Key takeaways
- A guillotine makes a 300 × 200 mm blank in 3 mm CRCA in about 36 s including handling, with no program and no nest, which is why it still owns the prototype rectangular blank.
- Shear force is L × t × τ; for 3 mm CRCA at 255 MPa shear strength a 300 mm cut needs 229 kN, so tonnage per millimetre of cut is 0.77 kN at that thickness.
- Sheared edges carry a rollover of 10–30% of thickness and a burr of 0.05–0.20 mm, and the burr must sit on the outside of the bend or it starts a crack.
- A nibbler removes a 3–6 mm kerf and leaves a scalloped edge whose cusp is set by the step, about 0.03 mm at a 1 mm step with a 10 mm tool and 0.23 mm at 3 mm.
- Neither operation appears in ISO 9013, which classifies thermal cuts only, so a sheared or nibbled edge has to be specified by burr height and squareness instead.
A prototype bracket is needed before the tool room closes, and the drawing is a rectangle with two holes. The laser queue is four hours deep. The guillotine can produce the blank in 36 seconds, and the two holes can be drilled in another two minutes. That is the case for shearing, and it has not changed in forty years.
Nibbling is the other half of the same argument. It can cut an opening that starts and ends inside the part, which a shear cannot do at all, and it can do it on a panel that is already formed and coated, which a flat-bed laser cannot reach. Both operations are slower and less accurate than a fibre laser on a flat blank, and both survive because the laser cannot always be the first tool to the part.
This page sets out the numbers behind nibbling and shearing, the edge each one leaves, and the prototype jobs where sending the part to the laser is the more expensive choice. The stamping and press shop hub covers the press side.
Nibbling and shearing: what each operation can actually do
Shearing cuts a straight line between two edges. Nibbling removes material along a path with a small punch, one stroke at a time, and can therefore cut a closed internal contour.
That is the whole distinction, and it decides most jobs before any tolerance is discussed. A shear produces a blank whose every edge is straight and runs to the sheet boundary. A nibbler produces any contour you can mark, at a kerf of 3–6 mm, with an edge made of overlapping punch marks.
The shear is fast because it stores energy in a flywheel and releases it in one stroke. Its cut is a controlled fracture: the blades penetrate about a third of the thickness, the material ahead of the blade fractures, and the slug separates. Force depends on the length of cut, the thickness and the shear strength, not on the area.
The nibbler is slow because it is a punch doing a milling job. A 10 mm round tool stepping 1 mm at a time removes a 10 mm wide strip at 0.3–0.6 m/min, and the operator steers it along a marked line. The stroke rate is high, in the hundreds per minute, but each stroke only advances the cut by the step.
Shearing and nibbling compared on the numbers
Values are for CRCA and mild steel. Shear strength is taken as 0.75 × tensile strength, which for a 340 MPa CRCA gives 255 MPa. The Ganesh column states what the plant runs.
| Criterion | Guillotine shear | Nibbler | Ganesh limit or note |
|---|---|---|---|
| What it cuts | Straight line, edge to edge | Any contour, including closed internal | Iron worker shears 30–40 mm plate |
| Cut speed | One stroke, under 1 s regardless of length | 0.3–0.6 m/min along the path | Laser at 3–7 m/min on the same thickness |
| Kerf or material loss | Zero, the cut is the line | 3–6 mm strip removed | Kerf is scrap, and it is not recoverable |
| Length tolerance | ±0.2 mm with a set backgauge | ±0.5 mm, operator-steered | ±0.5 mm typical on a cut part |
| Squareness | 0.2–0.5 mm per metre | 0.5–1.0 mm per metre | Checked on the first piece, not the last |
| Edge | Rollover 10–30% of t, then cut zone and fracture | Scalloped, cusp 0.03–0.23 mm | Both deburred before coating |
| Burr | 0.05–0.20 mm, one face | 0.05–0.15 mm on the cut side | Burr limit stated on the drawing |
| Minimum inside radius | Not applicable | Tool radius, typically 3–5 mm | A laser holds 0.5 × t or better |
| Thickness range | 0.5–6 mm on a guillotine | 0.5–4 mm by hand, more on a stationary machine | 0.8–40 mm overall part range |
| Force | F = L × t × τ | One punch, 20–30 T station | 0.77 kN per mm of cut at 3 mm |
| Distortion | Straightens as it cuts, low on thin sheet | Local, can ripple thin sheet | Both need a flat check before bending |
The force row is the one that catches people out. At 3 mm and 255 MPa, every millimetre of cut needs 765 N, so a 1,000 mm cut needs 765 kN, or 78 tonne-force. A shear rated at 100 T will not make that cut in 3 mm over its full bed length if the bed is 3,000 mm and the material is at the top of its strength range. The rating is per stroke, and the stroke is as long as the blade.
