What a Ringfeder Coupling Taught This Procurement Manager About Total Cost

Last September, the tube mill line stopped at 2:17 p.m. I know the exact time because my phone buzzed with three messages in thirty seconds. The maintenance supervisor had taken the guard off the motor coupling and found the key was gone. Not sheared—gone. The keyway had elongated into a shallow canyon, and the coupling hub was spinning on the motor shaft.

I'm the procurement manager at a 120-person metal fabrication plant. I've managed a maintenance budget of roughly $140,000 a year for the past six years, and I track every line item in our CMMS. This wasn't a normal "order a part" failure.

The motor was a 30 kW AC motor driving a gearbox that feeds the tube mill. The coupling hub was keyed to the motor shaft. (Should mention: the machine builder installed it originally, and they didn't include a shaft tolerance report in the manual.)

The part that looked simple

The production manager wanted a decision by 4:00 p.m. for next-day delivery. I had two hours to decide how to spend an unplanned repair budget. Had two hours to decide before the overnight freight cutoff. Normally I'd have compared three quotes and broken out total cost of ownership (i.e., not just the part price but installation, freight, overtime, and probable rework). There was no time.

Our regular supplier quoted a direct replacement coupling: $610, next-day freight $98, available by 10:00 a.m. The bearing distributor quoted a generic equivalent at $540. Then I called RINGFEDER Power Transmission on a hunch. Edward Cole, Ringfeder Power Transmission's applications engineer, answered and asked questions the other two vendors hadn't: "What's the shaft surface condition? What was the original alignment?"

The vendor who asked questions

Edward explained that a keyed connection concentrates torque in a single line along the key. If the key fits a bit loose, or a locknut backs off, the key starts to hammer. Then the keyway grows. He said RINGFEDER makes locking assemblies that fix a hub to a shaft without a key. Our existing coupling hub could be refurbished by fitting a RINGFEDER locking assembly, or we could install a new RINGFEDER coupling hub with no keyway.

The pricing was different. The direct replacement was $610. The RINGFEDER option was $1,048, including the locking assembly and a new machined hub. My first thought was "that's 70% more for the same job." What I mean is, it didn't seem like the same job—it's a different way of transferring torque, but the procurement line item looked similar.

Then I did the TCO math on the back of a shipping envelope: direct replacement meant a 2-hour install plus alignment, call it $700 labor. The RINGFEDER locking assembly still needs clean surfaces and proper alignment, but there's no keyway to fret. If it lasts even a third longer, the extra $438 pays for itself. If it avoids a repeat failure during a night shift, the savings are huge. Still, I almost went with the cheapest quote because it was familiar. (Mental note: familiar is not a reason.)

Edward sent the RINGFEDER catalog page for locking assemblies. At our shaft size, the rated torque capacity was roughly 2.4 times the motor's nominal torque. Not a guarantee—he didn't claim one—but enough for me to explain the decision to finance.

A bearing, a stepper motor, and a question I couldn't answer

While waiting for the quote, the maintenance tech doing the teardown found a second problem: a worn thrust bearing on the tensioner roll behind the gearbox. The plant runs thrust bearings on all three tensioner rolls, and this one had been running with end play for a long time. Was it the cause or an effect of the coupling failure? Honestly, I'm not sure. My best guess is the shaft misalignment loaded the bearing, but we replaced it anyway.

Then the controls engineer walked over and asked a completely different question: "How fast can a stepper motor turn?" It sounded random, but he was planning a retrofit of the indexing conveyor that feeds the tube mill. He wanted to replace a geared AC motor with a closed loop stepper motor. He'd heard that adding an encoder lets a closed loop stepper motor recover from missed steps, and he wanted to know if it could hit the 850 rpm output we needed.

I don't have hard data on the torque-speed curve for that application without loading the motor. My honest answer: it depends. A closed loop stepper motor can sustain more speed than the same motor in open-loop because the encoder gives the driver shaft-position feedback to adjust phase currents. But the limit is still the torque-speed curve, the available voltage, and the load inertia. The power transmission cost comes from matching those three, not from the motor price. We didn't buy a closed loop stepper motor that day. I told him to send me the load inertia and I'd ask for a quote.

"How fast can a stepper motor turn?" is the wrong question. The right question is: "What does the torque-speed curve look like for my load?"

The RINGFEDER coupling went in without a fight

Edward Cole, Ringfeder Power Transmission's applications engineer, sent a drawing of the hub and asked for the shaft diameter and tolerance. I didn't have the shaft tolerance on hand. I sent him a photo of the shaft and a caliper reading. He said that if the shaft was at nominal diameter (I think it was 60h7), the locking assembly would work with the existing hub. We ordered the complete kit: a new RINGFEDER coupling hub and locking assembly.

The most frustrating part of the sourcing process: the original coupling's part number was stamped on a collar, but the stamp had worn clean. You'd think a machine builder would engrave the model number somewhere more durable. No such luck. That's why we had to wait for Edward to identify the hub from dimensions.

We hit "confirm" and I immediately second-guessed. What if the hub arrived and the bore didn't match our gearbox input? The four days until delivery were stressful. When the box arrived, the fit was right. (Thankfully.)

Installation took less time than I budgeted. No key, no file work on the keyway, no shimming to compensate for a loose fit. The shrink disc went on, we tightened the screws in the pattern specified in the installation sheet, and the coupling was solid. Alignment still had to be done—there's no shortcut for that—but the total install was about 40 minutes, not the 90 minutes I had in the quote.

Never expected the "expensive" option to have a lower installed cost. Turns out the difference was only $438 on paper, but the RINGFEDER coupling killed the hidden cost of rework. That was the surprise: not the part price, but how much of our original budget was tied to the keyway's future failure.

What I'd tell my previous self

When a keyed coupling fails, don't automatically replace it with the same keyed coupling. That sounds obvious now, but I had to sit through a two-hour debate to get there.

The TCO equation isn't just part price plus labor. It's the probability that the same failure recurs, the cost of the night shift, and the value of not getting called at 2:17 p.m. I don't have hard data on how many coupling failures we've had in this plant, but my sense is this was the first in six years. One data point isn't a trend. Still, I've already added "keyless locking assemblies" to our standardized spare parts list.

And about the stepper motor question: if you're weighing a closed loop stepper motor, ask the vendor for the torque-speed curve at your actual driver voltage. "How fast can a stepper motor turn" has no single number. It has a curve.

Edward Cole at RINGFEDER Power Transmission taught me another thing: a good vendor asks about the shaft before asking for the purchase order. That's a signal. I now ask any new vendor two qualifying questions: "What data do you need?" and "What have you seen fail?" If they answer with a question about my application, they're probably going to keep my machines running.

The efficiency gain came from changing the problem, not repeating it. Doing the same thing faster is motion; doing the right thing once is progress. The mill hasn't lost a shift to that coupling since. We still have the envelope with the TCO math taped to the wall. (Mental note: put it in the CMMS before someone cleans the board.)