Scientific Molding & Process Validation — IQ/OQ/PQ for Medical and Automotive

An injection molding process that makes good parts when the setup technician is standing at the press is not validated. It is dependent. Dependent on that technician’s judgment about cushion position. On the barrel temperature profile that was adjusted last Tuesday. On the holding pressure that was increased because the previous lot of material was slightly more viscous.

Scientific molding replaces operator judgment with process data. Instead of “the parts look good,” it produces “at 345 bar peak cavity pressure with a 220°C melt temperature and 0.5 second gate freeze time, the Cpk on the critical dimension is 1.72 across 3 production lots.” The second statement survives an FDA audit. The first does not.

This guide covers how scientific molding works, how process validation is structured under the IQ/OQ/PQ framework, and what a buyer in a regulated industry should expect from a validated injection molding process.

1. What Scientific Molding Means

Scientific molding — also called decoupled molding or data-driven molding — is a methodology for developing and controlling the injection molding process based on measurable, repeatable process parameters rather than operator experience.

The core principle: the plastic determines the process, not the machine settings. Instead of setting a fixed injection speed and pressure and accepting whatever cavity pressure results, scientific molding measures what happens inside the cavity and adjusts the machine to produce consistent cavity conditions.

Traditional Molding vs Scientific Molding

The Three Stages of Decoupled Molding

Stage 1 — Fill (velocity controlled). The cavity is filled to 95–98% under velocity control. The transfer from fill to pack occurs at a specific screw position — not by injection time or pressure. The fill time should be consistent shot-to-shot (variation under 0.05 seconds). Consistent fill time means consistent material viscosity entering the cavity.

Stage 2 — Pack and hold (pressure controlled). After transfer, the cavity is packed under pressure control. The pack pressure is set based on cavity pressure data — enough to pack the part without overpacking (which causes flash, sticking, and molded-in stress). The hold time extends until the gate freezes, determined by the cavity pressure decay curve flattening out. Holding past gate freeze wastes cycle time and does nothing. Holding short of gate freeze allows material to flow back out of the cavity, causing sink and dimensional variation.

Stage 3 — Cooling. The screw recovers (plasticates the next shot) while the part cools in the mold. Cooling time is set to achieve a consistent ejection temperature — verified by part surface temperature measurement at ejection, not by a fixed timer.

The three stages are decoupled because the control variable changes at each stage — velocity during fill, pressure during pack, time during cooling — and the transitions are defined by cavity conditions, not machine timers.

2. In-Cavity Sensing

The instrument that makes scientific molding possible is the cavity pressure sensor — a piezoelectric or strain-gauge sensor mounted flush with the cavity surface, connected to a data acquisition system that records pressure at up to 500 samples per second throughout the injection cycle.

What the cavity pressure curve tells you:

• Fill time: the time from injection start to cavity pressure rise. Consistent fill time = consistent material viscosity. A fill time that drifts longer over a production run means the material viscosity is increasing — possibly moisture pickup in hygroscopic material, or barrel temperature drift.
• Peak cavity pressure: the maximum pressure reached in the cavity during pack. This is the pressure that packs the part. Consistent peak pressure = consistent part weight and dimensions. A declining peak pressure trend means the check ring is wearing or the non-return valve is leaking.
• Pressure at transfer: the cavity pressure at the moment the machine switches from fill to pack. If this varies, the part is being filled to different percentages before packing begins.
• Gate freeze time: the time it takes for the cavity pressure decay curve to flatten. This tells you when the gate has solidified and further holding pressure has no effect. Setting hold time based on gate freeze time rather than a fixed value typically recovers 1–3 seconds of cycle time.
• Cavity-to-cavity balance: in a multi-cavity tool with sensors in each cavity, the pressure curves reveal which cavities fill faster and which pack at lower pressure. Differences of >5% in peak pressure between cavities indicate fill imbalance that will show up as cavity-to-cavity dimensional variation.

3. Process Window Development

A process window is the range of each process parameter within which the process produces acceptable parts. It is developed by deliberately varying parameters to find the limits — not by recording what settings happened to work on the day the mold was sampled.

How it is developed:

1. Run the mold at the nominal settings. Record cavity pressure curves, part dimensions, part weight, and visual quality.
2. Vary one parameter at a time — melt temperature, mold temperature, injection velocity, hold pressure, hold time, cooling time — both above and below nominal.
3. At each variation, measure parts and record the point at which a dimension goes out of tolerance, a visual defect appears, or the cavity pressure curve deviates unacceptably.
4. The process window is the range between the upper and lower acceptable limit for each parameter.

Example process window for a PA66-GF30 automotive component:

A process window that is too narrow (melt temperature ±3°C) is not production-viable — normal machine variation will push the process outside the window routinely. A process window that is too wide (melt temperature ±30°C) suggests the part tolerances are generous or the process is insensitive — both acceptable, but the wide window should be understood, not stumbled into.

