Jump to content

[0.90] Procedural Dynamics - Procedural Wing 0.9.3 Dec 24


DYJ

Recommended Posts

One thing you can do is ALT-F2 to open the debug console and then mouse over a wing and hit o on your keyboard to print out some parameters. See if they differ from side to side. The actual wings are working for me though. It is just the control surfaces.

Link to comment
Share on other sites

  • 2 weeks later...

A possible bug found. Attaching any procedural wing to the another procedural surface results in moving the attached surface to tip of the other (that we want to attach it to). In other words - if I want to attach my elevators to a lower part of a vert stab it won't attach in the place I want but will stick to the tip of the stab instead. It's a little bit annoying, as it happens to hamper freedom of designing that the mod was supposed to raise. Could anybody do something about it? It would be so nice. :cool:

Link to comment
Share on other sites

With regards to the assymetric lift / CoM issues shown on the previous page. I've been noticing something odd with symmetry as well, and I wonder if its related:

If I attach two control surfaces, and use the tweakables menu to adjust the part thickness, the visual resolution shows that both sides have "Root" and "Tip" the same; they ought to be mirrored. By this, I mean, set the "Tip" thickness to "0.1" and leave the "Root" at 1.0. The pieces ought to be:

tip-left ROOT-LEFT ROOT-RIGHT tip-right

but what I get looks like:

TIP-LEFT root-left ROOT-RIGHT tip-right

If there's something subtly off about the mirroring calculations for control surfaces, it would explain both that symptom and the assymetric CoM or CoL for users with assymetric control surfaces.

Link to comment
Share on other sites

Without FAR installed the center of lift seems to render incorrectly at times. I can't be sure what is correct when mixing wings from stock and procedural dynamics. That is not much of a problem I can fine tune the CoL after a couple of short takeoffs. The real problem is asymmetric torque/lift/whatever. Using RCS Build Aid you can clearly see that placing symmetrical wings and control surfaces will very likely generate side to side torque. Also I've noticed inexplicable behavior when simply "pulling up". The aircraft in this case wanted to roll to the right. I don't have this problem with stock wings. Alas, wonderful mod but I had to uninstall.

Edit:

Ok, having FAR installed the way I have solved this so far, and I believe someone else earlier in the thread mentioned it, is by re-attaching the control surfaces after re-sizing them in addition to a few other things I will talk about later on. You will still have to even out the tips and roots to eliminate torque. You will notice that there are certain points where your vehicle mass changes when re-sizing the parts. You can use this fact to fine tune things. Install RCS Build Aid so you can see the torque forces. Disable snap mode and you can fine tune your wings/control surfaces to completely eliminate all torque.

My process is as follows with RCS Build Aid set to show engine torque: Attach wings and size them. Attach control surfaces and size them, make them close to a rectangle or you are going to run into problems. Now detach them and reattach them, also outside of snap mode. Notice that your placement changes the torque. If you have lateral torque, that is your red "axis" arrow from RCS Build Aid is not pointing to the side you will now need to size the tip/root to eliminate it (You can watch the actual torque number change as you hit the mass changing points of each part). Detach and reattach to be sure you are spot on. Once the red arrow showing the torque axis is pointing out the side you can maneuver the wings/control surfaces and other parts of your aircraft to achieve less than .10 kN of torque, I usually mess with it until I'm under 0.05 or preferably 0. Do attaching and sizing first, then fine tune placement. Something similar would likely work even without FAR installed.

I can confirm the above poster is correct, that the roots and wing tips do not appear to be properly mirrored when adjusting thickness. I also thing that this is causing the problems people are having as root and tip mass seem to be independent. The graphic appears symmetrical, but my guess is that the actual values are not mirrored correctly.

To reproduce this install RCS Build Aid, and your engineer type mod of choice, as well as Procedural Wings. Start a sandbox mode, enable CoG and select a solid rocket booster, add various things from the procedural wings mod and resize the tips and roots while making note of mass changes and torque changes. Try detaching and attaching various surfaces to watch the torque reset itself to the proper value except in the case of non-rectangular control surfaces. Also note that in an install without FAR the CoL seems to jump around especially when coupled with stock wings/control surfaces. I'm still not sure which CoL marker to believe if any but you can get by. Anyway I hope this helps someone.

Edited by Jikahn
Link to comment
Share on other sites

Anyone else having issues with this when using 0.23.5 and FAR? Wings don't seem to provide lift - or at least, CoL isn't affected by it.

I just flew a fresh build in .23.5 with all updates I could find, and it seems like things were fine. did you right click the wings in flight to see what the cL and cD are to check this?

