[WIP] R.E.L Skylon C2. Alpha Released. FAR config broken. (08 Dec 2014)

[CaptainKipard]

The fix is easy, you just have to make the cylinder and rod a tiny bit longer:

It's not that simple. I had to decompress the shock absorber as shown in that screenshot above. This shortens the hudraulics

One final thought: The mechanism at the top of the the leg looks suspiciously like it's designed to spread out left and right, essentially shortening the leg as it goes into the bay.

The blueprints and photos I have show it to be a rigid structure without any possibility of movement.

Duxford

I had to check on google what that is and now I'm jealous. I may take you up on your offer one day, right now though it's not necessary simply because the gear will need to be different anyway. The actual landing gear photos can only take me so far, but Skylon landing gear retracts to about 100 degrees away from vertical, which is more than Concorde which is pretty much horizontal when retracted. This presents some difficulties in rigging, and making it look plausible.

Edited August 9, 2014 by Cpt. Kipard

[ZodiusInfuser]

It's not that simple. I had to decompress the shock absorber as shown in that screenshot above. This shortens the hudraulics

Sorry, but I don't understand the difficulty. If you extend the compartment back a bit more as my image shows, the piston becomes longer without affecting the current position of the gear, suspension, or reducing the angle range from 100 degrees. Does extending it affect how other parts of the craft come together?

Even in the official Skylon User Manual on page 20 (28 in the pdf) there is a picture of the craft with its piston at a similar angle. Given the choice extending the piston like that is much more realistic than a telescopic one, particularly for this particular application.

[CaptainKipard]

Sorry, but I don't understand the difficulty. If you extend the compartment back a bit more as my image shows, the piston becomes longer without affecting the current position of the gear, suspension, or reducing the angle range from 100 degrees. Does extending it affect how other parts of the craft come together?

Firstly your image is misleading because it doesn't account for the size of the pivots. That means the actual overlap between the cylinder and piston is not as long as that image suggests.

Secondly placing the cylinder pivot further back not only allows for a longer piston/cylinder, but also increases the distance that hydraulic needs to extend to push the gear all the way. That means that it effectively makes no difference at all.

Even in the official Skylon User Manual on page 20 (28 in the pdf) there is a picture of the craft with its piston at a similar angle. Given the choice extending the piston like that is much more realistic than a telescopic one, particularly for this particular application.

So the reasonable conclusion is that either the artist of that model just phoned it in, or the proportions of my model are wrong. I don't have any blueprints for the Skylon gear because they don't exist in the public domain.

edit.

Yeah I can see my proportions are way of. I'll need to think about this.

Edited August 9, 2014 by Cpt. Kipard

[ZodiusInfuser]

Firstly your image is misleading because it doesn't account for the size of the pivots. That means the actual overlap between the cylinder and piston is not as long as that image suggests.

Secondly placing the cylinder pivot further back not only allows for a longer piston/cylinder, but also increases the distance that hydraulic needs to extend to push the gear all the way. That means that it effectively makes no difference at all.

The first is a valid point. I quickly mocked up that image during my lunch break, to see whether moving the piston along would cause the overlap to be reduced. As for the second point, there's probably some math that can be done to confirm or deny whether there is a solution, but I wouldn't know it. Without actually having the model I couldn't say either way, but I'd be curious to compare yours to that Lo-fi mocked up, so see what was actually altered, rather than just dismissing it.

So the reasonable conclusion is that either the artist of that model just phoned it in, or the proportions of my model are wrong. I don't have any blueprints for the Skylon gear because they don't exist in the public domain.

edit.

Yeah I can see my proportions are way of. I'll need to think about this.

It could well be an artist representation, and not the actual design, as I also saw this image that shows the gear flipped 180 degrees:

[lo-fi]

After having spoken to you in Orbitus' thread, my Unity crashed (which is the first time it's ever done so) and my entire rigged main gear disappeared. Annoying, but gives me a push to put it back together with a proper model if you'd like to send over. I'll rig it up and hand back over with source, plugin, parts and Unity package so at least you'll have something workable until Orbitus gets his method working. It gives options either way and adds an open-source landing gear module to the mix.

