Tooling gets discussed as a capital item, which makes it sound permanent. It is not. Every cycle takes something off a mould: a little polish from a surface, a little clearance from an edge, a little sharpness from a corner. The changes are small enough to be invisible between one part and the next, which is exactly what makes them worth watching.
What a shop is really managing is not whether a tool wears. It is whether the wear is measured before it reaches the part.
The first things to change are the ones that touch the most. Forming surfaces lose their finish where material slides across them, so parts begin to show drag marks that were not there on the approval sample. Edges that trim or shear lose their sharpness, and the cut face starts to show a burr or a tear instead of a clean break.
Clearances open up. Anywhere two surfaces are meant to meet at a controlled gap, wear makes the gap larger, and a larger gap shows up as flash, a heavier witness mark, or a part that sits looser in the tool than it used to. None of this announces itself. It arrives as a slightly different part.
A worn mould does not produce obviously bad parts. It produces parts that differ slightly from the ones approved, and the difference accumulates. A dimension that has moved inside its tolerance is still in tolerance. When it moves past the limit, the change is not dramatic. It is the same small movement that has been happening all along, one run too far.
This is why the comparison that matters is not against the drawing alone. It is against the first approved part, or against the last verified one, which is the same reference.
On a formed part it appears as a radius that has grown, a flange standing slightly taller, a hem that no longer closes flat. On a trimmed part it appears at the edge. On anything with a mating face it appears as a gap that needs more weld to fill, which quietly changes the heat input and therefore the distortion at the next operation.
Finishing hides a good deal of it. A polished part will look correct long after its geometry has started to move, because polishing removes the evidence of a small mismatch. That is why inspection on older tooling has to be dimensional rather than visual, and why a shop that inspects by eye will be the last to notice.
The practical method is to hold a small set of check dimensions and verify them on the first part of every run and again on the last. The first-part check catches a tool that was set up wrong. The last-part check catches a tool that moved while it was running.
Alongside the gauge there are signs that need no instrument: how much effort the tool takes, how the part releases, whether the operator has started doing something by habit that they did not do before. Someone quietly tapping a part out of a tool is telling you the tool has changed.
Most tooling wear is recoverable. Surfaces are repolished, edges are reworked, worn details are replaced, clearances are reset. The cost is time off the tool, which is why maintenance is planned against a production schedule rather than performed when a part finally fails.
The judgement call is when rework stops being worth it. At some point the accumulated corrections have changed the tool enough that a new one is the better answer, and that decision is usually made on the evidence of the last few runs rather than on a fixed life.
At the end of a run the tool has done the most work it will do before maintenance, so the last parts are the ones most likely to be out of family. A shop that samples only the first part is checking the tool it started with, not the tool it finished with.
For a buyer this is worth asking about directly. If a supplier can describe how they check the end of a run, they have thought about wear. If the answer only covers first-off approval, they have thought about setup, which is a different problem.
Tooling carries a record: when it was made, what it has run, what has been corrected, what is due. That record belongs with the customer as much as the tool does, because it explains why a part made on an older tool differs from one made on a newer tool of the same design.
Where tooling is owned by the customer, the record should travel with it. Where it is held by the supplier, the buyer should still be able to ask what condition it is in and what has been changed since approval.
The questions are simple, and the answers are informative even when they are vague.
A supplier who can answer these is describing a process. One who cannot is describing a tool they hope will keep working.
Wear is one of the reasons a mature part gets cheaper and a new one does not, and it is visible in the way a workshop plans its tooling. The titanium fabricated components we make are typical of the case, because each shape tends to start with its own tool. The customization page explains what a job needs before tooling begins, our plant describes where that tooling is built and maintained, and other parts and further fabricated parts show the smaller components where wear is easier to see on the finished face. Aluminium catch cans are a useful reference point for how a simple formed shape behaves once a tool has run for a while.
