Tool life and cost per hole in busbar punching

Ask a busbar shop what a punched hole costs and you usually get the price of the punch divided by a number somebody remembers. That figure is almost always too high as a share of the total, and it hides the two terms that dominate. Working the model through changes what you optimise.

The model

cost per hole = (tool cost + n × regrind cost) / total holes over tool life
              + machine rate × cycle time
              + tool-change downtime allocated per hole
              + scrap cost allocated per hole

Four terms. The first is what people call "tooling cost". The second is usually three to five times larger. The third is small but responds strongly to tool choice. The fourth is small on average and catastrophic in the tail.

A worked example

Take a 13 mm round punch and die set piercing M12 clearance holes in 10 mm half-hard copper on a machine running 150 hits per minute. Every rate below is illustrative and should be replaced with your own; the point is the structure and the relative sizes, not the absolute figures.

Start with the tooling. An M2 high-speed steel punch and die set at €220. Regrind at €30 including handling, inspection and re-shimming. Eight regrinds before the punch runs out of grind length. Life between regrinds of 30,000 hits, which is the figure that needs verifying with your tooling supplier for your material, machine and clearance.

total holes  = 9 intervals × 30 000 = 270 000
tooling cost = (220 + 8 × 30) / 270 000 = €0.0017 per hole

Machine time next. At €55 per hour all-in and 150 hits per minute, each hole occupies 0.4 seconds.

machine cost = 55 / 3600 × 0.4 = €0.0061 per hole

Then downtime. Nine tool changes at 12 minutes each is 1.8 hours of lost production.

downtime cost = 1.8 × 55 / 270 000 = €0.00037 per hole

Finally scrap. A 3 m length of 100 × 10 mm copper weighs 26.9 kg. At a copper price of €9/kg that bar is worth €242 as raw material alone, before any processing already invested in it. Suppose one bar in four tool lives is lost to a pulled slug, a collapsed edge or burr that misses the drawing limit.

scrap cost = 242 / 4 / 270 000 = €0.00022 per hole

That totals about €0.0084, or 0.84 euro cents per hole. Tooling is 20% of it. Machine time is 73%.

The punch is not where the money is. Sharper still: a single scrapped bar at €242 costs the same as the entire tooling consumption of about 142,000 holes. Tool life matters mainly because short tool life produces bad holes, and bad holes produce scrap and rework. It does not matter much because punches are expensive.

Punch materials

Four families cover essentially all busbar work.

D2 cold-work tool steel, typically 60-62 HRC. Cheap, dimensionally stable, adequate for thin or soft material and short runs. On thick copper the edge strength is marginal and chipping under snap-through shock is the usual failure. Suppliers offering both D2 and M2 steer D2 toward thin sheet, small quantities and cases where a design change is likely.

M2 conventional high-speed steel, 60-63 HRC. The default. Better abrasion resistance, impact resistance and toughness than D2, and it tolerates the frictional heat generated at the punch flank. Anchor Danly's punch material notes describe M2 as suited to long runs where abrasion resistance is the main requirement. For most busbar shops this is the baseline.

Powder-metallurgy high-speed steel: ASP 2023, PM 23, Vanadis grades and equivalents, many of them M3

chemistry to material number 1.3395. Chemistry similar to M2, but the carbides are far finer and more evenly distributed, which raises wear resistance without the toughness penalty you would normally pay. PASS Stanztechnik publish a tool-life factor of 6.0-10.0 for their PM punch steel against their baseline. That is a large multiplier from a component costing perhaps 1.5 to 2.5 times a conventional punch.

Tungsten carbide. Highest wear resistance by a wide margin, and the wrong answer for most busbar punching. Carbide is brittle, and snap-through on 10-20 mm copper delivers exactly the reverse shock loading it dislikes. It needs rigid guiding, accurate clearance and clean slug control before it will survive. It earns its place on very high volume at a single hole size, or on abrasive material, neither of which describes typical switchgear work.

Slot the choice into the cost model rather than arguing it in the abstract. In the example above, moving from M2 to PM steel at €420 for the set with a 6× life factor gives:

total holes  = 9 × 180 000 = 1 620 000
tooling cost = (420 + 8 × 40) / 1 620 000 = €0.00046 per hole

Tooling cost per hole falls by roughly three quarters, and the downtime term falls by the same factor because there are one sixth as many tool changes for the same output. The punch being nearly twice the price is irrelevant.

