Wheel-terrain interaction with Gazebo
Hello,
Up to now, I haven't found any evidence that Gazebo physics engine provides an accurate simulation of the wheel-terrain interaction between rigid wheels and soft soils. According to the ODE documentation, only the contact between rigid wheels and rigid soils is simulated through the Coulomb Friction Model.
I would like know if any of you has some knowledge on how to simulate the contact between rigid wheels and soft, deformable soils in Gazebo.
Thanks in advance !
I am not sure if that is possible. If it is, I am fairly certain that is going to run really slow, but if you don't need to run it in real time then ProjectChrono may have what you are looking for.
I am currently working on a terrain model to simulate rigid wheels in soft deformable terrain as part of my thesis. It currently exists as a ROS node that subscribes to the link states and joint states from Gazebo, then calls Gazebo services to apply the relevant forces and torques. I'd be happy to share what I have so far with you if you provided an email address
Hello! I would be happy get in touch with you. My nearest bet so far is an interface I found with Gazebo and Project Chrono, but as I need a real-time solution I don't know if it will work. Here's my email: francisco.de.castro@ist.utl.pt
Hi,
I'm also working to simulate wheel-soil interaction in Gazebo. I've came to this thread and wondered if there was any updates. Also, I was wondering if the source code was on github or somewhere else. I would be looking forward for your reply.
Hi Joshua, I am started working on wheel-terrain interaction of soft/loose soil. I will like to start my simulation in gazebo, but doesn't have any help. Please I will also like to share with me your model.
My emal: ahmadbukar90@gmail.com
Regards Ahmad
Hello, could you also share the code with me ? my email id is - a.pagariya@nisar.ai.
Just a side note: 1. I am also checking if I can get ROS and Chrono together.
Thanks in advance!
Hello Joshua, I work with Chrono in the las half year, but I was not able to speed up the simulation to get the necessary performance. I am a little bit disappointed. I would be happy to try your implementation as we are planning to use Humble for later on. mail: doba.daniel.karoly@kjk.bme.hu Thank you.