the parts . . . it is identical to flying it, you can just crash and not lose the funds or pilots . . .).

Contrast: I took off the tier 2 landing gear, and tried to fly it with the crappy stuff: crashed like a stock market. Funnily enough, it didn't fishtail so much though . . . the main problem was that the rear wheels (the fixed ones) were not well centered.

If there is a problem (and I suspect there is) I believe what it amounts to is one or more of the following:

1. Either the short or long (scissor shaped) starter landing gear are not behaving properly with the ground surface.

2. They are not behaving properly with the plane itself, causing it to be virtually impossible to get them to be attached in well-aligned fashion (even with that grid mod thing running).

[Diche Bach]

Not that this is really a "solution" but another point of note on the issue of the fish-taily, strange behaving landing gear.

In my career playthrough, I haven't yet got far enough to get to rovers, but I have all these nifty science instruments, and I wanted to try out the land navigation aspect of the game. So I improvised and mish-mashed together a "Science Jet Buggy" using jet aircraft, and space modules.

Partly this was me learning how to use the build UI, so it is a bit clunky, but it functions . . . or rather functioned, now that I've installed a couple more mods including "Stronger Joints" it does the bunny hop when it lands on the runway and explodes! so I guess I'll have to redesign it . . . not a big deal as it is anything but symmetrical and was mostly just experimental to see what I could slap together.

Netted 129.5 science in one ~1.5 hour journey to north, then into ocean (yes it works well in water), then back inland to highlands.

[Ozzallos]

I have the landing gear wobble blues as well. Not all the time, but more often than not. It's to the point where I don't rely on the runway and inherently design my aircraft to be short takeoff designs (ie; bigger landing gear in front than back). They still want to powerslide, but at least I can power out of it and get airborne this way. I did not know you could increase the landing gear drag, so something good came out of this thread for me, at least.

Edited July 22, 2016 by Ozzallos

[Daveroski]

The other day I had a space plane that was a bit bouncy when I went to launch it.

I raised the gear while it was still on the runway and then lowered them again.

It seemed to be fine after that.

 

[bewing]

On 7/16/2016 at 7:11 AM, pincushionman said:

EDIT: And for all of you insisting you can fly drive whith no real problems, can you link to one of yours for him to try? That could be helpful too.

Here's one of mine. Anyone who thinks that all planes fishtail on their system or in the game -- launch this either from the runway or the grass, and tell us all how much it fishtails before it gets into the air:

http://www.virtualrealitytoursllc.com/pix/ha_juno_lr.craft

 

Edited July 23, 2016 by bewing

[DaMachinator]

Wait, planes should take off by themselves? EVERY SINGLE PLANE I've made, even ones where the front wheels are higher than the rear wheels, requires me to pitch up at least a little to get airborne.

[swjr-swis]

On 7/16/2016 at 7:02 PM, bewing said:

Basic math, which includes physics simulations, is precisely identical across all machines.

Except when it is not identical across all machines...

[swjr-swis]

59 minutes ago, DaMachinator said:

Wait, planes should take off by themselves? EVERY SINGLE PLANE I've made, even ones where the front wheels are higher than the rear wheels, requires me to pitch up at least a little to get airborne.

It's made a bit hard by the difficulties of wheels getting blocked and/or huge discrepancies in sizes between nose wheel and rear wheels, but yes, you should try to aim for a plane that will lift off all by itself as soon as it gains enough speed, without having to touch the control surfaces at all.

You can do this two ways:

1. The easy way: make sure the nose wheel is just a notch lower than the rear wheels so the entire body of the plane is aimed slightly above the horizon as it rests on the tarmac.
2. Better: give the wings a few degrees of angle of attack so they are generating lift even when you are pointed exactly at the horizon. This will allow your plane body to keep pointed much more prograde, lowering body drag by a lot, and making your plane have much better performance. As a side effect, it should also cause your plane to lift off pretty much by itself without having to pitch up, just as soon as you hit the right speed.

[Foxster]

...also a good idea to set the rear wheels only just behind the CoM. 

Edited July 28, 2016 by Foxster

[bewing]

The further behind they are, the less fishtailing you get, though. So it's a compromise on that one.

[DaMachinator]

21 minutes ago, swjr-swis said:

It's made a bit hard by the difficulties of wheels getting blocked and/or huge discrepancies in sizes between nose wheel and rear wheels, but yes, you should try to aim for a plane that will lift off all by itself as soon as it gains enough speed, without having to touch the control surfaces at all.

You can do this two ways:

1. The easy way: make sure the nose wheel is just a notch lower than the rear wheels so the entire body of the plane is aimed slightly above the horizon as it rests on the tarmac.
2. Better: give the wings a few degrees of angle of attack so they are generating lift even when you are pointed exactly at the horizon. This will allow your plane body to keep pointed much more prograde, lowering body drag by a lot, and making your plane have much better performance. As a side effect, it should also cause your plane to lift off pretty much by itself without having to pitch up, just as soon as you hit the right speed.

Most of my aircraft are complicated enough that this becomes very difficult. 

I have a tendency to make 1.25m planes with massive swept wings in a canard + elevon configuration, especially for high-altitude craft. Effectively, a supersonic powered glider with a gullwing design, with in-wing or under-wing engines, a single rear-mount engine, and a lot of fuel.

[foamyesque]

A plane that takes off by itself is nice but don't let anyone tell you it's necessary. A smidge of pitch-up to induce a rotation into takeoff is just fine, and will usually let you take off at lower speeds than auto-liftoff.

[swjr-swis]

1 hour ago, DaMachinator said:

Most of my aircraft are complicated enough that this becomes very difficult.

I have a tendency to make 1.25m planes with massive swept wings in a canard + elevon configuration, especially for high-altitude craft. Effectively, a supersonic powered glider with a gullwing design, with in-wing or under-wing engines, a single rear-mount engine, and a lot of fuel.

Oh my, that does sound very complicated indeed - practically defying the laws of physics to get such things to fly.

Obviously way more complicated than getting a plane to lift off without needing to pitch up. My apologies for over-simplifying. I am not worthy.