Most rolling mill cardan shafts don't fail all at once. They go in stages. First there's a little backlash you can hear at low speed. Then a ring of grease appears around a seal. Then a vibration shows up that everyone stops mentioning because it's been there for months. Then a flange bolt lets go on a Saturday shift, and the mill sits idle while people hunt for a spare that nobody ordered.
If you work on a hot strip mill, a bar mill or a section mill, you know the sequence. The shaft is one of the lower-cost parts in the drive train, and one of the least forgiving when it's wrong. What follows is what we've learned building them — the sizing decisions that matter, where they actually wear, and what to check before the next roll change.
Follow the torque path: motor, reducer, pinion stand, spindle, work roll. That spindle is a rolling mill cardan shaft. Passing torque along is only part of the job. It also has to absorb the things the rest of the train can't:
Angular misalignment. Roll gap changes, and roll diameters shrink every time they're ground. The shaft keeps working at an angle that shifts through the campaign.
Axial movement. Roll changing slides the shaft in and out, so most designs carry a splined telescopic section.
Torque reversal. On a reversing mill, every pass flips the load direction. The shaft sees full reversal fatigue rather than steady load.
Shock. Biting a cold slab or a mis-fed billet sends a torque spike well above the running value.
Older mills managed this with slipper spindles or gear spindles. The cardan shaft won on angle capability and on how easily it can be repaired. In a lot of mill documentation you'll still see it written as a spindle or a universal joint spindle — same part, different name.
If you're diagnosing one, or picking a replacement, these are the points worth knowing.
The most common source of backlash. Grease thins out, mill scale and water get in, and the spline flanks wear into a step. You hear it as a clunk every time the drive reverses.
The trunnion bearings take a beating. Grease starvation, or a seal that let water through, and the races pit and spall. Once there's play here, the shaft goes out of balance and the vibration starts feeding back into the stand.
A cheap seal is usually what starts an expensive repair. When it fails, water and scale get into the joint and the grease gets out.
Under reversing duty the flange bolts see cyclical load. Cracked or stretched bolts, or bolts that have gone soft after repeated re-torquing, show up as growing runout before one lets go.
A shaft that was rewelded, or had a flange swapped without re-balancing, can run rough at full mill speed.
When people quote a cardan shaft, they often quote a peak or nominal torque. On a reversing mill that isn't the number that decides whether the shaft lasts. The design driver is the fatigue torque — the continuous, fully reversing load — measured against the shaft's rated fatigue limit. Quote on peak torque and a shaft can look fine on paper while it's under-sized for the duty it actually lives in.
Worth pinning down before you order:
Fatigue torque for your duty, not the peak figure. Reversing mills and continuous mills are not the same problem.
Tilt angle at the worst case. Torque capacity drops as the angle goes up, so size at the maximum working angle, not the installed one. Roughing stands that move a lot need more headroom than a fixed finishing stand.
Swing diameter and length against the space you really have, including the roll change stroke.
Balance grade if the mill runs fast. Above a certain speed, an unbalanced shaft keeps finding the weakest bracket.
Environment. Radiant heat, cooling water and scale all shorten seal and grease life. Say so when you order, because seals and lubricant get specified differently for a roughing stand than for a finishing stand.
Most shafts in the field are forged alloy steel — 42CrMo and 34CrNiMo6 are the usual grades — quenched and tempered, with the spline and journal surfaces hardened afterwards. Induction hardening or nitriding on the spline flanks is what buys the wear life. A soft spline is why a shaft comes back early with backlash. Forged beats cast here, because reversing load finds every inclusion in a casting.
The heat treatment spec isn't a formality. Hardness, case depth and core toughness all have to line up with the torque and the shock level of the stand. A shaft that's hard but brittle will crack at the flange. A shaft that's soft will wear out at the spline. The answer sits in between, and it depends on the mill.
Two families cover most of what a mill needs.
Flange type (SWC). Bolted flange connections at both ends. Standard for main drives where the flange pattern and torque are known.
Bearing block type (SWP). Uses a bearing block and journal, common where the mill was originally built for that arrangement.
A roughing stand with big angle changes and heavy shock is a different animal from a finishing stand running close to steady. It's worth specifying them separately instead of standardising one shaft across the whole line. The finishing stand usually doesn't need the same mass, and a lighter, better-balanced shaft is easier on its bearings.
You don't need a shutdown to catch most of this. Keep a simple record per stand and check the same items at each roll change:
Backlash at the spline by hand — measure it, write it down, watch the trend.
Grease at the joints — colour, water, scale.
Seal rings — any weeping.
Flange bolts — torque and a visual crack check.
Vibration trend at the stand.
Runout after any repair.
A shaft that drifts from two millimetres of backlash to eight over six months is telling you something. One that goes from eight to twenty in a fortnight is telling you to stop running it.
Send the supplier the drawing if you have it. If the drawing is lost — which happens more often than anyone admits — send the stand data instead:
Fatigue torque and peak torque
Working angle range
Length and swing diameter
Flange bolt circle and bolt count, or journal dimensions
Spline spec
Mill speed and the hot-rolled environment it lives in
Keep one spare per critical stand. The shafts that hurt most are always the ones on the stands everyone assumed would never go.
It depends on duty more than on the calendar. A finishing stand running steady might go years between overhauls. A reversing roughing stand with heavy shock and large angle changes is a much harder life. Track backlash and vibration, and let the trend set the interval instead of a fixed date.
A gear spindle uses crowned gear teeth to take the angle. A cardan shaft uses a universal joint with cross journals and bearings. Cardan shafts generally handle larger angles and are easier to repair in place, which is why most new and rebuilt mills use them.
Both are possible. Splines, seals, cross journals and flanges can be rebuilt on many designs. A shaft with a cracked flange or a bent body is usually finished, though, and a repair that skips re-balancing will cost you in vibration. Get the repair quoted against the price of new before you decide.
Use the fatigue torque of your duty cycle as the main figure, then check the peak against the shaft's peak rating. If a supplier only asks for a peak number, that's a sign to slow down.
Usually one of four things: play in the spline or the cross journals, a lost balance after repair, a bent shaft, or a loose flange. Work out which one it is before you re-torque anything, because tightening bolts fixes none of the first three.
If you're working out a replacement, or building a spare for a stand you can't afford to lose, send over the drawing or the stand data — torque, angle, length, flange pattern. We'll come back with a sizing check and a lead time.
Contact: Spark Wang
Phone: 86-19932127135
E-mail: admin@rongxingcoupling.com
Add: North side of Planning 4th Road, Southern Industrial New Town, Jizhou District, Hengshui City, Hebei Province, China