Injection Mold Trial & Acceptance: How to Approve a New Tool (T0, T1, T2, FAI)

Paying for a new injection mold is the easy part. The expensive part is approving it. A tool can pass a "looks good" first trial and still ship parts that fail assembly six months later — because the buyer signed off on a visual sample instead of measured data. This guide covers the full approval sequence: what happens at T0, T1 and T2, what a valid FAI package must contain, and exactly what to check before you release the final payment. It is written for sourcing and engineering teams buying plastic and silicone tools from Chinese mold makers.

What "Mold Approval" Actually Means

A new tool is not approved in one event. It is approved through a staged trial sequence, and each stage has a different purpose. Buyers who treat T1 as "the sign-off" are the same buyers who later argue about who pays for a fourth trial.

The three stages you should have in any tooling contract:

Trial Stages at a Glance

Stage	Purpose	What Is Checked	Typical Timing
T0 — first shot	Prove the tool physically works. Debug only.	Mold opens/closes, slides and lifters travel, ejection, cooling and hot-runner leaks, gates and vents clear, first shot fills.	Often not shipped to the buyer unless requested
T1 — steel trial	First corrected trial on production-intent material and process parameters.	Full dimensional layout (FAI), cosmetic audit, cavity-to-cavity variation, gate vestige, flash, warp, sink.	Samples + report to buyer
T2 — process window	Confirm the process is repeatable, not just capable at one setting.	Steel corrections verified, CTQ repeatability (Cpk), process window study, cycle time locked.	After buyer approves T1 correction plan
Run-off — pre-production	Verify the tool at production rate before release.	300–500 piece run at locked parameters, AQL cosmetic sampling, packaging check, mold PM completed.	Final gate before mass production

Two rules matter more than the labels: conduct every trial on production-intent material (not a random leftover resin), and at production cycle parameters (not a laboratory-optimised setting that will never be used again). A sample that only passes at a setting your molder cannot repeat in production is not a passing sample.

What a Complete T1 Sample Package Must Contain

When samples arrive, they should arrive with documentation — not just parts in a box with photos. Insist on all five items:

• Balloon-checked 2D drawing at the active revision, with every dimensional callout numbered so results can be traced feature by feature.
• Dimensional report covering 100% of drawing dimensions on 5 parts per cavity — measured on calibrated CMM or optical equipment, not calipers and a tape.
• Material certificate for the exact resin lot used in the trial, with the grade and colour confirmed against your specification.
• Process parameter sheet recording injection speed, hold pressure, melt temperature, mould temperature and cycle time.
• Cosmetic inspection report against your defined surface zones, with the lighting condition and inspection distance stated.

If any of these is missing, you cannot tell the difference between a good tool and a lucky shot. Ask for them before you approve corrections — not after.

Dimensional Acceptance: How Precise Is Precise Enough

Not every dimension deserves a CMM report. Splitting your callouts into three groups keeps approval fast and meaningful:

• Critical-to-quality (CTQ) dimensions — sealing faces, mating interfaces, datum features, anything affecting assembly fit. These need full measurement plus capability evidence. A widely used acceptance threshold is Cpk ≥ 1.33 on 30 consecutive shots, which corresponds to a process running roughly four sigma from the nearest tolerance limit. Features below that threshold usually need steel rework, not just a process tweak.
• Functional dimensions — checked at T1 and T2 against the drawing, with deviations recorded even when they pass.
• General tolerances — everything else, judged against your default tolerance class. Do not chase tight numbers here; it only adds cost with no benefit.

Cosmetic Acceptance: Define It Before Steel Is Cut

Most disputes at the approval stage are cosmetic, not dimensional — and nearly all of them trace back to a specification that was verbal instead of written. Before the tool is built, freeze these four items in writing:

• Surface zones (A / B / C) — which faces are visible to the end user, which are semi-visible, and which are hidden. Cosmetic requirements only apply to the zones you define.
• Defect allowances — sink marks, weld lines, flow marks, gate blush, ejector pin marks: on which zone, and to what degree.
• Gate vestige limit — stated numerically, for example "≤ 0.4 mm above the part surface". "As small as possible" is not a specification.
• Reference samples — a signed limit sample or controlled visual standard, photographed and dated, held by both parties.

Also fix the inspection conditions: the lighting, the viewing distance, and the surface classification. The same part can be judged "acceptable" and "rejected" under two different lighting setups. Deciding this in advance removes an entire category of later argument.

