Slot design for sheet metal: width, length and end radius

A slot that is 1.2 mm wide and 40 mm long will not come off a laser at the width the drawing asks for, and a slot with square ends will crack at the corner the first time the bracket is loaded. Both faults are in the geometry, not in the cutting, and both are cheap to fix in CAD and expensive to fix after the first article.

Slot design for sheet metal comes down to three numbers that are independent of each other. Width is set by the process: the kerf, the punch strength and the coating thickness. Length is set by what the slot has to do, either pass a stud or allow travel. End radius is set by how the part is loaded, and it is the number most drawings leave off entirely, so the CAD default radius is whatever the sketch offered.

The rule below covers each in turn, with the thickness limits we work to, the arithmetic for an adjustment slot, and the callouts that make the drawing unambiguous.

What is the minimum slot width for sheet metal?

The minimum slot width is 1.0 × t for laser cutting, 1.2 × t for punching, and never below 0.8 mm for a laser. For a 2 mm CRCA panel that means a 2.0 mm slot on the laser and a 2.4 mm slot in a press tool.

Laser width is limited by the kerf. A 12 kW fibre laser running 2 mm CRCA cuts a kerf of about 0.2 mm, so a slot drawn as 0.5 mm wide cannot be produced as drawn: the beam consumes the whole feature. At 0.8 mm the kerf is a quarter of the width, and the edges start to taper as the beam loses focus at the bottom of the cut. Below 1 mm wide, throughput also drops because the cut has to slow down to avoid dross welding to the lower edge, which is why very narrow ventilation slots cost more per metre than a wider slot array would.

Punch width is limited by the punch, not the sheet. A narrow punch buckles and snaps when the aspect ratio of the cutting edge to its length gets high. The working floor is 1.2 × t in mild steel, and higher in stainless because the shear strength is roughly 1.4 times that of CRCA, so the same punch sees more load.

Three width rules follow from those limits:

• Below 0.8 mm, stop cutting and drill. A row of 0.6 mm slots is a filter, and a filter is a different part made a different way.
• From 0.8 mm to 1.0 × t, laser only. Fine for airflow and light, not for anything a fastener passes through.
• At 1.2 × t and above, the slot is punchable, and whether it is punched or lasered becomes a volume decision rather than a geometry one.

Slot limits by width, length and material

The values below assume CRCA or mild steel unless the row says otherwise. Width is measured perpendicular to the long axis, and edge distance is from the slot wall, not the slot centre.

The edge distance and ligament figures assume the slot is not near a bend. Anything within the die deformation zone follows a different rule, which the hole-to-bend distance guide sets out: the clearance there is the bend radius plus half the die opening plus the thickness, not a multiple of t.

ISO 9013 classifies thermal cut quality by perpendicularity and dimensional tolerance range, and the width tolerance on a laser slot belongs with a stated class rather than standing on its own. The ±0.1 mm in the table is what we hold on thin sheet at a normal production class; a drawing that asks for better has to say so and accept the slower cut.

Why the end radius decides whether a slot cracks

A slot is an elongated opening, and an elongated opening concentrates stress in proportion to its aspect ratio. For a slot of length L and width W with the load running across the width, the stress concentration factor is:

Kt = 1 + 2L/W

A 20 mm by 5 mm slot therefore has Kt = 1 + 8 = 9. A round hole of any diameter has Kt = 3. Rotate the same slot 90 degrees so the load runs along its length and the factor becomes 1 + 2W/L = 1.5, which is below a round hole.

That single relationship drives the design decision. A slot in a load path should run parallel to the principal stress, not across it. On a bracket that carries tension along its length, a slot for a cable tie should run lengthwise for reasons of strength as well as routing.

End radius matters for the same reason at a smaller scale. A sharp-ended slot behaves like a crack of length equal to the slot length, and the theoretical factor rises further because the corner has no radius to spread the strain. A full-round end, radius equal to half the width, turns the geometry into an ellipse and delivers the Kt in the formula above. A radius of 0.5 mm on a 5 mm wide slot does not do that: it is a square end with a deburr.

Work the numbers on a mild steel bracket. Nominal tensile stress in the net section is 100 MPa. Across a 4:1 slot the elastic peak is 900 MPa, well above the 450 MPa tensile strength of CRCA, which means the material yields at both ends on the first load and the slot grows. Under a fatigue cycle the yielded end becomes the crack initiation site. The same bracket with the slot turned along the load sees a 150 MPa peak and stays elastic.

For stainless and aluminium, be stricter. Both work-harden, so the local yielding at a slot end raises hardness and lowers ductility exactly where the next cycle needs it. On 304 panels we keep slots out of the load path entirely and use a round hole with a slotted washer instead when adjustment is needed.

