A drawing and a 3D file describe a part, but they do not describe how it is made. The first run through the shop is where the sequence, the fixtures and the small corrections get settled: which bend is formed first, where the tacks go, which face is used to locate the next operation. All of it is knowledge that only exists after the job has physically happened.
That is why a repeat order is not simply the same job run again. It is the same part made with a process that has already been debugged, and the difference shows in both the time taken and the consistency of the result.
The second run starts by fixing the things that were settled by trial. The operation sequence is written down, the fixtures are brought out and checked, the welding settings that produced the accepted first batch are recorded, and the inspection points are defined. Anything that was decided by eye on the first run should be a number or a gauge position by the second.
Locking the setup is what makes a batch repeatable. A process that lives in one person's memory is not locked; it is personal, and it leaves when that person does.
It also makes the job easier to hand over. A settled method can be given to another operator with a reasonable expectation of the same outcome.
Not everything can be taken for granted. The first article of a repeat batch still gets measured, because a fixture can shift, a die can wear and a machine can be reset by someone else. The check is not suspicion of the previous batch; it is confirmation that the setup is where it was left.
Material is checked as well. A new delivery of tube or sheet can sit at a slightly different position against a forming tool, and small differences in temper change how much a part springs back when it is released.
Every tool changes in use. Forming dies lose a little of their edge, locating pins wear at the point of contact, and fixtures that are handled roughly lose the fit that made them accurate. None of that is dramatic, and all of it accumulates.
The practical answer is to keep tooling identified and inspected, so a fixture that has developed play is found before it produces a batch rather than after. Where a part has close alignment requirements, wear is worth measuring rather than estimating.
Drift has a second source, which is the person setting the job. A stop set by feel on the first run and by measurement on the second can give two different results, which is why the first run's settings are worth recording even when they seem obvious at the time.
Repeat batches rarely run from the same lot of stock all the way through. When new material arrives, the differences are usually small but real: surface condition, straightness, and how the material reacts to a bend or a roll. Where the part is welded, the filler wire and the consumables are part of the same question.
The safe approach is to treat a lot change as a reason to check the first part off the new lot rather than to assume equivalence. On a part with a tight forming requirement, that check is cheap insurance.
A repeat order should not quietly absorb changes. If the customer wants a modification, that is a revision, and the revision needs a drawing change, a fixture change if the locating features move, and a reset of the records. Running a modified part under the old part number creates exactly the confusion that revision control is meant to prevent.
Where a change is cosmetic only, such as a surface finish, it still belongs on the record, because it may not affect fit but it does affect what is delivered.
Repeat work is cheaper because the setup is already proven, not because the part is made to a lower standard. The tooling exists, the sequence is known, the first article has been checked, and the operator is not discovering problems as they go. The time that goes into the first run is the time that does not have to be repeated.
For a titanium muffler or a similar welded assembly, that is where the reduction actually sits: in the setup, the fixture and the settled method, not in the material or in the weld.
The same logic explains why the first order of a new part is always the most expensive one. It is carrying the development cost of everything the later orders inherit.
For a part that is going to run more than once, these points keep the second batch aligned with the first.
Repeat work is the backbone of what leaves the shop, whether that is finished titanium exhaust systems, formed tube runs or TIG welded assemblies. Our titanium mufflers show what a settled process produces, the custom work page explains how a first run is set up, and our material pages cover what to expect when the stock changes.
