Injection Molding Cracks, Crazing and Stress Defects: A Buyer’s Evidence Guide

01

Classify the symptom before assigning a cause

Use a controlled description rather than the word ‘crack’ for every visible line. A crack is a material separation that may be surface-only or extend through a section. Crazing is commonly a fine, localized network that can scatter light or appear whitish, but whitening, flow marks, knit lines, scratches, and surface residue can look similar at a distance. Molded-in or residual stress is a condition within a part; it can contribute to later distortion, crazing, or cracking, yet it is not necessarily visible on an as-molded part.

Ask for a defect map that identifies the drawing revision, resin and color, cavity where available, flow direction, gate and weld-line locations, section changes, ribs, bosses, inserts, ejector contact, and the exact position of each finding. Require matched photographs of suspect and accepted parts under an agreed light direction, background, viewing distance, and magnification where relevant. Record whether the condition was present at ejection, after a defined hold period, after assembly, during transport, or after exposure to a particular chemical or temperature.

This first distinction protects the investigation. A gate-adjacent surface craze on a transparent amorphous part, a fracture at a sharp boss during screw installation, and a line that appears only after detergent contact may warrant different hypotheses and tests. The buyer should not infer root cause from appearance alone, and the supplier should not describe a correction as verified until the observed symptom and its boundary conditions are defined.

02

Review likely causes across resin, geometry, mold, machine and service conditions

Material-related causes can include an incorrect resin grade or lot, contamination, unsuitable colorant or additive interaction, inadequate drying for a moisture-sensitive grade, excessive thermal history, prolonged residence, or an unverified regrind practice. These conditions may reduce strength or change chemical sensitivity, but none can be confirmed from a photograph. Request the material designation, supplier lot traceability, certificate or receiving record as applicable, dryer record when material guidance calls for drying, and the documented regrind or purge practice for the affected run.

Geometry and tooling can concentrate stress. Sharp inside corners, abrupt wall transitions, restricted flow around cores, gate placement, difficult packing paths, imbalanced filling, insufficient venting, local cooling differences, rough or damaged steel, and high ejection resistance are plausible contributors. Assembly loads can add to the problem at snap features, fastener bosses, press fits, and insert interfaces. A part that survives molding may still crack if it is forced into an assembly condition or restrained after ejection.

Process and machine conditions can also change orientation, thermal history, packing response, cooling and ejection behavior. Examples include a changed fill profile, melt or mold-temperature control issue, hold-stage change, unstable cushioning, non-repeatable transfer, check-ring performance, or a blocked cooling circuit. Treat these as testable cause families, not a list of universal setpoints. The significance of any setting depends on the resin supplier’s grade guidance, actual part thickness, tool thermal balance, gate freeze behavior, press capability and the measurement method used to compare runs.

Environmental stress cracking deserves its own branch in the cause map. A compatible-looking part can become vulnerable when service, assembly, packaging, cleaners, lubricants, adhesives, coatings or other media act with tensile stress. Chemical resistance is grade-, concentration-, temperature-, time- and stress-dependent. Ask the design owner to identify real contact media and exposures, then ask the material supplier or qualified laboratory whether the proposed test method and conditions are meaningful for that resin and application.

03

Require confirmation checks that can separate competing explanations

Start with evidence that preserves traceability: isolate the affected lot, retain cavity-coded parts where practical, and compare them with known acceptable samples from a defined reference lot. Review the part map, defect timing, production sequence, machine alarms, material lot, tool maintenance, process record and assembly history together. A cavity-specific pattern supports a different next check from a problem spread across all cavities or only one material lot.

Use the least destructive check that answers the question. Close visual inspection and suitable magnification can establish whether a line is open, superficial, or associated with a scratch or weld line. Cross-polarized viewing or photoelastic assessment may give useful comparative evidence of orientation and stress in suitable transparent or translucent materials and simple geometry; it is not a universal quantitative pass/fail method for opaque, filled, crystalline, or complex parts. A controlled thermal or assembly comparison may also be informative when its temperature, time, fixtures and acceptance observation are agreed in advance.

