What does spring force in the suspensions do with auto suspension?

[At4851]

What does spring force in the suspensions do and how does it work?  The suspensions are active so how does that affect it?  

edit:  if it has an auto suspension, it should change it on its own, but it has a spring force bar so what does that do?  If you set it low enough, itll bottom out, but I cant get it to go above half way.  

Also, does the auto suspension run all the time or just when you start or the rover changes?

And is the damper the number you can find on wikipedia, where 1 = no extra bounce, higher equals really slow 1st bounce, and lower equals some extra bounces?

Edited November 10, 2017 by At4851
more info

[Spricigo]

Real life springs obey Hook's Law: [Force]=-[constant]*[distance from rest position] 

If KSP springs are anything like that [spring force]  is related to the [constant]

[At4851]

It looks like it does that, but if it has an auto suspension, what does the spring force bar change?

[bewing]

3 hours ago, At4851 said:

It looks like it does that, but if it has an auto suspension, what does the spring force bar change?

Why do you think it has an auto suspension? I see no "auto" anywhere, except in the friction settings.

Or are you running mods?

 

[At4851]

I saw that here

and also here

https://www.scribd.com/document/342641909/Readme

and I have 2 cars that are really different weights that push the springs down the same amount

[At4851]

It said

* Suspensions are smart and self-adaptive, tuning themselves to always be smooth and springy independently of vessel mass and weight distribution.

That sounded like it changed the springs like an auto suspension some cars have today

[Spricigo]

That is a guess, hopefully an educated one:

KSP suspension is very simple with just two elements:

A)A spring that respond to compression/stretching  with a force proportional to [how much the suspension is away from its rest position ] , [spring force setting].    F=-kx

B)A damper that respond to suspension movement with a force proportional to [how fast the suspension is moving] , [damper strength setting]   F=-kv

No other adjustment except those two "k".

Suspensions are 'self-adaptive'  to how much it is away from its rest position and how fast it is moving, nothing else. And they are 'smart' because that behaviour don't requires user imput/oversight.

[At4851]

Usually in cars, adaptive suspension means it changes the springs while youre driving.  Using f = kx, I have 2 cars, one has a mass of 160 kg, without the wheels, and the other is 590 kg.  

The small one, with a spring of 1, pushes down the springs half way, .1m.  Using f = kx

1600n = k * .1m

 k = 16000 n/m

The other one, using f = kx again,

5900n = 16000n/m * x

x = .368m which is farther than the suspension can go, so its bottomed out.

It should be bottomed out, but its not, its pushed down .1m also.

Im just trying to figure out what its doing.  When I saw those posts I thought it had an auto suspension, but bewing said it doesnt, so if its not doing that, what is it doing? 

[bewing]

1 hour ago, Spricigo said:

A)A spring that respond to compression/stretching  with a force proportional to [how much the suspension is away from its rest position ] , [spring force setting].    F=-kx

 

Well, not quite. The source of all the problems with bouncy suspensions comes from the fact that the PhysX library seems to have two spring equations. For positive dx, F= -kx. For negative dx, F= -2kx. The fact that it's a non-energy-conserving system is what makes everything bounce perpetually, despite all the kludges the devs have put in to try to make it stop. Which is why we need Nvidia to fix their PhysX library.

2 hours ago, At4851 said:

It said

* Suspensions are smart and self-adaptive, tuning themselves to always be smooth and springy independently of vessel mass and weight distribution.

That sounded like it changed the springs like an auto suspension some cars have today

And yes, the system does have a "smart" aspect to it. The spring constant from the slider is modified by the mass of the ship. So a suspension with the same setting will be much stiffer on a high-mass ship than a low-mass ship. This happens when you initially load a craft. Any adaptation after that is very slow.

[Oiff]

47 minutes ago, bewing said:

And yes, the system does have a "smart" aspect to it. The spring constant from the slider is modified by the mass of the ship. So a suspension with the same setting will be much stiffer on a high-mass ship than a low-mass ship. This happens when you initially load a craft. Any adaptation after that is very slow.

I had wondered about that and figured it must be based on the mass of the ship.

 

This does seem to suggest though that the usefulness of larger landing gear is down to size, travel and impact resistance rather than the ability to support the craft.

[bewing]

There is an increasing max load for each size of gear. You don't want to max out your wheel stress, or bad things happen.

 

[At4851]

Awesome, thanks!

1 hour ago, bewing said:

Any adaptation after that is very slow.

Ive been wondering, lets say you have a lander sitting at an angle, will it try to level it, mabey to straight up from the angle its sitting at or to 0 degrees?

Ive also had some cars nose dive from braking and then stay there.  Is this the Physx bug that Im seeing?

[Oiff]

Yes I have exceeded the maximum load on wheels before, they poof resulting in unplanned disassembly.

 

The gear should attempt to reach equilibrium between the spring forces and the COG of the craft.  For example if you land on a slope the legs on the lower part will have more weight on them so will compress more.

In KSP though sometimes the gear just doesn’t find equilibrium and bounces around continuously.  Using SAS can aid the dampers here (or on one of my planes lowering the ramp worked well for this!).

 

I think my rovers have always returned to being level after braking.

[Spricigo]

5 hours ago, bewing said:

Well, not quite.

For an educated guess, close enough. 

Any idea of how far off is my supposition in regatd to dampers?

 

[bewing]

3 hours ago, Spricigo said:

Any idea of how far off is my supposition in regatd to dampers?

At4851 was correct in the OP -- 1 = critically damped.

[At4851]

9 hours ago, Oiff said:

The gear should attempt to reach equilibrium between the spring forces and the COG of the craft.  For example if you land on a slope the legs on the lower part will have more weight on them so will compress more.

I think it should yeah.  Ive landed on a slope and the lower legs were longer and the upper legs were pushed down more.  I wonder if part of the smartness of it is it auto levels since the post said its independent of weight distribution or if its the Physx issues coming through.  Im hoping if I can figure out exactly what its doing I can make it more stable when the bounciness comes through.  Ive seen a lot of things where its not returning to where I expect.  I wonder if its the 2 equations causing that

Does anyone know how the dampers were set up on the lunar lander?

[Oiff]

The thing is you don’t even need landing gear to land on the mun, you can just land nice and gentle on whatever is on the bottom.

 

Tourist missions to the moon I don’t use legs, just a pod, hitchhiker fuel tank and engine.  Let the engine graze the surface to complete the contract and blast off again.

I also had a minmus survey craft that had three winglets around the engine that it would land on while I hopped around collecting science.

[At4851]

Thats true, but if you want to land on a slope it could be a problem

[bewing]

Landing legs do not adjust to try to make you point straight up. They adjust until they are all the same length. So if you are on a slope, you will always end up pointing normal to the slope or worse. So using winglets to land on is always better, because they don't have compressible suspensions, and they have a higher impact tolerance.

 

[At4851]

So the odd leaning must be the unity bug then