10 Engineering Questions to Confirm Before Starting a Tooling Project

Nine times out of ten, a failed tooling project is traceable to information that was never exchanged before the purchase order.

Here is how a tooling project actually fails. The die set arrives. It runs 40,000 pieces instead of the 800,000 the schedule assumed. The line stops. Someone pulls the die, finds a crack running out of a hex corner, and the emails start: the supplier blames the wire, the wire supplier blames the coating, and the machine crew blames the die.

Nine times out of ten, the root cause turns out to be information that was never exchanged before the purchase order: nobody wrote down that the wire had changed from 1022 to 10B21 boron steel. Nobody confirmed whether the machine was the 4-station or the 5-station former. Nobody specified the carbide grade — the quotation just said "carbide."

The fix is not more emails. It is ten engineering questions, answered in writing, with numbers, before the first quotation. This guide goes through them one by one, with the actual values a fastener manufacturer should expect to see in the answers.

What this guide covers

• The 10 engineering questions, one by one, with real numbers
• Summary table: what each question decides, and what skipping it costs
• What a tooling supplier should do with your answers
• Red flags in the quotation phase
• Frequently asked questions

Question 1: Which Machine Will Run This Part — Exactly?

"A 7-station bolt former" is not an answer. The tooling needs: maker and model (a National bolt former, a Sacma SRB, a Nakashimada nut former, or a European multi-station parts former each have different holder standards), station count, punch holder and die holder bore sizes and fit classes (typically H6/H7), shut height, and the transfer system — because transfer fingers, not just dies, are part of a complete tooling set.

Tooling designed for a 5-station progression cannot be "adapted" to a 4-station machine. The forming sequence has to be redistributed, which changes every punch and die in the set. Confirm the machine first, with the holder drawing attached.

Question 2: What Wire Will Feed the Line?

Wire determines die pressure, galling risk, and carbide grade selection. Confirm: grade (1018/1022 low carbon, 10B21 boron, 4140 alloy, 304/316 stainless are the common fastener families), tensile strength of the incoming wire (annealed low-carbon runs roughly 400–500 MPa; alloy and work-hardening grades come in far higher and hit the dies harder), surface coating (zinc phosphate + soap lubrication, or bare), and diameter tolerance of the drawn wire.

A die set engineered for 1022 will underperform visibly on 10B21: higher forming load, faster corner wear, earlier crack initiation. If the wire may change mid-project, say so before the design is frozen — grade selection for carbide forming dies and for tool steel dies follows the material, not the other way around.

Question 3: What Volume, and What Tool Life, Are You Targeting?

Give the supplier two numbers, not one: annual volume, and expected pieces between regrinds. As a working reference in carbon and alloy steel wire:

These are industry-typical ranges, not promises — wire grade, coating quality, part geometry and machine condition move them up or down. But the ratio is what matters for the decision: on a high-volume line, premium carbide pays for itself by removing mid-run die changes from the schedule; on a 50,000-piece job, it never gets the chance.

Question 4: What Does the Part Geometry Actually Demand?

Review the drawing with the supplier's engineers, station by station, against the classic cold heading limits:

• Upset ratio: unsupported length-to-diameter roughly ≤ 2.5 per blow. A head that needs more than that needs an extra preform station — not more force on one blow.
• Extrusion reduction: forward extrusion is usually limited to roughly 25–35% area reduction per station. Beyond that, the tool pressure spikes and punch life collapses.
• Underhead fillets: small radii (0.2–0.5 mm class) concentrate stress in the die; they need to be a deliberate design decision, not a CAD default.

A part that looks "simple" but carries a severe extrusion ratio will punish a die designed like its neighbors. This is the question where an experienced supplier earns its keep — by pushing back on the drawing before anyone prices anything.

Question 5: Which Tolerances Are Functional, and Which Are Cosmetic?

This question saves more money than any other on the list. Fastener standards already define what matters: width across flats and thread concentricity for nuts (ISO 4032, ISO 898-2), head-to-shank eccentricity and head height for bolts (product grade per ISO 4759-1). Everything else on the drawing is negotiable.

Tightening every dimension to the tightest functional tolerance multiplies tooling cost and shortens regrind cycles for zero benefit. Mark the drawing: functional dimensions in one color, cosmetic in another. The supplier can then allocate precision where it matters — cavity tolerance held to ±0.01 mm and bore-to-OD concentricity to ≤0.005 mm on the dimensions that count, standard commercial tolerance everywhere else.

Question 6: How Many Stations, and What Does Each One Do?

