It was 4:47 on a Thursday afternoon in March 2024 when my phone rang. I had one foot out the door, and my plan for the evening did not include an industrial version of a rescue mission. The voice on the other end belonged to a maintenance manager at a cart manufacturing facility in Ohio. He didn't waste time with hello. "If we don't have 80 hub seals by Saturday, the line is dead on Monday."
By 'hub seals' he meant the O-rings that keep grease inside the bearing cavities in the pneumatic rubber wheel his company uses on its material-handling carts. Those carts are built for hospitals and factories, where the floor has expansion joints, thresholds, and all kinds of little obstacles. The air-filled wheel absorbs shock in a way a solid wheel can't. But it only works if the hub stays sealed. And the hub stays sealed because of a small, easy-to-overlook silicone O-ring.
The problem wasn't the wheel. It was the O-ring.
Four days earlier, this client's buyer had ordered replacement O-rings from a new supplier. The old supplier had gone out of stock, so the buyer did what buyers do: found another source. The parts arrived on Wednesday in a clearly labeled bag: TPU rubber O-rings. But the original spec said 70 Shore A silicone. When the mechanics tried to install the TPU parts, they wouldn't sit properly in the groove. They were too stiff, and one of them actually tore before it could be pressed into place.
So, what is TPU rubber?
The purchasing agent who called me asked a version of the question that I've heard more times than I'd like: "What is TPU rubber? Why can't we use it?" That's not a dumb question. TPU is short for thermoplastic polyurethane. It's a tough, abrasion-resistant elastomer that's used in hoses, tubing, cable jacketing, and lots of places where you need something that won't wear down easily. It processes like a plastic, which makes it cheap to form. But 'rubber' is one of those broad words that hides a lot of chemistry.
Silicone rubber belongs to a different family. It has a different backbone structure, and that changes the way it behaves. According to Dow's material resources (dow.com), silicone rubber is valued for flexibility across a wide temperature range, and it tends to bounce back after being compressed for a long time. TPU is a great material when you need toughness and abrasion resistance. But in this application, the spec was silicone because silicone handles the constant flexing and the temperature swings a wheel hub might see. Let me rephrase that: the spec wasn't arbitrary. It was there for a reason.
We didn't buy the TPU O-rings. We bought time instead.
The O-ring Handbook made us slow down
I pulled out our copy of the O-ring Handbook. If you've never used it, that book is basically the engineer's dictionary for O-ring sizes, groove dimensions, tolerances, and material compatibility. It's not glamorous. But when a line is down, it's worth its weight in shop manuals.
We measured the old O-ring with a caliper. The inside diameter was 34.8 mm, cross-section about 2.5 mm. The O-ring Handbook has a table for standard sizes, and the closest standard was a 35 mm ID with a 2.5 mm cross-section. The client's groove had been machined for that size. The handbook also lists recommended squeeze and fill percentages. Our numbers were tight but acceptable.
Then we checked the material chart. The client's spec said silicone, 70 Shore A. The TPU parts were hard and dense. The handbook confirmed the material choice. It wasn't a maybe call. It was the call.
Finding a Dow Corning silicone distributor who could actually do something
The client also needed a tube of Dow Corning 785 silicone sealant to reseat the hub cover, because the original cover gasket had been chewed up during disassembly. Oh, and the sealant was the small line item that almost got lost in the scramble. The O-rings were the center of attention, but without the 785, the hub still wouldn't be sealed.
I called our regular Dow Corning silicone distributor at 5:05 PM. The good news: they had the Dow Corning 785 in stock, in the same building. The less good news: they don't stock a standard 35 mm silicone O-ring in a box. But the distributor has a small rubber shop behind the counter. They stock silicone cord stock in various durometers and can splice it into custom O-rings. In an ideal world, you mold O-rings for a critical seal. In a 36-hour world, a properly spliced silicone O-ring is a legitimate option for a static seal—and this hub seal is static, not a dynamic shaft seal.
We picked the cord stock, calculated the splice allowance with the O-ring Handbook, and asked for 80 O-rings plus 2 tubes of 785. The distributor said Saturday morning, if we ship before 7.
I placed the order at 5:40. It wasn't cheap. The rush fee was $380 on top of the base order of $1,260. I paid it without negotiating, which is unusual for me. Honestly, even after hitting send, I kept second-guessing the groove width. What if the old O-ring was actually a metric special and I'd chosen the closest standard size that would technically fit but not last? That thought followed me through dinner and into a restless night.
The Saturday morning delivery
Saturday at 4:55 AM, a FedEx truck pulled into the client's parking lot. The package had 80 silicone O-rings, 2 unopened tubes of Dow Corning 785, and a printed copy of the material batch certificate. The maintenance manager later said the first O-ring went in smooth like the old one. They assembled a test wheel, put it under a slow leak test, and it held. By noon, the line was running.
The surprise wasn't that the distributor shipped on time. The surprise was that everything in the package was right. The sealant, the O-rings, the cert—all there. That shouldn't be surprising, but in my experience, rush orders are often the place where a small omission becomes a second emergency. This order had no omission.
What I'd do differently
The most frustrating part of this case wasn't the TPU bag. It was that the TPU bag had been sitting in the receiving area for a full day, clearly labeled, and nobody had opened it until the mechanics got to the line. The client didn't have an incoming-material check for rubber parts. The part number was right. The description said O-ring. The word TPU didn't trigger a hold because no process existed for a hold.
Since that week, that client uses a simple checklist for any incoming O-ring or rubber part: check the durometer, check the cross-section, check the material family, and demand the cert if it's going into production. It takes about five minutes per order. Five minutes of verification beats five days of correction. If they'd done that on Wednesday, they would have had time to get molded O-rings instead of spliced ones. The spliced ones worked, but I'd rather not rely on a seam when I don't have to.
As for the supplier question—yes, silicone and TPU are both elastomers, but they are not interchangeable. Keep the O-ring Handbook close, work with a Dow Corning silicone distributor who knows what can be rush-made and what can't, and check every rubber part before it gets near the line. The last one is the one that saves you.