Hardness is a measure of how a metal resists being deformed locally. A small indenter is pressed into a prepared surface under a controlled load, and the size of the impression that remains is converted into a value. A smaller impression means a harder material at that point.
That single number is useful because it correlates with other properties. A harder area generally has higher strength and lower ductility, which tells you something about how it will behave under load and, just as importantly, how it will behave when loaded repeatedly.
A weld is not one material. It is the filler and the melted parent metal in the bead, the heat-affected zone beside it, and the unaffected parent metal further out. Heating and cooling change each of those areas differently, so the same part can hold several different hardness values across a joint.
That is the practical reason for measuring at all. A weld that has cooled quickly, or a heat-affected zone that has been left in an awkward state, can be markedly harder and more brittle than the metal around it, even though it looks sound and passes a visual check without comment.
The surface must be flat, clean and smooth enough for the impression to be read accurately. Paint, scale, a coarse ground finish or a curved face all distort the result. This is why hardness is often taken on a prepared test coupon made alongside the part rather than on the finished component, where dressing the surface would be unacceptable.
Readings are taken in a pattern rather than as one spot: across the bead, into the heat-affected zone on each side, and out into the parent metal. The pattern is what makes the result interpretable, because a single value taken anywhere in that band could describe several very different conditions.
It does not prove a weld is sound. An unfused joint, an internal void or a crack can sit in metal whose hardness readings look entirely normal. Hardness describes the state of the material, not the continuity of the joint.
It also does not measure strength directly, because the relationship between the two depends on the alloying, the thickness and the way the part is loaded. Nor does it predict fatigue life. A hard, brittle zone can crack under cycling while a softer one merely deforms, and only a test that applies load can show that difference.
Custom work is where hardness checks earn their place, because the material and the process are both being defined for that job. When a new mould or drawing moves into production, knowing how the finished welds came out confirms that the process used is the one that was intended, rather than something that drifted between the sample and the batch.
It is also relevant when a part in service has failed. If a joint cracked where it should have deformed, the hardness pattern across that area is one of the few pieces of evidence still available, and it may point to a process problem rather than a design one.
The step sits in the middle of a longer chain, alongside dimensional measurement rather than instead of it. A sample is made from the mould, an engineer measures its sizes, hardness is checked, and the sample is then sent for confirmation before mass production begins. A part can be dimensionally perfect and still be wrong if its welds came out brittle.
For stainless work in particular, hardness is a useful early signal, because the material responds to heat in ways that are not always visible in the finished joint. That is why the check belongs in the wider sequence described on the custom work page rather than being treated as an optional extra, and why it is applied routinely across the TIG stainless steel welding side of the work.
