[1.10.x] Mark IV Spaceplane System (August 3)

[sh1pman]

5 minutes ago, Jarin said:

Hit alt+f12, go to Physics, under "Aero" select "show aero data in right click menu". Then right-click parts as you're headed down the runway and see what the individual part drags are.

About to crash:

[Jarin]

... *blinkblink* ... well, I feel kinda stupid now. Thank you, @sh1pman. In all my aero testing, I never once thought to just select the dang parts and pin them all before takeoff. You have just made my life so much easier.

Also, half throttle.

[sh1pman]

14 minutes ago, Jarin said:

... *blinkblink* ... well, I feel kinda stupid now. Thank you, @sh1pman. In all my aero testing, I never once thought to just select the dang parts and pin them all before takeoff. You have just made my life so much easier.

Also, half throttle.

yeah, I noticed that
More testing. Added more wings and engines to take off. Managed to break the sound barrier. Took screens at different speeds:

http://imgur.com/a/wzvdl

Edited February 2, 2017 by sh1pman

[Jarin]

I can confirm that the occlusion testing sphere fix took, at least. Now any part - including the very-short Mk4 probe core - placed behind the cargo bay gets unreasonable amounts of drag, rather than zero. Progress!

Edited February 2, 2017 by Jarin

[sh1pman]

4 minutes ago, Jarin said:

I can confirm that the occlusion testing sphere fix took, at least. Now any part - including the very-short Mk4 probe core - placed behind the cargo bay gets unreasonable amounts of drag. 

The fuselage before the first cargo bay also gets a significant drag increase, at least at lower speeds.

BTW, how do I embed imgur albums here? I swear I knew how to do it before.

Edited February 2, 2017 by sh1pman

[Jarin]

That plane's gonna be a flying wall of engines soon.

Also, Jiminy Christmas, you're getting nearly six thousand kilonewtons of drag on that tail.

5 minutes ago, sh1pman said:

BTW, how do I embed imgur albums here? I swear I knew how to do it before.

It broke months ago. Never fixed.

Edit: I suspect that extra drag increase on the forward fuselage is from your greatly increased pitch at low speeds, scratch that. It's probably registering the rear node as open. Which isn't as bad as the front node, but still a notable issue.

Edited February 2, 2017 by Jarin

[sh1pman]

3 minutes ago, Jarin said:

That plane's gonna be a flying wall of engines soon.

Also, Jiminy Christmas, you're getting nearly six thousand kilonewtons of drag on that tail.

It broke months ago. Never fixed.

Edit: I suspect that extra drag increase on the forward fuselage is from your greatly increased pitch at low speeds, but my derpy internet isn't even letting me load the full size images so I can't see all the numbers. How much difference are we talking?

Roughly two times more drag.
 

Also, behold:

[Jarin]

1 minute ago, sh1pman said:

Roughly two times more drag.

Yeah, I edited my post again. My internet's not the only thing a bit slow tonight, I was missing the obvious.

*CHOKE* twenty two thousand kilonewtons of drag what

Edited February 2, 2017 by Jarin

[sh1pman]

1 minute ago, Jarin said:

Yeah, I edited my post again. My internet's not the only thing a bit slow tonight, I was missing the obvious.

*CHOKE* twenty two thousand kilonewtons of drag what

That's five and a half "Mammoths" pushing my plane back. Something tells me it's not going to space today.

[Nertea]

I think the actual problem is (from much reading in the forums) that the occlusion system works by determining the drag area of the "back" of the cargo bay and subtracting that from the drag area of the "front" of the part behind it. This points to some kind of flaw in the way the object's drag cube is being calculated. However, I quite literally spent 4 hours last night testing, tweaking and testing again. I just can't seem to fix it. I guess I'll try again tonight, but I've literally rebuilt the unity file and cfg from scratch, both based off the mk3 bay now, and I have no answer. Even just forcing the drag cube to be the same as the fuselage doesn't seem to work!

About the only functional difference at the moment is the fact that the Mk4 bay opening animation is reversed compared to the Mk3 ones. That seems like it would be odd to have that cause the problem, but... I  guess? This is all weird.

