mecanum wheel

Typically: "How do I... ", "How can I... " questions
Post Reply
Posts: 3
Joined: 06 Apr 2021, 21:31

mecanum wheel

Post by ahmedelwaly »

I designed my robot and using mecanum wheel, how can i configure those wheel ?

Site Admin
Posts: 8724
Joined: 14 Dec 2012, 00:25

Re: mecanum wheel

Post by coppelia »


mecanum (or swedish or omnirirectional) wheels can be simulated in various ways via two main methods:
  • you can design your wheel to contain several free-rolling joints, that each hold a roller. The roller needs to be very carefully dimensioned/designed, since it needs to be convex, and have enough precision for a smooth operation. Additionally, depending on the physics engine you are using, you might want to give each roller enough mass and/or inertia, for stability reasons. See also dynamic design considerations 7-8.
  • the other way that we came up with is to design your mecanum wheel as a rigid entity (e.g. a compound shape). Then, in a hidden layer, add a motorized joint, on top of which an auxiliary shape (non-respondable) is attached. On top of that auxiliary shape, mount a free-rolling joint (typically with a 45 deg. angle), and on top of that joint add a respondable sphere. The tricky part is to constantly reset the orientation of the second joint, in order to emulate a mecanum wheel.
The first method is less stable, and also more computation intensive. The second method works fast and nicely, also on uneven ground. You might have to adjust the size of the sphere contacting the floor. You can find examples of this method if you look at the hidden layers, while simulation is running, of following models from the model library: KUKA Omnirob.ttm, KUKA YouBot.ttm and ball robot.ttm. Make sure to also inspect the child scripts attached to each of the wheels. Those scripts are in charge of constantly resetting the orientation of the free-rolling joints.


Post Reply