The number is never the whole story
A hardness callout can look precise on a drawing and still be incomplete. "6061-T6, 95 HB minimum" sounds specific, but it only works as a real requirement when the temper, the test method, and the test location are all tied to the same instruction. Leave any one of those out, and the number becomes easy to misread, easy to compare incorrectly, and easy to dispute later.
That is the part engineers miss most often: hardness is not a standalone property in aluminum extrusions. It is the result of processing history, and it is only meaningful when it is measured under a defined procedure. A profile made from 6063-T5, 6063-T6, or 6061-T6 may all be called "aluminum extrusion," but the hardness behavior is not remotely the same. A broader hardness reference guide makes that spread obvious across the common extrusion alloys.
Temper is the real driver
In aluminum extrusions, alloy chemistry sets the ceiling, but temper determines where the part actually lands. That is why two parts made from the same alloy can behave differently in machining, wear, dent resistance, and dimensional stability.
6063 is a good example. In T5 temper, it is commonly around 60 HB. In T6, it is higher, often around 73 HB. Nothing about the alloy designation changed. The difference came from thermal processing and aging response. The same pattern appears in 6061, where the T6 condition is substantially harder than the annealed or partially aged states. By the time you reach 7075-T6, hardness jumps again, because the precipitation-hardening response is much stronger.
That is why a spec that says only "6063" or "6061" is too loose for any part where hardness matters. It tells the mill what family to use, but not what property state to deliver. In production, that ambiguity shows up as parts that bend too easily, scratch too quickly, or machine differently from lot to lot.
A hardness value without a test scale is incomplete
Another common mistake is treating all hardness numbers as if they were interchangeable. They are not.
Brinell, Rockwell, and Vickers are all legitimate methods, but they answer slightly different questions and use different loads and indenters. A report that lists 95 HB is not directly comparable to one that lists 60 HRB unless the same test method and scale are being used and the values are being interpreted correctly. Even within the same alloy, the number on the report can shift depending on whether the lab used a Brinell ball, a Rockwell B ball, or a Vickers diamond pyramid.
That matters because a procurement note that simply says "hardness required" leaves too much room for interpretation. One supplier may certify Rockwell B because it is fast and practical for production QC. Another may use Brinell because the section is thicker. Both may be correct, but the results are not interchangeable without context.
The practical rule is simple: if the hardness method is not stated, the number is not complete enough to enforce.
The part itself can change the reading
Even when alloy, temper, and test method are all correct, measurement location can still change the story.
Extrusions cool unevenly. Surface zones, mid-wall zones, and core material do not always share the same response, especially before later heat treatment. On top of that, anodizing changes the surface entirely. A hardcoat layer can be dramatically harder than the substrate underneath, so a surface reading may describe the coating rather than the aluminum body you are actually trying to qualify.
That is why hardness specs should say where the test is taken:
- on bare metal or coated surface
- on the largest flat face or a defined wall
- at mid-wall, edge, or core, if the geometry allows it
- on a sample from each lot, each die, or each critical production run
If the drawing does not define location, a supplier can test a convenient spot that does not represent the part in service. Then the incoming inspection team gets a number that is technically valid but functionally useless.
The spec that actually works
A complete hardness specification is less about a single value and more about a controlled chain of information. The best callouts remove guesswork before production starts.
A practical requirement usually includes:
- Alloy and temper
- Applicable standard
- Hardness method and scale
- Test location
- Acceptance range or minimum
- Whether the coating counts or not
A sound purchase order might look like this in substance:
Alloy 6061-T6 per ASTM B221
Hardness minimum 95 HB per ASTM E10
Test on bare substrate at mid-wall on the largest flat surface
Report one reading per lot and identify the test location
That level of detail does two things at once. It gives the mill a process target that can actually be controlled, and it gives QA a standard that can be defended when results come back.
Without that structure, a hardness number is easy to misapply. A part can pass a loose inspection and still fail in service because the wrong location was tested or the wrong scale was used.
Why this mistake is expensive in production
Hardness problems rarely show up as dramatic failures on day one. They show up as slow, expensive friction.
A part that is softer than expected may dent during assembly, gall during sliding contact, or lose dimensional stability during machining. A part that is harder than expected may still be usable, but tooling wear rises, cycle times change, and forming margins disappear. In both cases, the root cause often traces back to a spec that named the alloy but not the temper, or named the hardness but not the method.
That is why the best engineers treat hardness as a process contract, not a lab result. The contract has to include how the material was made, how it was tested, and where it was measured. Once those three pieces are fixed, the number becomes meaningful. Before that, it is just a loose claim on paper.
The simplest rule to keep hard parts from becoming soft problems
When an aluminum extrusion must meet a hardness target, the real work is not finding a number. The real work is defining the conditions that make the number valid.
Alloy without temper is too vague. Hardness without method is too vague. A reading without location is too vague.
Put those together, and the specification starts to behave like an engineering requirement instead of a hope.
