How to Prevent Stamping Burr: Clearance, Wear and Setup Fixes

A connector OEM shipped 40,000 brass terminals before anyone noticed the burrs. The raised edge on the contact face was 0.08 mm tall. That is twice the 0.04 mm print limit, and it pierced the mating seal on 312 assemblies. The sort, rework and freight claim ran past $18,000. The root cause was a punch-die clearance that had opened from 6% to 11% of material thickness over 400,000 strokes without a regrind. Burr is the cheapest defect to prevent. It is also one of the most expensive to discover late, because it grows quietly with wear until it crosses a print limit nobody rechecks.

This guide covers the mechanism, the acceptance numbers, the five root causes and the fixes. It is written for die buyers, tooling engineers and stamping suppliers who want burr controlled at the source instead of removed at the deburr bench.

The Snapshot

• Acceptable burr height is typically 5-10% of material thickness for general parts, and often capped at 0.02-0.05 mm for connector contacts and safety-critical edges.
• Total blanking clearance ranges from 6-9% of stock thickness for soft aluminum to 14-20% for high-strength steels (typical industry figures).
• Regrind the cutting edge once the wear land radius reaches roughly 0.025 mm for precision work or 0.10 mm for general blanking. Burr height climbs steeply past these points.
• A ±10% hardness swing between coil lots can push burr height up 30-50% on the same die setup.
• Manual deburring costs $0.02-$0.10 per part in typical shops. At 500,000 parts per year that is $10,000-$50,000, often more than fixing the die.

Why Stamping Burrs Form: The Fracture Mechanics

Shearing does not cut sheet metal the way a knife cuts paper. The punch drives into the stock and the material passes through four zones on the cut edge: rollover, burnish, fracture and burr. Rollover is the drawn-in radius at the top, typically one third of the thickness. Burnish is the smooth vertical wall the punch polished as it traveled, and fracture is the rough tear where cracks ran out from both cutting edges. The burr is the metal left at the exit side when the final separation happens off-axis.

Two cracks start the separation. One initiates at the punch corner and one at the die corner, and in a healthy setup they meet in the middle of the thickness. When clearance, wear or alignment is wrong, the cracks miss each other. The material between them does not fracture cleanly, so it gets dragged and torn, and the leftover lip folds down into a burr. That is why burr is a geometry problem, not a material cleanliness problem.

The fracture angle tells the story. A clean cut shows fracture faces converging at roughly 5-8 degrees per side for mild steel with correct clearance. Oversized clearance widens that angle, deepens the rollover past 40% of thickness and lifts the burr. Undersized clearance forces the cracks to run parallel. The wall shows a double bright burnish band, and the edge grows a thin, sharp secondary burr that is easy to miss at 1x magnification.

Burr direction also carries diagnostic value. Burr normally forms on the die side for a blanked part and on the punch side for a pierced hole. If you find burr flipping sides or appearing on both faces, suspect punch-to-die misalignment beyond about 0.01 mm or a worn guide system. Section 7 covers the alignment numbers.

How Much Burr Is Acceptable? Height Limits by Thickness

There is no single global burr limit, but the industry converges on two rules. The general rule caps burr height at 5-10% of material thickness. The functional rule caps it by what the part must do: seal, mate, conduct, or stay safe to handle. Whichever number is smaller governs the print. The die must hold it for its full life between regrinds, not just at T1.

← swipe to scroll →

Table note: typical industry acceptance figures. Contact and safety limits reflect common OEM print practice. Your part print always governs.

Connector terminals sit at the tight end. A burr on a contact face breaks plating continuity and raises contact resistance by milliohms that matter in low-voltage circuits. Automotive sealing edges and airbag bracketry often call out 0.05 mm maximum with no broken-corner allowance. General structural brackets tolerate the 10% rule, but hand-safe edges still cap around 0.15-0.20 mm so assemblers do not get cut.

Write the limit into the die design brief, not only the part print. The die maker needs to know whether the burr target is 0.02 mm or 0.15 mm before selecting clearance, steel and grinding finishes. A target that tight usually pushes the design toward fine blanking or a shaved-edge station. Conventional blanking struggles to hold under 0.02 mm on stock thicker than 1.0 mm.