Ask the supplier to walk you through the proposed sequence in writing: cut-off → preform → extrude → head → trim, with what each blow does to the material and why the work is split that way. The rule of thumb from Question 4 applies here: if one station is carrying an upset ratio far above its neighbors, that station is where the tool will fail first — usually the second or third die insert, usually by longitudinal cracking under the highest load.

Question 7: Solid, Inserted, Assembled, or Segmented Structure?

Die structure is an economic decision, not a supplier default. In practice:

• Solid tool steel dies — prototypes, short runs, simple round geometry. Cheapest first article, worst cost-per-piece at volume.
• Carbide-inserted dies — carbide where forming happens, steel case everywhere else. The standard workhorse for round parts.
• Assembled dies — multiple functions or materials in one unit (see the assembled forming die line). Wins on complex multi-feature parts.
• Segmented dies — the cavity itself split into segments, so worn pieces are replaced or re-ground individually. The reference structure for hex nut tooling and other multi-corner cavities (see the segmented forming die line).

Question 8: Which Tool Materials — Named Grades, Not "Carbide"?

"Carbide" is not a material; it is a family. Ask the supplier to specify, per component:

• Dies and inserts: WC-Co grade with cobalt content — 15–20% cobalt grades (YG15, YG20C class) for high cyclic load and cracking resistance, finer-grain lower-cobalt grades where edge definition and wear life dominate.
• Punch points and pins: M2/1.3343 high-speed steel, PM high-speed steel, or carbide — moving up this ladder trades toughness for wear life.
• Cases and holders: alloy steel (4140/H13 class) with the shrink-fit interference calculated from the required assembly pre-stress.
• Steel dies: D2/1.2379 hardened to roughly 58–62 HRC, with vacuum heat treatment so decarburization does not eat the working surface.

A supplier who names grades and reasons per component — for the die, for the punch and sleeve, for every insert — is engineering your tooling. One who writes "imported carbide" is brokering it.

Question 9: What Spares, at What Ratio, Against What Lead Time?

The spare tooling policy is decided at project kickoff, not at the first breakdown. Agree on: which components are wear items (punch points, kick-out pins, trimming die edges, first-station inserts die first — in that order, usually), the spare ratio (10–20% of the wear items is a common starting policy for imported tooling), and replenishment lead time — for tooling shipped from China to the US or Europe, plan on weeks including logistics, not days.

The arithmetic is blunt: at 300 pieces per minute, one hour of line downtime is 18,000 pieces. A spare set that costs a few hundred dollars is cheaper than the first hour it prevents.

Question 10: What Are the Acceptance Criteria — In Writing?

Define before delivery, in the quotation itself:

• Inspection report with actual measured values — not "OK" stamps. Bore size and fit class, cavity dimensions against the drawing, concentricity figure.
• Surface finish: cavity forming sections Ra ≤ 0.4 µm; polished critical surfaces Ra 0.1–0.2 µm.
• Certificates: material certificates and hardness for steel components, carbide grade documentation for inserts.
• Regrind allowance and expected regrind count per tool — the number that turns a die purchase into a die budget.

Summary: What Each Question Decides

What a Supplier Should Do With Your Answers

The answers to these ten questions are the input to tooling engineering, and you should see them used. At Hongli, the engineering workflow runs from customer-assigned analysis and modeling, through simulation and detail drawings, to the complete tooling hardware, startup at the customer's site, and optimization until the set runs in real production — the same turnkey scope these questions feed into.

The manufacturing side has to hold up its half. Segments and inserts are cut on wire EDM machines running 0.1–0.3 mm wire with 0.05 µm positioning scales; complex pockets and profiles run on a five-axis machining center holding ±0.002 mm positional accuracy (calibrated per ISO 230-2); steel components are vacuum heat-treated with ±1 °C temperature control to keep working surfaces free of decarburization. These are the machine-level facts behind every answer to Questions 5 through 10 — feel free to ask any supplier for equivalents.

Red Flags in the Quotation Phase

• A quotation issued without asking for the machine data or the drawing's tolerance block — it means the tooling will be designed against assumptions.
• "Carbide" without a grade, or "imported material" without a certificate. That is a price, not an engineering decision.
• No regrind allowance or regrind count anywhere in the document.
• A price 30% below every other offer, with no engineering explanation of where the difference comes from.
• Refusal to walk through the forming sequence with your engineers before the order. A supplier who cannot defend the design will not defend it after the first failure either.

The Bottom Line

Every one of these ten questions costs one email before the RFQ. Every one of them costs a production run, an air shipment, or a customer delivery after the first tool failure. Work through them with your supplier in writing — starting with the engineering review, not the price list.