Two shops cut the same plate. One yields 88 percent of usable parts, the other 68 percent. The difference is rarely the machine — it is the nesting strategy. In CNC plasma cutting, material typically accounts for 45 to 60 percent of total part cost, so lifting yield by five percentage points is usually worth more than negotiating three percent off your steel price. This article covers the three techniques that move the needle most — common-line cutting, bridging and chain cutting — the process limits where they stop working, and a payback table you can copy directly.
1. Where material yield actually leaks away
Before optimising anything, break the loss down. Take a 1.5 m x 6 m sheet of 12 mm Q235 cutting flange-type parts. Losses come from five places:
| Loss source | Typical share | Can it be improved? | Notes |
|---|---|---|---|
| Gaps between parts (kerf + safety margin) | 6–12% | Largely yes | Common-line cutting removes the duplicated kerf between neighbours |
| Nesting voids (poor shape match) | 8–18% | Yes | Improved by automatic nesting and part grouping |
| Plate edge allowance | 2–4% | Partly | Clamping and pierce zones; depends on table design |
| The kerf itself | 1.5–3% | No | Set by nozzle orifice and current; compensate, do not try to eliminate |
| Skeleton scrap and rework | 2–8% | Yes | Distortion and poor cut sequencing |
The single largest controllable loss is the gap between parts — and that is exactly what common-line cutting targets.
2. Common-line cutting: two kerfs become one
In conventional nesting, two adjacent parts each carry their own contour, and the layout must leave at least twice the kerf width between them. The plate gets cut twice and a strip of material is thrown away. Common-line cutting makes two parts share a single cut line: one pierce, one cut, two finished edges.
The three variants
- True common line — the two part edges coincide exactly, the kerf centre sits on the design line and kerf compensation splits the tolerance between both parts. Saves the most material, demands the most accurate compensation.
- Bridge (web) — a 3 to 8 mm uncut bridge is left between neighbours. Parts stay attached, distortion drops, and the bridges are separated manually or by a second pass. Best for thin plate and small parts.
- Chain cutting — several outer profiles are joined into one continuous path, so a single arc start covers many parts. This saves pierce time and electrode life rather than material.
The three optimise different things, so real-world programs combine them: chain cutting on outer profiles to cut pierce count, true common lines on internal shared edges, bridging on thin-plate clusters to stop lift and warp.
3. Six moves that lift yield
| Move | How | Typical gain |
|---|---|---|
| 1. Common line first, fill second | Lay rectangular parts out as a common-line matrix, then fill the remainder with irregular shapes | +4–8% |
| 2. Group parts before nesting | Nest only parts sharing thickness, material and process; avoid mixing parameters on one sheet | +1–3%, fewer stoppages |
| 3. Manage a remnant library | Log skeletons and offcuts, and nest on remnants first | +3–6% over time |
| 4. Calibrate kerf compensation | Measure the real kerf instead of copying the process chart | +0.5–1.5%, better tolerance yield |
| 5. Optimise lead-ins | Short 3–6 mm arc lead-ins; never pierce on a critical edge | 20–40% less rework |
| 6. Sequence for heat | Work outwards from the plate centre and skip around to spread heat | 1–3% less scrap |
Moves 1 and 4 are one-off efforts that pay back indefinitely — do them first.
4. Where common-line cutting stops working
It is not universal. Avoid or limit it in four situations:
- Thick plate (above 25 mm) — the kerf is wide and heat input high, so the shared edge tends to remelt and collect dross.
- Tight-tolerance mating parts — a common line splits kerf tolerance between two parts. If the drawing calls for ±0.5 mm or tighter, keep a separate contour and finish-machine.
- Parts needing bevels or marking — a shared edge cannot be bevelled independently; split those parts out.
- Densely nested thin plate — concentrated heat warps the whole sheet. Use bridging instead of true common lines.
The test is simple: the material saved must exceed the rework it risks. At 4,000 CNY per tonne of steel and 60 CNY per hour of grinding, common-line cutting stays worthwhile as long as it pushes rework up by less than about 1.5 percentage points.
5. Implementing it in StarCAM
Icyrobot StarCAM nesting software ships common-line, bridging, chain cutting and a remnant library out of the box, and generates kerf-compensated programmes straight from DXF or DWG drawings. A sensible rollout order: import the part library, set material and thickness, calibrate kerf compensation, run automatic nesting, hand-tune the common lines, dry-run the simulation, then post to the CNC cutting controller. On the first job, run a controlled comparison — nest the same part set twice, once conventionally and once with common lines, and weigh the results.
6. What it saves in a year
| Item | Conventional nesting | Common-line + optimisation |
|---|---|---|
| Material yield | 68% | 83% |
| Plate consumed (for 20 t of parts) | 29.4 t | 24.1 t |
| Monthly material cost (at 4,000 CNY/t) | 117,600 CNY | 96,400 CNY |
| Pierce count (tens of thousands/month) | 3.2 | 1.9 |
| Electrode and nozzle use | Baseline | Roughly -25% |
| Monthly saving | about 21,200 CNY in material plus consumables | |
Across a ten-month production year that is over 200,000 CNY saved, while the nesting software itself typically pays back inside two months.
7. Frequently asked questions
Does common-line cutting make parts undersized?
Yes, if compensation is wrong. The kerf width is split between the two neighbours, so an incorrect offset leaves parts small by half to a full kerf width. Measure the actual kerf, set a half-kerf offset, and verify the first part with calipers before releasing the batch.
Will parts drop into the slag tray or lift?
On thin plate, often. Leave 3–8 mm bridges and separate afterwards, or reorder the programme so small parts are cut last and the surrounding skeleton stays intact as support.
Is every part a candidate?
No. Rectangular and straight-edged parts gain the most. Complex curved profiles rarely share an edge perfectly, so the payoff is small — improve those through the automatic nesting algorithm instead.
What kerf compensation value should I enter?
Never copy the process chart. Cut a 200 mm straight line on scrap with the same parameters and measure it. Typical plasma kerf on 12 mm carbon steel is 1.8–2.5 mm; fine plasma can reach 1.0–1.5 mm. Re-measure every time you change nozzle size.
Conclusion
Material yield is one of the few gains in CNC cutting that needs no capital equipment. Do common lines, bridging, chain cutting and remnant management properly and most of the gap between 50 and 80 percent yield is recoverable. To find out how much your current layouts could gain, talk to our engineers, or download StarCAM nesting software and run your own drawings through it. For related reading, see our CNC table vs handheld cutting comparison and the cut edge troubleshooting guide.