Coatings

A PVD coating is a few microns of ceramic on the flank and face. It raises surface hardness far above the substrate, and it changes the friction and adhesion behaviour at the interface. On copper, friction and adhesion are what you are buying.

Published data from a commercial coating house gives the following, and the numbers are representative rather than universal:

TiN is the general-purpose benchmark. TiCN is harder and better on abrasive work, but its 400 °C working limit is the lowest in the table. AlTiN's 900 °C capability is what makes it valuable in machining and it buys nothing at all on copper busbar, where the tool never approaches those temperatures.

For copper and aluminium the failure mode is adhesion and galling, not thermal softening. That points at the low-friction options: a chromium-based coating, or any coating carrying a molybdenum/tungsten disulphide dry-film top layer, which drops the coefficient of friction from around 0.35 to around 0.15. Less friction on the flank means less withdrawal force, and withdrawal is where much of the wear happens. PASS quote a tool-life factor of 2.0-4.0 for an appropriate coating, with the qualification that it should be matched to the workpiece material rather than chosen generically.

Two practical constraints. PVD deposits at around 375 °C, below the tempering temperature of hardened M2, so it does not soften the substrate; CVD deposits near 1050 °C and requires the tool to be re-hardened afterwards, which is why CVD lives on inserts and heavy forming tools rather than precision punches. And the face coating is lost the first time you regrind. The flank coating survives, which is where most of the benefit sits, so a reground coated punch still beats an uncoated one, but check the economics of recoating against the remaining grind life.

Regrind stock and how many regrinds you get

Grind life is the total length that can be removed from the punch point by sharpening. It is set by geometry, not by wear:

grind life = SBR − (material thickness + die penetration + stripper thickness)

where SBR is the straight-before-radius length, the distance from the tip back to where the punch begins to widen into its body. Grind past that and the punch will no longer pass through the stripper or will no longer produce the right hole. Die penetration is the amount the punch must enter the die at the bottom of the stroke to snap the slug free, usually a few tenths to about 3 mm depending on the tooling system.

The number of regrinds is the available grind life divided by the stock removed per regrind. Removal is typically 0.1-0.3 mm, enough to take out the wear land and a little clean metal beneath it, so a punch with 3 mm of usable grind length supports somewhere around 10-30 regrinds on paper. In practice you get fewer, because a punch run too far past its wear limit needs 0.5 mm or more taken off, and because each regrind requires the punch length to be restored with shims or head adjustment until the tolerance stack runs out. Every one of these figures depends on the tooling system and the punch geometry, so confirm them against your supplier's data rather than adopting the numbers here.

Grind wet, always. Dry grinding or inadequate coolant can crack or anneal the cutting edge and destroy the punch you were trying to save. Remove the grinding burr with a fine stone, clean off the coolant, lubricate on assembly and re-establish the punch length.

Wear land as the regrind trigger

The cutting edge does not stay sharp and then suddenly fail. It develops a small flat, the wear land, which grows steadily and changes the cut long before the punch is unusable. Once a measurable land exists, more of the material is being deformed rather than sheared, so force rises, rollover grows, burr grows and the hole drifts.

Measuring the land directly means taking the punch out and putting it under a microscope or comparator. Do that periodically to calibrate your judgement; it is not a production control. A land in the region of 0.2-0.4 mm is a common regrind trigger, and confirm the figure with your supplier for the specific tool. In production the practical triggers are the symptoms: burr height against the drawing limit, which suppliers are explicit is the normal determinant of regrind timing and is the most sensitive early indicator; rollover growth measured against a retained first-off sample; hole dimension and position drift; part warp; and rising press pressure or motor current on the stroke.

The discipline that matters is grinding early. A punch caught at 0.2 mm of wear land needs 0.15 mm removed and goes back to work. The same punch run until the edge collapses needs 0.5 mm or is scrap, and everything it made in between is suspect.

The clearance lever

Of everything in this article, punch-to-die clearance moves tool life further than any other single variable, and it costs nothing to get right.