What the drawing should say about the end

A slot dimensioned as a width and a length with "R0.5" in the corner note is ambiguous, because the CAD may model the radius either inside or outside the length. Dimension the slot as two dimensions plus a true position:

1. Slot width, with a tolerance that follows the process (±0.1 mm on laser).
2. Overall length, measured over the extreme ends of the full rounds, not over the flat sides.
3. End radius equal to half the width, stated as "full round ends".
4. Position of the slot centre, referenced to the part datum, not to the nearest cut edge.
5. A note where the slot is a load path: "slot axis parallel to load direction, full round ends, no sharp corners".

A worked example: sizing an adjustment slot

Adjustment slots fail in two ways: too short to absorb the tolerance they were added for, or too wide for the washer to bridge. Both are arithmetic.

Take an M10 bolt fixing a bracket to a frame, with 6 mm of adjustment needed to line up with a welded nut. The slot must pass the bolt, allow the travel and leave clearance so the bolt shank never bears on the slot end.

Width = bolt diameter + 1 mm = 11 mm.

Length = travel + bolt diameter + 2 mm of clearance = 6 + 10 + 2 = 18 mm.

Check the washer. A standard M10 plain washer is about 20 mm outside diameter, so it reaches 10 mm from the bolt centre. At full travel the slot end sits (18 − 6) / 2 = 6 mm from the bolt centre, leaving 4 mm of washer bearing on the sheet. That is enough. The useful relationship is that the end clearance is always (D + 2) / 2, which is 6 mm for an M10, no matter how much travel the slot carries, because the length grows with the travel. The washer has to beat that 6 mm by enough to spread the clamp load, and a 16 mm washer on the same slot leaves only 2 mm.

Check the width against the washer thickness. An 11 mm slot in 2 mm sheet under an M10 bolt at 40 N·m will dish around the bolt unless there is a washer of at least 2 mm thickness or a doubler plate. Slot width is not the only thing that has to be right; bearing area around it has to be as well.

Now the same calculation on a 3 mm CRCA arm that has to slide 25 mm. Width 11 mm, length 25 + 10 + 2 = 37 mm. The aspect ratio is 3.4:1, Kt is 7.7 across the load, and the arm will flex at the slot. The honest answer is a separate sliding plate or a longer arm, not a longer slot.

Slots in arrays: ventilation and weight reduction

A row of slots for airflow is a percentage problem. Open area per pitch is W/(W + ligament), so 3 mm wide slots with a 3 mm ligament give 50% open area, and 3 mm with a 6 mm ligament gives 33%. The ligament floor of 1 × t applies here as much as anywhere: a 1.5 mm bridge between slots on 1.5 mm sheet bows when the panel is handled and disappears when the panel is welded.

Two practical points about arrays. First, keep the first slot at least 1.5 × t from the panel edge, because the edge is where the sheet is handled and where the bend may start. Second, stagger adjacent rows by half a pitch if the panel has to pass an acoustic or EMI requirement; aligned slots form a straight leakage path and a visible moiré pattern at the same time.

We cut arrays on the 12 kW machine with common-line cutting where the geometry allows, which means two adjacent slots share a wall and the cut length drops. It also means the ligament must be dimensioned to survive shared cutting: the wall is cut once, from one side, not twice with two kerfs, so the finished ligament is the drawn value minus one kerf rather than two.

What Ganesh runs

Slots are cut on the 12 kW fibre laser with the 40 ft × 16 ft bed, in CRCA, mild steel, stainless, aluminium, brass and copper up to 20 mm, with the 6 kW and 3 kW machines handling thinner production work. Widths down to 0.8 mm in 1 to 3 mm sheet are routine; below that we say so at drawing review rather than cutting a feature that will not measure.

Where a part runs in volume, slots move into a press tool. The 20 T to 300 T mechanical presses and 250 T punching line cover the range, and the in-house tool room builds the progressive dies, so a slot that is too narrow to punch gets flagged when the tool is quoted, not when it is tried out. Formed parts then go to the 200 T, 4,450 mm, 5-axis brake, which holds ±0.5° on bend angle and ±0.5 mm on formed features, and to welding on three CLOOS cells or the 30-plus manual stations.

Slots that must stay clear are masked before seven-tank pretreatment and powder coating in the 18 ft × 14 ft booth. If you want the process detail, the design and CAD hub collects the design rules that sit alongside this one, and the laser capability page lists the machines and the thickness range.

When a slot is the wrong feature

If the slot exists to absorb a tolerance problem, fix the tolerance instead. A 15 mm slot on a bracket that should have been dimensioned to ±0.5 mm is a permanent concession that hides a stack-up error, and it weakens the part while doing it.

A slot is also the wrong answer for a rotating joint, for anything that must be sealed against water or dust, and for any edge that a customer will see under a coat of paint. A slot in a visible face shows the sheet thickness at the wall, and no amount of powder covers a 0.16 mm shadow line.

We do not cut slots narrower than 0.8 mm, we do not cut them after forming, and we do not produce square-ended slots to a drawing that states them as such without a query. Send the flat with the loaded slots marked, the adjustment travel and the bend lines, and the widths can be checked against the machine that will run them before the quote is written. Start with request a quote, or read the tab and slot guide if the slot is there to locate a panel rather than clear a fastener.