Some aero observations

[Chronosheep]

I've been experimenting a bit with the aerodynamics in 1.0.2, and found some interesting, and hopefully useful results.

1. Nosecones: All air intakes, all 1.25m nosecones, and the shielded docking port all produce the same amount of drag.

This doesn't seem quite right, since some of the parts should realistically be more aerodynamic than the others.

The 1.25m parachutes only produce slightly more drag than the nosecones.

2. Mk2 to 1.25m adapters: Contrary to what you might think, the long adapter produces (slightly) more drag than the short adapter.

3. Back-facing surfaces: A back-facing flat surface produces the same amount of drag as a forward facing flat surface.

4. EDIT: First assumed to be an aero bug, but turned out to be a rendering bug. Sorry about the confusion.

Drag arrows do not always render properly when objects are perpendicular to the air stream.

Edited May 16, 2015 by Chronosheep

[Yakuzi]

Great investigative work, have some rep! Devs, please address these issues in the next update.

Edit: about the perpendicular part drag bug. Have you tried at which angle they start producing drag? Alternatively, could it be that their drag is "added" to their parent in-line part (it doesn't look like it, but we might as well be sure)?

Edited May 16, 2015 by Yakuzi

[JokingHinge]

This desperately needs more attention, here's some rep.

[Volix]

Good things to know, and good screenshots to demonstrate them. How were the parts placed for the two screenshots under 4? Were they placed directly, or moved into place with the widgets? As far as I know things moved around using the translate widget are occluded according to their original placement (small fuel tanks radially attached then moved into a larger tank still create drag). Perhaps this is related? Strange behavior either way.

EDIT: Questions answered before post was approved (new member). Glad the bug is just visual.

Edited May 17, 2015 by Volix

[Hotblack Desiato]

wow, the bug with dragless perpendicular tanks just cries for abuse.

not as bad as the infinigliders, but close.

[cybersol]

MAJOR BUG: Most parts produce ZERO DRAG when placed perpendicular to the air stream.

I am unable to reproduce this bug in 1.0.2 stock aero, at least in a rocket version very close to your last picture. How are you attaching your tanks, as I have tried both using both surface attachment and cubics?

[SelectHalfling0]

Looking at your SLS rocket like oh.... oh my.

[Chronosheep]

I am unable to reproduce this bug in 1.0.2 stock aero, at least in a rocket version very close to your last picture. How are you attaching your tanks, as I have tried both using both surface attachment and cubics?

After some further testing, it seems like this bug is circumstantial, and I haven't yet been able to determine exactly what triggers it.

I've been using surface attachment, but using the radial attachment point will in some cases also do it.

In this case, some tanks are dragless, while others are not:

I suppose it's also possible that it's a rendering bug on the arrows. I guess it requires some more testing.

Edited May 16, 2015 by Chronosheep

[Temeter]

Activate aero stats in the debug menu, then you can check drag numbers.

[Chronosheep]

Activate aero stats in the debug menu, then you can check drag numbers.

Thankfully seems to be a rendering bug.

Will edit the first post to avoid misleading.

[cybersol]

Yeah, if I place the outer tanks just right and then offset them radially, then the drag arrows all seem to overlay the central arrow exactly. Therefore it looks like they are not producing drag, when they actually are using the action menus.

[Chronosheep]

Yeah, if I place the outer tanks just right and then offset them radially, then the drag arrows all seem to overlay the central arrow exactly. Therefore it looks like they are not producing drag, when they actually are using the action menus.

This doesn't seem to be the case. The arrows are still rendered (inside the tanks), but with zero length -- they look like red pancakes.

[NathanKell]

Note that the amount of drag the sides and back of a part make depend on mach number, so make sure all your tests are at the same mach number (the same speed at the same altitude at the same time of day, or be at about 10km, where temperatures are pretty constant).

[cybersol]

This doesn't seem to be the case. The arrows are still rendered (inside the tanks), but with zero length -- they look like red pancakes.

Yep, my bad. You are correct sir, and the drag errors appear correctly when you divert from prograde, so they are definitely in the right spot. Seems like a bug in the aero arrow code, a bad rotation matrix or cross product somewhere.

EDIT: More information. If I surface attach the side tanks using just an S rotation from default, I get the pancake/no arrows behavior. However, if I surface attach using just W they work normally.

Edited May 17, 2015 by cybersol

[Col_Jessep]

Thankfully seems to be a rendering bug.

And here I was thinking about how I could launch my space station sideways...

Good idea with the test series!

[Temeter]

Nice to see it's just a visual thing.