A hole on a drawing reads like a description of a void. In a shop it is a sequence: establish the position, get the tool in, cut the profile, get the slug out, break the edge. Each step of that sequence has a geometric requirement attached to it, and the drawing either supports the sequence or fights it. Most of the fights are about the material left around the hole rather than the hole itself.
That is why blanking and cutting problems so often trace back to the drawing rather than to the machine. A profile that cannot be cut cleanly at the intended size, a hole so close to an edge that the edge distorts, a slot with a square end that no process produces well: none of that is a machine fault. It is a requirement written without reference to how the shape gets made.
Learning to read a hole as a process statement changes what you ask for. Instead of a diameter, you start specifying a diameter, a position, a finish on the bore and an edge condition, because those are the things the cut actually has to deliver.
Every cutting process leaves an edge with a character. Thermal processes leave a heat-affected zone and a slightly tapered or roughened face. Mechanical processes leave a burr. Water-based processes leave a clean edge with little thermal effect but a different surface texture. The bore of a hole carries that character all the way through the thickness.
Whether that matters depends on what the hole does. A clearance hole for a bolt does not care much about bore finish. A hole that will be tapped, reamed, or used as a locating feature does care, and the difference decides whether the hole is cut to size or cut undersize and finished later. A drawing that specifies a finished bore without allowing an operation to produce it is a drawing that will be met by arguments.
The material between a hole and an edge carries load. When that ligament is too narrow, the edge pulls, the hole distorts, and a fastener that was meant to clamp instead starts to tear its way out. The problem is worse in thin sheet, where there is little thickness to distribute the load, and worse again in a part that sees vibration.
Edge distance is also a cutting constraint independent of load. A cut close to an edge has less material to hold its shape, and the edge can move as the cut releases residual stress. The result is a part that measures correctly in the model and slightly differently on the bench, with the error concentrated exactly where the ligament is thinnest.
The practical rule is that edge distance is a design value worth stating, not a leftover. It is cheap to give a hole more room and expensive to discover the need after the profile has been nested.
A slot is usually specified to solve a tolerance problem. It lets one part move relative to another during assembly, absorbing the accumulated variation that would otherwise have to be held by the parts themselves. Used that way, a slot is one of the cheapest design tools available, and it can remove a fixturing requirement entirely.
A slot is also a cutting statement, and its ends matter. A slot with fully rounded ends is what a rotating tool naturally produces, and it is also better for the material, because a sharp internal corner concentrates stress. A drawing that calls for a square-ended slot is asking for either a secondary operation or a corner that will crack. Rounding the ends costs nothing and removes both problems.
Where a slot is used for adjustment rather than clearance, its length is a requirement and should be dimensioned as one. A slot made longer than needed lets the assembly sit in a position nobody intended, and the part will be blamed for moving.
The difference between a hole and the fastener in it is a designed allowance, not a manufacturing error. Clearance exists so the assembly can find its own position. Sizing a hole to the fastener removes that allowance, and with it the ability of two parts to go together when their patterns were made separately.
The other direction is worse in practice. An oversized hole that is meant to be a clearance becomes a part that shifts under load, because clamping force is carrying the joint rather than the fastener's shank. On a fabricated bracket that shows up as fretting around the hole and a joint that loosens.
Getting that balance right usually means deciding which part is the master and which part is allowed to move. Once that is decided, the hole sizes follow from it, and the drawing carries an intent rather than a set of independent diameters.
In thin sheet, the dominant risk is distortion. Heat input from a thermal cut can bow a panel, and a hole pattern near a free edge can pull the edge as it is cut. Spacing holes away from edges and sequencing the cut to balance heat both help, and both are influenced by how the drawing arranges the features.
In thicker material the dominant risk is the bore itself. A thick hole is a deep cut, and the deeper the cut, the more the taper, the roughness and the heat-affected zone matter. That is where a specification for a finished bore, or a note that a hole is to be drilled after cutting, changes what the shop has to do rather than merely how it does it.
Treating a thin panel and a thick flange as the same kind of part, because both have holes in them, is how processes get chosen badly. The geometry of the section decides more than the drawing's general notes do.
A hole placed close to a weld sits in material that has already been through a thermal cycle, and the two effects combine. The heat-affected zone is harder and less ductile, the hole is a stress concentration, and the edge distance is measured from a surface that has moved. If the hole is also a locating feature, the combination is a reliable source of variation.
Where a hole must sit near a seam, the sequence is worth stating. Drilling after welding puts the hole where the material ended up rather than where it was drawn. Drilling before welding means the weld pulls the hole out of position by whatever the seam moves, which is often enough to matter on a thin part. Saying which one applies is a sentence on the drawing; leaving it open is a decision made by whoever gets there first.
Most cutting problems are decided before the nesting is done, so these are the points worth settling on the drawing.
Each of those is a sentence or a value, and each one removes a question that would otherwise be answered at the machine.
A hole pattern is a system, not a collection of holes. When every hole is dimensioned from the same reference, the pattern is rigid and predictable, and the mating part needs the same discipline. When holes are dimensioned from each other, the pattern stretches, and the stretch is largest at the far end where the last fastener sits.
That is why patterns often fail at one corner. The failure is not random. It is the accumulated tolerance arriving at the feature furthest from the reference, which is exactly where the arithmetic predicted it would. Recognising that lets you fix the drawing rather than tightening the last hole.
There is also a fabrication reason to think in patterns. A pattern that can be cut in one setup, or match-drilled against its neighbour, removes a positioning question entirely. You can see how that plays out in the TIG stainless steel welding work, where holes and seams are sequenced together, and in the titanium fabrication components range.
The exhaust pipe and stainless steel exhaust system pages show the same logic on tube work, where a flange pattern has to agree with a port face that was never in the same fixture. Smaller brackets and plates are collected under other parts, and the route from drawing to first article is set out under customization. A hole is a small feature that decides whether an assembly goes together on the bench or on the third attempt.