The clearance number that decides the edge
Blade clearance is 5–8% of thickness per side for mild steel, which is 0.08–0.12 mm on 1.5 mm sheet. Too little clearance and the blades rub, work-hardening the edge and doubling the force. Too much and the sheet tears ahead of the blade, leaving a rough fracture face and a burr that takes a file to remove.
That gap is set with a gauge and checked on scrap. It is also why a shear is not a universal machine: the same blade setting that gives a clean edge on 1.5 mm leaves a heavy burr on 3 mm, and a shop running both has to reset between jobs. A laser has no equivalent setup, which is part of why it wins on mixed prototype work.
A 300 × 200 mm bracket blank, both ways
Take a 3 mm CRCA blank, 300 × 200 mm, four holes of Ø8 mm and no other features. The blank is the stock for a bend trial, so the outline matters and the holes are there to check the fixture.
On the shear the four cuts run at 300, 200, 300 and 200 mm. At 765 N per millimetre, the longest cut needs 229 kN, well inside the machine. Four strokes plus handling and backgauge setting come to about 36 s. Add 90 s to drill the four holes on a pedestal drill, and the blank is ready in just over two minutes, with no program written.
On the laser the same blank is a 1,000 mm outline at 3.5 m/min, which is 17 s, plus four pierces at 0.4 s and a 30 s load and unload. The machine time is 49 s, which is competitive. The difference is everything before the machine: the nest has to be created, checked against the flat pattern, and queued behind work already on the bed. For a one-off, that is 20–40 minutes of engineering and waiting that the shear does not need.
When the blank is not rectangular
The moment the outline gains a curve, a notch or a 45° corner, the shear stops and the comparison changes completely. A nibbler could follow the contour, at 0.3–0.6 m/min and with a scalloped edge that needs filing. A laser cuts the same contour at 3–7 m/min with a square edge and no operator guidance.
For a 1,000 mm contour with two notches, that is roughly 3 minutes of nibbling against 20 seconds of laser cutting, plus the layout and marking time the nibbler needs. Beyond about 300 mm of non-straight contour, the laser wins on the clock as well as on the edge.
Where shearing still earns its place
Four prototype jobs keep the shear busy in a plant that owns a 12 kW laser.
Bend and weld trials. A strip 300 × 50 mm cut from the same coil as the production part is the only honest way to check springback, minimum flange and weld settings. The shear produces it in seconds, and its edge condition matches the production blank closely enough for the trial to mean something.
Test coupons. Tensile, hardness and coating adhesion coupons are rectangles by definition. Shearing them keeps the coupon edges parallel, which matters when the elongation result is going to be compared against a specification.
Shim and spacer packs. A pack of 0.5–2 mm plates in three thicknesses, cut to the same footprint, is faster to shear than to nest. The thickness tolerance from IS 513 is what sets the pack's accuracy, not the cut.
Blanks for a press trial. Before a tool is signed off, the tool room wants blanks at the exact developed size. Shearing them from the trial coil, with the grain direction marked, is quicker than cutting a nest and gives a cleaner edge for the first-off.
Where nibbling is the only option
A nibbler survives because it can start a cut in the middle of a part. Two situations use that.
An opening in a panel that is already formed. A laser bed is flat. Once a canopy panel is bent into a box, its cable entry cannot be offered up to the bed, and a nibbler or a hole saw is the tool that reaches. The same applies to a rack frame that has been welded and now needs an extra cable cut-out.
A field or fit-out modification. A part already coated and installed cannot go back through a laser without stripping and re-coating. A nibbler with a touch-up pen is the cheaper answer, even though the edge is scalloped and the kerf removes 3–6 mm.
Both cases share a property: the part cannot be flattened. That is the only reason to accept a nibbled edge. Where the part can be flattened, cutting the opening in the flat and forming afterwards is better on every measure, and it is what we would quote.
