Reading a wear-rate curve before it costs you a shutdown
Wear is rarely linear. Treating it as linear is how components fail between inspections.
Most wear planning assumes a straight line: measure thickness, divide by elapsed hours, project forward to the minimum allowable. It is simple, it is defensible, and for a lot of components it is wrong in the direction that hurts.
Why the curve bends
A liner, an impeller or a chute plate usually wears slowly at first, while the original profile still directs material the way it was designed to. As the profile degrades, flow patterns change, turbulence increases, and the wear rate rises. The last twenty percent of material often disappears in far less than twenty percent of the life.
A linear projection taken early therefore over-predicts remaining life, and the error is largest exactly when you are relying on it to schedule a shutdown.
What to do instead
Take at least three measurements before projecting anything, and plot them. If the interval-on-interval rate is increasing, project from the most recent interval rather than from the average. It is a cruder method and it is closer to reality.
Where a component has a history, use it. Two previous lives measured properly will tell you more about the shape of the curve than any amount of theory about the material.