Blanking Clearance: The Number One Burr Root Cause

Clearance is the gap between the punch and the die opening, expressed as a percentage of material thickness. It controls where the two fracture cracks meet. Get it right and the cracks converge cleanly, leaving a small, predictable burr under the print limit. Get it wrong and no amount of deburring downstream will fix the economics.

Excessive clearance lets the material bend into the gap before it fractures. Rollover grows past 40% of thickness, the fracture face tilts, and the exit-side lip stretches into a tall, ragged burr. The burr height roughly doubles when total clearance opens from a correct 8% to a worn 16% on 1.0 mm mild steel. Excessive clearance also thins the cut wall, which ruins press-fit features and welded-edge quality.

Insufficient clearance looks better at first glance and is worse over time. The cracks cannot converge, so the material shears twice. You see a second bright band on the cut wall and a thin, sharp, sometimes hooked burr. The extra shear force can rise 20-30% above the correct-clearance baseline, which accelerates edge chipping on hard steels and overloads slim punches. Small clearance trades a visible defect for hidden die damage.

Clearance is a total value across the cut, split between both sides. For round punches, the die opening is larger than the punch by the full clearance. A common spec error is quoting per-side and total values interchangeably on one drawing. That mistake yields dies cut at double or half the intended gap. Always state total clearance and the percentage basis on the die drawing. Our own practice at DieStrike holds key die dimensions to ±0.005 mm and die component geometry to ±0.002 mm. The as-built clearance then matches the design intent. DFM feedback returns within 24 hours when a clearance callout is ambiguous.

Clearance Settings by Material and Thickness

Clearance scales with material behavior, not just thickness. Soft, ductile metals need less gap because their cracks run straight. Hard and high-strength grades need more gap so the fracture has room to converge before the material work-hardens against the edges. The table below lists typical total clearance ranges used across the stamping industry.

← swipe to scroll →

Table note: typical industry clearance ranges for conventional blanking. Fine blanking uses under 1% clearance and is a different process. Verify against your material supplier data for exact grades.

Thickness multiplies everything. A 1.0 mm stainless sheet wants about 0.10-0.14 mm total clearance, while a 4.0 mm stainless plate wants 0.40-0.56 mm. The same die drawing that says "10% of t" covers both, but a fixed absolute number does not. When a program mixes gauges, design the die for the thickest stock. Accept a small clearance penalty on thin gauges, or build gauge-specific insert sets.

If your clearance data lives only in the original die design and the tool has been reground five times, re-derive it. Measure the punch and die opening with an optical comparator or CMM at three heights, compute the actual gap, and compare to the table above. Our punch clearance maintenance guide walks through the measurement and shim-correction procedure station by station.

Edge Wear Radius: When Sharp Edges Start Making Burrs

Every cutting edge starts sharp, with a ground radius under 0.005 mm, and every stroke rounds it off. Wear builds a small radius on the punch and die corners, called the wear land. As that radius grows, the edge stops initiating a clean crack and starts pushing material aside instead. The pushed material tears late, off-axis, and lands as burr. Wear is the reason a die that passed T1 fails burr checks at stroke 500,000.

The regrind trigger depends on how tight your print limit is. For precision connector terminals and laminations, regrind once the wear land reaches about 0.025 mm radius. For general blanking under the 10% rule, shops typically allow up to 0.10 mm before the burr climbs past the print. The burr height curve between those points creeps slowly, then rises steeply, so waiting for visible burr means you are already late. Measure the land under a microscope instead, at a typical cadence of every 100,000-250,000 strokes for D2 on mild steel.

Edge steel choice sets how fast that curve runs. D2 at HRC 58-60 holds a wear land for roughly 300,000-800,000 strokes on low-carbon steel, a wide band that depends on lubrication and material cleanliness. Powder-metallurgy HSS and carbide extend the interval by 3-5x on abrasive or stainless grades. Carbide is effectively mandatory on silicon electrical steel, where D2 might survive only 50,000-100,000 strokes. DieStrike builds stamping dies and cutting components to ±0.005 mm with cavity heat treatment at HRC 62. Our stamping die steel guide covers the selection logic in detail.