Link to comment
Share on other sites

I just flew a fresh build in .23.5 with all updates I could find, and it seems like things were fine. did you right click the wings in flight to see what the cL and cD are to check this?

Good call on checking the cL. Wings are providing lift but CoL in SPH not responding to wing placement. I did a bit of random testing in the SPH, clicked a random pod, added a single fuselage and then added pWings - works as normal. Made an actual craft with fuselage, fuel tanks, cargo bay etc, added pWings - no joy this time. I'm suspecting my B9 install is funked as I now can't r-click and toggle cargo bays and I think Ferram added some mojo to the latest FAR to detect shielding/drag dependent on animation state, so I'm suspecting that it's picking the wings up as 'shielded'. Going to nuke my B9 install from orbit in the morning to verify, but for now I think it's safe to assume my bugs are PEBKAC in origin. *facepalm*

Link to comment
Share on other sites

Good call on checking the cL. Wings are providing lift but CoL in SPH not responding to wing placement. I did a bit of random testing in the SPH, clicked a random pod, added a single fuselage and then added pWings - works as normal. Made an actual craft with fuselage, fuel tanks, cargo bay etc, added pWings - no joy this time. I'm suspecting my B9 install is funked as I now can't r-click and toggle cargo bays and I think Ferram added some mojo to the latest FAR to detect shielding/drag dependent on animation state, so I'm suspecting that it's picking the wings up as 'shielded'. Going to nuke my B9 install from orbit in the morning to verify, but for now I think it's safe to assume my bugs are PEBKAC in origin. *facepalm*

Ah! Have you updated firespitter? B9 is totally dependent on the firespitter plugin for most of its functionality. Give that a shot if you've not already.

Link to comment
Share on other sites

A possible bug found. Attaching any procedural wing to the another procedural surface results in moving the attached surface to tip of the other (that we want to attach it to). In other words - if I want to attach my elevators to a lower part of a vert stab it won't attach in the place I want but will stick to the tip of the stab instead. It's a little bit annoying, as it happens to hamper freedom of designing that the mod was supposed to raise. Could anybody do something about it? It would be so nice. :cool:

This is by design. Hold down the right mouse button before you click left to place the wing, and it won't snap to the top of the P-wings.

Link to comment
Share on other sites

So, I'm having a little issue, well kind of a major issue actually. My wings keep flexing in every single plane that I make, even tiny wings flex/have really weak node strength and just flex so much in large planes that I can barely control the goddamn thing. It happens in every single craft I make that I put pWings on.

Link to comment
Share on other sites

PWing also does not set the node size at all. Perhaps addition of this will fix the wobble. I have only just started with KSP plugin coding but i could give it a try.

Btw. KJR seems to ignore PWings completely. I mean there is no debug output coming from it. I see only stuff from other parts.

Edit: Nevermind. I saw i can enable PWings in KJR's config. Then you get "[Warning]: KerbalJointReinforcement: extents could not be properly built for part Mark2 - PWing", without doubt being the consequence of PWings using SkinnedMeshRender instead of a static mesh.

Also: PWing attached to big new ARM tank -> perfectly ridgid. PWing attached to girder -> wobbly. Which is funny because i didn't even know that it was possible to get such ridgid connections with PWings at all.

Edited by DaMichel
Link to comment
Share on other sites

Been hacking the PWing code ... this came out of it (plus some fixes) :D

9mcw9pA.jpg

You get the idea. I have to redo the models though, and it's just for esthetics.

Edited by DaMichel
Link to comment
Share on other sites

Yep i like where that's going dude! maybe holla atcha boy here about which parts of the code you're changing up to get that kinda effect? Also, if you'd like to get some help with the whole texturing thing lemme know, here is my pwing texture resume`:

Javascript is disabled. View full album

Edit: @nathankell: still workin on yours hehe, been busy building dirt jumps lately. But ill have something nice before too long.

Link to comment
Share on other sites

Thanks. I didn't expect such nice fast response :)

@Dirt_Merchand: Nice planes & textures! How are you getting all of them on one craft? I'm not done with my own textures though. Usually i can come up with something good :wink:

My code uses 4 bones in total, 2 bones for leading and trailing edge respectively. This way i can move the edges instead of scaling the whole thing. That is why i said the models need to be redone. Basically it's the weightings which need to be adjusted and then everything needs to be exported as .mu again. I will do this for the B9 wing. And i'm going to as DYJ if i'm allowed to make a fork on Bitbucket.

Edit: Doing the B9 wing. Carefully adjusting things to get matching parameters.

rYsJUIV.jpg

Edited by DaMichel
Link to comment
Share on other sites

Guest
This topic is now closed to further replies.
×
×
  • Create New...