Passing final inspection doesn’t mean quality control succeeded. Quality control succeeds when bad parts can’t be made, not when they’re caught afterward. The distinction matters because catching defects at the end of the line is expensive: the defective parts consumed machine time, material, and labor that can’t be recovered. Preventing defects from forming in the first place captures that value.
This perspective shifts quality control from an inspection function to a process control function. Inspection remains necessary to verify that controls are working, but the primary effort goes into making processes incapable of producing defects rather than filtering defects out of production.
Incoming Material Control
Quality problems often trace back to incoming materials. Material that doesn’t meet specification produces parts that don’t meet specification, regardless of how well the molding process runs.
Certificate of analysis (COA) verification confirms that each material lot meets specification. COAs document melt flow index, density, moisture content, and other relevant properties. Comparing COA values to specification catches out-of-spec material before it enters production. Some operations simply file COAs; effective quality control actually reviews them.
Incoming inspection goes beyond COA review for critical materials or suppliers with variable history. Physical testing of incoming samples, such as melt flow verification or moisture content measurement, confirms that the material matches its documentation. Testing protocols should specify sample size, test methods, and acceptance criteria.
Lot traceability links every part produced to the material lot used in its production. If field failures occur, lot traceability enables investigation of whether a particular material lot contributed. If a material lot is later found defective, lot traceability identifies which parts were affected. Without traceability, problems cannot be contained, and investigations cannot be completed.
Moisture testing for hygroscopic materials should occur before processing, not just upon receipt. Material can absorb moisture during storage even if it arrived dry. Verifying moisture content, ideally less than 0.02 percent for most hygroscopic materials, before processing prevents moisture-related defects.
Process Monitoring
Consistent parts come from consistent processes. Monitoring process parameters identifies variation before it produces defects.
Critical parameters vary by application but typically include temperatures (barrel zones, nozzle, mold), pressures (injection, pack, back), speeds (injection, screw recovery), and times (injection, pack, cooling, cycle). Identifying which parameters most affect part quality for each specific application focuses monitoring effort.
Cavity pressure monitoring provides the most direct measurement of what the plastic experiences during molding. Unlike machine parameters that indicate what the machine is doing, cavity pressure indicates what’s happening inside the mold. Cavity pressure sensors installed in the mold transmit real-time data showing fill progression, pack pressure transmission, and cooling behavior. Parts molded at consistent cavity pressure profiles show consistent dimensions and properties.
Statistical process control (SPC) uses statistical methods to distinguish normal variation from significant changes. Control charts plot parameter values or part measurements over time, with calculated limits indicating when the process has shifted. Cpk measures how well the process stays within specification limits; values above 1.33 indicate capable processes.
| Cpk Value | Interpretation | Action |
|---|---|---|
| Below 1.0 | Process incapable | Immediate improvement required |
| 1.0 to 1.33 | Marginally capable | Improvement recommended |
| 1.33 to 1.67 | Capable | Maintain and monitor |
| Above 1.67 | Highly capable | Consider reducing inspection |
Machine monitoring systems capture process data automatically, enabling analysis without manual recording. Modern systems log hundreds of parameters per cycle, creating databases for troubleshooting and capability studies. Data without analysis has limited value; effective systems include tools for visualization and statistical analysis.
Alarm thresholds trigger alerts when parameters exceed acceptable ranges. Setting thresholds too tight creates nuisance alarms that get ignored; setting them too loose allows defects before detection. Effective threshold setting requires understanding normal process variation and defect correlation.
Response to out-of-control conditions must be defined before conditions occur. When a control chart signals that the process has shifted, what happens? Who gets notified? What containment actions are required? Defined response procedures ensure consistent reaction to detected problems.
First Article Inspection
First article inspection (FAI) verifies that production setup produces conforming parts before full production begins. Problems found during FAI are far less costly than problems found after thousands of parts are produced.
Dimensional verification checks critical dimensions against specification. First articles should be measured thoroughly, not just checked with go/no-go gages. Actual measurements provide data for process capability assessment and documentation for customer approval.
Appearance inspection evaluates visual characteristics against defined standards. Reference samples showing acceptable and unacceptable conditions eliminate subjectivity. Light box inspection under specified illumination conditions provides consistent evaluation.
Functional testing verifies that parts perform their intended function. This may include fit checks with mating components, assembly trials, or performance tests. Functional failures that escape dimensional and visual inspection get caught here.
Documentation requirements typically include completed inspection reports, dimensional data, photographs, and customer approval signatures. FAI packages become permanent records demonstrating that conforming parts were produced at production start.
In-Process Inspection
In-process inspection catches variation before it accumulates into large quantities of defective parts.
Inspection frequency balances detection speed against inspection cost. High frequency catches problems quickly but consumes inspection resources. Low frequency is economical but allows more defects before detection. Frequency should be set based on process stability history and defect cost.
Critical versus non-critical dimensions receive different attention levels. Critical dimensions, those affecting function, safety, or assembly, warrant more frequent inspection than non-critical dimensions. Inspection planning identifies which characteristics require which level of attention.
