Hi all,

I have a similar problem with the simulation of a modified pioneer on an
inclined plane. If I make a pioneer with a mass of 21 kg to crash against a
box of mass of 3000 kg, then the box starts rolling over the plane. My
suggestion is that the moment of inertia of all the objects is set to a very
low value (about 10e-10) and then even a small torque on an object cause it
to roll. The same happens to you when the rod rolls, so that it applies a
torque to the other objects (due to friction). Given the very low moment of
inertia of teh table, it starts rolling.
I even tried to set the inertia, but nothing changed.

I post here because I think the two problems are related.

I attach the files of my simulation.
http://dl.dropbox.com/u/20433489/gazebo/pioneer3d.model
http://dl.dropbox.com/u/20433489/gazebo/pioneer3d.world

Any suggestions?
thanks in advance

Paolo




2011/6/16 Rein Appeldoorn <[email protected]>

> No, that's the weird fact, it doesn't change at all when I'm varying
> different parameters like mass, friction etc.
>
> What do you mean with "off the trunk"?
>
> I attached the piece of code for moving the rods:
> http://80.56.19.206/files/movingrodscode.txt
>
> More and more, I start to think that my Gazebo isn't set-up properly. First
> on 0.10, I got strange behavior when using the act-array interface. I
> couldn't event rotate a single hinge. Now, on 0.9, other thinks occur which
> aren't really explainable.
>
> Is there a way to check if your Gazebo is set up properly?
>
> Thanks again for helping me!
>
> Rein
>
> On Thu, Jun 16, 2011 at 10:42 AM, John Hsu <[email protected]>
> wrote:
>
> > Hi Rein,
> > does the strange behavior change at all when you increase/decrease mass
> of
> > the walls?
> > Have you tried off the trunk to see if this problem still exists?
> > lastly, can you attach the code you used to move the rods too?
> > thanks,
> > John
> >
> > On Thu, Jun 16, 2011 at 1:08 AM, Rein Appeldoorn <[email protected]>
> > wrote:
> >
> > > I'm using Gazebo version 0.9. This is because I had a problem with
> > > actuating
> > > the hinges via the generic-actarray in Gazebo version 0.10. Therefore,
> > I'm
> > > using version 0.9 now.
> > >
> > > I'm not using a wall mesh; the wall consist of multiple box geometries.
> > >
> > > The movement of the table, what you saw in the movie, happens when
> > rotating
> > > the rod. This is weird because the static part of the table (walls)
> have
> > > the
> > > following parameters:
> > >
> > > <mu1>0.3</mu1>
> > > <mu2>0.3</mu2>
> > > <mass>1e5</mass>
> > >
> > > I've added my ODEPhysics.cc and my .world-files:
> > > http://80.56.19.206/files/ODEPhysics.cc
> > > http://80.56.19.206/files/simulation_010.world (Gazebo 0.10)
> > > http://80.56.19.206/files/simulation_090.world (Gazebo 0.09)
> > >
> > > Thanks for responding!
> > >
> > > Rein
> > >
> > > On Thu, Jun 16, 2011 at 9:45 AM, John Hsu <[email protected]>
> > > wrote:
> > >
> > > > ah, sorry, I just went back to see the video again, and apparently I
> > saw
> > > > the
> > > > wrong video earlier!
> > > >
> > > > Ok, the bouncing table problem is most likely a problem with max
> > contacts
> > > > specified.  I've seen this when max contacts is set too low for the
> > mesh
> > > > involved.  What version of gazebo are you running?  can you attach or
> > > point
> > > > me to ODEPhysics.cc?
> > > >
> > > > One way to fix this (not the best way) is, try putting a small box
> > block
> > > > under each corner of your "wall" mesh.
> > > >
> > > > John
> > > >
> > > > On Wed, Jun 15, 2011 at 2:50 AM, Rein Appeldoorn <[email protected]>
> > > > wrote:
> > > >
> > > > > I checked my gazebo settings in /server/physics/ode/ODEPhysics.cc
> and
> > > > they
> > > > > were as follows:
> > > > >
> > > > > contact.surface.mode = dContactSlip1 | dContactSlip2 |
> > > > >                               dContactSoftERP | dContactSoftCFM |
> > > > >                               dContactBounce | dContactMu2 |
