Human Ragdoll Specs / Examples

Miscellaneous Forums/General Discussion/Human Ragdoll Specs / Examples

No comment yet on what new tangent of an invisible tangent I'm running along except that it is not in the slightest bit timely and it will be done in a day or not at all.

So... The first and hopefully nearly last step:

There have been a few requests here and there (can't find them though) for help with humanoid ragdoll physics, and I have surprisingly found myself asking for the same thing.
What have you found on the topic to be of use for doing a very simple physics model of a human?

No comment yet on what new tangent of an invisible tangent I'm running along except that it is not in the slightest bit timely and it will be done in a day or not at all.

Whaaat?

Anyway, this....

http://freespace.virgin.net/hugo.elias/models/m_ik2.htm

its pretty easy,
you go through the joints in the skeleton with findchild and create primitives and joints with ODE or tokamak.
You place the joints and primitives at the skeletal joints(children).
Then the most difficult part is connecting the primitives together in the right way so that the leg is connected to the body etc.

and thats it

IK isn't what you want, it's the other way round.

If you're not using ODE or tokamak...

*rummages*
veeery simple example here - santas. Written in BlitzMax.

Well, thanks for the help; it's always useful to see how everyone does physics...

But, I should clarify:
I'm thinking more along the lines of something that could describe (probably with percentages) the weight of various limbs and the strength/limits of various muscle groups and joints.

That Santas thing is great, though! :)

I thought that might be the case...
If you can wade through this it might be some use to you.

Can't find anything about muscle strength, though.

I've done alot of research recently on this type of thing, specifically for 2d ragdoll stuff. Maybe you remember my old demo's. As you'll know, it's pretty complex stuff and as with most physics simulations it requires alot of fiddling to find that 'sweet spot'.

I found that topical such as ...

procedural animation
bipedal physics
robotics
bipedal balance
bipedal dynamics
simulating muscles

.. provided with a wealth of information but hardly anything which was actually useful in terms of the contex I needed it. Actually getting a 3d figure to balance is extremely difficult which I'd imagine is why it was left out of that fighting game someone posted a week or so ago.

Check out also . http://www.sodaplay.com/

I'd recommend trying 2d first with jacobsens verlet methods and building on that.

Stevie