What we run, and what we do not
Ganesh does not run a nibbling machine. That is a real limit, and it means we do not take on a job whose only route is an opening cut into an assembled frame. What we do have is an iron worker that shears and punches to 30–40 mm, and a 12 kW fibre laser with a 40 ft × 16 ft bed for everything from 0.8 mm to 20 mm plate, plus two 6 kW and two 3 kW machines for thinner work.
For a prototype, that combination gives two honest routes: sheared rectangular blanks from the iron worker, or laser-cut outlines and features from the flat. The laser route is the one that carries the holes, the notches and the corner radii in a single setup, and it is the route that produces a blank matching the production flat pattern, which is what a bend trial actually needs.
The consequence for a prototype drawing is simple. If the part needs an internal opening, put it in the flat and let the laser cut it before forming. If it needs a rectangular blank for a trial, say so and the shear will make it. The laser cutting capability page lists the thickness range and bed size, and laser cutting for prototypes and production covers the quantity at which a prototype route should change.
What to put on the drawing
A sheared or nibbled prototype needs less detail than a production part, but not none. Six fields prevent the usual round of questions.
- Material, grade and thickness, with grain direction if the part will be bent across it.
- Whether the blank may be rectangular, or the outline is final.
- Burr limit and which face must be burr-free.
- Any internal opening, with its minimum corner radius stated so a tool radius can be chosen.
- The bend lines, marked on the flat, so the burr side can be oriented to the inside of the bend.
- Whether the part is a trial blank or a fit-check part, because that changes how much deburring is worth doing.
When this is the wrong route
Shearing is the wrong route for any part with a curved outline, an internal opening, or a tolerance tighter than ±0.2 mm on a feature that is not a straight cut. Nibbling is the wrong route for anything that can be cut flat, anything that will be powder coated without filler, and anything where the kerf loss would change the fit.
Both are also wrong above 6 mm for a guillotine and above 4 mm for a hand nibbler, where the force and the operator effort stop being sensible. Above those thicknesses the iron worker or the laser takes over, and above 20 mm plate the profile leaves our plant.
If the prototype is the first of a family, the more useful question is what the production route will be, because a prototype made on a different process teaches the wrong lessons about springback and burr. Send the drawing and the annual quantity with it, and we will quote the prototype on the route production will use wherever the lead time allows, through the quote request.
Related reading: the Aluminium stamping: formability and tooling considerations and the Blanking, piercing and forming: what each operation does take the same problem further.
Frequently asked questions
Can you shear a hole in sheet metal?
No. A guillotine only cuts a straight line from one edge of the sheet to another, so every cut has to run out to a boundary. A hole or an internal opening needs a punch, a nibbler or a laser, and a sheared edge can never close a loop inside the part.
What tolerance does a sheared blank hold?
About ±0.2 mm on the cut length with a set backgauge, and ±0.3–0.5 mm on squareness over a metre. That is looser than a laser and it is usually irrelevant on a prototype blank, because the blank is the stock for a bend trial rather than the finished part.
Is nibbling still used in production?
Rarely. A nibbler removes material at 0.3–0.6 m/min and leaves a scalloped edge, while a laser cuts the same contour at 3–7 m/min with a square edge. Nibbling survives where the part cannot go on a flat bed, such as an opening cut into an assembled frame.
Why does my part crack at the bend after shearing?
The burr. A sheared edge has a burr on one face, and if that face ends up on the outside of the bend the burr acts as a stress raiser and the bend opens a crack. Deburr the blank, or orient the burr to the inside face, and the crack stops.
Does shear blade clearance matter on thin sheet?
It decides the edge. Clearance is normally 5–8% of thickness per side; too little and the blades rub and work-harden the edge, too much and the sheet tears instead of shearing, which leaves a heavy burr and a rough fracture face. On 1.5 mm that is a 0.08–0.12 mm gap.
Sources and further reading
- IS 513 — Cold reduced carbon steel sheets and strips, including thickness tolerances that set the shear setup · reviewed 3 Nov 2026
- ISO 9013:2017 — Thermal cutting: classification of thermal cuts, which covers laser, plasma and oxy-fuel edges and not mechanical shearing · reviewed 3 Nov 2026
- ISO 2768-1 — General tolerances for linear and angular dimensions without individual tolerance indications · reviewed 3 Nov 2026
Need a prototype blank this week?
Send the flat pattern, thickness and the features you need. We will tell you whether it comes off the shear, the laser or a press, and what the edge will measure.
Request a quote