SKH High-Speed Steel or Tungsten Carbide: Choosing Your Drill Point Die Material

Two die sets, same size code, same screw, same coil of wire. One runs a long, quiet campaign and comes off the pointer with its flute lands still crisp. The other loses a corner before the first shift is out, and nobody notices until a customer complains that the screws skate on 1.2 mm deck. The choice between high speed steel and tungsten carbide drill point dies gets argued as a hardness contest, and argued that way it has a clean winner and a wrong answer. The real variable is usually sitting three feet away from the die: the pointing machine, and how consistently it presents a blank.

This piece is for buyers specifying die sets rather than building them — a screw maker replacing tooling, a trader sourcing dies for a customer's line, an engineer writing the material line on a purchase order and not sure whether the carbide premium is money or insurance.

What the die is actually being asked to survive

A drill point die is cold-forming tooling, not a cutting tool. A matched pair closes on the end of a headed, unthreaded blank and squeezes the drilling point into existence — flutes, chisel edge, point diameter, the whole geometry — in one blow. Then it opens, a fresh blank arrives, and it does it again, thousands of times an hour.

That puts three separate loads on the same piece of steel or carbide:

• A compressive peak at closure. The die is momentarily carrying the full forming load of the blank material.
• Sliding abrasion. Displaced material flows across the land and the flute-forming surfaces every cycle. This is what quietly eats geometry.
• Impact and off-axis shock. If the blank arrives off-centre, long, short, or with a burred cut-off face, the first contact is a hammer blow on one edge instead of a squeeze on the whole face.

Here is the detail that separates people who have run a pointer from people who have only bought dies: point tooling rarely announces that it is finished. It does not snap in half and stop the line. The point diameter creeps up, the flute land rounds off, and screws that used to punch through 3 mm structural steel in a few seconds start taking noticeably longer, then start walking on the surface before they bite. The line does not find out. The customer finds out, on a roof, in the wind.

ZLD Precision Mold, a Dongguan die maker whose whole range is drill point dies, states the choice plainly on its own spec system: dies are made in SKH high-speed steel (M2 / M9 / M51) or tungsten carbide, selected by screw material, output volume and die-life target. That is the correct three-variable framing. It is also incomplete, and the missing fourth variable is the one this article is about.

The trade-off, stated without the marketing

High-speed steel: tougher, and forgiving of a bad day

High-speed steel is the more ductile of the two. It absorbs shock. When you overload it, it tends to deform, upset or wear rather than fracture — and deformation is a failure mode that gives you warning in the part before it gives you a catastrophe in the tool. A worn HSS die makes progressively worse screws; you can catch it with a gauge on the finished screw and pull the set at a planned point.

It also grinds more easily, which matters for anything custom: a special point profile, an odd flute length, a size between two standard codes. And a set that costs less to replace changes behaviour on the floor — a supervisor who knows spares are cheap changes tooling on schedule instead of nursing a marginal die through one more shift.

Tungsten carbide: harder, and holds the geometry that actually matters

Carbide's advantage is not that it "lasts longer" in the abstract. It is that it resists the specific failure that ruins drill points: abrasive rounding of the land and growth of the point diameter. Carbide holds the formed geometry through campaigns where high-speed steel would have drifted out of tolerance long before it looked worn.

The cost is fracture toughness. Carbide has very little. It does not bend and then complain; it is fine, fine, fine, and then a corner is gone. Everything that puts an off-axis load into the die — a blank that is not centred, play in the ram or the gibs, a die holder that has worn its seat, a double feed, a mistimed closure — becomes a chipping risk instead of a wear event.

And a chipped carbide die is more expensive than the die. The real bill is the sorting: the screws made between the chip and the moment somebody noticed. Depending on your inspection interval that can be a few hundred pieces or an entire tote, and if the batch already shipped it is a customer conversation, not a scrap ticket.

Your pointing machine decides more than your screw material does

This is the part most material comparisons skip. Carbide converts machine condition into money, in both directions. On a tight, well-maintained pointer with repeatable blank presentation, carbide is close to a free upgrade over the life of a program. On a machine with play in it, carbide shortens the mean time between disasters instead of lengthening the run.

Before you pay the premium, walk the machine with this list:

1. Blank centring. Are feed fingers, guides and the transfer worn? Does a blank arrive on axis every time, or on axis most of the time?
2. Cut-off quality from the header. Angled or burred blank ends load one side of the die first. Check a handful of blanks, not the spec sheet.
3. Blank length variation. Consistent overfill is a forming variable; inconsistent overfill is a shock variable.
4. Ram and slide play. Gib wear and bearing clearance show up as a die that closes slightly differently on every stroke.
5. Die holder seat and shims. A hammered seat or an improvised shim stack means the die is being asked to hold itself square. Carbide will not.
6. Closure timing and parallelism. If the two halves do not meet flat, the edges take the load.
7. Lubrication delivery. Intermittent lube on a fast line is an abrasion and heat problem before it is a die problem.

If more than two of those come back "mostly", specify high-speed steel and spend the difference on the machine. That is not a compromise; it is the cheaper repair. Buyers who want the wider version of this argument — where tooling money goes and what it buys — will recognise the pattern from how tooling and mold costs actually break down in Chinese manufacturing.