 

[sh1pman]

28 minutes ago, Nertea said:

I think the actual problem is (from much reading in the forums) that the occlusion system works by determining the drag area of the "back" of the cargo bay and subtracting that from the drag area of the "front" of the part behind it. This points to some kind of flaw in the way the object's drag cube is being calculated. However, I quite literally spent 4 hours last night testing, tweaking and testing again. I just can't seem to fix it. I guess I'll try again tonight, but I've literally rebuilt the unity file and cfg from scratch, both based off the mk3 bay now, and I have no answer. Even just forcing the drag cube to be the same as the fuselage doesn't seem to work!

About the only functional difference at the moment is the fact that the Mk4 bay opening animation is reversed compared to the Mk3 ones. That seems like it would be odd to have that cause the problem, but... I  guess? This is all weird.

 

Maybe the game thinks that the cargo bay is open, when it's actually closed?

[Nertea]

18 minutes ago, sh1pman said:

Maybe the game thinks that the cargo bay is open, when it's actually closed?

Already tried that, even removed the whole cargo bay module. 

[sh1pman]

13 minutes ago, Nertea said:

Already tried that, even removed the whole cargo bay module. 

But other mk4 parts are fine, right? If there is no cargo bay module, it should behave just like every other mk4 part, like fuselage for example.

[Nertea]

Nah, it has a different cross sectional shape. In theory it should be overridden by my manually specified drag cubes, but it isn't! That's what I can't figure out. 

[Jarin]

1 hour ago, Nertea said:

Nah, it has a different cross sectional shape. In theory it should be overridden by my manually specified drag cubes, but it isn't! That's what I can't figure out. 

I really wish I knew more about the guts of this game to help troubleshoot here. Might check against the models OPT uses? I don't think there's any drag issues on those bays.

[sh1pman]

4 minutes ago, Jarin said:

I really wish I knew more about the guts of this game to help troubleshoot here. Might check against the models OPT uses? I don't think there's any drag issues on those bays.

OPT, Mk2-exp, Modular Rocket Systems, SpaceY, etc. Many mods add custom cargo bays, but I've only seen this bug with MkIV.

[Hellbrand]

any chance of a non opening cargo fuselage section? or maybe an addon for a high mount wing? or maybe even a front loading ramp/cockpit?

 

[K.Yeon]

On 03/02/2017 at 6:17 AM, Nertea said:

I think the actual problem is (from much reading in the forums) that the occlusion system works by determining the drag area of the "back" of the cargo bay and subtracting that from the drag area of the "front" of the part behind it. This points to some kind of flaw in the way the object's drag cube is being calculated. However, I quite literally spent 4 hours last night testing, tweaking and testing again. I just can't seem to fix it. I guess I'll try again tonight, but I've literally rebuilt the unity file and cfg from scratch, both based off the mk3 bay now, and I have no answer. Even just forcing the drag cube to be the same as the fuselage doesn't seem to work!

----

 

 

On 03/02/2017 at 9:11 AM, Jarin said:

I really wish I knew more about the guts of this game to help troubleshoot here. Might check against the models OPT uses? I don't think there's any drag issues on those bays.

OPT had this massive drag problem last patch as well.

It was fixed by making the cargo bay's custom drag cube same as if the part is solid.

When game makes the drag cube calculation (the one it puts in PartDatabase.cfg) for hollow/cargobay parts, it automatically made the cubes Y face have less area (to reflect the hollowness) but for some reason in order for cargo bay drag to be correct, the custom drag cube have to be 'solid' for it to work.

If you take a look at mk2 cargobay, it's custom drag cube is almost exactly the same as the mk2 fuel tank.

hope this helps!

 

Edit:
I just saw:  "Even just forcing the drag cube to be the same as the fuselage doesn't seem to work!"
I remember the depth factor in the drag cube values is very important, if it's not correct it will mess up the whole drag cube... For some of OPT hollow parts, i made solid model of the hollow part, and used it to get those values.

Edited February 6, 2017 by K.Yeon

[Nertea]

14 hours ago, K.Yeon said:

 

OPT had this massive drag problem last patch as well.

