Procedure qualification begins with a piece of metal that looks like waste. Two plates or two tube sections are joined exactly as the production joint will be joined, and then the result is deliberately destroyed. Viewed from the office, the exercise seems backwards: the shop has just spent material and a welder's afternoon producing something it immediately cuts apart.
The point is that the coupon is the only place where the truth about a procedure can be established cheaply. A production part that fails in service costs a customer, a return and a reputation. A coupon that fails in the test shop costs an afternoon and produces a fact. The discipline exists to move failure upstream to where it is affordable.
A coupon is only worth what it represents. If it is welded flat while production is welded out of position, the test measures the wrong thing. If it uses a heavier wall because that was what sat on the rack, the heat flow is wrong and the results do not transfer.
Representativeness means the coupon carries the same material condition, thickness band, joint form, filler and essential variables that production will use. Anything that changes how heat leaves the joint, or how much metal must be deposited, belongs in the coupon rather than in a footnote.
Titanium demands more care here than stainless. A coupon welded in open air with a casual trailing shield can look perfect on the face and be contaminated underneath. Our TIG titanium welding parts work is built around that constraint, because a titanium coupon that was not properly shielded is not a conservative test, it is a meaningless one.
The bend is the crudest and most informative test in the set. The weld is cut into strips, the reinforcement ground flush on the face that will be stretched, and the strip folded over a former until the weld section has been pulled through a severe radius. The test looks for ductility across the joint, and it exposes anything that behaves like a crack.
Lack of fusion at the root opens into a visible tear. Trapped oxide or slag in the wall shows as a split along the fusion line. A weld that is sound but too hard, usually because it cooled quickly or was over-alloyed, cracks at the edge of the heat-affected zone rather than in the deposit. Where the failure sits tells you which variable to look at.
Root bend and face bend pull opposite sides of the joint through tension, so they interrogate different regions. A procedure that passes only one is qualified on one side of the weld, and that asymmetry is worth noticing when you read a record.
The macro is a cross-section cut from the coupon, polished and etched so the weld profile appears as shapes. It answers questions the bend cannot. The bend tells you whether something failed; the section shows the geometry that produced it.
On a good macro you can see penetration into the root, the fusion boundary on both sides, the width of the heat-affected zone and the shape of the reinforcement. You can see whether the root is concave, whether there is a notch at the toe, and whether the weld bit into one side more than the other. Every one of those conditions can pass a visual check on the outside of a production part.
A hardness traverse taken in steps from base metal through the heat-affected zone and into the deposit maps how the thermal cycle changed the material. On stainless the interesting region is just outside the fusion line, where the steel went through a temperature excursion without melting and where the resulting structure can be harder and less forgiving than either parent metal or weld.
The traverse is not there to produce a verdict on its own. It shows whether the cooling rate implied by the procedure produced something the parent metal can live with. A hard band beside the fusion line is a signature of fast cooling, and fast cooling on thin wall often means the welder travelled quickly to compensate for low current.
Most qualification failures are not mysterious. They arrive in a small number of recognisable shapes, and each points at a part of the procedure rather than at the welder's character.
Incomplete fusion at the root usually traces back to a gap too tight for the current used, or a root left without purge so the pool never wet in. Scattered porosity traces to gas that was too little, too turbulent, or contaminated by moisture on the filler. Cracking traces to restraint and cooling rate together, which is a fixture question as much as a welding one.
After a failure, the shop has to change one variable at a time. A coupon that fails, is re-run with different current, gas and gap, and then passes has taught nobody anything. The record needs to show which change fixed it, or the next part fails for the original reason. You can see how that discipline shows up in production on our stainless steel exhaust mufflers and titanium fab components pages, where the same joints repeat across a product family.
Procedures are not permanent. Any change to an essential variable invalidates the qualification for that joint, and the shop has to judge whether the change is essential. Thickness moving outside the qualified band, a different material family, a different filler classification, or losing access for a second side: those are the usual triggers.
Non-essential changes, such as a different brand of the same filler or a new fixture, normally do not require re-testing. The judgement sits in the middle, where a change looks small but alters heat flow. Adding a chill bar is the classic case, because it is a fixturing decision that changes cooling rate, and cooling rate is exactly what the hardness traverse measured.
The qualification record is what makes the coupon worth the material. It must identify the coupon, the material, the filler, the variables used, the results and who witnessed them. Without that, the test happened and left nothing behind.
Photographs of the macro belong in the record, and so does the failed specimen if there was one, because a failure described only in prose loses the detail that makes it useful years later. Shops that keep specimens find that the failure they are investigating is often one they solved before.
When a supplier says a joint is qualified, the useful follow-up questions are short and specific. They are also the questions a good supplier enjoys being asked.
Tigindustry has been making welded parts in Shanghai since 2009, and the qualification habit is one reason the same joint can be ordered again years later and come back the same. You can see the range on our stainless steel exhaust system and exhaust pipe pages, or browse by alloy under shop by material.
