From the LMM Rolls maintenance blog. This post is the on-blog edition of our field guide. The same article lives on our main site at lmmrolls.com/backup-roll-axial-displacement/ — that URL is the canonical version. If you arrived here from search, feel free to bookmark the main site for future updates and printable PDF exports.
Backup Roll Keeps Crawling Out Axially? Don't Just Torque the Lock Nut — Check These 4 Clearances
Backup roll axial displacement is a headache on hot strip mills, plate mills, and section mills alike. If you need a refresher on how backup rolls and work rolls share the load in a mill stand, this introduction to mill rolls covers the basics.
One end of the barrel wears shiny, the strip shape drifts to one side, and periodic edge marks show up on the strip. You shut down and find the lock nut has backed off half a turn. The reflex of every repair crew: torque it back on, add a little extra.
A few days later, loose again. Tighten, loosen, tighten. The nut gets cranked harder and harder, the chock threads are close to stripping, and the roll still creeps.
Plenty of crews work this way: "If it creeps, tighten. If tightening doesn't hold, swap the nut." But any engineer with a decade on the floor knows better. In eight cases out of ten, backup roll axial displacement doesn't start at the lock nut. It lives in one of four other clearances.
Blind tightening doesn't solve anything. It can crush the thrust bearing and strip the chock threads — a far more expensive failure.
So let's take these four clearances apart one by one. Each one can be measured and judged right at the mill. Follow the sequence and you'll find the root cause.
First, be clear on what "creeping out" actually means
A backup roll is designed to allow a small amount of axial float. Locking it up dead solid is actually wrong — with no room for thermal expansion, the thrust faces get pounded.
Normal range: design float for backup roll axial displacement generally sits between 0.3 and 0.8 mm (values differ by mill type; the equipment manual is the authority). Within that band, the barrel can float slightly and release thermal stress.
Signs the float has gone beyond spec:
- Visible end play at the barrel. Push the roll end by hand after a shutdown — you can feel movement beyond 1–2 mm.
- The lock nut shows marks of having rotated. The match marks on the nut face and the chock face no longer line up, which means it came loose in operation.
- Rubbing marks on the chock face. An abnormal shiny wear band where the chock meets the housing window — the chock has been shifting inside the window.
- Periodic edge defects in strip shape. Indentations or waves appear along the strip edge at a spacing that matches the roll's rotation period.
Any two of these together means the axial float is out of spec — and tightening the nut alone won't fix it.
The four clearances, in the order you should check them
1. Chock-to-housing-window clearance — the big looseness everyone overlooks
Backup roll chocks sit in the housing window and locate against its side faces and sill. Once the window wears, the chock gains room to move.
During rolling, axial force travels through the chock into the housing. Any clearance between chock and window lets the chock itself shift slightly, and that shift adds directly to what you see as roll "creep."
How to measure:
- Feeler-gauge all four sides of the chock against the window. Give the axial locating faces on the drive side and operator side the closest look.
- A single-side clearance over 0.5 mm must be dealt with — shim it or repair the window.
- Check contact between the chock bottom and the window sill with a 0.05 mm feeler; it should not go in deeper than 20 mm.
2. Lock cylinder pressure / disc spring preload decay — not enough force to hold
Hydraulically locked backup rolls get their axial preload from a lock cylinder. Mechanically locked ones rely on a Belleville disc spring pack. Both fatigue over time.
Hydraulic lock cylinders:
- Aging seals leak internally; pressure-holding capacity drops, and the setpoint decays while the mill runs.
- A bent or sticking piston rod never reaches full stroke, so preload falls short.
- A leak anywhere in the hydraulic line keeps system pressure down.
Disc spring packs:
- Years of compression leave permanent set. Free height drops and preload fades.
- One disc fractures and the whole pack loses force.
- Discs installed facing the wrong way (the cone direction matters) compress by the wrong amount.
How to measure:
- Hydraulic: read the system gauge. Hold at locking pressure for 5 minutes — the drop should stay within 10% of the set value.
- Disc springs: measure pack compression with a dial indicator and compare against factory data. Replace the pack if preload has decayed more than 15%.
- On disassembly, inspect every disc for deformation, cracks, and corrosion.
3. Thrust bearing wear or burnout — the last line of defense gone
Both ends of a backup roll usually carry thrust bearings (thrust ball or thrust roller types) whose only job is axial force. Wear them out or burn them, and nothing limits axial float anymore.
Common failure causes:
- Poor lubrication. The thrust bearing sits far outboard; oil or grease never quite arrives, and dry friction burns the bearing.
- Abnormal axial loads. A worn window or weak locking force dumps shock loads onto a bearing that was never meant to see them.
- Excessive mounting preload. The bearing is pressed too tight, no oil film forms between rolling elements and raceways, and early burnout follows.
