search

RobotLocomotion / drake

Model-based design and verification for robotics.
https://drake.mit.edu
Other
3.25k stars 1.26k forks source link

urdf parser does not parse dynamics damping tag #11294

Closed RussTedrake closed 5 years ago

RussTedrake commented 5 years ago

Using http://underactuated.mit.edu/src/double_pendulum/double_pendulum.urdf from MBP results in different dynamics than RBP, because the MBP parser is not reading the <dynamics damping="0.1" /> tags.

It seems like we certainly could: https://drake.mit.edu/doxygen_cxx/classdrake_1_1multibody_1_1_revolute_joint.html#aad91b55e6b714db4ba96190e6fab07ac

RussTedrake commented 5 years ago

Update: The sdf parser seems to ignore the tag as well.

sammy-tri commented 5 years ago

I'm not sure this is a parser problem. Both the URDF and SDF parsers are tested to properly parse the damping in the acrobot model at https://github.com/RobotLocomotion/drake/blob/master/multibody/parsing/test/acrobot_parser_test.cc#L119 (I'd added an additional test to detail_urdf_parser_test.cc based on the double pendulum to verify before I noticed the existing test, so I'm pretty confident of the parsing).

Will ping the dynamics team to see if they've got any ideas.

sherm1 commented 5 years ago

Any thoughts on this, @amcastro-tri ?

RussTedrake commented 5 years ago

i've managed to resolve it. apologies. (my urdf had a drake_ignore tag earlier that was cancelling the effect, and my sdf had the dynamics tag outside of the axis tag -- a perfect storm!).

RussTedrake commented 5 years ago

and rigid body tree includes the damping forces in the "bias term", but MBP does not (it comes as a separate external force). sigh...