[stephm]

hello !

I see on other post you need example with PISTON:

I have reworked the rear gear Mockup with two type animations :

• one gear bay is now managed is blender by Armature (manage the C Bay or trappe_C in blender. )
  
• And Piston+cylinder to simulate main Actuator for Gear.
  

Archive contain the blender File, and the FBX export and UNITY scene.

REAR MOCKUP

Nested Cylinder-(blender file)

In blender for the Export TAB in FBX

If you click under Export Fbx on option : Operator Preset

By clicking this option a list appears , and I have made a preset for you :Export Unity

click this preset and the tab is configured for UNITY.

Now look how part are managed in blender by animation (part have KEY), and look how

the actuator is parented is contain no animation just Constraint, parent move The cylinder And Piston

In unity it's automatic key are created

For the ARMATURE, I have choose something a bit complicated, but in unity almost all the animation is recreated

you have absolutely nothing to do .

In this Case they are just one position not managed on static for the Trappe_C object.

• in UNITY click on the Trappe_C element and notice in the Inspector the Rotation line is bold because
  
• I have tweaked the static position
  
• Right click Rotation , and chose revert to value Prefab
  
• The static Position is not the same as blender ! (in blender Trappe_C is constantly managed by armature)
  
• Run animation and look how the Trappe_C move during animation
  

to Correct the glitch between static and starting animation just tweak Y rotation to manage the same position

as the main starting animation in UNITY.

And that all, all Key are created automatically

But if you don't want export the armature, it's easy in blender use position created by armature to create

Key on the part after that delete the armature.

edit :

Of course Actuator don't need animation because, in part CFG you can use MODULE (Fxlookat..), but it's just

an example here, and for nested piston you have just to ADD the intermediate animation for the nested piston.

Keep the two main part Piston and Cylinder managed by Constraint in blender, and for the nested part

create key and position manualy to match the whole movement, it's easy only on one AXIS (Z for example, because

it's the Fxlookat axis).

I have forget this example sorry . (remake a nested piston for you )

Reworked the main actuator is now in 3 part (one nested piston) link actualized

Edit 11/08/14

Added link for an another example of nested hydraulic cylinder ! (blender format only)

I'm better to do this in 3DS by scrip or function, but for blender in this example I use again

a Damper constraint an the intermediates cylinders are managed by KEYframe, but they are

under Damper constraint too because you have only one axis to manage when you create

the intermediate Key frame.

@+

steph

Edited August 11, 2014 by stephm

[lo-fi]

Nice work, Steph!

@Cpt. K:

Just took a peek over here and found the render in the OP: http://forum.kerbalspaceprogram.com/threads/72802-Main-Landing-Gear-Rigging-Animating

Yep, the hydraulics will be fine - very simple to set up.

Regarding the nose gear alignment issues, I can see why you've had a challenge getting the proportions right. I've found so many different variants on the theme I wouldn't know which to go with myself! I think (actually I'm sure) the conceptual artists don't think about the finer points of the mechanical geometry provided it looks about right for that particular drawing. Currently the bay doors are too short or the gear is too long depending on how you want to look at it. :/

Is the nose gear you posted also modelled in the fully compressed state?

Edited August 11, 2014 by lo-fi

[stephm]

hi! thank lo-fi, I hope the example are not to bad and will be useful.

-The state of the gear seem to be compressed on ground by vessel mass, or fully decompressed.

if find that the proportion of the skylon are good according to the first picture

on this post, it's look great ..

[CaptainKipard]

Thanks Stephm

Is the nose gear you posted also modelled in the fully compressed state?

The state of the gear seem to be compressed on ground by vessel mass, or fully decompressed.