Regrinding has its own burr risk. Removing stock from the punch face changes the press-fit relationship and, on clearance-controlled pairs, shrinks or grows the effective gap. A regrind plan should specify how many millimeters come off each face per event, typically 0.05-0.15 mm, and whether shims compensate to restore clearance. Our companion guide on setting a die regrind schedule gives the interval math and the shim-restoration procedure.

Chipping is the failure mode to separate from wear. A worn edge rounds and produces growing burr gradually. A chipped edge produces sudden, localized burr spikes at one spot on the perimeter, often with a visible nick under 10x magnification. Chipping comes from excessive hardness above HRC 62 on shock-loaded edges, thin sections, misalignment or double-hits. If burr appears at the same clock position every part, stop counting strokes and inspect for a chip. A chipped edge needs weld repair or insert replacement, not a routine regrind.

Material Hardness and Thickness Variation

The die is only half the burr equation. Sheet metal arrives with its own scatter, and a setup tuned at the coil's median will drift at its extremes. Two variables matter most: hardness and thickness tolerance.

Hardness changes how the crack initiates. A harder coil resists crack initiation longer, the crack runs at a shallower angle, and the fracture misses its partner, leaving taller burr. Industry experience puts the effect at a 30-50% burr height increase across a ±10% hardness swing. That is a low-carbon steel specified at HRB 60-70 arriving at HRB 75.

Thickness works the other direction from clearance: the same absolute gap is a smaller percentage of a thicker coil, so an upper-tolerance coil runs under-cleared and a lower-tolerance coil runs over-cleared. On a 1.00 mm nominal with ±0.05 mm mill tolerance, an 8% design clearance swings from 7.6% to 8.4% effective, usually acceptable. On a 4.00 mm plate with ±0.20 mm tolerance, the same swing is far wider, and burr variation shows up lot to lot.

Control the incoming material the way you control the die. Require the mill cert with every coil and check three fields: hardness, thickness tolerance and surface condition. Run a first-off burr check at each coil change and record the value, because the coil-to-coil delta often exceeds the stroke-to-stroke drift. If burr jumps when the coil lot changes and the die has not moved, you have a material problem. The fix belongs in the purchase spec, not the tool room.

Temper matters as much as grade. Quarter-hard and half-hard tempers shear with progressively more burr than dead-soft stock. Expect 20-40% more at identical clearance, because ductility drives the crack path. If your product needs hard-temper strength with soft-temper edge quality, blank soft and harden after. Otherwise budget a deburr station and say so in the cost model.

Setup, Feed and Alignment Errors

A perfect die in a sloppy setup still makes burrs. Three setup errors dominate: punch-to-die misalignment, feed misregistration in progressive dies, and speed-lubrication mismatch.

Misalignment shifts the clearance unevenly around the cut perimeter. If total clearance is designed at 8% of a 1.0 mm sheet, that is 0.08 mm total, 0.04 mm per side. A lateral punch shift of just 0.02 mm cuts one side to 0.02 mm and opens the other to 0.06 mm. One flank shears under-cleared while the opposite flank tears over-cleared. The part shows burr on one side only, a classic tell.

Die sets should hold punch-to-die alignment within about 0.01 mm for precision work, using quality guide posts and bushings. Shimless die sets with ball-bearing guides make that repeatable after every teardown. Check alignment at every die re-install, not just at build.

Feed misregistration attacks pilot-located progressive dies. When the strip over-feeds or under-feeds past the pilot allowance, typically ±0.05-0.10 mm, the pilot pins deform the pilot holes before the cut station fires. The cut then lands off-position relative to the previous stations. The result is uneven wall thickness and burr that wanders along the edge instead of sitting in one place. Servo feeds hold registration near ±0.02 mm at speed, while worn mechanical feeds drift to ±0.10 mm or worse as the machine heats up. Verify feed accuracy at operating temperature, not cold at startup.

