If you’re looking to create an airless tire, you’ll need a material with both flexibility and strength. A non-pneumatic tire, after all, must withstand punctures and pressure while also providing a smooth ride.
With such opposing design objectives, engineers from the University of Illinois turned to simulation to narrow down the options and solve the puzzle.
Kai James, assistant professor in the Department of Aerospace Engineering at U of I, and graduate student Yeshern Maharaj used computer algorithms to determine a variety of suitable structural patterns for the non-pneumatic tire.
The design duo focused on an area of the tire that is very open to design, according to James: the shear layer.
“The shear layer is where you get the most bang for your buck from a design perspective,” said James . “It’s where you have the most freedom to explore new and unique design configurations.”
The shear layer, located just below the tread and just above the tire’s flexible spokes, offers a space for material possibilities.
An airless tire ultimately needs to handle strain under pressure – a quality known as shear – while also being stiff in the axial direction. The University of Illinois engineers tried different designs that provided a varying priority to constraints like buckling, stress, and stiffness.
James and Maharaj employed computer simulation to model the elastic response on the shear layer and calculate the material’s ability to stretch and twist.
“Beyond a certain level of stress, the material is going to fail,” James said. “So, we incorporate stress constraints, ensuring that whatever the design happens to be, the stress doesn’t exceed the limit of the design material.”
The computer simulation begins with a computer-simulated block of the bulk material, and then the non-optimal patterns are systematically eliminated. Because a solid block lacks elasticity, the material is virtually cut away, leaving spaces for flexibility.
“If you carve holes in the material until it is something like a checkerboard pattern, with half of the material, you’d also have half of the original stiffness,” James said. “Now, if you do a much more complicated pattern, you can actually tailor the stiffness.”
Gradually, each new design is an improvement on the previous one, says the U of I professor, until ultimately you end up with a design that is near optimal.
In the edited interview with Tech Briefs below, James explains why this method for creating an airless tire may be the beginning of new wave of design that relies more on automation than the human engineer.
Tech Briefs: How does your tire design change how the shear layer is made?
Kai James: The shear layer is the layer beneath the tread that connects essentially the outer portion of the tire with the layer that functions as a series of flexible spokes.
The shear layer goes from a curved shape to essentially a flat shape that conforms with the ground, due to elastic compliance. It’s a fairly thick layer, and it’s pretty open to design. It’s voluminous and gives us a large area within which we can explore different geometries and configurations.
Tech Briefs: How did you use computer simulation to explore these different configuration possibilities?
Kai James: We used what’s known as finite element analysis. We take the equations that describe the way the raw material, the rubber essentially, will deform elastically. The code simulates the way that the material bends, changes shapes, and responds to loading.
We break the domain – the region that we’re simulating – into a grid. And then we can solve the equations, point by point, within that grid. When we assemble all that information, it gives us a picture of how the tires are going to perform.
Tech Briefs: What are the advantages of an airless tire?
Kai James: The main advantage is failure. It can’t be punctured. You can drive over a nail, and it will function in just about the exact same way. You don’t have to re-inflate it. Tires lose air over time. That’s not going to be an issue here.
Tech Briefs: Based on this work, do you envision mainstream use of airless tires?
Kai James: Possibly. I think there are still quite a few challenges. I think it’s probably unlikely that you’re going to see them on commercial vehicles like the car that you’d buy at the dealership. I think there are still a lot of performance issues that prevent it from being really competitive with pneumatic tires.
You’re going to sacrifice a lot of fuel efficiency with an airless tire, because it’s not as energy efficient. Airless tires deform substantially, and as a result, you’re moving energy that’s going into actually deforming the tire. A similar principle is what causes mechanics to make sure that you always have the right tire pressure. If your tire is deflated, then you’re going to lose fuel efficiency. That problem is kind of inevitable with the non-pneumatic tire.
But I do see it being a good option for less widely used vehicles, smaller vehicles, small drones, or vehicles the size of a golf cart where fuel efficiency isn’t as critical. I do think the airless tire is a very useful option for those less common, less widely used applications.
Tech Briefs: Based on the analyses and the computer simulations, what is the ideal design for this kind of tire?
Kai James: We actually came up with essentially a family of different designs as we determined a pattern for the shear layer. This is what’s known as a metamaterial approach. You design a repeating unit cell; then, when you put a bunch of cells together in a grid, the cells forms a periodic material.
Depending on if the cells are in a perfect checkerboard pattern, or if they’re sort of off-set one row from the other, you’ll get different results.
And it depends on how you prioritize the different design objectives. In one case, you might want a high degree of shear flexibility, so you want the periodic material to be very compliant to shear deformation, but you also want it to be very stiff in compression. The way that you weight those different design objectives is going to determine what your ultimate pattern is going to look like.
The design task is what we refer to as non-convex, which means there are many different optima that you can arrive at, and most of them will perform fairly well. There wasn’t one specific design; there were really several that were able to satisfy our design objectives.
Tech Briefs: What other constraints did you consider?
Kai James: We also considered buckling constraints. If you look at the patterns that the optimizer came up with for the periodic material, there are a lot of slender members – these little struts that comprise the lattice of the material. Those can be subject to buckling, so we included analysis in the computational model that would modify the design to make it more robust to that particular failure mode.
Tech Briefs: What inspired this work?
Kai James: The non-pneumatic tire problem has been around for quite some time. In fact, there have been other researchers who have approached this problem from a computational standpoint. We thought that we could lend a little bit of originality by including stress and buckling constraints.
Creating a non-pneumatic tire is very challenging because you have these competing design objectives. On one hand, you want it to be flexible in shear, but at the same time, you want the structure to be strong enough to not buckle, and to stay within the stress limits of the material. When you have that type of problem, algorithms are best suited to come up with the solution, because these computer-based algorithmic design processes are very mathematical and systematic, as opposed to being intuitive or based on a human designer's experience and rule of thumb. We thought it would be a very interesting design challenge and one that was very well suited to our design approach, which relies heavily on computers. We wanted to try to get the computer to essentially do the thinking for us.
Tech Briefs: What are you working on now?
Kai James: Right now, we are working on a similar computational design framework, but one that will synthesize mechanical assemblies. Starting from a black box, we provide the algorithm with an elastic model that can simulate formation and material stiffness, and we want to see if we can get the algorithm, without any further input from the human designer, to synthetize, for example, a wheel mechanism.
I think this really gets to a basic challenge in computational design: Can we create entirely novel design concepts using algorithms, with very little human input?
We essentially want to automate the design processes and, if possible, get algorithms to actually invent entire design ideas on their own, just by telling the algorithms: “This is what we want the system to do. This is the mechanical function that we’re seeking.” We’re not just trying to synthesize the tire but also the entire wheel and the hinge as well.
It’s almost like a way of reinventing the wheel, but doing it with computers.
What do you think about automating entire design ideas? Share your comments and questions below.