A tool that fails its first article is a process event, not a complaint. What separates a supplier who resolves it from one who negotiates it is the paperwork that arrives in the first week.
This page sets out what a buyer should receive after a first-article failure, how the eight disciplines of 8D translate into cutting tool terms, the three nonconformances that account for most cases, and a note template that starts the process properly.
After a first-article failure the buyer should receive three things: a written acknowledgement that the result is disputed or accepted, a containment statement covering tools in transit and in stock, and a preliminary root cause direction with the evidence behind it.
The 8D structure works for tools if D2 is a measured description rather than an adjective, and if D4 separates tool design, tool manufacture and tool use, because those three have different owners and different corrective actions.
Wear criteria, which decide whether a tool failed or simply reached the end of its life, are covered in drill and tool wear patterns and in the regrind and requalification guide. This page covers the process after the failure is established.
What you should receive, and when
The first deliverable is a written acknowledgement. It should state whether the supplier accepts the result, disputes it, or needs more data, and it should arrive quickly enough that the buyer can decide what to do with the parts already made.
The second is containment. Tools in transit, tools in the buyer's stock, and tools at any subcontractor have to be identified and held, or released on a defined basis. Containment is not a root cause activity; it exists to stop the same result reaching the next part.
The third is a preliminary direction with evidence. Not a final cause, but a statement such as "the bore is tapered and the taper follows the pre-hole, so the pre-hole operation is the first suspect", together with the measurements that support it.
Timing is usually expressed in industry terms rather than as a promise: acknowledgement within a day or two, containment within the first week, and a full 8D within a period agreed in the purchase order. Where a supplier states a shorter period, the useful check is whether the previous cases met it.
The most common source of dispute at this stage is that both sides measured and got different numbers. That is why the sampling and retest arrangement has to be fixed before the result is argued.
Our commitment is narrower and firmer. A nonconformance notice sent by email is answered the same working day, and the case is supported by phone, with on-site attendance where the part and the machine justify it.
Support does not stop at the reply. The engineer who answers stays with the case: diagnosis runs by phone and by photograph, and where the part or the machine cannot be assessed that way, an engineer attends.
The aim is that a problem is settled as promptly as the case allows, rather than parked in a queue with a day count attached.
The containment, root-cause and closure intervals for a specific order are set in the purchase agreement.
They depend on how the failure was found and how far the parts have travelled.
8D applied to a cutting tool
The eight disciplines are the same as in any manufacturing environment, but each one has a tool-specific content that is easy to leave out.
| Step | What it means for a cutting tool |
|---|---|
| D1 Team | A named owner on the supplier side and a named counterpart on the buyer side, with the authority to hold a batch |
| D2 Problem description | Measured values, not adjectives: bore size against the tolerance band, roundness, taper, edge condition, and the number of pieces affected |
| D3 Containment | Hold stock and transit, add a 100 percent check on the next delivery, quarantine the affected batch at both ends |
| D4 Root cause | Separate tool design, tool manufacture and tool use, with a discriminator for each rather than a single narrative |
| D5 Corrective action chosen | State what changes and why that change addresses the cause, not just what will be inspected more |
| D6 Implementation | The date the change took effect, the first batch it applied to, and how that batch was identified |
| D7 Prevention | Which control changed: the drawing, the process sheet, the inspection plan or the supplier's incoming check, and where the record now lives |
| D8 Closure | The buyer confirms the result on a defined quantity, and the case closes against that evidence rather than against a date |
Two steps do the most work. D4 is where a supplier can hide a use problem inside a design narrative, and D7 is where a fix becomes permanent or evaporates when the next order starts.
D2 deserves its own caution. "Bore oversize" is not a problem description. "Bore 0.012 mm over the upper limit on 18 of 20 pieces, roundness within 0.004 mm, taper 0.006 mm over the bore length" is a description that points at a cause.
Three typical nonconformances
Most first-article failures fall into three groups, and each has a discriminator that decides whether the tool or the process owns the problem.
| Nonconformance | First discriminator | Root cause direction | Evidence required |
|---|---|---|---|
| Size out of tolerance | Run a second tool from the same batch on the same pre-hole | Tool if the second tool repeats the size; pre-hole or machine if it does not | Two tools, one pre-hole, one measurement method, recorded by both sides |
| Edge chipping or breakage | Inspect the fracture surface and replay the cutting parameters | Material or interruption if the chip origin is at an inclusion or a cross hole; geometry if it starts at the edge | Photographs of the edge at magnification, the parameter record, the workpiece condition |
| Tool life short of expectation | Compare the wear land form with the expected wear mode | Abrasion if the wear land is uniform; adhesion or built-up edge if the wear is localised and smeared | Wear land measurement at intervals, coolant condition, material certificate |
The second tool test is the cheapest experiment in the whole process and it is the one most often skipped. It separates a batch problem from a process problem in one cut, and it produces evidence both sides can see.
For chipping, the position of the chip origin matters more than its size. A chip that starts at the cutting edge points to geometry or edge preparation; a chip that starts behind the edge points to a mechanical event such as a cross hole or a hard inclusion.
Tool life disputes are usually disputes about the criterion. A tool is not "worn out" in the abstract; it is worn out against a stated wear land, a surface finish limit or a size drift. Without the criterion, both sides are describing different end points.
The nonconformance note template
A single page with these rows gives the supplier everything needed to start, and prevents the two most common delays: a missing parameter record and a missing measurement comparison.
