Is CoL being calculated correctly?

[Redshift OTF]

So, I'm tweaking my Space Van concept so it doesn't flip out when it's near empty of fuel and re-entering the atmosphere. I'm adding more wing on the back of the plane so the CoL matches the CoM however I noticed this when adding wing pieces:

Without adding the piece I have this:

Adding the piece at the back hardly moves the CoL at all:

However if I add it here the CoL moves backwards so it is in line with the CoM:

Shouldn't the first placement be more effective than the second? Symmetry is enabled on both placements and I'm not using any part clipping for this placement, (not that I think that would matter).

[Redshift OTF]

Here's a more obvious example.

CoL is in-line with CoM:

Placing more wing at the back doesn't change the CoL:

Placing more wing in this configuration changes the CoL significantly:

Am I going crazy?

[gbleck]

I think your first wing example has the wing on sideways. I'm not sure but I think the sp wings may be directional.

[Redshift OTF]

Directional? Thanks for the theory although I wasn't aware that wings had to be placed in a certain direction to generate lift. That doesn't strike me as being a good thing, lol.

[Franklin]

Is CoL impacted more with a thinner wing (example vs. a shorter wing (example A), rather than just more wing parts in general?

Similar to something like this: http://www.pilotfriend.com/training/flight_training/aero/drag_red.htm

[Rhomphaia]

Directional? Thanks for the theory although I wasn't aware that wings had to be placed in a certain direction to generate lift. That doesn't strike me as being a good thing, lol.

Not a theory, just the way the stock aero model works. wing pieces that are end on to the air flow don't generate any lift.

[Red Iron Crown]

Wow, learn something new every day. Am I right in thinking the correct orientation is with the attachment nodes to the sides rather than forward or aft?

This information seems to really limit what can be made usefully with wing parts.

Edit: Behavior confirmed. Pretty sure half the wing parts I've used on my planes are functionally useless unless the plane is flying sideways. :/

A craft with no lift.

Add wings with nodes to the side, lift is added.

Add wings with nodes to the front, no lift.

Edited November 5, 2014 by Red Iron Crown

[Jimbimbibble]

If you get FAR then wings will be modeled correctly and you won't get weird effects like this.

[Redshift OTF]

Thanks for the confirmation guys. OMG, this can't be right. Half the wing I've placed on vessels might well be useless now. I hope Squad will fix this!

[Rickenbacker]

That's odd. Though to be fair, neither of those wings should add any lift whatsoever, as they're placed on and partly inside already existing wings .

[Redshift OTF]

That's very true in my case. It's pretty likely they will stop that if they update the aerodynamics model in future updates. In Red Crowns example it is very wrong.

[numerobis]

Real airfoils have a direction too. Put your Cessna wing on backwards and you'll have a bad time of it.

It would be handy if the direction were clearly indicated, true.

[cantab]

This information seems to really limit what can be made usefully with wing parts.

That wings are orientation-sensitive is why there are pairs of parts that seem nearly identical, such as two equal-sized oblong wing pieces - one has the long dimension fore-aft, the other has the short dimension fore-aft, and you need to keep them that way.

[Bilfr3d]

I think we need to uhh..... TEST the craft... yes that's right maybe you could upload it so we can TEST it? nothing else but TESTING of course

[Red Iron Crown]

Real airfoils have a direction too. Put your Cessna wing on backwards and you'll have a bad time of it.

It would be handy if the direction were clearly indicated, true.

Cessna wings aren't made Lego-style out of modular parts, either. I agree that the lack of any indicator or documentation of how it works in game is pretty frustrating, there's no expectation that orientation matters because the parts are visually symmetric.

That wings are orientation-sensitive is why there are pairs of parts that seem nearly identical, such as two equal-sized oblong wing pieces - one has the long dimension fore-aft, the other has the short dimension fore-aft, and you need to keep them that way.

I get that now, though I always assumed the pairs of parts were to allow differing node orientations, not that the orientation of the part itself mattered so much. It is still a bit limiting, the design pictured below has nearly half of its wing area made useless by this rule and I don't see a way it could be done with wing parts only:

Guess I need to rethink how I make wings, or install Procedural Wings again.

