Rover Wheels: Convex or Flat?

Question

I was comparing the wheels of the Curiosity rover and the Perseverance rover and noticed that the latter had a very flat cylindrical design as compared to the former.

This got me thinking about which wheel profile was sturdier. That got me thinking about various things that might be different. Intuitively, I felt that the convex wheel would be better off because it would be able to transfer some of the applied force into normal components, which would keep the shear stresses lower (Aluminum 6061 has a shear strength of only 207 MPa, compared to the tensile yield strength of 276 MPa).

So I decided to test it on Ansys:

From these, I can visibly see that the regions of maximum stress are lower in the convex wheel. But I am starting to doubt if I'm misreading something. Could someone knowledgable help?

---

TL;DR

Among convex and flat profiles, which one is sturdier? (If you feel the word sturdy is ambiguous, please interpret it as "less metal fatigue", perhaps.)

Answer 1

I found a JPL article on it, but did not locate a CAD file (let us know if you find one). From the article it sounds like both wheels are quite close to the same with the main change being the tread. I think the image is likely photoshopped to be side by side rather than rendered side by side; so the shape change is likely less drastic than it looks in the photo.

Illustrated here, the aluminum wheels of NASA's Curiosity (left) and Perseverance rovers. Slightly larger in diameter and narrower, 20.7 inches (52.6 centimeters) versus 20 inches (50.8 centimeters), Perseverance's wheels have twice as many treads, and are gently curved instead of chevron-patterned. Credit: NASA/JPL-Caltech

Your ANSYS FEA makes sense. A curved or dished surface is stronger than a flat surface of the same thickness. When something is rounded or dished its has a larger area moment of inertia compared to the flat version. But as far as representing the rover wheels you would certainly need to model the spokes to get the whole picture. And without the CAD model, you wouldn't know for sure.

Here are some of my thoughts on the rover wheel design change:

1. If the wheels are in fact less spherical, the designers may have found that the rover wheels did not need as much surface angle variation than previously thought. The main reason a wheel is rounded is so it can keep tread in contact with the surface at various angles as opposed to a perfectly vertical force. That is why motorcycle tires are round, and car tires are more flat.
2. If the wheels are in fact less spherical, the designers may have found that they were not steering torque limited. Rounded tires take less torque to rotate because there is a smaller distance between the center of rotation and the furthest tread touching the surface.
3. The tread change from chevrons to straight perpendicular treads, is likely because they found that during operation the wheels were never subject to any loads perpendicular to the direction of rotation. This would be a disastrous design for a car on earth because a car is frequently subjected to cross winds or lateral gravitational loads when traversing a ramp perpendicularly (side-hilling)(think driving a NASCAR track at 5mph). However, an expensive rover on Mars that has 6 wheels and every movement planned by a team, has no need to attempt a side-hilling maneuver. Consequently all the treads should be oriented for maximum traction in the direction of rotation. This tread shape also increases the strength of the wheel by increasing the area moment of inertial in the axial direction (like lots of little angle iron reinforcements).