Reading wear patterns

A worn liner leaving a mill carries more process information than most of the instrumentation around it. It has spent months integrating everything the plant has done — feed changes, throughput increases, flow disturbances — and the record is written on its surface.

It is usually scrapped without anyone looking at it.

Uniform wear is the exception

The default assumption is that a component wears evenly and is replaced when it is thin. In practice the loss is almost always uneven, and the shape of that unevenness is the diagnostic.

Concentrated in a narrow band. Flow is being directed at that line. This is geometry, not material — and the answer is to move the stream, fit a sacrificial element there, or accept and plan for it. A harder material in this position buys some time and leaves the cause untouched.

Shifted from where it used to be. Something has changed: feed point, particle size distribution, throughput, a partial blockage upstream. The part is reporting a process change that may not have been noticed elsewhere.

Heavier on one side of a symmetrical assembly. Alignment, distribution or an installation issue. Worth resolving, because it will repeat with the replacement.

Even and symmetrical. Normal duty. This is the case where a material comparison is genuinely the right lever.

Where the pattern is most informative

Mill liners — the wear profile changes grinding action long before the liner is thin. Lifters that have worn rounded no longer lift the charge as designed, and throughput and grind size drift with them. Replacement is often justified by the profile rather than by remaining thickness, and a plant that replaces on thickness alone may be running months of degraded grinding it has not attributed to anything.

Pump impellers and casings — wear degrades hydraulic efficiency continuously. The energy cost of that degradation accumulates quietly and frequently exceeds the cost of the wear part itself. The pattern also distinguishes causes: cavitation damage looks different from abrasion, sits in different places, and calls for a different remedy.

Chutes and transfer points — wear marks the impact zone precisely. It shows where the stream actually lands, which is regularly not where the design intended.

Pipe bends — loss on the outer radius is expected. Loss elsewhere indicates the flow regime is not what was assumed, often because velocity or solids concentration has moved away from design.

Measure a few points consistently, not many points once

Detailed one-off surveys are less useful than a simple set of measurements taken the same way at every change.

Define a small number of points — four to six is enough for most components — mark them so the same locations are used every time, and record remaining thickness at each along with hours or tonnes in service. The absolute values matter less than the trend, and the trend only exists if the method stays constant.

That trend answers questions that no single inspection can: whether wear rate is stable or accelerating, whether a material change actually helped, and whether a process change had a maintenance consequence nobody linked to it.

Wear rate is not linear

Extrapolating early measurements to predict a replacement date almost always overestimates remaining life.

The reason is that wear changes flow. As a profile degrades, the path shifts, turbulence increases, impact angles worsen, and the rate rises. The last portion of a component’s life is shorter than the first — sometimes considerably.

Planning on linear extrapolation produces a replacement date that arrives after the component actually needed changing. In a shutdown-governed plant, that error surfaces as an unplanned stop.

The question worth asking at every change

Before ordering a replacement, one question separates a fix from a repetition: is this wear pattern what I expected?

If yes, the decision is about material, interval and cost. If no, something has changed, and fitting an identical part means the same outcome will arrive on the same schedule.

Most upgrades that disappoint were answers to the first question, applied to a situation that was actually the second.