How to judge:
- After a shutdown, barring the roll by hand, a distinct click-click play when pushing axially means internal wear.
- On disassembly, check raceways for spalling, rolling elements for fracture, and the cage for deformation.
- Inspect the thrust bearing seat faces for brinelling or distortion.
4. Thermally induced axial force — the invisible force
During rolling, the barrel surface and the roll core sit at different temperatures, and the two ends of the barrel can differ too. Uneven thermal expansion turns into axial thermal stress.
When is the thermal push worst:
- Unsteady rolling rhythm. Fast-slow pacing swings the barrel temperature widely.
- Uneven cooling. A blocked or badly distributed roll cooling water spray leaves one end of the barrel cool and the other hot.
- Fresh roll after a roll change. A cold roll thrown straight into rolling takes a heavy thermal shock — the guide to the use of work rolls walks through correct handling and warm-up practice.
Thermal axial force can reach tens of kN, sometimes above a hundred. With locking force and thrust bearings healthy, the stack holds it. But with clearance in the window and decayed preload, the same force pushes the roll right out.
How to blunt it:
- Warm up a new roll at low speed before ramping up — let the barrel heat evenly.
- Keep the cooling water system clean, so the barrel cools evenly along its full length.
- Wind speed and load down gradually before a shutdown — no hard stops.
The order of work: measure first, tighten second
The sequence decides how fast you find the problem. Work down this list:
- Chock-to-housing-window clearance (feeler gauge)
- Locking force — lock cylinder pressure or disc spring preload
- Thrust bearing condition
- Thermal axial force — review rolling rhythm and cooling practice
Working principles:
- Fix the clearance problems (window, locking) before replacing any bearing.
- Torque the lock nut to the value in the equipment manual. Don't pile it on. Excessive torque means excessive bearing preload, which means heat, which means burnout.
- Log the numbers after every intervention and keep a trend record. Next time the roll creeps, comparing against history points you to the cause much faster.
The one-paragraph takeaway
Backup roll creeping out axially? Don't reach for the torque wrench first.
Check whether the window clearance is excessive, verify the locking force, inspect the thrust bearing, then ask whether thermal push is at work.
Those four steps beat breaking ten lock nuts.
Frequently asked questions
How much axial float is normal for a backup roll?
Most mill designs allow 0.3–0.8 mm of axial float on a backup roll (exact values vary by mill type — always follow the equipment manual). A small, controlled float gives thermal expansion somewhere to go and protects the thrust faces. A roll that is locked up completely solid will damage its thrust faces during warm-up.
Why does retightening the lock nut fail to stop backup roll axial displacement?
Because the lock nut is rarely the root cause. In roughly 8 out of 10 cases the real problem is one of four clearances or force losses: chock-to-housing-window clearance, hydraulic lock cylinder pressure or disc spring preload decay, thrust bearing wear, or thermally induced axial force. Over-torquing the nut does not fix any of these — it can crush the thrust bearing and strip the chock threads.
How do you measure the clearance between the roll chock and the mill housing window?
Use a feeler gauge on all four side faces of the chock, with special attention to the axial locating faces on the drive side and operator side. A single-side clearance above 0.5 mm must be corrected by shimming or window repair. Check the chock bottom against the window sill with a 0.05 mm feeler — it should not insert deeper than 20 mm.
How do you test a hydraulic lock cylinder or Belleville disc spring pack for preload loss?
For hydraulic locking: bring the system to locking pressure, hold for 5 minutes, and watch the gauge — the pressure drop should not exceed 10% of the set value. For disc springs: measure pack compression with a dial indicator and compare against factory data; replace the pack if preload has decayed more than 15%. During disassembly, inspect every disc for permanent set, cracks, and corrosion, and confirm the discs are not installed facing the wrong way.
What are the signs of a failing thrust bearing on a backup roll?
Barring the roll by hand after a shutdown, you feel a distinct click-click play when pushing axially. On disassembly, look for spalled raceways, broken rolling elements, or a deformed cage, and check the thrust bearing seat faces for brinelling. Typical causes are starved lubrication at the outboard position, abnormal axial shock loads caused by window clearance or weak locking force, and excessive mounting preload that prevents oil film formation.
What is thermally induced axial force, and how can it be reduced?
During rolling, the roll surface and core — and the two ends of the roll body — sit at different temperatures. Uneven thermal expansion generates an axial force that can reach tens of kN or more. If locking force and thrust bearings are healthy, the stack holds it; if clearances are worn, the roll gets pushed out. Reduce it by warming up a new roll at low speed after a roll change, keeping roll cooling water evenly distributed along the barrel, and ramping speed and load down gradually before a shutdown instead of stopping abruptly.
