Wheesley Drag bug?

[KerikBalm]

Has anyone else noticed that Wheesley engines, regardless of whether they are attached to a 1.25m or not?

I build a panther powered plane, at over 300 m/s (no afterburner), drag is only about 4kn:

At about half that speed, switching to wheesley resulted in over 50kn of drag!

anyone else have this problem?

[linuxgurugamer]

You need to compare them at the same speed

[KerikBalm]

Hint: increasing speed does not decrease drag, drag increases roughly according to v^2. When getting into high machs, it starts to diverge.

The wheesleys couldn't get going much above 200 because of the 50+ kN of drag, meanwhile the panther's in dry mode blew right past 200 shortly after taking off (the test plane used had a very high TWR).

I watched the drag as the speed increased, at around the same speed, it was well under 4kN, around 2kN iirc.

Drag doesn't go from 50+ kN at 180, to 4 kN at 330

Edited February 2, 2020 by KerikBalm

[Vanamonde]

There could be several reasons for this, but the question has been moved to modded support for the time being. 

[OHara]

18 hours ago, KerikBalm said:

anyone else have this problem?

Not me; I looked but I don't see the problem in my install of 1.8.1 (nor earlier versions). 

For me, the drag-debug information in that right-click menu says at YP and YN : 1.25m², or less depending on to what I attach the Wheesley.   Your screenshots showed 4--5m². My KSP also shows a smaller effective areas for the sides of the Wheesley, XP : 1.11m², etc.   Maybe a mod or its leftovers has given KSP a different set of dimensions for the Wheesley? The PartDatabase.cfg holds the numbers in hard-to-read un-commented form, as in the spoiler below.

Spoiler

url = Squad/Parts/Engine/jetEngines/jetEngineBasic/JetEngine
DRAG_CUBE {
cube = A, 1.249,0.7622,0.7222, 1.249,0.7622,0.7222, 1.214,0.9558,0.1621, 1.213,0.3323,1.428, 1.249,0.7608,0.7222, 1.249,0.7621,0.7222, 0,0.3405,-3.576E-07, 1.25,1.39,1.25
cube = B, 1.114,0.7593,0.7222, 1.114,0.7593,0.7222, 1.214,0.9558,0.1621, 1.213,0.3323,1.428, 1.114,0.7584,0.7222, 1.114,0.7599,0.7222, 0,0.3405,-3.576E-07, 1.25,1.39,1.25
}

 

  Version 1.8.0 had a bug where PartDatabase.cfg was not being generated correctly, but no-one noticed any problem then with the Wheesley in particular (and we looked for the sake of the bug report).

[KerikBalm]

2 hours ago, OHara said:

Not me; I looked but I don't see the problem in my install of 1.8.1 (nor earlier versions). 

For me, the drag-debug information in that right-click menu says at YP and YN : 1.25m², or less depending on to what I attach the Wheesley.   Your screenshots showed 4--5m². My KSP also shows a smaller effective areas for the sides of the Wheesley, XP : 1.11m², etc.   Maybe a mod or its leftovers has given KSP a different set of dimensions for the Wheesley? The PartDatabase.cfg holds the numbers in hard-to-read un-commented form, as in the spoiler below.

  Hide contents

url = Squad/Parts/Engine/jetEngines/jetEngineBasic/JetEngine
DRAG_CUBE {
cube = A, 1.249,0.7622,0.7222, 1.249,0.7622,0.7222, 1.214,0.9558,0.1621, 1.213,0.3323,1.428, 1.249,0.7608,0.7222, 1.249,0.7621,0.7222, 0,0.3405,-3.576E-07, 1.25,1.39,1.25
cube = B, 1.114,0.7593,0.7222, 1.114,0.7593,0.7222, 1.214,0.9558,0.1621, 1.213,0.3323,1.428, 1.114,0.7584,0.7222, 1.114,0.7599,0.7222, 0,0.3405,-3.576E-07, 1.25,1.39,1.25
}

 

  Version 1.8.0 had a bug where PartDatabase.cfg was not being generated correctly, but no-one noticed any problem then with the Wheesley in particular (and we looked for the sake of the bug report).

Hmmm, I remember there was something wrong with the terrier according to someones post (it should have been shielded, but was overheating).

I'm not using any mods except for scatterer right now, so... maybe I'll just try redownloading

*edit*, so I looked in my file, and this is what I've got:

Quote

    url = Squad/Parts/Engine/jetEngines/jetEngineBasic/JetEngine
    DRAG_CUBE
    {
        cube = A, 1.376,0.8489,2.833, 1.376,0.7819,2.833, 5.082,0.9755,0.2241, 5.488,0.9774,0.2259, 1.327,0.8279,2.7, 1.327,0.8254,2.7, 0,-0.01703,0, 2.633,0.25,2.5
        cube = B, 1.376,0.8489,2.833, 1.376,0.7819,2.833, 5.082,0.9755,0.2241, 5.488,0.9774,0.2259, 1.327,0.8279,2.7, 1.327,0.8254,2.7, 0,-0.01703,0, 2.633,0.25,2.5
    }

 

I wonder why its so different

Edited February 3, 2020 by KerikBalm

[OHara]

Quicker than re-downloading, if you remove PartDatabase.cfg, KSP will regenerate it from the models of the parts.

Somehow your current PartDatabase thinks the Wheesley is 2.5 meters in diameter and 25cm long, as if KSP looked at the wrong model.

I have sympathy for the lego-style method they used starting in KSP 1.0.5, computing components of drag from the shape of each part and combining the drag of parts in a simplified way.  But, the long list of numbers in which the components are stores (decoder in the spoiler here) makes it look very mysterious and obscures problems from modders and Squad themselves, so bugs are not noticed until the game is out.  For purposes of reentry heating, KSP still thinks the Terrier is 21 meters long as you mentioned.