Gas Porosity or Shrinkage? Diagnose Before Changing HPDC Parameters

Short answer: do not treat porosity as one defect. Smooth, near-spherical voids often support a trapped-gas hypothesis; irregular, branching or interdendritic cavities more often support a shrinkage hypothesis. Mixed gas-shrinkage pores also exist. Diagnose morphology, location and process history before changing velocity, vacuum, intensification or cooling.

Why the distinction matters

The same X-ray indication can lead to opposite process actions. If gas entrapment dominates, the useful evidence may be in slow-shot control, fill pattern, venting, vacuum timing and leakage. If solidification shrinkage dominates, the useful evidence may be in hot spots, gate freeze, intensification timing, local cooling and feeding. Treating every pore with more pressure or less speed can hide the mechanism and move the defect elsewhere.

A practical diagnostic sequence

1. Confirm the measurement

First confirm that the indication is repeatable. Compare X-ray or CT with sectioning when practical, preserve cavity and location identity, and separate surface-connected defects from internal pores. Average density can quantify total void content, but it does not reliably separate trapped gas from shrinkage by itself.

2. Read morphology and location together

A smooth, rounded pore is consistent with gas pressure shaping the cavity. An angular, dendritic or connected cavity near a thermal hot spot is more consistent with inadequate feeding during solidification. Location matters as much as shape: flow ends, overflows, thick transitions, ribs, bosses and last-to-solidify zones create different hypotheses.

3. Connect the defect to the process timeline

For a gas hypothesis, review shot-sleeve filling, slow-shot transition, gate velocity, vacuum level and timing, vent condition, spray and leakage. For a shrinkage hypothesis, review metal and die temperature, gate freeze, intensification onset and duration, local cooling balance and section changes. Use actual curves and timestamps, not nominal setpoints alone.

4. Test one bounded hypothesis

Change a small number of physically related factors inside approved limits. Keep alloy, die, cavity, measurement method and inspection window traceable. A short controlled DOE is safer than a sequence of undocumented adjustments, especially when gas and shrinkage signatures overlap.

What to investigate when gas is the leading hypothesis

Check vacuum integrity and valve timing, vent and overflow effectiveness, plunger acceleration, sleeve fill ratio, lubricant or moisture contribution and turbulence at the gate. Intensification can compress trapped gas, but it does not remove the source. A smaller X-ray indication after pressure increase is not proof that air entrapment has been solved.

What to investigate when shrinkage is the leading hypothesis

Map last-to-solidify regions, gate freeze and local thermal balance. Then test intensification timing and duration, pressure transmission, local cooling, die temperature and geometry-dependent feeding. The correct direction is component-specific; a result from one alloy, wall thickness or die cannot be copied as a universal recipe.

Where AI can help without replacing metallurgy

AI can connect defect labels and locations with shot curves, vacuum traces, temperatures, cavity, alloy batch and time. It can rank hypotheses, find recurring combinations and propose a bounded next trial. It should not infer pore mechanism from one KPI or prescribe an unrestricted parameter change. The engineer still verifies the mechanism, safety window and quality result.

Evidence boundaries

Pore morphology is evidence, not a unique fingerprint. Gas and shrinkage can interact, and 2D sections can misrepresent a 3D cavity. The cited studies use specific magnesium alloys, geometries and process windows; they support the diagnostic logic, not a universal parameter recipe.

How Castella supports the decision

Castella can structure the investigation around defect mode, location and time-aligned process data, then compare controlled parameter windows. Its role is traceable decision support: preserve the baseline, explain why a trial is proposed, and require production validation before adoption.

Related Castella guide

Use the casting defects guide to connect porosity evidence with a wider defect-diagnosis workflow.

Technical sources

Li, Xiong and Guo: X-ray tomography classification of gas, gas-shrinkage and shrinkage pores in HPDC AM60B — https://doi.org/10.1016/j.jmst.2015.10.002

Gokhale and colleagues: quantified effects of gate velocity, intensification pressure and melt temperature on AM50 porosity — https://doi.org/10.1016/j.msea.2006.04.082

Frequently asked questions

Can X-ray alone prove the pore mechanism?

Not always. X-ray shows location and projected morphology; CT, sectioning, metallography and process evidence may be needed to distinguish mechanisms.

Does a round pore always mean gas?

No. Roundness supports a gas hypothesis, but pressure, solidification and sectioning can alter shape. Use several independent signals.

Which parameter should be changed first?

The parameter linked to the best-supported mechanism, inside an approved window and with a predefined quality check. There is no universal first adjustment.