Visual inspection standards define what operators look for and how to evaluate what they see. Written standards describing defects, boundary samples showing acceptable limits, and training for inspectors create consistent evaluation. Without defined standards, visual inspection becomes subjective.
Automated inspection using vision systems or measurement automation provides consistent, high-frequency inspection without labor cost of manual inspection. Return on investment depends on volume, defect cost, and capability of available systems. Automation is effective for consistent, defined defects; complex or subtle defects may require human judgment.
Final Inspection
Final inspection is the last verification before parts ship. It must be thorough enough to prevent escapes while efficient enough to avoid becoming a bottleneck.
Sampling plans determine how many parts to inspect from each lot. AQL (Acceptable Quality Level) plans specify sample sizes and acceptance/rejection criteria based on lot size and acceptable defect rate. Common AQL levels include 0.65, 1.0, and 2.5 percent, with lower AQLs requiring larger samples and stricter acceptance criteria.
| Lot Size | AQL 1.0 Sample | Accept/Reject | AQL 2.5 Sample | Accept/Reject |
|---|---|---|---|---|
| 151-280 | 32 | 1/2 | 32 | 2/3 |
| 281-500 | 50 | 1/2 | 50 | 3/4 |
| 501-1200 | 80 | 2/3 | 80 | 5/6 |
| 1201-3200 | 125 | 3/4 | 125 | 7/8 |
Inspection procedures document exactly what to check and how. Standardized procedures ensure consistent inspection regardless of who performs it. Procedures should specify equipment, methods, acceptance criteria, and disposition of nonconforming material.
Accept/reject criteria must be unambiguous. When is a scratch a defect? When is a dimension out of tolerance? Defined criteria, reference samples, and calibrated gages eliminate judgment calls that create inconsistent outcomes.
Traceability through shipping connects final inspection records to shipped lots. If a customer quality complaint arises, traceability enables investigation of what was inspected, what results were obtained, and where the breakdown occurred.
Documentation and Traceability
Quality records provide evidence of conformance and enable investigation when problems occur.
Production records document what was made, when, by whom, with what materials, and under what process conditions. These records enable reconstruction of production circumstances when investigation is needed.
Inspection records document what was checked, what results were obtained, and what disposition was assigned. Retention of inspection data provides evidence for customer quality demonstrations and internal trend analysis.
Lot identification marks parts or packaging with codes linking to production records. Bar codes, date codes, and lot numbers enable tracing from finished goods back to raw materials and process conditions.
Record retention periods depend on industry requirements and customer contracts. Medical device records may require retention for product lifetime plus several years. Automotive records may specify 15 years or more. Define retention requirements at project start.
Standards and Certifications
Industry standards provide frameworks for quality systems, and certifications demonstrate compliance to those standards.
ISO 9001 is the foundational quality management system standard, applicable across industries. It establishes requirements for documented procedures, process control, and continuous improvement. ISO 9001 certification indicates that a quality system exists and is audited by a third party. The standard doesn’t specify what to measure or what tolerances to hold; it requires that whatever quality criteria are established be documented, followed, and improved.
IATF 16949 adds automotive-specific requirements to ISO 9001. It includes requirements for APQP (Advanced Product Quality Planning), PPAP (Production Part Approval Process), FMEA (Failure Mode and Effects Analysis), and SPC. Automotive OEMs and major Tier 1 suppliers typically require IATF 16949 certification from their molders. The standard emphasizes defect prevention, variation reduction, and waste elimination throughout the supply chain. PPAP documentation alone can run to hundreds of pages for complex parts.
ISO 13485 specifies quality system requirements for medical device manufacturing. It emphasizes risk management, design control, and traceability throughout the product lifecycle. Medical device molders typically hold this certification. Additional requirements may apply depending on device classification: Class I devices face fewer requirements than Class II or Class III. FDA registration and compliance with 21 CFR Part 820 add further requirements for devices sold in the United States.
AS9100 covers aerospace and defense applications, adding requirements for configuration management, counterfeit parts prevention, and supply chain controls beyond ISO 9001. The standard reflects the safety-critical nature of aerospace applications and the extended product lifecycles common in that industry.
Certifications indicate capability, not guarantee of results. A certified supplier has demonstrated system compliance to auditors. Whether that system produces good parts for your specific application depends on how well they execute. Certification is necessary but not sufficient for supplier qualification.
Customer-specific requirements often extend beyond standard certifications. Automotive OEMs add their own requirements. Medical device companies add validation requirements. Understanding customer-specific requirements before quoting prevents surprises during qualification.
Quality systems prevent defects from reaching customers. The sophistication required depends on the application, but the principle applies universally: build quality into the process rather than inspecting it into the product.
Sources
- ISO 9001:2015. “Quality Management Systems Requirements.”
- IATF 16949:2016. “Quality Management System Requirements for Automotive Production.”
- ISO 13485:2016. “Medical Devices Quality Management Systems.”
- ANSI/ASQ Z1.4. “Sampling Procedures and Tables for Inspection by Attributes.”
- Automotive Industry Action Group. “Statistical Process Control Reference Manual.”
- RJG Inc. “Process Monitoring Best Practices.” https://rjginc.com/