A decision matrix you can take to a quote

Read it as a tally, not a formula. Four or more in one column is a decision; a split down the middle usually means you should buy one set of each and settle it with data.

What is not published — and the questions that close the gap

ZLD publishes the material options and the selection logic. It does not publish the numbers underneath them, and neither do most die makers in this niche. That is not a red flag by itself, but it does mean the specifics belong in your enquiry rather than in your assumptions:

• Carbide grade. "Tungsten carbide" is a family, not a specification. Ask for the ISO K-classification, the cobalt binder percentage and the grain size, and ask for the grade to appear on the order acknowledgement so the second set matches the first.
• Heat-treated hardness of the steel dies. Ask for the target HRC range and whether hardness is recorded per heat-treat lot. A hardness range on paper is also the thing you point at if set two behaves differently from set one.
• The grade naming itself. SKH is the JIS prefix for high-speed steels; M2, M9 and M51 are AISI/ASTM series numbers. A supplier listing "SKH M2 / M9 / M51" may be using a common shop shorthand, but do not assume a one-to-one SKH-to-AISI equivalence on your own authority. Ask which designation governs, and ask for the material certificate that goes with it. This is a two-minute email that prevents a six-week argument.
• Price ratio. Ask for both materials quoted at the same size code, at the quantity you actually buy. Ratios quoted at one-set volumes rarely survive to production volumes, and the gap between two quotes for the same spec is itself informative — see why China quotes for the same spec differ.
• Salvage. Ask whether a worn set can be re-ground back into tolerance, and at what fraction of new cost. This changes total cost of ownership more than the headline price does.

No measured die-life comparison is published for this factory, and the company has said it cannot supply trial-run test figures — so treat any life number you are quoted anywhere in this category as a claim you generate yourself, on your machine, or not at all.

A qualification run you can specify in one paragraph

The good news about drill point dies is that the experiment is cheap. ZLD states that it accepts trial orders from a single set, with standard sampling in 3–4 days and custom sampling in 5–7 days, and production in about 10 working days. That is enough to run a real comparison instead of a debate:

1. Order one set in each material at the same size code — the same line on the L1–L7 die chart, same die style, same point geometry.
2. Log the machine state before you start — the seven checks above, written down, dated.
3. Run them on the same machine, same operator, same blank lot, alternating rather than sequentially so tooling wear and coil variation do not confound each other.
4. Every fixed interval, pull ten screws and measure point diameter and drill diameter with a gauge, not by eye.
5. Time drilling into a fixed-thickness coupon — pick a thickness at the top of your range, such as the 3–6 mm structural steel that heavier self-drilling programs live in — and record it as a number each interval.
6. Define end of life before you start: a point diameter limit, a drilling-time limit, or a screw that fails your own pull-out check.
7. Keep the failed sets. A chipped carbide corner and a rounded steel land tell you different things about your machine.

Common questions

Is carbide always the better die if I can afford it?

No — and that framing is what causes the chipping complaints. Carbide rewards a machine that presents blanks the same way every stroke. On a pointer with play in the slide or worn feed fingers, high-speed steel will out-run carbide in practice, because it survives the shocks that end a carbide die early.

Does a PVD coating change the material decision?

It changes the wear picture, not the toughness picture. A coating adds surface hardness to whatever is underneath it; it does not make high-speed steel harder in bulk, and it does not stop carbide from chipping when a blank arrives crooked. Choose the substrate for your machine's condition first, then decide about coating.

How do I know the second batch matches the first?

Write the grade, the hardness range and the key dimensions into the order, and ask for them back on the order acknowledgement. Repeatability between batches is a documentation problem more often than a metallurgy problem — the same discipline that applies across sourcing industrial fasteners and their tooling from China.

What should I measure to prove a die is worn out?

The screw, not the die. Point diameter growth and drilling time into a fixed coupon are the two numbers that predict field behaviour. A visual check of the die tells you what happened; a gauge on the screw tells you what your customer is about to receive.

What to ask the supplier next

Before you place a die order, send these questions and keep the reply:

• Which material do you recommend for my screw size, material and monthly volume — and what is the reasoning?
• What is the carbide grade (ISO classification, cobalt content, grain size), and will it be stated on the order?
• What hardness range are the high-speed steel dies heat treated to, and is it recorded per lot?
• Is your "SKH M2 / M9 / M51" naming JIS, AISI or a house convention, and can you supply the material certificate?
• Quote both materials at the same size code at my real quantity, and tell me the re-grind or salvage option and its cost.
• Will you trial-run the die before shipment and send me what you measured?
• If I send a worn die or photos of my current screw, can you match it?

That last pair is worth pushing on. ZLD Precision Mold, a Dongguan maker working only in this product, publishes a size system where each die code maps to a screw size across IFI #4–#14 and DIN ST2.9–ST6.3 with drill diameter, point and wire diameters, flute length and die style; it accepts an enquiry as a drawing, a screw sample or photographs of your old die; and it states that key dimensions, point geometry and surface condition are checked before packing, with in-house test equipment used to trial-run dies before shipment. Those are the four things worth demanding from any die maker — a size system you can read, an entry point that takes your old die as the specification, a dimensional check before packing, and a trial run you can ask to see. The ZLD Precision Mold factory profile lists its stated terms and what was and was not verified.