It was fixed by making the cargo bay's custom drag cube same as if the part is solid.

When game makes the drag cube calculation (the one it puts in PartDatabase.cfg) for hollow/cargobay parts, it automatically made the cubes Y face have less area (to reflect the hollowness) but for some reason in order for cargo bay drag to be correct, the custom drag cube have to be 'solid' for it to work.

If you take a look at mk2 cargobay, it's custom drag cube is almost exactly the same as the mk2 fuel tank.

hope this helps!

 

Edit:
I just saw:  "Even just forcing the drag cube to be the same as the fuselage doesn't seem to work!"
I remember the depth factor in the drag cube values is very important, if it's not correct it will mess up the whole drag cube... For some of OPT hollow parts, i made solid model of the hollow part, and used it to get those values.

Maybe I'll post some pictures when I get home, but in my case the fuselage has exactly the same proportions as the cargo bay. It's just a like a solid version. 

[sh1pman]

@Nertea So how's it going? I'd love to test some dev builds if there are any. Wish I knew how these drag calculations work, I'd mess with it myself . Just wanna help if I can.

Edited February 7, 2017 by sh1pman

[Nertea]

5 hours ago, sh1pman said:

@Nertea So how's it going? I'd love to test some dev builds if there are any. Wish I knew how these drag calculations work, I'd mess with it myself . Just wanna help if I can.

Nothing to test, I'm afraid, I have made no progress.

I'll let you know when/if I do. 

[blowfish]

On 2/6/2017 at 10:59 AM, Nertea said:

Maybe I'll post some pictures when I get home, but in my case the fuselage has exactly the same proportions as the cargo bay. It's just a like a solid version. 

Where do the drag cubes on the cargo bay come from?  Maybe there was an error in copying them?

[Nertea]

8 hours ago, blowfish said:

Where do the drag cubes on the cargo bay come from?  Maybe there was an error in copying them?

I just copy pasted them from PartDatabase.cfg entry for the Mk4 fuselage of equivalent size... I even tried pasting them for both the A and B cubes.

The database entry:

PART
{
	url = MarkIVSystem/Parts/Fuselage/mk4fuselage-3/mk4fuselage-3
	DRAG_CUBE
	{
		cube = Default, 10.29,0.6791,4.019, 10.29,0.6774,4.019, 23.47,0.9617,0.3096, 23.47,0.9611,0.2989, 19.63,0.8969,1.774, 19.63,0.8502,2.534, -0.0003667,0.002363,-0.3819, 7.87,2.522,4.108
	}
}

The part entry

	DRAG_CUBE
	{
		cube = A, 10.29,0.6791,4.019, 10.29,0.6774,4.019, 23.47,0.9617,0.3096, 23.47,0.9611,0.2989, 19.63,0.8969,1.774, 19.63,0.8502,2.534, -0.0003667,0.002363,-0.3819, 7.87,2.522,4.108
		cube = B, 10.29,0.6791,4.019, 10.29,0.6774,4.019, 23.47,0.9617,0.3096, 23.47,0.9611,0.2989, 19.63,0.8969,1.774, 19.63,0.8502,2.534, -0.0003667,0.002363,-0.3819, 7.87,2.522,4.108
	}

 

Edited February 8, 2017 by Nertea

[Nertea]

Sooo... I think I fixed it.It was B9PartSwitch procedurally re-rendering the drag cubes with fuel switching. Easily fixed when I traced it, just needed to add a leaveDragCubes flag...

@sh1pman, @Jarin Test link:

https://www.dropbox.com/s/6qsphec151nc9g0/MkIVSystem2_3_5_X2.zip?dl=0

[blowfish]

22 minutes ago, Nertea said:

Sooo... I think I fixed it.It was B9PartSwitch procedurally re-rendering the drag cubes with fuel switching. Easily fixed when I traced it, just needed to add a leaveDragCubes flag...

@sh1pman, @Jarin Test link:

https://www.dropbox.com/s/6qsphec151nc9g0/MkIVSystem2_3_5_X2.zip?dl=0

Oh, yeah probably should have warned you about that  (and probably also thought of it myself when these problems started arising).  Sorry for all the trouble!