Last April, our supplier quality lab put 6,000 wheel hub assemblies on hold. I remember looking at the pallets and thinking, “These won't all be scrap. But they also aren't ready for customers.”
I'm a supplier quality manager at Dorman. That means I'm the person who reviews parts before they're allowed into the product line. I look at maybe 200 different part numbers in a year, from wheel hub parts to throttle bodies and drive shaft assemblies. It's not about finding someone to blame. It's about keeping a part from failing after a few thousand miles.
The supplier who said yes to everything
The story starts about eight weeks before that hold. One of our aluminum casting suppliers told us they had bought a new CNC cell and wanted to expand into finished components. “You send us the casting, we do the machining, we do the assembly,” their sales manager said. “One stop. It saves you freight and lead time.”
I wanted to say yes. When a capable vendor asks for more work, it can be the start of a great relationship. But something didn't sit right. This supplier had deep experience in aluminum casting, but no history with tight-tolerance wheel hub machining. The casting and the finish machining are two very different skills.
The frustrating part was that they weren't trying to pressure us into anything fraudulent. They genuinely believed a 5-axis machine and a quality lab would be enough. That belief—that “we can add it because we bought the equipment”—is exactly what causes quality problems.
When the first pilot batch arrived, the CMM report showed a pilot bore offset of 0.0017 inch. I know, in absolute terms, that's less than two thousandths of an inch. But this is the bore that centers the ABS tone ring. That offset would make the tone ring wobble under load, which can create an intermittent speed signal at highway speed. The seal journal also had a faint but measurable tool mark. We red-tagged the whole batch.
“They're within general machining tolerance,” the supplier argued. “They'll fit.”
They fit. But we weren't buying “fits.” We were buying 6,000 parts that need to behave identically in the field. Without process control, the second batch might be worse. We sent them back for a root-cause analysis.
The Dorman fuel injection throttle body boundary
A broad process capability isn't necessarily bad. The Dorman product catalog includes stamped brackets, forged control arms, aluminum extrusions, and CNC-machined parts. Those require different processes. The problem starts when a company treats capability as permission to cross into a whole new failure mode.
The same casting supplier had also pitched to supply the complete Dorman fuel injection throttle body assembly. On paper, it made sense: they could cast the housing, buy the motor and sensor, assemble it, and ship it. But a throttle body is not an aluminum housing that happens to have a butterfly plate. It's a calibrated electromechanical assembly with a DC motor, position feedback, return springs, and airflow curves that have to match combinations on particular vehicles. A motor can fail electrically. A position sensor can drift. The electronics have to be validated at temperatures, vibration levels, and voltage conditions that a casting supplier never sees.
Could they make a decent housing? Probably. Could they buy the same motor we use? Yes. Could they test the calibrated assembly the way a specialized throttle body maker can? Not in the first year.
That's when I said no. I'd rather work with a specialist who knows their limits than a generalist who overpromises. It sounds like a cliché until you're choosing between money saved on integration and customers stranded with a delayed-pedal fault.
Reading drive shaft product reviews differently
The rejected wheel hub batch forced us to upgrade our first-article process, and it also changed how I read customer feedback. When a customer searches for dorman rear drive shaft product info and reviews, they're usually trying to answer one question: “Will this fix the vibration?” The info and reviews can tell us a lot about field performance. If installers consistently mention a clean spline fit and no vibration at highway speed, that tells me the supplier's balance process is stable. If reviews mention vibration that isn't caused by U-joint angle, we open an investigation.
After the hub rejection, I started asking every drive shaft or rotating-part supplier for three things: runout data, balance machine calibration records, and a clear answer about which operations they perform in-house. If they can't answer all three, I don't care how nice their brochure is.
The same test applies to sensors
That mindset extends to electronics, too. Take a part like the Denso 234-4220 oxygen sensor. It's a popular replacement because Denso has spent decades making oxygen sensors, and the part has a specific calibration. A metal former might be able to stamp the outer shell, but the sensing element is a zirconia ceramic with printed electrode layers. Unless the manufacturer owns that sensing element and the calibration process, they aren't making an oxygen sensor—they're assembling other people's components.
I'm not saying assembled components are automatically bad. Some of the best products in the aftermarket use carefully selected parts from multiple vendors. But if a supplier claims they can “make an oxygen sensor” because they bought a roll of shells, that's a red flag. A specialist who says “we don't make the sensing element—we source it from a qualified supplier” is giving you an honest boundary. That honesty is valuable.
What changed after the rejected batch
We didn't end the relationship with that casting supplier. After the rejection, they went back to their cell, found the offset issue in a fixture, corrected it, and ran a new first article. The replacement batch passed. And they later told us they were going to stick to castings plus simpler machining for now, not the throttle body. That was the answer I'd been looking for all along.
Since then, our vendor approval process starts with a question that gets a lot of blank stares: “What don't you do?” The best suppliers answer quickly. They know their core process, and they know when to recommend someone else. That's not just humility—it's an engineering judgment. The FTC advertising guidelines say claims have to be truthful and substantiated. I apply the same idea to supplier capability claims. If a supplier says “we can do that,” I ask for the data that proves it. If they say “we can't do that well,” I trust them more.
The lesson from six thousand wheel hub parts
Quality isn't only about hitting a tolerance. It's about knowing where a boundary lies and respecting it. A wheel hub, a fuel injection throttle body, a rear drive shaft, and an oxygen sensor all look simple from the outside. Each one hides a different set of failure modes. You need a supplier who knows what those failure modes are and who has the test data to show they've addressed them.
My perspective is based on aftermarket drivetrain, fuel system, and sensor components, not on every low-tolerance product category. If you're buying simple brackets, the risk calculation is different. But I've watched enough “yes” quotes turn into quality escapes to prefer a specialist with self-awareness over a generalist who says yes to everything.
(And if you're searching “is radiator springs racers open,” I can't help with that one—that's a completely different kind of “open” than the supplier capacity I deal with. But if you're checking radiator springs or wheel hubs before a long trip, the questions above are still worth asking.)