ragdoll engineering = good old fashion pain.

ragdolls are nothing more than rigid bodies connected through x-degree-of-freedom constraints. So solving constraints. so lots of phyiscs. So you'll need a head shaped like a huge lightbulb and a PhD in Rocket Science to understand and apply all of it.

Mr Rowl has some demos. Look for Tomas Jakobsen's physics particle article, although it's incomplete and the ragdolls you'll get won't be of the highest quality (Hitman ragdolls).

No tutorials that I know off. Nothing that would give you UT2004, HL2 ragdolls (which are pretty high tech). I guess if I get my arse in gear and work on a decent ragdoll demo, I'll do one.

Also, there is a doc by Baltman & Rick about rigid bodies and rotational constraints (based on jakobsen's linear constraints and verlet integration).

I'll dig out the docs when I'll get some sleep.

Ask the Najdorf. His post is near the top of this forum at the time of writing. He just made a shareware (I think) game using ragdoll physics, developed himself. I'm so proud of him! :)

Ragdoll physics is when you basically have a skeleton and you must find the correct joint values based on forces (velocity or object collision are two examples).

I don't have much experience with ragdoll physics but I imagine that it would be similar in mathematical complexity to inverse kinematics. Try searching for Cyclic Coordinate Descent, inverse kinematics and ragdoll physics on [google].

I don't know whether it would be possible to do ragdoll physics using inverse kinematics, but it would be interesting.