For suspected chemical interaction, do not substitute an arbitrary solvent dip for application evidence. Define the actual medium or a justified surrogate, concentration, exposure time, temperature, stress or restraint, sample conditioning, sample count and failure definition. ISO 22088-1 provides general guidance on choosing an environmental stress-cracking method for thermoplastics. ASTM D1693 is specifically for ethylene plastics and warns that its results are not for direct application to engineering problems, so it should not be imposed on an unrelated resin or used alone to qualify a finished component.

Where a change may have affected mechanical performance, agree a part-representative test or a clearly limited material test. ISO 527-1 sets general principles for tensile properties under defined conditions, but a tensile coupon result is not proof that every molded feature or assembled part will perform in service. The supplier’s report should state what each test can confirm, what it cannot confirm, the sample identity, equipment or fixture, conditioning, result, deviation and comparison baseline.

04

Contain first, then correct the demonstrated mechanism

Before a root cause is established, place the defined suspect population on hold and protect parts needed for comparison. Confirm whether the defect is limited by cavity, date, shift, material lot, tool state, assembly operation, packaging batch or customer exposure. If product risk requires sorting, define the inspection method, lighting, viewing distance, trained inspector criteria, sampling or 100% scope, disposition of ambiguous parts, and traceability of released stock. A broad visual sort may not detect latent stress, so do not present it as a complete remedy unless its limits are acknowledged.

Corrective actions should match the evidence. A verified ejection drag issue may call for repair, polish or release-condition review; a sharp, highly stressed feature may need a design or tooling change; an incorrect material condition may require controlled material handling and a new trial; a service-fluid interaction may require material, design, assembly-load or exposure-control review. When several plausible factors exist, change one meaningful condition at a time where feasible and retain samples in sequence. Multiple simultaneous adjustments can improve appearance while concealing which mechanism mattered.

Ask for a corrective-action record that names the symptom, containment, tested hypothesis, controlled change, result, residual risk and revalidation decision. It should distinguish a temporary deviation from a permanent correction. For a permanent resin, mold, process, machine, inspection or assembly change, require the applicable controlled revision and evidence that the agreed visual, dimensional, assembly and functional criteria were reviewed again. A repair that removes a visible line is not automatically evidence that strength, chemical resistance, fit or long-term performance remains acceptable.

05

Put acceptance criteria and supplier evidence into the sourcing agreement

Write the acceptance condition in product terms. For a safety-, sealing-, load-bearing-, optical- or cosmetic-critical zone, identify the zone on the drawing or approved visual standard, the permissible condition, inspection surface, lighting, viewing distance, magnification if used, sample size or sorting scope, and disposition authority. ‘No cracks’ may be an appropriate requirement in a defined critical area, but a useful criterion also says how an open crack, a craze, a scratch-like mark and an uncertain indication will be judged. Appearance-only acceptance should not replace a defined assembly or functional requirement when the application needs one.

During supplier selection or pre-production review, ask: Which resin grade, colorant, regrind limit and drying controls apply? Where are stress concentrators and ejection loads on the CAD and tool? Which cavities, gates, cooling circuits and tool-maintenance points could make a local defect repeat? What production records link a part to its material, press and cavity? Which service chemicals, assembly loads, temperatures and packaging contacts must the part tolerate? The answers should inform a part-specific control plan rather than a generic promise of crack-free production.

For release after a crack, craze or stress investigation, request a concise evidence pack: controlled part and tool revision; material and lot identification; defect map and photographs; containment scope; relevant process and machine records; mold or assembly inspection findings; trial plan and changed condition; traceable before-and-after samples; agreed test or inspection results; and a signed disposition with open risks. Set the evidence depth according to part risk, volume and consequence of failure. The goal is an auditable decision, not a claim that one trial establishes universal settings or lifetime performance.