Speed and lubrication change the fracture, too. Higher stroke rates heat the cutting edges. Edge temperature above roughly 150-200 °C softens the temper of standard tool steels and accelerates wear, pulling the regrind interval forward. Lubrication reduces the friction that drags material into the clearance zone, and running dry can raise burr height 20-50% on stainless and aluminum. If burr grew after a productivity push from 300 to 600 SPM, check edge temperature and lube coverage before you blame the die.

One more setup variable deserves its own line: stripper and pad pressure. Insufficient strip pressure lets the material lift with the punch on retraction, stretching the exit burr. Typical practice holds the stripper force at 5-10% of the strip force for general blanking and higher for thin, springy stock. Nitrogen spring pressure should be checked with a load cell at maintenance intervals, since gas springs lose 3-8% of rated force over millions of cycles.

Deburring Cost vs Fixing the Die

Every stamping program eventually asks the same economic question: remove the burr downstream, or fix the die upstream? The honest answer needs numbers, and the numbers usually favor the die.

Manual deburring with hand tools or abrasive pads runs $0.02-$0.10 per part in typical shops once labor, fixturing and inspection are loaded. Tumbling and vibratory finishing drop the per-part cost to $0.005-$0.02 but add 2-24 hours of batch cycle time and cannot selectively protect finished surfaces. Brush deburring lines cost more to install but suit high volume at $0.002-$0.01 per part. At 500,000 parts per year, manual deburring alone consumes $10,000-$50,000 annually, recurring forever. A die correction, whether insert replacement, clearance rework or an added shave station, typically costs a few thousand dollars once.

The crossover rule of thumb: if deburring costs exceed the die fix within 6-12 months, fix the die. Add the hidden costs and the case strengthens. Deburring adds handling marks, adds WIP queues that stretch lead time, and cannot fully recover parts whose burr tore the cut wall below tolerance. For connector OEMs and automotive suppliers under IATF 16949 controls, a deburred edge also complicates process validation. The defect is managed downstream rather than prevented at the source.

There are legitimate keep-deburring cases. Prototype runs under 5,000 parts, where a die correction would not amortize. Mixed low-volume families sharing one die, where no single burr source justifies rework. And materials like some aluminum tempers that form soft, foldable burrs a downstream wipe operation flattens for under $0.005 per part. Document the decision either way, with the annualized math, so the choice survives audits and management changes.

When the decision is fix, rank the fixes by cost. Re-set clearance with shims first, often under $500 of tool-room time. Replace or regrind worn inserts next, typically $200-$1,500 per set depending on steel. Upgrade edge steel from D2 to PM-HSS or carbide inserts, $1,000-$5,000, when abrasion is the wear driver. Add a dedicated shave or fine-blanking-style station last, at $5,000-$20,000. Reserve it for edges that must hold under 0.02 mm burr on volumes above one million parts.

Burr Inspection at Die Acceptance

Burr control fails most often at the handover gate, because acceptance testing checks dimensions and rarely quantifies the edge. A die that ships with 0.09 mm burr against a 0.05 mm limit will produce scrap from its first production shift. Build burr measurement into the die buy-off the same way you build trial and first-article approval into a mold buy-off.

The measurement method depends on the limit. For general parts at 0.05-0.15 mm, a height gauge or calibrated burr comparator reads burr height to ±0.01 mm in seconds. For connector-class limits under 0.03 mm, use an optical comparator at 20-50x or a CMM edge scan, measuring peak burr height above the nominal face plane. Measure at minimum four positions per cutting station, including the start and end of cut where cracks close last, and record values, never pass/fail alone.

Sampling should span the run, not just the beginning. Take burr readings at stroke 0, 1,000, 10,000 and 50,000 during acceptance, then plot the trend. A flat line under the limit means the die holds. A rising line that crosses the limit before the next planned regrind is a red flag. The acceptance should then set the regrind interval to that crossing point, minus a 20% safety margin. This single plot converts burr from a mystery into a maintenance schedule.