Yes, I'm afraid so. I need some accurate blueprints if I'm going to redo it all correctly. The main gear I posed in the other thread today is also fully compressed.

Looking over my reference material I can see that the torque arms are much smaller than I made them. Actually someone already pointed this out before. I'll try shrinking it. Maybe that will give me some options.

[stephm]

hi !

Cpt.kipard: Can I have download you rear gear sample to see how is he ? (not sure if my grammar is good it's a question !!)

(see the download link on 747 pluggin gear)

-Just to try and understand the very interesting lo-fi's explanation ! (it's about the two wheel collider)

@+

[CaptainKipard]

I don't understand what you're asking.

If you want to download it then go ahead.

[stephm]

hi!

ah ok and sorry for bad English !

An another usefull link credit to dielenator

Method used is stretched armature, you don't have need key for the intermediate piston.

On you're rear gear, I prepare this example and to see if the FBX record the full animation in KEY !

And this is the gear rigged with two type off actuator :

MainGear4Tutorials.zip

-simple constraint Damped tract (or track to it's you're choice)

-nested with the dielenator method, but only one bone is needed .

- In blender notice the two little empty object :

• attach_strech_Armature
  
• Controller
  

these two empty, drive the entire nested cylinder, and if you unlink both of them you have

possibility to place or duplicate the nested cylinder on a different position.

In unity one glitch only the mesh : GearCylinderD_NestedB, in static position come

in a weird position, but if you start the animation everything is good and recorded like in blender.

You have just to tweak position of this mesh and voilà !.

In my opinion in real plane they don't use big cylinder to move gear, too big and not enough resistant !

they use animated strut, and the cylinder gear is only used for moving gear into the raised position (or lowered), no reel constraint

on this piston, just the weight of the gear .

In fact Constraint are redirected in strut.

steph

Edited August 13, 2014 by stephm

[stephm]

hi !

I'm back here, because for my personal learning too

I'm looking in the source code of Firespitter pluggin, to see How too !!

I think they are lots of possibility with this pluggin and instead of understand how it work

I find better to try the pluggin with you're main Gear and use it!

If Snjo look here and confirm !

-possibility multiple wheel and collider (wheelColliderName = "wheelCollider,wheelcollider1 etc.)

-Wheelcollider parameters are accessible by cfg (if I look code )

-And many other feature very interrsting to test !

So I'm starting with you're main landing gear and the module FSwheel (firespitter)

a new test !

@+

EDIT : 20h37

I have good new , the gear work (Module Fswheel firespitter):

the gear use three wheel collider :

1 for the AxleBeam animation up and down, and for the the torque link movement

and one on each wheel, wheel are independent !

AxleBeam is controlled by a Fxlookatconstraint , the target is the rear wheel !

For now the plane roll very well on runway I need to conduct some test

on ground to see how the gear react and if it's realistic

steph

Edited August 15, 2014 by stephm

[stephm]

Hello !

After rolling on the land near mountain (30 minutes),I find the gear enough realistic for KSP purpose !!

it's again the same gear , (not symmetrical), but there is no problem for rolling the whole thing :

the file contain the folder : Skylon_nose_test , put this folder in Gamedata

the two blender recent file.

And skylon_fullConfig.craft the thing on the picture , put the file in SPH folder !!

the archive -> see next post for upgraded archive !

My suggestion, leave the runway and roll near the mountain and look the gear .

The advantage of Firspitter module (Fswheel) give you the ability to modify the wheelcollider directly

with the cfg file of the part.

In SKMainGear_FRS.cfr look at the Module party section :

MODULE

{

name = FSwheel

for example :

suspensionSpring = 15.0 -> this parameter change the spring compression

wheelColliderRadius = 0.50 -> it's diameter of the wheelcollider

}

And all function are not represented in the cfg, you have the possibility to have sound and smoke and light !!!!

Try it !!

Thing to do add an inclined position for the AxleBeam object to mimic the gear without charge (in the air)

and a little animation to retrieve the horizontal position when the gear raise .

