Defect Investigation
Terms such as porosity, cold shut, flash, blister, shrinkage, drag mark, and warpage describe an observed condition. They do not, by themselves, identify the root cause. Similar-looking conditions can result from different combinations of part geometry, alloy condition, tool design, thermal balance, process settings, trimming, handling, machining, or finishing.
Effective problem solving starts with a controlled sample and a clear defect definition: where it appears, how often it occurs, when it began, which cavity or tool location is involved, and how it affects the drawing or function. The investigation should preserve the part revision, material lot, tool state, process record, and downstream history.
Surface & Filling Defects
A cold shut forms where metal fronts meet without fusing. Surface flow marks should be assessed separately; a visible line alone does not establish the defect mechanism. Possible contributors include an unfavorable fill path, restricted gate area, local heat loss, venting limitations, or process variation. Prevention may involve reviewing geometry, gate and overflow strategy, thermal condition, and the process window together.
Flash forms when metal enters an unintended gap, often near a parting line, slide, insert, or ejector. The response should consider tool condition, alignment, locking and support, local pressure, and trim requirements. Define the trimming requirement and investigate the underlying tool or process condition to reduce recurrence.
Drag, scuffing, soldering, and ejector marks relate to contact between the solidifying part and the tool. Draft, surface condition, local temperature, release practice, ejection timing, ejector distribution, and part rigidity may all contribute. Cosmetic acceptance should identify which process marks are permitted and where.
Internal Casting Defects
Porosity is a broad term. Gas-related voids, shrinkage-related voids, and interconnected leakage paths have different mechanisms and consequences. A void visible in a section does not automatically predict leakage, and an acceptable external surface does not prove internal integrity.
The engineering team should connect the requirement to the risk. Pressure-containing or sealing parts may need a defined leak test and acceptance limit. Machined sealing faces or bores may require local controls because machining can expose subsurface conditions. Radiographic or other non-destructive methods can provide useful evidence when the application justifies them, but the method, sensitivity, sampling, and acceptance criteria must be agreed.
Inclusions and contamination may be influenced by incoming material, melt handling, transfer, tool condition, and process discipline. Material and melt controls should be evaluated with the actual failure evidence rather than assumed from appearance alone.
Dimensional & Geometric Defects
Warpage, mismatch, shrink variation, and dimensional drift can arise from uneven sections, thermal imbalance, ejection, trimming, residual stress, handling, or clamping during machining and inspection. A measurement result must be interpreted with its datum alignment and free-state or restrained-state condition.
When a casting is later machined, variation in the as-cast locator can move every machined feature. The corrective plan may require a change to the casting, fixture, machining sequence, or datum strategy—not simply a tighter machining offset.
Defect Prevention
Prevention begins before tooling with a DFM review of wall transitions, draft, fillets, ribs, bosses, undercuts, parting, filling, venting, overflows, thermal control, ejection, trimming, machining stock, and cosmetic zones. Trial results then provide evidence for refining the tool and process.
During production, controls should focus on inputs and signals that relate to the known risks. These may include material identity, tool condition, temperature-related controls, process parameters, first-piece checks, periodic dimensional checks, visual standards, and maintenance observations. The exact control plan should be part-specific.
Defect Inspection Methods
No single inspection detects every defect:
- Visual inspection supports surface, trim, contamination, and cosmetic assessment.
- Gauges and dimensional measurement evaluate size, position, form, and feature relationships.
- Leak or pressure-decay methods evaluate an agreed sealing requirement.
- Radiographic methods can reveal certain internal discontinuities under a defined technique.
- Sectioning, microscopy, or mechanical testing can support investigations but may be destructive.
- Coating, adhesion, roughness, or cleanliness checks address downstream requirements.
Inspection should have a defined purpose, sampling plan, method, acceptance rule, and record. More inspection is not a substitute for a capable process.
Acceptance Criteria
Requirements such as “no porosity” or “perfect surface” are difficult to apply consistently because the method, location, scale, and functional consequence are undefined. A practical specification identifies the controlled zone, defect type, permitted size or reference level, evaluation method, sample frequency, and disposition authority. Cosmetic standards can use zone drawings and approved boundary samples. Internal-integrity requirements may use an agreed test method or reference standard tied to the functional area.
Acceptance should also distinguish conditions that are relevant before and after machining or finishing. A small as-cast surface indication may disappear during machining, while a subsurface condition may only become visible after material is removed. The control plan should follow the delivered state and actual failure risk.
Corrective Action
When a nonconformance occurs, contain affected material, verify the measurement method, compare good and affected samples, review recent changes, and test likely causes. Corrective action should state what changed and how effectiveness will be verified. Retain evidence that links the action to later production.
Document the observed condition, confirmed cause, corrective action and subsequent verification results so the findings can support future production controls.