Aluminium extrusion cutting waste usually isn't a material problem — it's a combination problem. Window, door, and curtain-wall systems typically need several different piece lengths cut from a small set of standard stock bar lengths per profile, and the amount of waste that plan produces depends heavily on which stock lengths get used for which combinations of pieces — not just on how carefully each individual bar is packed.
Why mixed lengths matter more here than in single-length jobs
A profile system rarely has just one required length. A single window run might need frame uprights, transoms, and cills, each a different length and quantity — and most fabricators stock more than one bar length per profile (commonly something like 5.6m/6.2m/6.5m/6.6m runs, depending on supplier and system). The waste in the final plan comes down to how well those different required lengths get matched to the available stock lengths, not to any one bar in isolation.
Cutting the job stock-length by stock-length (use up all the 6.5m first, then move to 6.2m) tends to leave awkward remainders, because it never actually compares "this combination on a 6.5m bar" against "this combination on a 6.2m bar" — it just uses whichever length is next. An optimizer that's actually allowed to choose between stock lengths per pattern, rather than being fed one length at a time, finds combinations a fixed-order approach won't.
A worked example
Cutting list for one profile: 2150mm × 6, 1400mm × 8, 980mm × 10. Stock available: 6500mm and 6000mm lengths, unlimited quantity. Kerf 4mm, end trim 20mm, minimum useful offcut 250mm.
Run through Barcut Pro's actual engine, this returns a 6-bar plan:
| Stock length | Cuts |
|---|---|
| 6500mm | 2150 + 2150 + 2150 |
| 6500mm | 2150 + 2150 + 2150 |
| 6500mm | 1400 + 980 + 980 + 980 + 980 + 980 |
| 6500mm | 1400 + 980 + 980 + 980 + 980 + 980 |
| 6000mm | 1400 + 1400 |
| 6000mm | 1400 + 1400 + 1400 + 1400 |
24 real cuts, all 24 required pieces placed, 89.2% material yield across the 6 bars used. Notice the 2150mm pieces and the mixed 1400/980mm pieces each landed on their own dedicated pattern on the 6500mm stock (three 2150s fit almost exactly; a run of 980s alongside one 1400 makes better use of the length than either would alone), while the shorter 6000mm stock got reserved for pure 1400mm runs where its extra length wasn't needed. That's a pattern a length-by-length manual pass is unlikely to land on by inspection, especially once a real job has more than two or three required lengths.
What this means for reducing waste on real jobs
- Enter every stock length you actually stock for that profile, not just the one you'd default to — the optimizer can only choose between lengths you give it.
- Set kerf and end trim to your actual saw's numbers, not a rounded estimate — see what saw kerf is and why it matters.
- Let repeated combinations show up as repeated patterns rather than planning bar-by-bar — a job like the one above reduces to a handful of distinct patterns even though it places 24 individual pieces, which is what actually keeps it fast and low-error to run at the saw.
Try the 1D cutting optimizer on your own profile cutting list — enter every stock length you hold and let it decide the combination, rather than committing to one length at a time.