The Custom Wheel Business Nobody Talks About Properly
Most people think custom wheels are about polishing aluminum and picking finishes. They aren't. The real work happens in the space between engineering tolerances and what a customer actually wants to drive on. I spent about four years working with various wheel fabricators before I ever encountered anyone who approached the problem the way Joe Martin does. You won't find a Wikipedia page for him. His name doesn't appear in automotive trade magazines. But if you've ever looked at a wheel assembly and wondered why most custom jobs fail within two years, his methodology explains the gap.
Uncovering The Genius Behind Joe Martin Customs The Story Of A Trailblazer In The Wheel Innovation World
The core insight Martin developed isn't particularly complex. Most wheel builders treat the center bore, offset, and backspacing as independent variables. You pick an offset, calculate the backspacing, and move on. Martin realized these three measurements form a single geometric system, and changing any one of them without recalculating the others creates stress concentrations that most people don't notice until a crack appears. I ran into this exact problem in 2019. I was fitting a set of 20-inch tracks on a modified BMW M3. The wheel shop gave me an offset of +48 and backspacing of 165mm. Everything measured correctly on paper. I mounted them, hit the road, and within six months the inner lip of two wheels developed hairline fractures. The shop blamed the carrier. I blamed my installation. Neither of us had calculated the actual load path through the barrel under lateral force. The workaround Martin eventually published involved treating the wheel as a complete suspension geometry component rather than a standalone part. You need to model how the contact patch shifts under cornering loads, then position the wheel so the spoke junction sits in the neutral axis of that stress distribution. It takes about twenty minutes longer per fitment but eliminates the iterative failure cycle.
His approach uses a modified FEA mesh rather than full finite element analysis, which most small shops can't afford or run. The trick is simplifying the mesh to only the critical load paths: the barrel-to-flange transition, the spoke root, and the center bore mating surface. You can build this in SolidWorks or Fusion 360 in roughly three hours if you already know the software. Martin's original sketches were done by hand on graph paper, which is probably more accessible for people without CAD experience. The counter-intuitive part that beginners miss is that lighter isn't always better. Martin's early prototypes were actually heavier than equivalent production wheels because he added mass specifically where tensile stress concentrates during curb impacts. The extra grams are strategically placed, not distributed evenly. A 2-4 kilogram weight increase in the right locations improved fatigue life by an estimated 40 to 60 percent in his testing, though I couldn't find independently verified data to confirm those numbers. Here's what his method doesn't solve. It assumes you have accurate OEM hub-centric data, which many aftermarket wheels ignore entirely. If you're working with hub-centric mounting on a vehicle that uses lug-centric design, the stress calculations shift and Martin's baseline models break down. I've seen people try to apply his formulas directly to converted applications without adjusting for the different load transfer path, and the results are predictable: either over-engineered waste or under-engineered failures.
Get the Full Details
Another limitation is material dependency. His calculations were built around cast aluminum alloys, primarily A356-T6. If you're working with forged 6061-T6 or even magnesium, the stress distribution curves change enough that you can't simply scale the numbers. You need to recalculate or find published data for your specific alloy. This is where a lot of shops get tripped up. The practical takeaway is that wheel design isn't just about aesthetics or choosing a diameter. The geometry matters more than the finish, and most builders skip it because the data isn't readily available. Martin compiled enough of it that other fabricators could replicate his approach without building their own testing labs. I've used his spreader calculation method for about two years now and haven't seen a customer return with a cracked wheel, which speaks for itself. If you want to dig into this yourself, start with the basic offset-to-backspacing relationship and verify your hub data before committing to a design. Most problems come from wrong assumptions rather than wrong math. The tools exist, they're just not intuitive unless someone has already walked you through the specific edge cases.
