CNC Sheet Nesting Optimizer
Packs your parts onto sheets and exports SVG — mixing sizes in a row saves a third of the material, rotation only 3.6 per cent.
The layout
What the gap costs, by part size
Where that lands in your job
The lever is mixing sizes in a row, not a clever algorithm
Across 400 randomised jobs on a 1220 × 610 sheet, laying parts out one size per row — the obvious approach, and what most people do by hand — ran 67.6% over the area floor, where a plain shelf pack ran 26.1%. That's a third more sheets for the same parts, and the grid was worse in 61% of jobs. Nothing sophisticated recovers it: sorting tallest-first and filling each shelf with whatever fits captures nearly all of the gain.
And rotation is the minor lever, which matters because it's the one people reach for first. Allowing 90° turns saved 3.6% of sheets across the same trial and helped in only 9% of jobs. That's worth having on material with no grain — and worth knowing when there is one, because keeping every part aligned on plywood, veneer or brushed aluminium costs you a few per cent of material rather than the large penalty people assume. Mixing sizes is worth roughly nine times what rotation is.
The gap between parts costs area in proportion to perimeter, so small parts pay enormously. A 3 mm gap inflates a 15 mm part's footprint by 44% and a 400 mm panel's by 1.5% — a factor of twenty-nine across the parts in an ordinary job. And because the cost goes as the perimeter, doubling the gap on a small part more than doubles the overhead.
But the job-level figure is modest, and saying so is the honest part. In a typical mixed job those small parts are a small share of the area, so a 3 mm gap costs about 4% of the sheet overall even while it's costing 25% on the spacers. The dramatic per-part number is real and it isn't the number that decides anything — if you're going to spend effort somewhere, spend it on the layout rather than shaving the gap.
- The area floor is unreachable. It assumes parts can be liquefied and poured. A shelf pack landing within about a quarter of it is doing well; if you're far off, the usual cause is one part nearly as wide as the sheet forcing a shelf of its own.
- Keep the gap at least twice the kerf. Below that, two parts share a cut edge — sometimes deliberate as a common-line cut, which halves the cutting on a grid of squares, but a mistake when it happens by accident. Common-line cutting also means a slip on one part ruins two.
- Leave a sheet-edge margin and don't fight for it. Edges are rarely square, clamps and hold-downs live there, and material bows most at the perimeter. A 5 mm border costs about 1.6% of a half sheet — less than one failed part.
- Nesting saves material, not time. A tighter layout has the same total cut length, and often runs slower because the head travels further between parts. If a job is cutting-time bound rather than material bound, the two goals genuinely conflict.
How to use
- List each part as name, width, height and quantity.
- Set the gap to at least twice your kerf.
- Turn rotation off if the material has a grain direction.
- Download the SVG to check the layout before cutting.
Frequently asked questions
How much material does nesting actually save?
About a third, and almost all of it comes from mixing part sizes within a row rather than from any sophisticated algorithm. Across 400 randomised jobs on a half sheet, laying parts out one size per row ran 67.6 per cent over the theoretical area floor while a plain shelf pack ran 26.1 — a third more sheets for the same parts, with the grid worse in 61 per cent of jobs.
Is rotating parts worth it in nesting?
Less than people expect. Allowing ninety-degree turns saved only 3.6 per cent of sheets across a 400-job trial and helped in just 9 per cent of jobs. Mixing sizes in a row is worth roughly nine times as much. That matters most as a price: keeping every part grain-aligned on plywood, veneer or brushed aluminium costs a few per cent of material rather than the large penalty people assume.
Should I let the nesting rotate parts on plywood?
Usually not. Plywood, veneer and brushed metal all have a visible direction, and a rotated part shows it immediately on a finished piece. Since rotation is worth only around 3.6 per cent of material, the trade is a few per cent of sheet against every part looking right — which is an easy call for anything visible and a different call for internal parts nobody sees.
How much does the gap between parts cost?
It costs area in proportion to perimeter, so small parts pay enormously. A 3 mm gap inflates a 15 mm part footprint by 44 per cent and a 400 mm panel by 1.5 — a factor of twenty-nine across the parts in an ordinary job. And because the cost follows the perimeter, doubling the gap on a small part more than doubles the overhead.
Does the gap between parts matter for the whole job?
Much less than the per-part figure suggests, and this is worth being clear about. In a typical mixed job the small parts are a small share of the total area, so a 3 mm gap costs around 4 per cent of the sheet overall even while costing 25 per cent on the spacers. The dramatic number is real and it is not the number that decides anything — layout is the better place to spend effort.
What gap should I leave between parts?
At least twice your kerf plus a little, so the two cuts either side of the gap do not meet. Below that the parts share a cut edge, which is sometimes deliberate — a common-line cut halves the cutting on a grid of squares — but is a mistake when it happens by accident. Common-line cutting also means a slip on one part ruins two, so it suits simple shapes rather than fiddly ones.
Why can nesting never reach the theoretical minimum?
Because the area floor assumes the parts can be liquefied and poured into the sheet. Real parts are rigid rectangles that leave gaps wherever their proportions do not match, and one part nearly as wide as the sheet forces a row of its own with waste beside it. A shelf pack landing within about a quarter of the floor is doing well on a mixed job.
Does better nesting make a job cut faster?
No, and it often makes it slower. A tighter layout has exactly the same total cut length because the parts have not changed, and packing parts efficiently by area tends to scatter them so the head travels further between cuts. If a job is bound by cutting time rather than material, sorting for travel beats sorting for area, and the two goals genuinely conflict.
How big a margin should I leave at the sheet edge?
Around five millimetres on a small sheet, more if your machine clamps at the edges. Sheet edges are rarely square, hold-downs and honeycomb clips live there, and material bows most at the perimeter. On a half sheet a 5 mm border costs about 1.6 per cent of the area, which is far less than one failed part near the edge.
What is a shelf or skyline packing algorithm?
The simplest nesting approach that works: sort parts tallest first, then lay them into horizontal bands, opening a new band when the current one runs out of width. It leaves some waste above the shorter parts in each band, but it is fast, deterministic and captures nearly all the gain over laying out one size per row — which is where the real saving is.
Why did my part not get placed?
Because it does not fit the usable area in either orientation once the edge margin is taken off. A 600 mm part on a 610 mm sheet with a 5 mm margin has only 600 mm of usable height, so it fits by a hair or not at all. Parts like that are reported rather than silently shrunk or dropped, since a layout that quietly omits a part is worse than one that refuses.
Should I nest by hand or use software?
By hand is fine if you mix part sizes within each row — that is where a third of the saving lives, and a person eyeing a sheet does it naturally. What people actually lose to is the tidy instinct to group identical parts together, which looks organised and wastes a third of the material. Software mainly saves the fiddling rather than finding layouts a careful person could not.
🔒 This tool runs entirely in your browser. Nothing you enter is uploaded, logged, or stored.