Rover Suspension Design - Unbalanced bearings?


#1

Hi Everyone,

I’m new to the site, and I don’t have much background in making/designing small-scale rover-type of things… but one thing about the Suspension Design for Rover has been bugging me. I have a little experience in off-roading and mini-baja build… so some things I say might not be so valid for this smaller design haha. Might just be over-thinking it.

As a preface, lets recap the ideal bearing:
pic1
An ideal bearing should only take load/force in the x-y plane, where it rotates freely about the z-axis. It should not be subject to any axial force in the z-axis, nor should it be subject to any torques along x or y axes… torque in the z-axis is of course allowed, and that is how it spins. When I say “should not”, i’m talking about potential for warping the bearings and causing un-even wear… that is why you should always get your new tires balanced by a mechanic before installing them on your car.

Rover’s suspension geometry of course, has roughly speaking, 2 bearing-type interfaces per independent axle:
Capture2

Just considering one of the bearing-type interfaces, you can see that a force on the tire, results in a moment T1 about the x-axis, since the wheel is off-set from the bearing by a distance d:
image

Is there any way to improve this design? I’ve been playing around with fixed-axle geometry and 4-arm linkage designs in my head… also Macpherson struts and whatnot… but I don’t have any time to CAD or sketch them out.

It is also entirely possible that I am over-thinking this wayy too much… especially since proof-of-concept was shown in Taylor’s video post a few months back. My intuitive concern is that the material of the suspension is 3D printed plastic, and i’m not too familiar with the yield strength of 3D printed materials. My other intuition is that since the drive-train is now effectively contained within each wheel via hub motor (as compared with a normal internal combustion engine driven car), this allows a lot more space to play around with a geometrical solution to this problem. I had a colleague from my previous company, a designer, who said he loved to approach all engineering problems assuming that there is always a geometrical solution to anything.

Would love everyone’s thoughts and opinions on this. Thanks!

Cheers,
Kai
(ps. Hope I figured out how to upload photos nicely here… if there’s an error in the pictures i’ll try post them in the comments or something. )
(pps. Also just thought of a “forked” suspension design as follows, but it really increases the bulkiness of the overall build, so i’m inclined to shy away from this solution: )
image


#2

Hey Kai,

This is a great question! Thanks for asking.

I think the following info might answer your question. I updated the assembly to show the current parts with all bearings and took a screenshot of a section view:

rover_section

One thing to note about both joints you highlighted is that they are “double supported” to spread the loads created by torques in the x or y axes. In this annotated view below, you see that on each joint there is one 6817 and two 6808 bearings, for a total of 2 and 4 respectively.

rover_section_labeled

Note that in the bottom, where the red ring gear attaches to the pink suspension arm, the interface is secured with typically 12 strong screws, so it is a very rigid joint.

I need to post this as its own topic, but I have a bunch more photos relevant to this discussion here:

Does this help answer your question?


#3

Hi Taylor,

This is great! The cross-section view definitely helped :slight_smile:

I certainly see now how the double-bearing design increases the robust-ness of the design tremendously… i’ve attached a sketch i’ve done showing the how the forces roughly balance out… do note that Red arrows are force, Blue arrows are torque, and the Green circle is my guess at the highest stress point after all the force-torque balancing. Also, note that the red arrows I drew on the free body diagram have different bodies that they are acting on… but i’m lazy to make a detailed and proper free body diagram now.

image

Thanks so much for your input, Taylor :slight_smile: I look forward to following up and see how the progress for Rover 2.0 goes!

Cheers,
Kai