Content Menu
● Why Hole Quality Actually Matters More Than Most People Think
● Mechanical Punching – Still the Fastest and Cheapest for Round Holes in Volume
● When Punching Stops Making Sense – Drilling, Flow Drilling and Orbital Drilling
● Fiber Laser Cutting – The Default Choice for Everything Else
● Abrasive Waterjet – Zero Heat, Maximum Flexibility
● Plasma – Only When You Really Need Cheap Thick Holes
● Choosing the Process – Quick Reference Table I Use on the Shop Floor
● Final Thoughts After 20+ Years Doing This
● Q&A – Questions I Get Asked Every Week
Cutting a hole in sheet metal sounds straightforward until the part comes back from the floor with torn edges, oversized dimensions or cracks around the hole. I’ve spent years programming turret punches, fiber lasers and waterjet machines, and the same questions keep coming up on every new job: which process actually makes sense for this material, thickness and quantity? What does the edge really look like when it comes off the machine? How much secondary operations hold up? This article pulls together everything I’ve learned the hard way, plus real data from recent papers, so you don’t have to learn it the hard way too.
A hole is never just a hole. The edge condition controls hole expansion ratio in forming, fatigue life in structural parts, thread strength in tapped holes, sealing in gaskets, and paint adhesion on coated material. A rough fractured edge on AHSS can drop hole expansion ratio from 100 % down to 20 % and turn a part that should last a million cycles into one that cracks at 80 000 cycles. That’s why process selection starts with the final use of the part, not with whatever machine happens to be free that afternoon.
Nothing beats a modern servo-electric turret punch when you have hundreds or thousands of round holes in mild steel or aluminium up to about 6–8 mm thick. A 30-ton Amada EM-3612 or Trumpf TruPunch 5000 will hit 900 strokes per minute on 25 mm pitch and hold ±0.05 mm on hole position all day long.
Clearance is everything. Standard rule of thumb is 10 % of material thickness per side for mild steel, 6–8 % for stainless, 12–15 % for aluminium. Too little clearance and you gall the tools and get massive burr. Too much and you get a huge fracture zone and poor dimensional accuracy.
Real example from last year: 1.5 mm DX51D+Z galvanised panels for telecom cabinets, 180 holes per sheet (mostly 4.2 mm and 6.8 mm). We ran a 58-station thick-turret machine with cluster tools and finished a 1250 × 2500 mm sheet in 18 seconds including clamping and repositioning. Tool life on Wilson Tool HP2 coating was 280 000 hits before sharpening. Edge burr stayed under 0.08 mm, perfect for M4 and M6 form tapping straight after.
Another case that surprised a lot of people: 5 mm S700MC high-strength steel brackets for truck chassis. Normal punching gave terrible fatigue results because of micro-cracks on the sheared edge. We switched to a tapered punch with 3° relief and a 0.5 mm stepped die (basically semi-fineblanking). The edge became almost 100 % burnish and the rotating-bending fatigue limit jumped from 180 MPa to 420 MPa. The paper that describes the exact geometry is listed in the references below.
As soon as thickness goes above 8–10 mm or the material is too brittle for shearing, we move to rotational processes.
Flow drilling (also called thermal drilling or Formdrill) is brilliant on thin gauges when you need a strong thread. The tool friction-heats the material and extrudes a bushing three to four times the original thickness in one stroke. 1.2 mm 5052.5 mm 5052-H32 aluminium becomes a 4 mm tall M8 bushing with full thread engagement. No chips, no secondary bushing needed.
Orbital drilling on CNC machining centres is the go-to for titanium and stacked CFRP/aluminium in aerospace. The tool spins on its own axis and simultaneously orbiting around the hole centre, so radial forces drop by 70–80 % compared to normal drilling. We routinely hold H7 on 8 mm Ti-6Al-4V at 12 mm depth with surface finish Ra 0.8 µm.
By late 2025 pretty much every job shop has at least a 12 kW fiber laser with nitrogen assist. Speeds are ridiculous: 4 mm 304 stainless at 35 m/min for contours, round holes down to 4 mm diameter cut at 15–20 m/min with wobble head for perfect cylindricity.
Rule of thumb for good round holes: – Hole diameter ≥ material thickness (ideally ≥ 1.5 × t) – Use high-pressure nitrogen (20–25 bar) and beam wobble or pulse ramping on small holes – Lead-in/lead-out geometry matters – micro-joints or overburn on skeleton side prevent dross spikes
Example: heat-exchanger end plates, 3 mm 316L, 3200 holes 5.0 mm diameter. A Bystronic ByStar Fiber 15 kW with “CleanCut” mode and 3 mm wobble amplitude produced holes with 0.03 mm roundness error and zero burr, ready for orbital welding without any cleanup.
Thicker material gets trickier. At 15 mm mild steel we still use oxygen-assist laser for speed, but dross on the bottom is inevitable. Acceptable if the hole is just for a bolt, not acceptable if it has to seal against pressure.
When heat is absolutely forbidden (hardened tool steel, armour plate, copper bus bars, laminated materials), abrasive waterjet is the only realistic option. Modern KMT or Flow 90 000 psi intensifiers with dynamic piercing hold taper under 0.1 mm on 25 mm aluminium.
Downside is speed and garnet cost. 3 mm stainless runs about 600–800 mm/min, 25 mm aluminium maybe 80 mm/min. Starting holes is the critical part – vacuum assist or low-pressure pre-pierce prevents delamination on composites.
New Hypertherm XPR300 with True Hole technology actually produces bolt-ready holes in mild steel up to 25 mm, but you still get 0.5–1 mm dross and a heat-affected zone. Fine for structural fabrication, terrible for stainless or anything corrosion-critical.
| Thickness | Material | Quantity | Tolerance | Edge Requirement | Process I Pick 90 % of Time |
|---|---|---|---|---|---|
| 0.5–3 mm | Mild steel | >500 | ±0.1 mm | Low burr | Servo turret punch |
| 1–6 mm | Stainless | 10–5000 | ±0.1 mm | Oxide-free | 12–20 kW fiber laser + N2 |
| 8–50 mm | Aluminium | <200 | ±0.2 mm | Zero HAZ | Abrasive waterjet |
| 4–10 mm | AHSS / UHSS | >1000 | ±0.15 mm | Fatigue critical | Special thickened-edge punch |
| >20 mm | Any | Any | ±0.5 mm | Structural only | Hi-def plasma or waterjet |
The biggest money losers I see are shops that bought a big laser and now try to laser everything, or shops that still think punching is dead. Reality is hybrid is king: keep a good servo punch for round holes in mild steel and aluminium up to 6 mm, keep a modern fiber laser for stainless, complex contours and medium runs, and outsource or own one waterjet for the thick and exotic stuff.
Pay attention to edge condition from the start, because fixing it later with deburring, reaming or hand grinding eats profit faster than anything else. And always always test the actual process on scrap from the same coil the production parts will come from – coil-to-coil variation will bite you otherwise.
That’s it. Go make good holes.