The mechanism is at withdrawal, not at the cut. When clearance is too tight, the copper springs back onto the punch flank after snap-through and every stroke ends in an interference fit that has to be forced out. Dayton Lamina's field data attributes up to two thirds of total punch wear to withdrawal, and describes tight clearance producing a hole smaller than the punch point, with frictional heat discolouring the metal behind the tip and damaging the heat treatment. Correct clearance produces a hole slightly larger than the punch, converting a press fit into a slip fit. Their published comparison of a traditional 5% per side clearance against their engineered clearance on the same material reports 41 hours of maintenance per million parts for the tight setting, with a case study on 0.060 in cold-rolled steel showing a three-fold increase in punch life from clearance alone.

Stack that against the other multipliers. Correct clearance contributes roughly 2-3×. PM substrate contributes 6-10×. An appropriate coating contributes 2-4×. Running with lubricant rather than dry avoids a 0.4-0.6 penalty. These do not multiply cleanly, and nobody should claim a hundredfold improvement, but the spread between a badly specified tool in a badly clearanced die running dry and a properly specified one is close to an order of magnitude. Most of that gap is closed by decisions that cost nothing at the point of order.

Two specific busbar traps. Small holes in thick bar carry a heavy penalty, with PASS publishing derating factors of 0.6-0.8 for a cut feature smaller than 1.5 times material thickness and 0.3-0.5 for one smaller than 1.0 times thickness; an M12 hole in 12 mm copper sits squarely in that second band. And hard-temper bar is harder on tooling than half-hard, so a temper substitution by procurement surfaces as a tool-life problem weeks later. That is one reason to hold the supplier to a stated temper under EN 13601. The clearance side is treated in full in punch-to-die clearance for copper and aluminium busbar, and the force side in the busbar punching force calculator.

Lubrication

Running dry costs roughly half your tool life. PASS's published factor for no sheet lubrication is 0.4-0.6, and copper adheres readily to tool steel.

There is a conflict here. Oil film adhesion between the slug and the punch face is one of the recognised slug-pulling mechanisms, and "do not use processing oil" appears on tooling suppliers' own lists of slug-pulling remedies. Removing the lubricant to cure slug pulling works, and it is the wrong trade. Fix slug retention mechanically, keep the lubricant, take the tool life. Then account for the downstream consequence: lubricant residue on a joint face is a contamination problem before plating, tinning or assembly, and the cleaning step belongs in the process rather than being discovered at inspection.

Slug pulling

A pulled slug is the most expensive routine event in punching. It rides up with the punch, lands on the bar and is driven into the surface on the next stroke. Best case, a marked bar and a rework. Worst case, a chipped punch, a damaged die and a scrapped length.

The mechanisms on copper are vacuum adhesion between the bowed slug and the punch face, oil film adhesion, and edge pressure bonding from burrs. Magnetic adhesion, the fourth on the standard list, does not apply. The countermeasures, roughly in order of how often they solve it:

• Spring-loaded ejector pin with a side vent hole in the punch. The pin pushes the slug off; the vent breaks the vacuum. This is what makes generous clearance usable.
• Slug-retention die geometry, a short cutting land followed by a slightly narrower relief that grips the slug so it cannot climb back.
• Shear angle on the punch face, which bows the slug so its own springback releases it.
• Restricting punch entry into the die at the bottom of stroke, since deep entry creates more vacuum on withdrawal.
• Compressed air through the tool centre.

Increasing clearance to buy tool life increases the slug-pulling tendency. The ejector pin is what you buy to take that trade.

What to do with the model

Populate the four terms with your own machine rate, tooling prices and hit rate, then look at where the money is. On most busbar work machine time dominates, which means hit rate and setup time matter more than punch price. A 24-station processing centre running 200 hits per minute moves cost per hole further than any tooling decision available to you, and a 400 or 600 kN punch and shear machine with a fast tool change moves it further than a cheaper punch. Tooling choice then matters for the second-order reason: it keeps holes inside the drawing limits for DIN 43673-1 patterns without stopping the machine, and it keeps bars out of the scrap bin.

Track three numbers per tool and the model stays honest: hits since last regrind, burr height at first-off and at the trigger point, and grind length remaining. For a shop that has never measured any of them, start with burr height at first-off. It is the cheapest measurement in the building and it drives the rest.

Where the burr limit is tight, take it out downstream rather than chasing it with tooling. A deburring pass on a dedicated machine costs less per metre than the tool life you would sacrifice trying to punch a burr-free hole, and it is repeatable.