The Hidden Risk: Cavity-to-Cavity Variation

This is the single most commonly skipped check, and the most expensive one to discover late. Buyers frequently review one or two random good-looking samples and approve the tool. But a small, consistent deviation on one cavity of a four-cavity mould will still pass a fit check at T1 — and then grow as the tool wears, until a fraction of production parts start failing assembly with no obvious root cause and no baseline data to compare against.

The fix is simple: require dimensional data per cavity at T1, measured on all cavities across a consecutive shot run, and store that as the tool's reference baseline. At every future maintenance cycle, compare current samples against the T1 baseline. Wear is then visible early, while it is still cheap to correct.

Is FAI the Same as PPAP?

No, and mixing them up causes real problems. A First Article Inspection (FAI) is the dimensional and material verification of the tool's output: every drawing feature measured and reported. PPAP is a broader production-part approval submission — dimensional results plus process capability, measurement system analysis, control plan and failure-mode documentation.

Practical advice for buyers sourcing offshore: even if your customer does not formally require a PPAP level, collect the FAI and the process-window data anyway. If you ever move the tool to a second moulder, qualify a new factory, or face a customer audit two years later, that documentation is the difference between a smooth transfer and rebuilding the evidence from scratch — which typically costs several thousand dollars in metrology time alone.

Approval Checklist: Sign Off Only When All Lines Pass

• ☐ Trial run on production-intent resin grade and colour, lot certificate provided
• ☐ Trial run at production cycle time and parameters, not laboratory settings
• ☐ Minimum 15 consecutive shots run; dimensional data recorded per cavity, not per batch
• ☐ 100% of drawing dimensions measured on 5 parts per cavity, ballooned report returned
• ☐ Cpk ≥ 1.33 documented on all CTQ dimensions, 30 consecutive shots
• ☐ Cosmetic audit against written A/B/C zone rules, at the agreed lighting and distance
• ☐ Gate vestige, flash and ejector marks measured against numeric limits
• ☐ Assembly fit and function tested with your mating parts or hardware
• ☐ Open non-conformances categorised as process-only or steel-change, with a written correction plan and lead time
• ☐ Every steel change approved by you before it is cut (steel-safe corrections confirmed where possible)
• ☐ T1 per-cavity baseline recorded and archived for future wear comparison
• ☐ Pre-production run-off completed: 300–500 pcs at locked parameters, AQL cosmetic sampling passed
• ☐ Mould preventive maintenance completed and documented before shipment
• ☐ Golden sample signed by both parties and retained
• ☐ Tool transfer pack prepared: final 2D/3D, process sheet, FAI records, spare-parts list

Print this list. Walk it line by line with your supplier on a call. It takes under an hour and it is the cheapest insurance you will ever buy on a tooling project.

Where Buyers Usually Lose Money at the Approval Stage

• Approving on photos. A good photograph hides a 0.15 mm dimensional deviation and every sink mark that only shows under grazing light.
• Accepting samples with no data pack. You cannot audit a tool you have no measurements for — and you have no argument if the parts later fail.
• Changing the part design after T1 without a formal change record. Design drift between trial and production is the most common cause of "the tool doesn't make the part we approved".
• Releasing the final payment before the run-off. Payment milestones should follow trial gates, not calendar dates.
• Ignoring the process window. A tool that only produces good parts in one narrow setup will produce scrap the moment the resin lot, ambient temperature or operator changes.

A well-run trial sequence also protects your schedule. Because T1, T2 and run-off each consume days, they need to be visible in the tooling plan from day one — our guide to mould lead times from China breaks down how trial stages fit into the overall timeline, and our article on choosing a mould supplier covers the contract terms that make an approval sequence enforceable rather than optional.

Building Trials Into the Tooling Contract

Approval discipline is a purchasing decision as much as an engineering one. When you request quotes, ask every supplier three questions: how many trial stages are included in the price, is a full dimensional report included at T1 or billed separately, and what happens if the tool fails the run-off. Suppliers who answer precisely are the ones whose tools arrive with data. Suppliers who answer vaguely are the ones whose tools arrive with a folder of photographs.

Need Your Trial Sequence Planned Properly?

Precimold designs and builds injection moulds for plastic and silicone parts with a documented trial sequence: T0 debug, T1 steel trial with a full dimensional and cosmetic report, T2 process-window verification, and a pre-production run-off before shipment. We send samples with the data pack, not just the parts. Send us your 3D model, 2D drawing and annual volume, and we will return a tooling quote that states exactly which trials and reports are included. We reply within 12 hours.