> > > > > dContactApprox1;
> > > > >
> > > > > So this can't be the problem in my opinion. The parameters for the
> > > > > geometries of the table are as follows:
> > > > >
> > > > > <mu1>0.3</mu1>
> > > > > <mu2>0.3</mu2>
> > > > > <mass>1e5</mass>
> > > > >
> > > > > Any idea what we are doing wrong?
> > > > >
> > > > > Thanks,
> > > > >
> > > > > Rein
> > > > >
> > > > >
> > > > > On Fri, Jun 10, 2011 at 1:08 PM, Rein Appeldoorn <
> [email protected]>
> > > > > wrote:
> > > > >
> > > > > > John,
> > > > > >
> > > > > > Thanks for your reply.
> > > > > >
> > > > > > What do you mean with the friction coefficients of the "wheels"?
> > > > > > And how do I check my Gazebo-settings if they are right?
> > > > > >
> > > > > > Regards,
> > > > > >
> > > > > > Rein
> > > > > >
> > > > > >
> > > > > > On Thu, Jun 9, 2011 at 8:01 PM, John Hsu <
> [email protected]
> > >
> > > > > wrote:
> > > > > >
> > > > > >> Hi Rein,
> > > > > >> Not sure if this will help, but have you tried reducing the
> > friction
> > > > > >> coefficients of the wheels <mu1> and <mu2> to something less
> than
> > 1?
> > > > > >> also, make sure that in the version of gazebo you are running,
> > > > > >> contact.surface mode has "dContactApprox1" in it.
> > > > > >>
> > > > > >>        contact.surface.mode = dContactSoftERP | dContactSoftCFM
> |
> > > > > >> dContactApprox1 | dContactMu2;
> > > > > >>
> > > > > >> John
> > > > > >>
> > > > > >> On Thu, Jun 9, 2011 at 5:51 AM, Rein Appeldoorn <
> > [email protected]>
> > > > > >> wrote:
> > > > > >>
> > > > > >> > Code:
> > http://www.student.tue.nl/W/r.p.w.appeldoorn/simulation.txt
> > > > > >> > <
> http://www.student.tue.nl/W/r.p.w.appeldoorn/simulation.world>
> > > > > >> >
> > > > > >> > On Thu, Jun 9, 2011 at 2:43 PM, Rein Appeldoorn <
> > > [email protected]>
> > > > > >> wrote:
> > > > > >> >
> > > > > >> > > Hi all,
> > > > > >> > >
> > > > > >> > > We are working on a simulation model of an autonomous
> football
> > > > > table.
> > > > > >> We
> > > > > >> > > already build our model but we're struggling with the
> realism.
> > > We
> > > > > >> solved
> > > > > >> > a
> > > > > >> > > lot of problems but to this one, we can't find an answer to.
> > > Here
> > > > is
> > > > > a
> > > > > >> > movie
> > > > > >> > > of the weird/unrealistic behavior:
> > > > > >> > >
> > > > > >> > > Movie:
> > > > > >> >
> > > http://www.student.tue.nl/W/r.p.w.appeldoorn/gazebo_tablesoccer.avi
> > > > > >> > >
> > > > > >> > > Has someone perhaps experienced similar problems or perhaps
> > does
> > > > > >> someone
> > > > > >> > > want to take a look at our code because we're probably
> making
> > > > > trivial
> > > > > >> > > mistakes.
> > > > > >> > >
> > > > > >> > > Code:
> > > > http://www.student.tue.nl/W/r.p.w.appeldoorn/simulation.world
> > > > > >> > >
> > > > > >> > > Thanks in advance,
> > > > > >> > >
> > > > > >> > > Rein
> > > > > >> > >
> > > > > >> >
> > > > > >> >
> > > > > >>
> > > > >
> > > >
> > >
> >
> ------------------------------------------------------------------------------
> > > > > >> > EditLive Enterprise is the world's most technically advanced
> > > content
> > > > > >> > authoring tool. Experience the power of Track Changes, Inline
> > > Image
> > > > > >> > Editing and ensure content is compliant with Accessibility
> > > Checking.
> > > > > >> > http://p.sf.net/sfu/ephox-dev2dev
> > > > > >> > _______________________________________________
> > > > > >> > Playerstage-gazebo mailing list
> > > > > >> > [email protected]
> > > > > >> >
> https://lists.sourceforge.net/lists/listinfo/playerstage-gazebo
> > > > > >> >
> > > > > >>
> > > > > >>
> > > > > >>
> > > > > >> --
> > > > > >> John Hsu, Ph.D.
> > > > > >> Senior Research Engineer
> > > > > >> Willow Garage Inc.
> > > > > >> +1-650-475-2809
> > > > > >>
> > > > > >>
> > > > >
> > > >
> > >
> >
> ------------------------------------------------------------------------------
> > > > > >> EditLive Enterprise is the world's most technically advanced
