Mold Trial T0, T1, T2: Correction Logic and Control

Most tools do not produce a sellable part on the first shot, and a tool that takes six trials to get there has already cost more than the corrections it needed. The difference between a tool that converges in two trials and one that wanders through six is rarely the skill of the toolmaker — it is whether the trial sequence is controlled and whether process tuning is kept separate from steel correction.

Figure 1. T0, T1 and T2 mold trial stages with the output each trial must produce.

This guide sets out what each trial stage is for, what has to come out of it, and the discipline that keeps a tool project on schedule.

What the trial stages mean

Terminology varies between companies, so agree on definitions at project kickoff and put them in the project plan.

• T0 — the first trial after mold assembly is complete: fit, function, fill and basic part formation. T0 answers "does the tool work as a mechanism and can it make a recognizable part?"
• T1 — the first trial after the corrections raised at T0 have been implemented. T1 is expected to produce parts close to the drawing, and it is where the process window is established.
• T2 — the trial after T1 corrections; used to confirm the corrections and to run the capability study, cycle-time optimization and pre-production validation.
• T3 and beyond — normally only for cosmetics, tight tolerances, or features that could not be resolved by analysis. Each additional trial should require a written justification, because it costs press time and delays the program.

The important point: a trial stage is not a schedule slot. It is a correction milestone. If nothing was corrected in the tool, there is no reason for the next stage to have a new number.

What every trial must produce

A trial without data is just a run. Each trial should deliver:

• A trial report — date, press, operator, resin and lot, dryer setting and actual moisture, melt temperature, mold temperature, cycle time, clamp force, injection and pack profiles, and the parameters that were frozen.
• Sample sets — retained samples labeled with cavity number, condition and time, plus the golden sample when one is approved.
• Full dimensional report — every drawing dimension, every cavity, taken on parts that have stabilized. Not the parts that came off the press five minutes ago.
• Appearance and cosmetic assessment — weld lines, flow marks, gate blush, sink, flash, ejector marks, color match.
• Fill and balance study — short-shot series to confirm fill order and cavity-to-cavity balance, plus a cavity pressure or weight study where the tool allows.
• A correction list — every issue, with a proposed action, an owner, and an estimate of whether it is a steel change or a process change.

The rule that keeps tools converging

Separate process parameters from tool corrections, and treat them as two different conversations.

Process parameters live in the machine: velocity profile, switchover point, pack pressure and time, cooling time, mold temperature, clamp force. They are free to change, they are recorded, and they should be set by a logical method — cavity pressure, part weight, or short-shot progression, not by feel.

Tool corrections live in the steel: gate size, runner balance, cooling circuit changes, venting, ejector placement, wall thickness, draft, shut-offs. They cost money and time, so each one needs a reason, an expected effect and a measurement to prove it worked.

The failure mode to avoid: using process to compensate for a steel problem, then forgetting which one was the fix. A dimension that is brought into tolerance by adding pack pressure will drift as soon as the material lot, the ambient temperature, or the machine changes. Correct the steel and keep the process inside a defensible window.

Running T0 properly

• Confirm the tool before you judge the parts. Check shut-off, slides, lifters, ejector return, hot runner heating and thermocouple readings, water flow and leaks, and safety interlocks.
• Dry the resin properly and record actual moisture. Most "bad tool" verdicts at T0 come from wet or cold material.
• Start conservative. Low velocity, modest pack, short cooling, mold protection on. Do not start at the settings from a different tool.
• Short-shot deliberately. A stepped short-shot series shows the fill pattern, the last place to fill, and any hesitation or race-tracking.
• Run long enough to be meaningful. The first twenty shots of a new tool are not representative. Let mold temperature stabilize, then sample.
• Log everything, including what you did not fix. The T0 report is the input to the correction list, so an unrecorded observation is a lost correction.

Running T1 and T2 properly

• Verify the corrections first. Measure the specific dimensions or features that were corrected, in isolation, before looking at the whole part.
• Then build the process window. Establish a window around the nominal setting, not a single point: vary melt temperature, pack pressure and cooling time within sensible limits and record the effect on critical dimensions. A process with no window is a process that will not survive production.
• Stabilize before dimensioning. Semi-crystalline and hygroscopic parts, especially PA and POM, need to reach equilibrium. Define conditioning time and conditions in the trial plan and use them consistently.
• Run a capability study. Take a statistically meaningful sample from every cavity, at the same stabilized condition, and calculate Cpk for the critical dimensions. Cpk should be calculated on parts made inside the process window, not on hand-picked good parts.
• Qualify the cycle time. Report the cycle time achieved, and the cycle time the tool can realistically hold in production. Do not report the best one-off cycle.
• Freeze a validated setup. Record the approved parameters in a setup sheet tied to the tool number, and keep a golden sample with a written description of what makes it acceptable.

Documentation and accountability

• Correction log. One line per issue: identified at which trial, root cause, action, steel or process, cost owner, and status. This is the single document that decides whether the tool is on schedule.
• Photo evidence. Before-and-after photographs of corrected areas and of defects are worth more than paragraphs of description.
• Approval gate. No trial is scheduled until the previous correction list has been signed off and the parts are available for the trial review.
• Link trials to the schedule. A trial stage that slips a week usually slips the customer's tool approval and the production start.

Figure 2. Quick reference: the key checks and reference values from this guide.

Checklist

• Trial stage definitions agreed in writing at project kickoff
• Trial plan issued: objectives, machines, material, sample size, conditioning
• Resin dried to specification, actual moisture recorded
• Tool mechanically verified before parts are judged
• Stepped short-shot study completed
• Full dimensional layout on stabilized parts, all cavities
• Fill balance and cavity-to-cavity weight or pressure data captured
• Process frozen inside a validated window, not at a single point
• Capability (Cpk) study on critical dimensions from all cavities
• Golden sample approved and stored with its condition of approval
• Every issue logged with root cause, action and owner
• Each correction classified as steel or process, with expected effect
• Setup sheet issued for the tool; parameters locked

A mold trial is a controlled experiment, and the correction list is its result. Run each stage with a stated objective, produce the data that stage is supposed to produce, correct the steel when the steel is wrong, and keep the process inside a window you can defend in production. Tools that follow this discipline reach T1 or T2 with parts that are genuinely close — and stay there.