Closed azadm closed 9 years ago
Sorry for the late reply @azadm .
For the fact that the contact points are not diffent between the right foot and the left foot @jeljaik and @DanielePucci can explain their tricks in having stable contacts.
On the topic of the direction of the wrench, I guess that the model present in icub-gazebo
lacked behind the one properly generated using our two tools for generating icub models (see https://github.com/robotology/codyco-superbuild/issues/55). On newly generated models, the l_foot and r_foot frame are coincident with the ankle ft sensor frame, and so the measurement should be consistent.
As a quick test you can try the properly generated model for iCubGenova01 by following the "
Use of generated models in Gazebo" instruction in https://github.com/robotology-playground/icub-model-generator .
Any update on this? Can we close this issue?
Closing this issue. Ready to re-open it in case it is needed.
:+1:
I am trying to find out the contact points between iCub's feet and the ground and also the contact forces. When I put an iCub model in Gazebo and type
gz topic -e /gazebo/default/physics/contacts
in a terminal I get a list of the contact points (at each foot) and contact wrenches for every millisecond. For example:which shows that at that specific time there are 5 contact points for the right foot and one point for the left foot. The problem is that for the left foot the force in the
z
direction is positive whereas for the right foot it is negative for one of the points. Also the results are different from what I get bygz topic -e /gazebo/default/iCub/l_foot_ft_sensor/left_foot_ft/wrench
orgz topic -e /gazebo/default/iCub/r_foot_ft_sensor/right_foot_ft/wrench