Direct answer: Thin-sheet weld gap troubleshooting starts by separating assembly gap from process parameters, before the machine is adjusted at all. Incoming blank size and thickness, and the forming step that closes the shell, decide the fit-up the arc receives. A parameter window absorbs only a bounded amount of gap variation, so the troubleshooting order is incoming material and forming first, welding parameters second.
Who this is for: This guide is written for process and quality engineers troubleshooting an automatic thin-sheet welding line, and for the people who have to decide whether the machine, the material, or the forming step is at fault.
Scope: It covers how to separate assembly gap from process parameters, what evidence each explanation requires, and how to record tolerances so the argument does not repeat. It does not cover parameter development for a specific material and joint, nor the mechanical design of forming equipment.
Thin-sheet weld gap troubleshooting starts with two lookalike failures
Burn-through, lack of fusion, an inconsistent bead, and a wandering seam all present as welding defects, and the first response is usually to adjust welding parameters. Sometimes that is correct. Often the parameters were adequate and the joint presented to the torch was not the joint the parameters were developed for.
EVST separates these two before touching a parameter set. Thin-sheet weld gap troubleshooting depends on that distinction, because the two causes behave differently over time. A parameter problem is stable: the same setting produces the same result on every part. A fit-up problem is variable: the same setting produces different results as the incoming material and the forming step drift. If your defect rate moves without anyone changing the machine, look upstream first.
This is more pronounced on thin section than on plate. A gap that a heavy-plate procedure tolerates can be most of the material thickness on light gauge, and the process has correspondingly less room to compensate.
Start at the blank
The first measurable is what arrives. Blank length sets the developed circumference and therefore whether the shell closes with a gap, an interference, or a mismatch at the seam. Thickness variation across the blank changes both the forming result and the heat balance at the joint. In the reference footage for this guide the blank is measured for size with a tape and for thickness with a dial gauge before it enters the machine.
That check is worth formalising rather than leaving to whoever is on shift. Recording incoming dimensions against the tolerance you actually purchased — not the nominal figure — gives you the input distribution. Without it, every later argument about fit-up is anecdote.
Material condition matters alongside dimensions. Surface coating, oil, scale, and edge condition from the cutting operation all affect either the forming behaviour or the arc, and edge quality in particular is a common hidden variable when blanks are cut by different processes or suppliers. If cutting is upstream of this line, long sheet cutting acceptance is the neighbouring question.
Forming decides the fit
Rolling a blank into a shell is where the seam gap is created. Roll setting, the number of passes, springback for the specific material and thickness, and how the leading and trailing edges are pre-bent all determine whether the two edges meet cleanly along the full length or only in the middle. In the reference footage the blank is rolled on a four-roll machine and the formed shell shows its longitudinal seam.
The characteristic failure is not a constant gap but a varying one — tight at one end, open at the other, or open only at the extreme ends where pre-bending was insufficient. That signature is diagnostic: a defect that appears at the same position on every shell is a forming signature, while a defect that appears randomly along the seam is more likely to be a process or material variation.
EVST asks for this measurement first. So the most useful single measurement in this whole investigation is the gap profile along the seam after forming, taken on several consecutive parts. It costs very little and it usually settles the argument. The same reasoning applies to circular joints, where cylinder path boundaries for welding and cutting have to account for the shape the forming step actually produced.
Decision table: symptom evidence mapped to the likely cause
Each row is settled by measurement rather than by discussion. Where the confirming measurement has not been taken, the row is a hypothesis and should be labelled as one in the report.
| What you observe | Most likely explanation | Measurement that confirms or refutes it |
|---|---|---|
| Defect at the same position on every part | Forming signature — springback or pre-bend at the edges | Gap profile along the seam on several consecutive shells |
| Defect rate changes between material batches | Incoming variation in thickness or mechanical properties | Incoming dimensional record by batch, against purchased tolerance |
| Same setting, different result on identical-looking parts | Fit-up variation the parameter window cannot absorb | Gap measured immediately before welding, part by part |
| Consistent defect on every part regardless of fit | Parameter or joint design mismatch | Procedure record, joint preparation, and how the parameters were established |
| Problem appears after tacking or clamping | Fixture or tack sequence distorting the joint | Gap measured before and after clamping and tacking |
| Seam wanders relative to the torch path | Positioning or datum problem rather than parameters | Shell placement repeatability in the welding fixture |
How much gap a parameter window can absorb
Every welding procedure has a fit-up range within which it produces an acceptable result. That range is a property of the process, the joint design, the material, and the thickness, and it is established when the procedure is qualified — not asserted afterwards. Within the range, adjusting parameters is legitimate. Outside it, adjusting parameters converts one defect into another.
