Spring-mass damper

[thedmd]

Hello.

I wonder where is the difference between regular spring-mass damper quations and NewtonCalculateSpringDamperAcceleration() function. Can You explain me that?

[Julio Jerez]

no difference, the joint one is special in the sence the it only can be used to calculate the joint acceleration, you can use the normal one for the joint, but the one in the joint can only be used in a joint callback.

[thedmd]

Thanks for fast replay.

So if it evaluate regular Hooke's Law. is time delta even used for something? Maybe it will be used in future?

[Julio Jerez]

The function obey regular Hook law, you can ignored and use your own if you like.

[thedmd]

Ok. I thounght that this time value were used to increase stabilitiy in some way. Bud since there is no any feedback loop I think this cannot be done (at least not in an easy way).

Thanks for replay.

[Julio Jerez]

the time used for stability, check out implicit spring mas systems.
you can set the damper to zero it will still oscilate like a perfect stable harmonius system.
The time take out the extra energies added stepping with a numerical integration.

although some expert do think and spread the idea nothing in Netwon have value because it is close source.
It will be silly of me to write a funtion to calculate somethimg like a = - (ks * x + kd * s)
don't you think?

In fact a funtion like, although a more advanaced one teh will inclide cpling derivaives, will be at the core of the next Newton particle system for soft bodies in Newton.

What are you planning to do?

[thedmd]

although some expert do think and spread the idea nothing in Netwon have value because it is close source.
It will be silly of me to write a funtion to calculate somethimg like a = - (ks * x + kd * s)
don't you think?

For me it is close source form is ok. I cannot agree with criticism of closed source form of Newton. But this is not proper thread to such discussion.
Probably You have right, defining such kind of functions can be treated as silly in some way. I'm rather curious how it behave, because there is no documentation on Wiki. If it has any special features (now I know it have) I prefer to ask before I broke something. : )

In fact a funtion like, although a more advanaced one teh will inclide cpling derivaives, will be at the core of the next Newton particle system for soft bodies in Newton.

Soft bodies sounds very good.

Anyway, not all Newton functions are obvious for first look. NewtonCalculateSpringDamperAcceleration() is one of them. Since I know what kind of result can I expect, the time parameter was mysterious. Better to ask, how do You think?

What are you planning to do?

In case of Newton I'm looking forward for complete release with hardware acceleration support. I must say: great job Julio.

[Julio Jerez]

check out 2.04 teh car is more polish and use car dynamics curves.
you can even opin a when and thr car move.
I still ave to elavorate more, bu preti much over 99% of the raycast car joint is in, all it need is wriking.
The Ray Cast car uses that function for suspemsion, you can take a look at teh joint

[thedmd]

Yes. I have seen that car is using spring acceleration function. This is reason why I asked about it.

I just notice about 2.04 release. Many things change, I will try to understand all changes fully.

Curves are great enchancement. I actually using them in vehicle model.

Maybe You can give me a hint, so I will then know that I'm going into good direction. At beginning I must say that API for custom joints is not what I'm familiar with very much. I guess function NewtonUserJointAddLinearRow() defines a joint like breakable slider along given axis. Am I right? I wonder if there is a reasonable way to define a traction circle without creating joints.

Since I trying to write my own implementation with let say limited physics resources, a few attempts to solve problems may seems to be weird. I don't rewrite Newton vehicle but I treat it like a reference and source of hints.

Sorry about offtopic.

[Julio Jerez]

I wonder if there is a reasonable way to define a traction circle without creating joints.

That is one for the new feature of the modified DGRayCast car,

It still need some more refinement, but you can do stuff like this:



here is the emulation, still needs some more work but you can get the idea, I will get there after I competre the wiki toturials.


it is very diffucvlt to get a stable circle of friction with out a contraint solver, becaus eteh circlei of frition is define for a single tire onder contraoll conditions in a lab.
when you have more than one tire influencing other tires, then the only thing you can do is set the forces on each tire and have trh constraint solver determine the distributions of the forces.

[Julio Jerez]

Let us start over. I deleted all the post with reference to other libraries.
What you do with Newton is up to you, I do not need to know how is compared to anything else out there.

You are welcome to learn from Newton, but my suggestion is to learn from text books, and from the Scholars papers like Baraf, Mirtich, and Min C Lin.

[thedmd]

Let us start over. I deleted all the post with reference to other libraries.

Thank you. I understand.

You are welcome to learn from Newton, but my suggestion is to learn from text books, and from the Scholars papers like Baraf, Mirtich, and Min C Lin.

Actually I use both. A few years ago I even wrote my own engine. Well, actually just rigid body motion engine with very poor collisions. Well, not all things are clear just after reading some papers and/or already working solutions.

Since this topic was about spring-mass damper, I think we can close it. I got an answer to my question. Discussion about vehicles will find a better place on this forum, I think.