Sheet Metal Nesting: Optimizing Layouts for Metal Fabrication

Sheet metal nesting is the practice of arranging parts on a metal plate or sheet so the cuts leave as little scrap as possible. For a woodworker or maker cutting rectangular parts - brackets, plates, covers, base panels, gussets - the math is the same bin-packing problem you already solve on plywood: fit the pieces, respect the kerf, count the sheets. CutPlan is a panel optimizer, so it handles the rectangular side of that problem well. What it does not do is true-shape (contour) nesting, where curved and irregular outlines interlock like puzzle pieces. This guide draws that line honestly, gives you the metal-specific numbers to plug in, and shows a worked example so you can see where a rectangular optimizer earns its keep and where you genuinely need dedicated CAM software.

If you have never run a layout before, start with the complete guide to cut list optimization for the fundamentals, then come back here for the metal-specific adjustments.

Two kinds of nesting, and which one you actually have

The word "nesting" covers two very different problems, and picking the wrong tool wastes either money or material.

Rectangular nesting treats every part as its bounding rectangle and packs those rectangles onto the sheet. This is exactly what a panel-cutting optimizer does. It suits any part whose outline is a rectangle, or whose rectangular blank you will later drill, notch, or fold. A surprising share of fabrication work is rectangular: electrical enclosure panels, machine guards, shelf plates, mounting brackets, hearth plates, and the flat blanks for folded boxes all start life as rectangles.

True-shape nesting (also called contour or freeform nesting) packs the real outlines together at any rotation, tucking one part's concave notch around another part's bump. On a job full of L-shapes, discs, triangles, or organic brackets, true-shape nesting can lift material yield by 10 to 25 percent over a rectangular layout, because all the space a bounding box wastes becomes usable. The catch is that it needs CAD geometry (DXF/DWG) and CAM software with shape recognition, which is a paid, specialized category.

A quick test: if you can write your parts as a list of "width x height x quantity," rectangular nesting will serve you. If your parts only make sense as drawn outlines, you want true-shape CAM. CutPlan sits firmly in the first camp, and this post stays there too.

What changes when the material is metal, not plywood

The packing algorithm does not care whether it is arranging oak or aluminum. But four practical inputs change, and getting them wrong is what makes a layout fail on the shop floor.

• Kerf depends on the process, not a blade. A saw has one kerf. Metal cutting has a wide range depending on how you sever the material, and the value you enter directly resizes every part. Guessing here is the most common mistake, covered in the table below and in our kerf explainer.
• Rotation is usually free. Wood has grain, so panels often lock to one orientation. Most metal has no directional structure, so you can let the optimizer rotate parts 90 degrees at will, which typically buys 5 to 12 percent tighter packing. The exceptions are brushed, grained, or directionally patterned finishes on stainless and anodized aluminum: treat those exactly like wood grain and keep orientation fixed.
• Heat needs breathing room. Plasma and laser cutting dump heat into the sheet. On thin gauge (under 3mm) parts crowded together can warp or the second cut can distort the first. Adding 1 to 3mm of part spacing beyond the kerf keeps the heat-affected zone of one edge away from its neighbor. Waterjet is cold-cutting and needs no thermal gap.
• The clamp or grip zone is off-limits. Whatever holds the sheet during cutting cannot have parts under it. On a CNC bed that is often 10 to 25mm along the gripped edges. Enter it as edge trim so the optimizer never places a part where the machine cannot reach.

Kerf by cutting method (enter this, do not guess)

Kerf is the width of material the cut removes. Set it too low and finished parts come out oversized and jam against each other; set it too high and you waste stock and may misreport how many sheets you need. These are typical working ranges - always confirm against your own machine and a test cut.

Two nuances the table cannot capture. First, kerf usually widens as material gets thicker, so a 2mm plasma kerf on 3mm steel may open to 3.5mm on 20mm plate. Second, thermal kerf is not symmetric around the cut path, but for rectangular layout purposes entering the total kerf width is close enough. If your machine software already compensates the toolpath, enter a kerf of zero in the optimizer and let the machine own it, or you will double-count.

Standard metal sheet sizes to set as stock

Before packing anything, define your stock accurately. Ordering the right sheet size is often a bigger lever than the packing itself, the same way it is for wood - see the full standard sheet sizes guide.

• Mild steel: 2440 x 1220 mm (the 4x8 ft sheet) is the global default. 3000 x 1500 mm is equally common in Europe and Asia and nests larger parts more efficiently.
• Aluminum: 2440 x 1220 mm and 2500 x 1250 mm are the primary sizes, with wider specialty stock available.
• Stainless steel: mirrors mild steel at 2440 x 1220 mm and 3000 x 1500 mm. Because stainless runs 3 to 5 times the price of mild steel, every percent of yield matters more.
• Gauge: anything from 0.5 mm sheet up to 25 mm+ plate. Thicker plate is pricier and slower to cut, so the payoff from tight nesting climbs with thickness.

A worked example: 40 brackets from 6 mm mild steel

Numbers make the trade-offs concrete. Say you need 40 rectangular mounting plates in a mix of sizes, cut from 6 mm mild steel on a plasma table. Stock is 2440 x 1220 mm at roughly $120 per sheet. Your plasma kerf is 2.5 mm, you add 2 mm thermal spacing, and the bed grips 20 mm along two edges.

• Eyeballed manual layout: laying parts out by hand in neat rows, ignoring rotation, you fit the job on 7 sheets. That is $840 of steel with roughly 28 percent of it going to scrap and the drop bin.
• Rectangular optimizer, rotation on: letting the software rotate parts and mix sizes across each sheet, the same 40 parts drop onto 5 sheets. That is $600, about 14 percent scrap, and a $240 saving on one job.
• What rectangular nesting cannot recover: if those plates had large corner cut-outs, true-shape CAM might interlock them onto 4 sheets. That last sheet of yield is the gap between a panel optimizer and dedicated metal CAM - real, but only worth chasing when the geometry and the material price justify the software.

Run that weekly and the rectangular optimization alone is over $12,000 a year in mild steel. Swap in stainless at $400+ a sheet and the same layout choice swings closer to $40,000. The point is not the exact figure, it is that the biggest, cheapest win - moving from eyeballed rows to an honest rectangular pack - is available for free before you ever consider paid CAM. For the wider comparison of automated versus by-hand layout, see CNC nesting vs manual cut list.

Setting metal up in a panel optimizer, step by step

Here is the exact sequence to translate a metal job into a rectangular optimizer like CutPlan:

• Define stock first. Enter your real sheet size and, if you keep drops, add usable offcuts as extra stock so the optimizer consumes them before opening a fresh sheet.
• Set the clamp zone as edge trim. Trim the gripped edges (say 20 mm) so no part lands where the machine holds the sheet.
• Enter kerf from the table, or zero if the machine compensates. This is the single value most likely to bite you, so confirm it against a test cut.
• Add part spacing for thermal cutting. 1 to 3 mm on plasma and laser; leave it at zero for waterjet.
• Enable rotation unless the finish is directional. Turn it off only for brushed or grained stock.
• Review, then export. Sanity-check the layout, then send it on. For getting the geometry into a controller, our DXF export for CNC guide walks through file prep.

One honest limitation worth stating plainly: a rectangular optimizer outputs a packing plan and a cut list, not a machine-ready toolpath with lead-ins, tabs, and pierce points. It tells you how many sheets and where each part sits. The controller or CAM step still owns the actual toolpath. That division of labor is fine for most rectangular jobs, and is why a free panel tool is enough to settle the sheet count and layout while the toolpath stays where it belongs.

Frequently Asked Questions