Plasma nesting in practice: pierces, heat and the skeleton.

On a plasma table the layout decides more than material yield. It sets how many times you pierce, where heat builds up, and whether the leftover skeleton stays stiff enough to finish the sheet. These are the trade-offs the plasma community actually argues about, with sources, and an honest statement of what nesting software does not decide.

Quick answer: every part costs at least one pierce, and piercing is the hardest thing you do to your consumables. Common-line cutting can halve the pierce count, but each shared edge costs half a kerf per part unless the CAM compensates. Heat wants distance and a sensible cut order, especially on thin gauge. Thick plate needs extra room for pierces and lead-ins. A nesting tool controls distances and sheet count; pierce count, cut order and shared edges are CAM decisions.

Pierce count versus common-line cutting

Piercing is the expensive moment of a plasma program. The arc blows molten metal back at the nozzle, consumables wear fastest right there, and the thicker the plate, the worse it gets. A layout that needs fewer pierces saves consumables and time directly, which is why common-line cutting is the classic production technique: neighbouring parts share an edge, and one pass cuts both.

The catch is dimensional, and Langmuir table owners discuss it openly on their forum: the kerf splits across the shared line, so each part comes up about half a kerf short along that edge unless the CAM compensates the geometry beforehand. For brackets that get welded into a frame, nobody notices. For parts with a fit or a tolerance, someone will.

The honest statement for this page: NestForge does not do common-line cutting. Parts always keep their full outline and full spacing. If your goal is fewer pierces through shared edges, deepnest-next (the maintained fork of Deepnest) offers line merging, and SheetCAM-side techniques can produce shared cuts too; those are the right tools for that job. NestForge optimizes material, not pierce count.

Heat, warp and cut sequence

Thin gauge warps when too much heat lands in one area, and Langmuir forum threads describe exactly that happening when parts sit too close together. The trade press covers the systematic side: Canadian Metalworking's plasma nesting article and The Fabricator's smart-nesting article both recommend spreading parts on thin material and sequencing cuts so the torch is not working in metal that is already hot.

Note that those are two different levers. Distance is a nesting decision. Sequence is a CAM decision, made where the toolpaths are generated. NestForge handles the first and not the second: it produces layouts, not toolpaths, so it has no concept of cut order. If your CAM offers sequencing control, that is where the heat path gets managed.

The skeleton has to stay a skeleton

As parts drop free, the skeleton is all that holds the remaining sheet together. Webs thinner than one full kerf can burn away entirely, because the two neighbouring cuts overlap. A floppy skeleton lets finished parts tip up, and a tipped part in front of a moving torch is how machines get damaged. The trade-press advice is consistent: keep webs at least one full kerf wide everywhere, and give the skeleton more meat where it turns sparse. This one genuinely is a nesting decision; it is your spacing value, and the numbers per process are in the part spacing guide.

Lead-ins need room, and thick plate needs more

A pierce does not happen on the part contour. The CAM places it in waste material and approaches the contour with a lead-in, and that arc or line needs clearance. Kinetic's Primecut CAM documentation recommends a plate gap of about 10 mm (3/8") as a starting point, more on thick plate precisely because pierces and lead-ins need more room there, less on thin sheet.

Which brings up this page's second honest limitation: NestForge has no idea where your pierces and lead-ins will go, because it produces no toolpaths. If your CAM uses 6 mm lead-ins, that length has to fit in the gap between parts, so you add it to the spacing value yourself. The layout then has room wherever the CAM decides to pierce.

What the layout decides, and what it does not

A nesting layout controls material yield, the distances between parts, the sheet count, and which parts share a sheet. Your CAM controls kerf compensation, pierce placement, lead-ins, cut order, and any shared-edge technique. G-code always comes from the CAM. When a tool promises to optimize your plasma cutting, check which side of that line each promise sits on. NestForge sits entirely on the layout side, and the section above says plainly what that excludes.

What NestForge contributes on a plasma table: layouts where kerf and your chosen spacing are provably respected. An independent exact-arithmetic validator re-checks every placement, so overlapping kerfs never reach your machine, and the exported DXF keeps true arcs instead of polygon soup. It runs entirely offline in your browser. The limits, once more: no common-line cutting, no cut sequencing, no G-code; pierce and lead-in allowance goes into your spacing value. The free tier nests up to 10 parts on 1 sheet, with no time limit.

Nest a plasma sheet, free

Free tier: 10 parts, 1 sheet, no time limit. No account, files never leave your machine.