A void, surface line, or deformation can arise from more than one mechanism. Conversely, one process issue may create several different symptoms. Naming a defect from appearance should therefore never replace defining the problem and gathering evidence.
Investigation begins with a repeatable observation: where is the defect, at what stage was it found, which requirement does it violate, and in which samples or production lots does it recur? Machine settings and acceptance limits are valid only for the specific part and conditions.
Define the Observation Before Guessing the Cause
Instead of “poor quality,” write “a void exposed on the sealing surface after machining” or “a section of the edge that did not form completely.” Record location, size, orientation, stage of appearance, and observation method. Photographs should include scale and location markers, and conforming and nonconforming samples should be retained with production-lot identification.
The stage at which the issue appears also matters: before or after trimming, shot blasting, machining, heat treatment, coating, or assembly. This information does not yet prove the cause, but it narrows the hypotheses and clarifies the next test.
Do Not Treat Gas Porosity, Shrinkage Porosity, and Oxide Damage as the Same
A void may be associated with entrapped gas, insufficient feeding during solidification, or a combination of mechanisms. A rounded or irregular shape is a clue, not a definitive diagnosis. To distinguish gas and shrinkage porosity in real castings, Tanaka and colleagues compared CT data before and after a blister test; the paper itself notes heat-induced sample deformation as a method limitation. [2]
In a study by Wan and colleagues on ADC12 specimens, three-dimensional reconstruction showed that pore shape, size, and location matter in fatigue behavior. [3] The values and relationships in that study are limited to its alloy, geometry, and test. Project acceptance limits should derive from part performance and the agreed inspection method.
Voids are not the only possible discontinuity. In HPDC Al-Si-Cu specimens, Olofsson and colleagues observed hidden oxide damage alongside pores and sometimes independently of the areas dominated by porosity. [4] This finding belongs to a specific study, but it is a reminder that “not seeing a pore in one image” is not proof that all discontinuities are absent.
- 1Define the observation
- 2Gather evidence
- 3Form a testable hypothesis
- 4Make a controlled change
- 5Evaluate the result
- 6Confirm effectiveness
Distinguish Misruns from Cold Shuts
In a misrun, part of the expected volume or edge has not formed. In a cold shut, two metal fronts meet without sufficient metallurgical bonding and may appear as a seam or fold. Not every surface line is a cold shut; parting lines, flow marks, scratches, and marks from later operations must be considered separately.
In experiments on aluminum alloys, Iwata and colleagues examined increased solid fraction, flow resistance, temperature loss, and pressure transmission in relation to cold-shut and misrun formation. [1] The research demonstrates possible mechanisms under defined conditions; it cannot provide a universal pressure or temperature for every part. For each casting, the filling path, temperature, pressure, venting, and geometry must be assessed with data from that process.
Flash, Distortion, and Cracking Require Different Evidence
Casting flash is excess metal that has passed through die-closure boundaries, the parting line, or around moving components; do not confuse it with machining chips or burrs. Review die condition, locking, contact surfaces, and the history of process changes together with the flash location.
For distortion, first verify the measurement method, datum, part temperature, and how the part is supported or clamped. For cracks, the time of appearance, direction, fracture surface, and relationship to ejection or later operations matter. Document the sample before grinding or cosmetic correction so evidence of the defect origin is not destroyed.
Select a Test for a Specific Question
Visual inspection, sectioning, CT, leak testing, metallography, and dimensional measurement do not answer the same question. Method, resolution, sample location, and preparation can change what is observed. One test alone cannot confirm every aspect of quality or the origin of a defect.
| Observation | Next Question |
|---|---|
| Void exposed after machining | At what depth and location does it appear, and in how many parts does it recur? |
| Leak during assembly | Does the leak path pass through the body, or through the joint, seal, or contact surface? |
| Incomplete edge | Did the geometry truly fail to fill, or was material later detached or damaged? |
| Dimension out of specification | Do the datum, instrument, measurement temperature, and clamping method match the agreement? |
| Surface line | Do sectioning, the filling path, or process evidence support that classification? |
Test Hypotheses with Controlled Changes
After recording the evidence, prioritize possible causes from the data: material and melting, die condition, venting, injection profile, thermal control, interruptions, and finishing operations. For each test, state which variable changes, which conditions remain fixed, and how the result will be measured.
If several changes are made at once and the defect decreases, the contribution of each change remains unknown. Simulation can refine a filling or thermal hypothesis for a better-targeted test, but it does not replace samples and process data. For the limits of model use, see the Die-Casting Simulation Guide .
Close Corrective Action by Confirming Effectiveness and Scope
An investigation does not end with a setting change or die repair. Determine which parts and lots were affected, how inventory is handled, what record proves implementation, and how effectiveness is monitored using the same defect definition. A single conforming sample does not prove that the action is repeatable.
Link photographs, inspection reports, change revisions, and test results to the production identifier. If the acceptance criterion was ambiguous, clarify it with the customer; do not change it unilaterally in the middle of the investigation. Quality Control pageandStart a Project provide relevant routes for discussing project requirements.
Information Needed to Investigate a Defect
- The location, size, and stage at which the symptom appeared are recorded.
- The drawing or functional requirement related to the nonconformity is identified.
- Conforming and nonconforming samples are traceable to production lots.
- The test method, resolution, and sample location suit the investigation question.
- In the cause test, the changed variable and fixed conditions are documented.
- The scope of affected items and action effectiveness are monitored against the same criterion.
Frequently Asked Questions
Can the cause of porosity be diagnosed from a photograph?
A photograph is useful for recording location and appearance, but it generally cannot distinguish conclusively among gas, shrinkage, or combined origins. The sample, process records, and a suitable test must be interpreted together.
Is every line on the surface a cold shut?
No. Parting lines, flow marks, scratches, and marks from later operations can look similar. Classification should be supported by geometry, sectioning, or process data.
If radiography shows no void, is the part defect-free?
No. The result depends on resolution, thickness, viewing direction, and discontinuity type, and every method has its own detection scope. Acceptance should be based on the part's functional requirements and inspection plan.
References and Further Reading
- Iwata et al. (2012) — Variation of Molten Metal Pressure and Its Effect on the Defects of Aluminum Alloy Die Castings ↗Open-access experimental research on filling, pressure transmission, cold shuts, and misruns in aluminum die casting.
- Tanaka et al. (2016) — Porosity Discrimination System for Aluminum Alloy Die Castings Using X-ray CT Imaging ↗Open-access experimental research distinguishing gas and shrinkage porosity using CT and blister testing.
- Wan, Zhao & Zou (2014) — Effect of Micro-porosities on Fatigue Behavior in Aluminum Die Castings by 3D X-ray Tomography Inspection ↗Open-access experimental research on ADC12 specimens examining three-dimensional pore shape and location in fatigue behavior.
- Olofsson, Bogdanoff & Tiryakioğlu (2024) — On the Competition between Pores and Hidden Entrainment Damage during In Situ Tensile Testing of Cast Aluminum Alloy Components ↗Open-access experimental research on HPDC Al-Si-Cu specimens covering pores and hidden oxide damage in tension.
Prepare a Clear Defect Description and the Available Evidence
For the technical order review, bring together the defect location, stage of observation, and inspection results so the investigation can be built on traceable information.
View the Related Capability → Send Project Information and Start a Review