[pellinor]

Do you know these small thrust arrows on engine or RCS nozzles? They are probably done by the RCS build aid mod.

It would be useful to show lift arrows in the same fashion on every lifting surface. So a wing piece put sideways would show up as a small blue ball without an arrow. If the arrow contains a magnitude, you would also notice if the actual lift rating of a surface is different from what its mesh model would suggest.

[cantab]

Oooh, fair point. On looking, it seems all the wings are designed to be attached to the parent part on their sides. Which I suppose if this direction-sensitivity must exist is the sensible way for them to be. To fix one wing piece behind another and have both generating lift will require an intermediate structural piece to provide that side attachement.

For example

The wing C gives no lift in forward flight, but provides a side for the rear delta and the wing A to attach to, meaning the rear delta does give lift. Tightened up to remove the gaps this will look good and won't add much weight or drag, but it might affect aerodynamics when the plane sideslips. I-beams would avoid that issue but be more visible and add more mass and drag. Stacks of cuboct struts would give no mass or drag penalty but spike the part count.

[Red Iron Crown]

I think it might be easier to just stick to wing parts attached the proper way moving forward and not use that style of attachment. I'm too stingy with mass and drag to resort to the (clever) workaround to the attachment issue.

Between learning about this wing orientation thing and numerobis' discovery that two half size Mk2 tanks have half the drag of a full size one, most of my spaceplane designs are going in the circular file.

[Wanderfound]

TLDR version: is stock aero broken, producing many completely ridiculous results?

Why, yes it is.

Fortunately Squad know this and are working on fixing it. And, in the meantime, Ferram to the rescue.

[numerobis]

Red Iron Crown: you want mass-efficient lift? Take a control surface (any of them really, but the higher lift:mass ratio the better). Tilt it 60 degrees. Use that as a wing -- turn off all three control inputs. With one unit of lift per four tonnes of mass, you'll start to float off the runway at 50 m/s.

You can tilt it more to get even more lift, but I prefer to keep to 60 so that the lift increases as my angle of attack increases, until I'm at 30 degrees AoA.

[KerrMü]

thx guys, I didn´t know that either, but it explains a few problems I had with my planes in the past.

[Redshift OTF]

TLDR version: is stock aero broken, producing many completely ridiculous results?

Why, yes it is.

Fortunately Squad know this and are working on fixing it. And, in the meantime, Ferram to the rescue.

I think there is a difference between the stock aerodynamics being a bit funky and wing surfaces having zero aerodynamic effect when rotated 90 degrees. No offence

[Red Iron Crown]

Red Iron Crown: you want mass-efficient lift? Take a control surface (any of them really, but the higher lift:mass ratio the better). Tilt it 60 degrees. Use that as a wing -- turn off all three control inputs. With one unit of lift per four tonnes of mass, you'll start to float off the runway at 50 m/s.

You can tilt it more to get even more lift, but I prefer to keep to 60 so that the lift increases as my angle of attack increases, until I'm at 30 degrees AoA.

If I was a purely by-the-numbers guy I might do that (and use command seats for all manned missions, too). It just has two problems: 1. It looks horrible, I like my spacecraft to look at least vaguely plausible. 2. It's hard to avoid the infiniglide exploit when using control surfaces for primary lift, even if they're disabled.

[numerobis]

I've yet to get a clear explanation of the physics of infiniglide. Do you know of one?

There's an exploit I worked out a while ago about how you get forward force when activating a control surface, but it requires activating control surfaces.

[Red Iron Crown]

I've yet to get a clear explanation of the physics of infiniglide. Do you know of one?

There's an exploit I worked out a while ago about how you get forward force when activating a control surface, but it requires activating control surfaces.

Not overly clear, but basically it boils down to control surfaces generating thrust when changing AoA. The effect can be reduced by disabling the control surfaces but it is still present to a degree (see example here, SAS trying to keep a tail-sitting VTOL pointed up during descent is enough to keep it aloft even with disabled control surfaces).

The effect can be exploited to do some interesting things, but it's a bug to my mind that feels a bit cheaty.