Acceptance paperwork should carry four burr-specific fields. They are the print limit with its measurement method, the measured value at each sampled stroke, the derived regrind interval, and the as-built clearance at each station. Without all four, the receiving plant inherits a die whose burr behavior is undocumented and whose maintenance schedule is guesswork. DieStrike ships every stamping die with dimensional reports on the ±0.005 mm die tolerances. Trial sampling establishes the burr trend before the tool ever reaches your press. Die lead times run 2-5 weeks, and standard components like punches ship in 3-7 days.

Burr Prevention Checklist for Design and Maintenance

The five root causes map to a checklist you can run at die design, at every setup, and at every maintenance window. Each line has a number attached, because a checklist without thresholds is just a wish list.

Design phase. State total clearance as a percentage of thickness on every station drawing, using the material table above. Specify edge steel against the material's abrasiveness, carbide for silicon steel and volumes over 5 million strokes. Put the burr print limit and its measurement method in the die design brief. Size guide systems for 0.01 mm alignment on precision stations.

Setup phase. Verify punch-to-die alignment within 0.01 mm at install. Confirm feed registration within ±0.05 mm at operating temperature. Check stripper force against the 5-10% of strip force baseline. Run first-off burr measurement at every coil change and record it against the coil cert.

Maintenance phase. Measure the wear land every 100,000-250,000 strokes, regrind at 0.025 mm for precision or 0.10 mm for general work. Re-derive actual clearance after every regrind and shim back to design. Inspect for edge chipping whenever burr localizes at one clock position. Log nitrogen spring force and replace at 10% loss.

Escalation triggers. Burr crossing 80% of the print limit between scheduled regrinds means shorten the interval or upgrade edge steel. Burr that flips faces means realign before anything else. Burr that only appears on new coil lots means tighten the purchase spec. When two triggers fire at once, pull the die, because running marginal produces sorting costs that exceed downtime within weeks.

FAQ: Stamping Burr Questions

Q1. What burr height is acceptable in metal stamping?

The general industry rule is 5-10% of material thickness, so 0.05-0.10 mm on 1.0 mm stock. Connector contacts and safety edges typically cap at 0.02-0.05 mm regardless of thickness. The part print always governs, and the die must hold the limit at end of regrind life, not only when new.

Q2. Why did burrs suddenly appear after months of clean production?

Sudden localized burr usually means a chipped edge, often from a slug pull-up, double-hit or hardness spike. It repeats at the same clock position on every part. Gradual burr growth across all edges means wear, and the fix is regrinding at the 0.025-0.10 mm wear-land threshold. Burr that appeared at a coil change points to material hardness or thickness variation.

Q3. Does more clearance reduce burr?

No. Burr rises on both sides of the optimum. Correct total clearance is about 7-10% of thickness for mild steel. Going to 15-16% roughly doubles rollover and burr height while thinning the cut wall. Under-cleared dies make thin, sharp secondary burrs and raise cutting force 20-30%, which damages the edges faster.

Q4. Is deburring cheaper than fixing the die?

Rarely at volume. Manual deburring costs $0.02-$0.10 per part, so $10,000-$50,000 per year at 500,000 parts. A clearance rework or insert replacement typically costs a few hundred to a few thousand dollars once. If the deburr bill exceeds the die fix within 6-12 months, fix the die.

Q5. How do I check burr without expensive equipment?

Run a fingernail or cotton wipe across the edge: burr above about 0.05 mm catches cotton fibers visibly. For numbers, a $200-$500 burr comparator gauge or a depth micrometer against a surface plate reads to ±0.01 mm, enough for the general 10% rule. Reserve optical comparators and CMM scans for limits under 0.03 mm.

The Bottom Line

Burr is a geometry problem with five measurable causes: clearance, edge wear, hardness swings, misalignment and feed drift. Set the burr limit in the die brief and regrind on numbers, not eyesight. DieStrike builds IATF 16949 stamping dies to ±0.005 mm in 2-5 weeks. Talk to our die engineering team before your next coil runs.