(but I'm not sure, need to test if it's possible...)

voilà

@+

steph

Edited August 16, 2014 by stephm

[CaptainKipard]

Merci Beaucoup Mon Ami.

I have started remodelling the gear this morning. I will incorporate your work after I'm done. I'm a little confused about the axle beam situation. Orbitus says he's not working on his plugin anymore. What is the situation there? I need to know this before I start remodelling the rear gear.

edit

I have tested the gear a little bit. I have three issues.

1. The wheels rotate too slowly

2. The front gear doesn't compress under the weight of the craft.

3. After retracting the front gear the wheel inside snaps into a compressed state. When deploying it snaps back into the uncompressed state and then comes out.

Besides that I'm very happy. Thankyou.

Edited August 15, 2014 by Cpt. Kipard

[stephm]

hello !

and this why the firespitter pluggin is very good is tweakable :

For problem 1 :

For your training and because it's cfg manipulation, in Fswheel code we have this KSPfiedl !

[KSPField] // Adjust the roll speed of the wheel if it looks off
public float [B]rotationAdjustment[/B] = 1f 

so in cfg in the Fswheel module section add this :

rotationAdjustment = 2.0 (place it just under retractedDrag = 0.1 for example) (correct the problem 1)

the cfg is already configured but if you want try some test ....

(Tested, and it's look like better)

For problem 3 :

I remember you how wheelcollider work :

• oriented Yup
  
• and it move down with the suspension distance parameter
  
• he need to start in Full compressed position to work !
  

(Fswheel use the UNITY collider physic, same as stock ModuleLandingGear !)

this the rule, but to fight this starting position (full compressed), and I know the suspensionParent

don't like direct animation, i'll find a way to have a correct position in every state !

It's just a little animation :

key 1 position Fully Compressed , key 2 position Fully decompressed

same as wheelcollider without charge !(correct the problem 3)

this animation is not on the suspensionParent (look in blender animation is on Mover_SUSparent)

---

And other tweak to approach a realistic movement , I 'll have tweaked two parameter :

• wheelColliderSuspensionDistance = 0.30 (augmented give more range movement on ground)
  
• suspensionSpring = 10 (reduced a bit to have a suspension more dynamic correct the problem 2)
  

it's look very good now

---

this is the new configuration I have mad to test an another bug (and it's very bad !!)

---

The G sequence Action key for gear broke the gear or do weird thing in game :

tell me if you have the same bug, I have absolutely now idea where's the problem !!!!!!! need more advice here !

New archive for test : Firespitter_Test.rar

(erase all previously file, and replace with this new test file)

For rolling I think is good now, I have just to find why the gear animation do weird thing !!!

steph

@+

Edited August 16, 2014 by stephm

[CaptainKipard]

Thankyou Stephm. Could you add size 2 stack nodes to the gear and remove surface attachment please? Maybe it will be more rigid, and wont flip over.

[stephm]

Could you add size 2 stack nodes to the gear and remove surface attachment please.

-I have a little problem never do stack thing yet...

I know it's strange I make a big landing gear and, I don't know how to for stack node !

But it's just a cfg manipulation I'm right ?

(but don't worry I'm looking how to )

steph

[lo-fi]

I had some diabolical problems animating the nose gear, I was assuming it was to do the the blender to 3DS export. Does the animation play ok in unity?

What are your grip settings for the wheelcolliders, Steph?

Was this all modelled in Blender originally, Cpt. Kipard?

[stephm]

@lo-fi

I presume all model's Cpt.kipard come from blender, and I work only

with blender (I have 3ds 2011 on an other work station, but never use yet for here)

Yes no problem at all for FBX to UNITY:

1 : for object with animation axis are not important

2 : for object managed by cfg you have to keep the rule axis needed

by the module.

In UNITY wheelcollider work Yup so you have to put in blender the same orientation

than unity .

