“Good or bad” is three different questions

A profile can be beautiful and wrong. It can have a flawless surface, correct dimensions and still be in the wrong temper, which means it will take a permanent set under a load that the drawing said it would carry. Or it can be right in every measurable way and arrive with paperwork that makes it impossible to prove anything later.

So the useful framing is not a single quality judgement. It is three separate questions, each with its own detection method and its own cost of failure:

Is it the right material? Alloy and temper. Invisible. Requires testing, and the test result has to be tied to the specific batch in your hands.

Is it made to the drawing? Wall thickness, slot geometry, envelope, straightness, coating thickness. Measurable on site, and this is where most real non-conformance sits.

Is it documented and traceable? A certificate that reports actual measured values for that batch, and marking that lets you go from a finished assembly back to the production lot. This is what turns a suspected problem into a bounded one instead of a warehouse-wide recall.

The popular five-step checklists start with visual inspection because it needs no equipment. That ordering is honest about convenience and misleading about value. A surface can be perfect on material that is under-aged, and a wall can be thin behind a coating that looks uniform.

Tier 1: ten minutes at the dock

These checks need a caliper, a steel tape, a flat surface and a scale. They catch the majority of dimensional and handling problems, and they cost almost nothing to run on every delivery.

Take the measurements from more than one piece and from more than one place in the bundle. The top length of a bundle is not representative, and a single sample tells you nothing about consistency, which is usually the actual complaint.

Tier 2: four measurements worth owning

Beyond the basic dimensional checks, four measurements earn their keep in a receiving area because each one catches a defect that the dock checks structurally cannot see.

The cheapest and most underrated of these is the scale. Weigh a known length, divide by the length, and compare with the theoretical mass from the drawing. Aluminium alloys sit at roughly 2.7 g/cm³, so a shortfall in measured mass per metre maps directly to missing cross-sectional area. It will not tell you which wall is thin, but it is a whole-length average taken in under a minute, and it is the single fastest way to find the defect that visual inspection is blind to.

Coating thickness is worth owning for the same reason, because it converts a word into a number. “Anodised” is a process; a thickness class in micrometres is a specification. An eddy-current gauge is the normal method, and it has to be calibrated on bare substrate and read in several places, including areas that are harder to coat such as recesses and internal corners.

Portable hardness is an indication, not a verdict. It is good for screening lots against each other and for catching material that has clearly not been aged correctly. It is not a substitute for a tensile test when a specification limit is being enforced.

The functional gauge is the one that catches what drawings miss. Fit the actual connector, tighten to the intended torque and look at what happens to the slot. A profile that measures correctly and still spreads its slot walls under a tightened fastener is a real failure, and no dimensional check on its own will find it.

Tier 3: what has to go to a laboratory

Some things cannot be settled on a receiving dock, and pretending otherwise is how bad material gets signed off. These are the items where a competent buyer specifies the test, the standard and the frequency in the purchase order rather than hoping.

Chemical composition. Optical emission spectrometry gives the actual alloying elements. This is the definitive answer to “is it the alloy that was ordered”, and it is the only answer that will hold up in a dispute.

Tensile properties. Proof stress, tensile strength and elongation from a machined sample. This is what a mechanical specification is actually written against, and it is the only way to confirm a temper claim.

Metallographic examination. Grain structure, and for hollow sections the quality of the longitudinal seam welds formed as the metal splits and rejoins around the die. Seam quality matters when the section will be bent, or loaded across the seam, and it cannot be seen from outside.

Coating durability tests. Salt spray exposure and cross-cut adhesion. Salt spray is a comparative test: it is useful for comparing one lot or one supplier against another, and it is not a prediction of service life outdoors. Treating it as a life prediction is one of the most common misreadings in incoming quality.

The certificate: 2.1, 2.2 and 3.1 are not the same document

“We have certificates” is one of the least informative sentences in purchasing. Under EN 10204 the common document types mean materially different things, and only one of them reports measurements from your batch.

For anything structural, ask for 3.1 by name. Then make the document usable: require that the certificate carries a heat or lot number, and require that the same number appears on the bundle tag or the stencilled marking on the material. A certificate without a matching physical mark is a document about some aluminium, not about your aluminium.

The same logic applies to the definition of a lot. If you cannot say which lengths belong to which lot, then a single failed sample either condemns everything or condemns nothing. Agree what constitutes a lot before the first delivery, and mark accordingly.

Temper is the invisible variable

If there is one thing to take from this article, it is that temper is a heat treatment state, not a surface feature. Two lengths of the same alloy in different tempers are indistinguishable by eye, by weight and by dimension, and they behave differently under load.

The consequence of getting temper wrong is not usually a dramatic break. It is a frame that deflects more than calculated, a joint that loses preload and works loose, or a bracket that takes a permanent bend during assembly. All three look like a design problem from the outside, which is why the wrong temper can go undiagnosed for a long time.

The practical controls are simple: name the alloy and temper on the order, require a 3.1 certificate that states them, screen lots with portable hardness so that a change is noticed, and escalate to tensile testing when a change appears or when the application is structural. The guaranteed minimum values for each alloy and temper are tabulated in the relevant material standard, so the order should reference the standard rather than a number remembered from a datasheet.

What certifications prove, and what they do not

Certificates get asked for and rarely get read, which is why they are worth separating into what they actually carry.

A supplier can hold ISO 9001 and still ship the wrong temper, because the certificate is about the system, not about the delivery. The documents that carry product data are the mill certificate, the coating report and your own measurements. Treat management-system certificates as one input to supplier selection, and never as evidence about a specific consignment.

Write the checks into the purchase order

Inspection only works if there is something to inspect against. Most disputes trace back to an order that said “aluminium profile 4040” and left everything else to convention. Eight lines turn that into something enforceable.