> > content
> > > > > >> authoring tool. Experience the power of Track Changes, Inline
> > Image
> > > > > >> Editing and ensure content is compliant with Accessibility
> > Checking.
> > > > > >> http://p.sf.net/sfu/ephox-dev2dev
> > > > > >> _______________________________________________
> > > > > >> Playerstage-gazebo mailing list
> > > > > >> [email protected]
> > > > > >> https://lists.sourceforge.net/lists/listinfo/playerstage-gazebo
> > > > > >>
> > > > > >
> > > > > >
> > > > >
> > > > >
> > > >
> > >
> >
> ------------------------------------------------------------------------------
> > > > > EditLive Enterprise is the world's most technically advanced
> content
> > > > > authoring tool. Experience the power of Track Changes, Inline Image
> > > > > Editing and ensure content is compliant with Accessibility
> Checking.
> > > > > http://p.sf.net/sfu/ephox-dev2dev
> > > > > _______________________________________________
> > > > > Playerstage-gazebo mailing list
> > > > > [email protected]
> > > > > https://lists.sourceforge.net/lists/listinfo/playerstage-gazebo
> > > > >
> > > >
> > > >
> > > >
> > > > --
> > > > John Hsu, Ph.D.
> > > > Senior Research Engineer
> > > > Willow Garage Inc.
> > > > +1-650-475-2809
> > > >
> > > >
> > >
> >
> ------------------------------------------------------------------------------
> > > > EditLive Enterprise is the world's most technically advanced content
> > > > authoring tool. Experience the power of Track Changes, Inline Image
> > > > Editing and ensure content is compliant with Accessibility Checking.
> > > > http://p.sf.net/sfu/ephox-dev2dev
> > > > _______________________________________________
> > > > Playerstage-gazebo mailing list
> > > > [email protected]
> > > > https://lists.sourceforge.net/lists/listinfo/playerstage-gazebo
> > > >
> > >
> > >
> >
> ------------------------------------------------------------------------------
> > > EditLive Enterprise is the world's most technically advanced content
> > > authoring tool. Experience the power of Track Changes, Inline Image
> > > Editing and ensure content is compliant with Accessibility Checking.
> > > http://p.sf.net/sfu/ephox-dev2dev
> > > _______________________________________________
> > > Playerstage-gazebo mailing list
> > > [email protected]
> > > https://lists.sourceforge.net/lists/listinfo/playerstage-gazebo
> > >
> >
> >
> >
> > --
> > John Hsu, Ph.D.
> > Senior Research Engineer
> > Willow Garage Inc.
> > +1-650-475-2809
> >
> >
> ------------------------------------------------------------------------------
> > EditLive Enterprise is the world's most technically advanced content
> > authoring tool. Experience the power of Track Changes, Inline Image
> > Editing and ensure content is compliant with Accessibility Checking.
> > http://p.sf.net/sfu/ephox-dev2dev
> > _______________________________________________
> > Playerstage-gazebo mailing list
> > [email protected]
> > https://lists.sourceforge.net/lists/listinfo/playerstage-gazebo
> >
>
> ------------------------------------------------------------------------------
> EditLive Enterprise is the world's most technically advanced content
> authoring tool. Experience the power of Track Changes, Inline Image
> Editing and ensure content is compliant with Accessibility Checking.
> http://p.sf.net/sfu/ephox-dev2dev
> _______________________________________________
> Playerstage-gazebo mailing list
> [email protected]
> https://lists.sourceforge.net/lists/listinfo/playerstage-gazebo
>
------------------------------------------------------------------------------
EditLive Enterprise is the world's most technically advanced content
authoring tool. Experience the power of Track Changes, Inline Image
Editing and ensure content is compliant with Accessibility Checking.
http://p.sf.net/sfu/ephox-dev2dev
_______________________________________________
Playerstage-gazebo mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/playerstage-gazebo

Reply via email to