A pair of flanges with matching holes will always line up while they are clamped in the fixture. That is not evidence of anything. The question is what happens when the clamps come off and the assembly is offered up to its mating part, and by then the welding is done and the geometry is set.
Bolt-hole alignment across a fabricated joint is therefore a pre-weld problem disguised as a post-weld one. By the time the bolt will not pass, the causes were decided earlier: in the hole position, in the sequence, or in how much the joint was allowed to move while it cooled.
Every bolted joint has clearance between the bolt and the hole, and that clearance exists so the joint can be assembled without force. It is not there to absorb fabrication error. A hole that is generous enough to swallow a misalignment is also a hole that lets the joint slip under load.
The distinction matters in thin-wall work, where a slotted or oversized hole has very little material around it. Opening a hole to make a joint fit reduces the bearing area that carries the load, and on a bracket in a vibrating assembly that is how a hole becomes a crack.
First, hole position: drilled from a mark rather than from a jig, or drilled before a joint was closed, so the hole followed the part as it moved. Second, joint movement: the weld pulled the two halves apart or across as it cooled, carrying the holes with it. Third, stack-up: each part within tolerance, and the sum of the tolerances outside it.
Separating them is the useful part of the job. If the holes are in the right place relative to their own part but the parts are in the wrong place relative to each other, the problem is movement or stack-up. If one hole is out on its own, the problem is drilling, and no amount of sequence work will fix it.
Drilling in place guarantees alignment for that one assembly and nothing else. It is the right answer for a repair, a one-off, or a joint that has to be matched to an existing vehicle. It is the wrong answer for anything that has to be repeated, because the position lives in the part rather than in a tool.
Drilling to a jig does the opposite: it makes the position repeatable and it makes the first article slower. The compromise most shops use is to drill one hole to locate, assemble and tack, then drill the rest through both parts once the geometry is settled, with the assembly still held in the fixture.
Welding moves holes because welding moves the material around them. A joint welded from one end to the other pulls progressively in one direction, and a pair of holes either side of that joint will end up closer or further apart than they started. Balanced sequence, alternating sides and backstepping all reduce the movement without changing the design.
Heat input matters for the same reason. A hot, fast pass on thin material shrinks more than a cooler one, and the shrinkage acts across the joint where the holes are. This is where the stainless TIG welding approach of controlled, evenly distributed passes pays for itself in parts that bolt up first time.
Slots are a legitimate design feature when the joint is meant to adjust: a bracket that has to accommodate tolerance in a body shell, a mount that has to be aligned on assembly. They are a bad repair when they are added after the fact, because the slot is usually cut in the direction that was convenient rather than the direction that should float.
When a slot is designed in, it needs a hardened washer or a captured nut to carry the clamp load across the opening, and it needs a specified travel. A slot with no stated travel is a slot whose size will be decided by whoever is filing it.
The check is to assemble with the fasteners just snug and see whether every bolt drops through by hand. If one needs a drift, the joint is misaligned regardless of whether it can be pulled together. Measuring the offset between the two hole centres before tightening gives a number to work with.
It is also worth checking alignment after the assembly has been through a heat cycle, if it will see one in service. A joint that lines up cold and not hot was stressed by the thermal expansion, and that is a design issue rather than a fabrication defect.
Both checks belong in the record for a custom project, because they are the only evidence that the joint was right before it left the shop.
The choices are to open the hole, to slot it, to remake the part, or to weld up and re-drill. Opening and slotting are fastest and worst, because both remove material from the loaded section. Welding up and re-drilling restores the section and lets the hole be placed correctly, but it puts more heat into the joint that moved in the first place.
The decision usually comes down to whether the joint carries load and whether the part is visible. A hidden bracket on a low-load circuit can tolerate a sensible slot. A visible mount on a stressed assembly should be remade, because the repair is a permanent compromise on a part that will be looked at.
Alignment is decided by the drawing, the fixture and the sequence together, so all three belong in the quotation.
None of this is difficult, and all of it is cheaper than a return. A joint that bolts up on the bench is the result of a hole position someone controlled and a sequence someone thought about, not of a bigger drill. The cut and drilled sections and the straight tube range both show how much of the finished fit is decided before the first arc, and the other parts pages carry brackets where hole position is the whole specification. Matching a joint to an existing vehicle is covered in the bent pipe work, and the customization notes explain how a sample is gauged when the mating part is not available to measure.
