# Still runaway simulations in gazebo

Hello List,

UPDATED: I completely rewrote the question to make it more clear.

I am having difficulties in getting stable behaviour with my simulated rowing boat. The model can be found on my github page in the boot3_description/urdf directory.

A simplified version can be found in the testing directory Please see crash.mp4 for how the simulation (and simulator) blows up.

Sometimes there is an error message like:

gzclient: /build/ogre-1.9-B6QkmW/ogre-1.9-1.9.0+dfsg1/OgreMain/include/OgreAxisAlignedBox.h:252: void Ogre::AxisAlignedBox::setExtents(const Ogre::Vector3&, const Ogre::Vector3&): Assertion (min.x <= max.x && min.y <= max.y && min.z <= max.z) && "The minimum corner of the box must be less than or equal to maximum corner"' failed.


The model is crash_boot.xacro and for testing I use the (normally) generated crash_boot.sdf. The model now contains only 3 revolute joints. Playing around with friction and damping does not help.

What does help is reducing the max step size` by a factor of 10. But it then becomes impractically slow, especially when using the full blown model with almost 40 joints.

Hopefully someone can help, as I am really stuck here.

Thanks again, Sietse

PS. I also asked this question on gazebo issues, but did not yet got a response.

edit retag close merge delete

Sort by » oldest newest most voted

That type of behavior can come from a couple places. Have you double checked your inertia values? Look at this tutorial.

Are your joints configure properly? Take a look at axis and limits.

Look at the effort you apply to the joints. Very large forces can be unrealistic and in turn cause incorrect behavior.

Out of curiosity, are you trying to row on a flat ground plane? If you want to row in water then you should look at this repository.

more

Thanks, see my edit above. And also thanks for the links! I want to do experiments with rigging of race-shells, in order to make better discussions among rowing coaches possible. This is only the very first start. Later I will add a rower in the form of a "robot" with all his joints. And then be hopefully able to optimize the rigger setting for that rower, optimizing it for maximum speed given a power output of the rower.

( 2018-11-13 10:23:24 -0500 )edit

The inertia and mass values look okay. You could start debugging your problem by inspecting the forces you are applying to the joints. If you haven't done so, you could also reduce the friction coefficients on the boat body.

( 2018-11-14 09:12:07 -0500 )edit