As far I know there are no other complication ..

what do you mean by "grip", in just set up Yup both unity and blender.

Is declared In blender by an empty named wheelCollider.

For animation all key are from blender.

I hope that will help you !

steph

[lo-fi]

I meant the grip settings for the wheelcollider. The defaults are ridiculous: 20000. Its like having a wheel made of Velcro!

[stephm]

ha ok the parameter so for the rear gear I have this :

MODULE

{

name = FSwheel

wheelColliderName = wheelCollider

boundsCollider = Bounds

wheelMeshName = Wheel

suspensionParentName = suspensionParent

numberOfWheels = 3

wheelRotationAxis = 1,0,0

disableColliderWhenRetracted = false

//disableColliderWhenRetracting = true

//disableColliderAtAnimTime = 0.01

animationName = SkylonFRS_ANI

customAnimationSpeed = 1

brakeTorque = 8

brakeSpeed = 0.3

hasMotor = false

motorEnabled = false

overrideModelFrictionValues = true

overrideModelSpringValues = true

overrideSuspensionDistance = true

forwardsStiffness = 0.2

forwardsExtremumSlip = 1.0

forwardsExtremumValue = 5000.0

forwardsAsymptoteSlip = 2.0

forwardsAsymptoteValue = 2000.0

sidewaysStiffness = 0.2

sidewaysExtremumSlip = 1.0

sidewaysExtremumValue = 5000.0

sidewaysAsymptoteSlip = 2.0

sidewaysAsymptoteValue = 2000.0

wheelColliderMass = 0.025

wheelColliderRadius = 0.51

wheelColliderSuspensionDistance = 0.30

suspensionSpring = 10

suspensionDamper = 0.1

suspensionTargetPosition = 0.0

deployedDrag = 0.6

retractedDrag = 0.1

rotationAdjustment = 2

}

In yellow this parameter according unity doc is a multiplier

set to 0 he kill the value in blue ! (the defaut parameter are to hight in KSp in my opinion !)

Starting with value around 5000 is better an if the gear stay stick on the ground

change the value to a lower value !

@Cpt.Kipard : it's ok for node, I know how to I prepare a new model with only the node_stack_top

back later

[CaptainKipard]

Was this all modelled in Blender originally, Cpt. Kipard?

yes4567890

[stephm]

hello I'm back

(need to learn stack thing and how they work, it's ok , and fast lol)

this is the new model the rear part have no mesh collider and only one node stack up.

delete all files and replace .

the Mockup aircraft is build now with two stock connector on each side so it's easy to put the gear on.

and in the right place.

Firespitter_Test.rar

On runway the gear work now, no problem with animation (ah a little glitch , between the static position and start animation)

a shift is visible, I don't know if it caused by node or other .

Nothing was modified on animation and position, again an another glitch.

And I don't why the animation bug when the gear have a mesh collider.

that work with the stock module , but with Fswheel not , or I have missed something in the firespitter doc.

steph

EDIT:

I have a little anxiety concerning the weight of the craft, because with only one node the gear wobble du to Mass

so now the node is aligned on the wheelcollider axis to keep Force aligned, and even that the need strut to

stay in place ... May two node ?

@+

[CaptainKipard]

Two nodes wont help. KSP doesn't allow node loops. In other words even if you have two nodes, only one of them will attach.

I'm pretty sure the reason for the big wobble is that you've attached the small girder segment to the surface of the tank. I have created my own test craft with only node attachments (using my structural hubs) and the amount of wobbling is a lot smaller, and virtually non existent if you handle the craft realistically, i.e. NOT turning wildly. Problem solved (mostly). When it's time to created the craft file for the final release we might try some strong struts, but I doubt we'll need them.

Next problem is, we need to somehow remove breaking force from the nose gear. Only the main gear should break, otherwise the craft does a somersault. I guess BrakeTorque needs to be lowered?

Edited August 16, 2014 by Cpt. Kipard