EVST writes the fit-up range into the procedure documentation for that reason. This is the practical reason to write the fit-up range into the procedure documentation rather than leaving it implicit. When the range is documented, an out-of-range gap becomes an upstream action rather than an argument with the welding engineer, and when it is not documented, the welding step absorbs blame for variation it never controlled.
Sensing changes the boundary but does not remove it. Seam tracking and adaptive control extend the range a line can tolerate, at a cost and with their own limits. They are a legitimate engineering answer once the upstream variation is characterised, and a poor substitute for characterising it.
Recording tolerances so the argument stops repeating
Most recurring fit-up disputes exist because nobody wrote down what fit-up was required. According to ISO 13920:1996, general tolerances for welded constructions cover dimensions for lengths and angles as well as shape and position, which is the right vocabulary for stating what the assembly has to deliver before welding.
The acceptance side has equally settled language. According to ISO 5817:2023, quality levels for imperfections are defined for fusion-welded joints in steel, nickel, titanium and their alloys, so the required level can be named rather than described. According to ISO 3834-2:2021, comprehensive quality requirements for fusion welding cover the control of fit-up, procedures and personnel, which is the framework that ties the two together.
None of these standards state a gap value for your part. They provide the structure in which your specific limits are declared and agreed. That is exactly what is needed to stop the same argument from recurring every quarter, and it is a normal part of comparing automatic welding system solutions on a like-for-like basis.
Fixturing and tacking between forming and welding
Between forming and welding there is usually a clamping or tacking step, and it can either correct or create a gap. Clamps that close a shell locally can open it elsewhere. A tack sequence that starts at one end can drive the mismatch to the other. Both are common, and both are invisible unless the gap is measured before and after.
That measurement — gap profile after forming, then again after clamping and tacking — separates the two stages cleanly. It is the difference between saying that the forming step is the problem and knowing it.
Fixture wear belongs in the same review. A fixture that was correct at installation and has since worn produces exactly the pattern of slowly rising, position-dependent defects that gets blamed on material.
What to send for a forming and welding review
Send the blank specification with the incoming tolerance you actually receive, the formed shell dimensions and the seam gap you measure after forming, the joint design and the tack or clamping arrangement, the current parameter set with how it was established, and the defect record including where on the seam the problem occurs. That set is normally enough to say which stage is responsible.
Expect the review to ask for the gap profile if it does not already exist. It is the single measurement that most often resolves the question, and it is the one most often missing. The footage behind this guide shows measurement, forming, and the resulting seam; it does not demonstrate a defect rate or a tolerance result, and none is asserted here.
Frequently asked questions
Where should thin-sheet weld gap troubleshooting start?
Not first. Check whether the fit-up presented to the arc is inside the range the procedure was qualified for. If it is not, parameter changes convert one defect into another rather than fixing the cause.
What single measurement helps most?
The gap profile along the seam after forming, taken on several consecutive parts. A defect at the same position every time points to forming; a random position points to material or process variation.
Can seam tracking solve a fit-up problem?
It can extend the range a line tolerates, at a cost and within limits. It is a reasonable answer once upstream variation is characterised, and a poor substitute for characterising it.
How do we stop this argument from recurring?
Write the required fit-up and the accepted quality level into the documentation. ISO 13920:1996 covers general tolerances for welded constructions and ISO 5817:2023 covers quality levels for imperfections, so both sides can be named rather than debated.
Project inputs for an application review
Send the following and the fault can be attributed to fit-up or to parameters with evidence rather than opinion:
- the blank specification: material, nominal thickness, and the incoming tolerance you actually receive
- the formed shell dimensions and the seam gap you measure after forming
- the joint design and the tack or clamping arrangement used before welding
- the current welding parameter set and how it was established
- the defect record with where on the seam the problem occurs
Send the blank specification with the incoming tolerance you actually receive, the formed shell dimensions with the measured seam gap, the joint design and tacking arrangement, the current parameter set and how it was established, and the defect record with its position on the seam. That set is enough to attribute the fault to fit-up or to parameters with evidence. Related reading: long sheet cutting acceptance, cylinder path boundaries for welding and cutting.