4. IQ/OQ/PQ Validation Framework

IQ/OQ/PQ is the validation framework required by FDA for medical device manufacturing (21 CFR Part 820) and expected by automotive OEMs under the PPAP framework (specifically the process validation elements). It provides documented evidence that the manufacturing process consistently produces parts meeting specifications.

IQ — Installation Qualification

What it answers: Is the equipment installed correctly?

For injection molding, IQ documents that:

• The mold is installed in the specified press (make, model, clamp force, screw diameter)
• The auxiliary equipment (temperature control unit, dryer, robot/granulator) is installed and connected per the equipment layout
• The cavity pressure sensors are installed, calibrated, and connected to the data acquisition system
• The mold is functioning — opens, closes, ejects — without interference

Documentation: Equipment serial numbers, calibration certificates for sensors and instruments, installation checklist signed by the setup technician and verified by quality.

OQ — Operational Qualification

What it answers: Does the process operate within the defined parameters?

OQ is where the process window is established (see Section 3 above). It demonstrates that:

• The process runs stably at nominal parameters
• The upper and lower limits of each parameter have been tested
• Parts produced at the limits of the process window still meet specifications
• The cavity pressure sensors are recording data that correlates with part quality

Documentation: Process characterization report with cavity pressure curves at each parameter setting, dimensional data for parts at nominal and limit conditions, process window summary table.

PQ — Performance Qualification

What it answers: Does the process consistently produce acceptable parts under production conditions?

PQ is a production run — or multiple runs — under normal production conditions with normal operators, normal material handling, and normal shift patterns. It demonstrates that:

• The process produces parts within specification across multiple production lots (typically 3 lots minimum for medical)
• The process capability (Cpk) meets the requirement (Cpk ≥ 1.33 minimum, ≥ 1.67 for critical characteristics)
• Lot-to-lot variation is within acceptable limits
• The measurement system (Gage R&R) is capable of detecting out-of-specification parts

Documentation: Multi-lot capability study, Cpk analysis per critical dimension, Gage R&R study results, production run summary including any deviations and their disposition.

The Difference Between IQ/OQ/PQ and a Normal Production Run

A normal production run produces parts. IQ/OQ/PQ produces parts AND documented evidence that the process will continue to produce acceptable parts when nobody is watching. The documentation — not the parts — is the output of validation. The parts are merely proof that the documentation is honest.

5. Process Capability — Cpk and Ppk

Cpk (Process Capability Index) measures how well a process can produce parts within specification, accounting for both the process spread and how centered the process is on nominal. It answers the question: “If we keep running this process as it is running now, what percentage of parts will be in specification?”

Ppk (Process Performance Index) is the same calculation but uses the overall standard deviation (all data points pooled) rather than the within-subgroup standard deviation. Ppk includes lot-to-lot variation and long-term drift that Cpk filters out. Ppk is always ≤ Cpk — the gap between them measures how much long-term variation exists beyond short-term process capability.

Interpreting Cpk values:

The most common Cpk mistake: Calculating Cpk on three shots from one cavity at one moment in time and declaring the process capable. Real Cpk requires data across multiple cavities, multiple lots, and multiple time points. A Cpk of 2.3 on a 5-part sample from T1 is a report. A Cpk of 1.55 on 125 parts across three production runs is a capability statement. The second one means something.

6. What to Require by Industry

Medical Device (FDA / ISO 13485)

• Full IQ/OQ/PQ documentation package
• Cpk ≥ 1.67 on all dimensions identified as critical to quality (CTQ) in the design FMEA
• Multi-lot PQ: minimum 3 production lots, minimum 300 parts per lot
• Gage R&R ≤ 10% for the measurement system (≤ 30% is marginally acceptable)
• Process validation maintained as a living document — re-validation triggered by any material, tooling, or process change

Automotive (IATF 16949 / PPAP)

• PPAP Level 3 including process capability study (element 12) and measurement system analysis (element 13)
• Cpk ≥ 1.67 on all SC (significant characteristics) and CC (critical characteristics)
• Initial process study: minimum 100 parts from a significant production run (minimum 300 consecutive parts or 1 hour of production, whichever is larger)
• Annual re-validation: process capability data submitted with the annual PPAP (the automotive equivalent of on-going process validation)

General Industrial / Consumer

• Cpk ≥ 1.33 on functional dimensions
• FAI report with dimensional data from each cavity
• No formal IQ/OQ/PQ required unless specified by the customer quality agreement
• Process parameters documented and controlled, but validation documentation less extensive than medical/automotive

Scientific molding and process validation add cost to the mold qualification phase — instrumented tooling, sensor integration, extended sampling, documentation hours, and a validation engineer who writes reports instead of making parts. For a medical device component that will be implanted in a patient, or an automotive bracket that will hold a brake line, the cost of not doing it is a recall. For a consumer product enclosure, the cost of doing it fully may exceed the value of the program. The judgment is in knowing which category your part falls into.

See our quality assurance systems →

Request a quote with your validation requirements →