Tool identification. Part number, revision, serial or batch, and the delivery note it arrived on.
First-article data. The measured values against the drawing requirement, with the instrument used and its calibration status.
Failure phenomenon. What was observed, where on the tool or the bore, and how many pieces.
Cutting parameters. Speed, feed, depth, coolant type and pressure, and the machine.
Pre-hole condition. How the bore was made before the tool ran, with its size and roundness.
Both parties' measurements. The buyer's values, and the supplier's values if a returned tool was measured.
Requested disposition. Replacement, rework, credit, or technical support on site.
Requested response date. What the buyer needs, and by when, to keep the line running.
Sending this note before returning the tool is faster than sending the tool and waiting. The data usually identifies the cause, and the tool then confirms it. Where the case needs an escalation path, use the contact page rather than the sales channel.
Who measures what
Two measurements of the same bore that disagree are a process problem in their own right, and they are the most common reason a case stalls.
The agreement should state three things: who measures the failing feature, with which instrument, and against which reference. Where the two parties use different instruments, the difference between them has to be established before the result is argued, not after.
A practical arrangement is that the buyer measures and the supplier reviews the record, with the supplier free to re-measure a returned tool on its own equipment. The returned tool measurement answers a different question, which is whether the tool left the factory within its own tolerance.
Where the instrument question cannot be settled, a third-party measurement is cheaper than an unresolved dispute. The measurement-system reference listed below explains why the instrument and the method belong in the agreement rather than in the argument.
The capability data that makes a first-article judgement defensible, including the sample size and the gauge record, is described in capability data to request from a tool supplier.
Prevention on both sides
Prevention is not one action. The supplier controls what leaves the factory, and the buyer controls what the tool meets when it arrives.
Supplier side. A tighter first-article check on new or changed tools, a retained sample from each batch so that a later dispute has a physical reference, and a change record that ties a drawing revision to a manufacturing date.
Buyer side. A recorded parameter set for each tool, a controlled pre-hole with its own measurements, and a wear criterion written into the process sheet rather than held in an operator's experience.
Both sides benefit from a shared failure vocabulary. If the buyer reports "poor finish" and the supplier investigates geometry, the case takes a week longer than it needed to.
The audit questions that test whether a supplier's prevention system is real, rather than documented, are in the supplier audit checklist for cutting tools. The quality evidence a buyer can ask for before the first order is on the manufacturing quality evidence page.
If the failure is part of a channel dispute rather than a single order, the model that applies is described in distributor territory and exclusivity models.
A nonconformance is one outcome of the trial gate. How that gate is set up, and what the other four gates must produce, is in how nonconformance fits the procurement sequence.
For the wear call itself on a two-stage tool, the criteria are in wear criteria for combination drill-reamers.
What published sources say about this
These pages support the measurement logic, the capability definitions and the tool behaviour referenced above. Each is cited for what it states, under its own conditions.
- Dimensioning and Tolerancing (ASME Y14.5): the definitions that make a nonconformance a specific failure against a specific control.
- What is a measurement system? (NIST/SEMATECH e-Handbook): measurement variation as a component of observed variation, the reason two parties can measure the same bore differently.
- What is Process Capability? (NIST/SEMATECH e-Handbook): process spread against specification, the basis for a first-article judgement on a quantity rather than a piece.
- What are control charts? (NIST/SEMATECH e-Handbook): the time-ordered view that distinguishes a step change from a drift, which is the difference between a batch defect and a wear trend.
- Bias and variability (NIST/SEMATECH e-Handbook): separating a repeatable offset from a random spread, which is how a size offset is told apart from a scattered result.
- Reamers, machine reamers (toolmaker technical overview): reamer construction and guidance, the reference for what a tool can and cannot hold.
- Reaming and fine boring (toolmaker competence page): the finishing families, including the adjustable tool that makes a size correction possible without a redesign.
Outbound references are listed because the reasoning depends on them, not as endorsements. Response intervals quoted on this page are industry practice. The commitment we publish is the same-working-day reply; containment, root-cause and closure intervals are set in the purchase agreement.
Frequently asked questions
Should the tool be returned before the data is sent?
No. Send the measurements, the parameters and the photographs first; the data usually identifies the cause.
Return the tool when the supplier asks for it or when the dispute requires a physical check.
How long should an 8D take?
Long enough to test a root cause, which is usually weeks rather than days for a tooling case.
What matters is that containment is immediate and that the closure is against evidence from a defined quantity, not against a calendar date. Our own commitment is the response, not a calendar: same working day by email, then phone and on-site support until the problem is settled.
What if the supplier disagrees with the failure?
Then the measurement method becomes the subject. Agree who measures, with what, and against which reference before arguing the result.
A third-party measurement is often cheaper than the meeting that would otherwise follow.
Is a second tool from the same batch a fair test?
It is the fairest quick test available, because it holds the pre-hole, the machine and the operator constant and changes only the tool.
If the second tool repeats the result, the tool batch is the suspect; if it does not, the process is.
What belongs in a containment statement?
Which batches are held, where they are, what check is applied to the next delivery, and who releases them.
Containment should be specific enough that a warehouse can act on it without asking a question.
Updated 14 November 2026: first publication. Scope: what a buyer should receive after a first-article failure, the eight disciplines applied to cutting tools, three typical nonconformances with their discriminators, the note template, the measurement arrangement and prevention on both sides.
Response times are industry practice, set by agreement.