Think Plastics

Notes on injection moulding: part design, tooling and cost

Front page / How do I plan a pilot run before full production of a moulded part?

How do I plan a pilot run before full production of a moulded part?

A pilot run, called a run at rate in automotive work, is a production trial under real series conditions. It proves that the approved mould, press, process, gauges and people can make the required volume at the required quality. Plan it after the mould and process are approved, run it for a set time at the planned cycle, log every stop and the good parts made each hour, and compare the demonstrated output with your demand. QM Bench summarises the AIAG PPAP expectation as a significant production run lasting 1 to 8 hours and producing a minimum of 300 consecutive parts, except where the customer sets its own figure.

What makes the run valid

QM Bench lists the series conditions as production moulds and tools, gauges that have been released, operators who are trained, the intended cycle time, and the control plan actually being followed. If a critical condition is missed, the run is not valid, however good the numbers look. For a moulded part, add:

  • the production resin grade and colour, from a normal lot, dried as specified
  • the production press, or one named in the approval, with the approved process sheet
  • normal regrind practice, if regrind is allowed
  • the real packing method, so handling damage shows up

What to log

RecordWhy
good parts per hour, by cavityshows whether the rate held through the run and flags a weak cavity
each stop: start, end and reasonQM Bench suggests categories such as breakdown, changeover, material, quality and waiting
scrap by defect typeshort shots, flash and sink each point to different fixes
key dimensions over timechecks the process stays centred as the mould reaches steady temperature
process readingscycle time, cushion and mould temperature, for the process record

Worked example (our numbers)

A four-cavity mould on a 30-second cycle can make 4 × 3,600 ÷ 30 = 480 parts an hour, or 3,840 in an 8-hour pilot run. Say the run produces 3,300 good parts after stops and scrap: that is about 86 percent of the theoretical output, or about 412 good parts an hour. If demand is 15,000 parts a week and the press is available for 40 hours a week, the requirement is 375 parts an hour, so the demonstrated rate passes with about 10 percent margin. If one cavity caused most of the scrap, fix it before the next order, because the margin is thin.

QM Bench gives a similar example: 470 good parts in 8 hours, or 58.8 an hour, gives 4,406 parts over a 75-hour week. That is 92 percent of a 4,800-part requirement, so the run is not approved.

After the run

  1. Write a one-page result: approved, approved with actions, or not approved, with the bottleneck and the biggest losses.
  2. Use pilot parts for assembly trials, packaging and drop tests, and early customer or field feedback.
  3. Agree a period of extra inspection for early production. Some buyers' supplier manuals call this a safe launch plan.
  4. Repeat the run if the result failed, keeping earlier results in the record.

Summary of public guidance on run-at-rate trials. Your customer's requirements decide run length, part count and the volume used for judging.

Questions buyers ask

Is a pilot run the same as a T1 trial?

No. A T1 trial checks the new mould; a pilot run checks the full production process at rate once the mould is approved.

How many parts should a pilot run make?

Enough to cover your needs: at least the customer's minimum, enough parts for assembly and packaging tests, and long enough to see the process at steady state.

Can pilot parts be sold?

Often, if they pass inspection and were made under approved conditions. Agree this in advance.

What if the run fails?

QM Bench advises finding the bottleneck and the biggest losses in the downtime log, fixing them and repeating the run, keeping earlier runs on record so progress is visible.

Sources