Ask three shops to cut the same bracket from the same sheet and you get three edges that look similar in a photograph and behave differently under a torch. The outline is a shape. The edge is a metallurgical condition, and it is created by whichever process was chosen to make it.
Waterjet, laser and plasma all separate metal. They differ in how they deliver energy. One erodes the material with an abrasive carried in water. One melts and vaporises with a focused beam. One melts with an arc and blows the melt out with gas. Everything that follows for the finished part — the heat-affected zone, the edge squareness, the dross, the hardness at the cut face — comes from that difference rather than from the machine's price.
Plasma is an arc constricted through a nozzle, hot enough to melt any alloy a fabricator is likely to hand it, and the gas jet that forms the arc also clears the kerf. It is the process that cares least about what it is cutting. Rust, mill scale, paint and a less than perfect fit against the table are all tolerated.
What it leaves behind is a cut face that has been through a thermal cycle. There is a heat-affected zone at the edge, a recast layer where metal melted and froze again, and on stainless and aluminium a surface that has lost its oxide protection and will need attention before it welds or before it goes into service. On thick sections the kerf also leans. The top of the cut is narrower than the bottom, which matters the moment the part has to fit something.
None of that makes plasma wrong. For plate where the edge will be machined or buried in a weld, cutting speed and thickness capacity are worth more than edge cosmetics. It is the wrong choice when the cut face is the visible edge or a mating surface.
A laser delivers energy into a very small spot, so the melt zone is narrow and the cut is fast. On thin sheet the result is a clean edge, tight geometry, and a heat-affected zone shallow enough that it rarely changes how the part behaves. Holes come out round, corners come out sharp, and nesting gets efficient because the kerf is small.
The trade is thickness and reflectivity. As section thickness climbs, the process slows and edge quality falls away; on thick stainless and aluminium the cut face can become rough and the kerf wider than the drawing assumed. Assist gas matters too. Cutting stainless with an inert assist gas gives a clean, bright edge; cutting with oxygen-assisted gas is faster but leaves an oxidised face. Both are legitimate, but they are not interchangeable when the next operation is a weld.
An abrasive waterjet is a cold process in the sense that matters. It erodes a narrow slot through the material without a thermal cycle, so there is no recast layer, no heat tint, no hardening at the edge, and no change to the metallurgy of the parent metal. Whatever the sheet was before the cut, it is the same at the edge afterwards.
That is why waterjet suits material that is expensive to spoil. Titanium, thick stainless, aluminium plate, clad sheet and anything already heat treated all come out with their properties intact. It also cuts without a fixturing argument, since the forces are low and the part does not need to be clamped against distortion.
The costs are speed and edge character. Waterjet is slower than a laser on thin sheet, and the cut face carries a fine, matte, slightly tapered texture rather than a polished one. In most fabrication that texture is irrelevant. In a sealing face or a bearing bore it is not.
Every cut removes material, and the amount removed is not in the drawing. A part drawn to a finished size comes out undersize by roughly the kerf width if the cut path is programmed on the line. Shops compensate, but compensation is an assumption, and assumptions differ between processes and between materials. When a laser-cut part and a plasma-cut part of the same nominal size meet, they can disagree by more than either edge looks like it should.
Taper is the second hidden dimension. Laser taper is small and consistent. Plasma taper grows with thickness. Waterjet taper depends on speed and on how the jet loses coherence through the depth of the cut. On a lap joint or a butt weld none of this is visible. On a flange face or a slip fit it is the difference between a part that assembles and one that has to be dressed.
A hole is a harder cut than an outline because the process has to enter, travel a closed path and exit without leaving a witness mark. Laser pierces and cuts small holes well in thin material. Plasma struggles with small holes in thick plate; the arc has to lead in, and the resulting hole is often tapered and slightly out of position. Waterjet makes accurate holes in any thickness at the cost of time, and it can start in a drilled pilot to avoid the pierce altogether.
The practical rule is that hole quality should be specified by what the hole does. A clearance hole for a bolt has different requirements from a hole that will be tapped, reamed or used as a locating feature. On a weldment, a hole that is going to be filled by weld metal is barely a hole at all.
Whatever process cut the part, the face that meets the arc has to be clean. Laser-cut stainless with an oxidised assist-gas edge, plasma-cut stainless with a recast layer, and waterjet-cut anything carrying abrasive residue all need the same thing before welding: the contaminated layer has to come off.
There is also a geometry effect. A cut edge is square within the process's own tolerance, and a joint that depends on two square faces meeting will show that error as a variable gap. A fabricator who checks the fit before tacking and adjusts the bevel is doing the work the cut did not do. This is routine in the kind of work that passes through our TIG stainless steel welding bay, where a gap that varies along a joint shows up immediately in the bead profile.
The honest way to pick a process is to start from the operation after the cut rather than from the cut itself. If the edge will be welded, the priority is a clean face and predictable fit. If the edge will be seen, the priority is appearance. If the part will be machined, the priority is that the cut leaves enough material and does not harden the surface the tool has to enter. If the material is expensive or already heat treated, the priority is not damaging it.
Where several of those apply, the cut is usually split: an outline on the fast process and the critical features left for a second operation. That is not indecision. It is matching the cost of the cut to the cost of the consequence.
When a cut part is quoted, the process is often implied rather than stated. A handful of questions resolve most of the ambiguity before any metal is cut.
We cut and weld to drawings and to samples, and the route is agreed before the first part is made — see customization for how that conversation runs. The tube work that follows cutting is shown under exhaust pipe, the titanium side under TIG titanium welding parts, and the fabricated assemblies under titanium fabrication components and stainless steel exhaust system.
