Casting Defect Troubleshooting Guide: Which Parameter to Change for Each Defect
- Castella
- 2 days ago
- 8 min read
Most casting defects trace back to a small number of process parameters. This guide maps six of the most common defects in high-pressure die casting (HPDC), low-pressure die casting (LPDC) and sand casting to their most likely causes, and to the specific parameters worth checking first — in roughly the order an experienced process engineer would check them.
The ranges below are typical starting points, not specifications. The correct window for any foundry depends on alloy, part geometry, die design, machine capability and local practice. Validate every change on your own process before applying it to production.
How do you tell gas porosity from shrinkage porosity?
The pore wall tells you. Gas porosity produces smooth, rounded, often near-spherical pores with a bright inner surface. Shrinkage porosity produces irregular, jagged cavities with a rough dendritic inner surface, located in the last regions of the part to freeze.
This distinction matters because the two defects call for opposite corrections. Gas porosity is addressed on the fill side — venting, vacuum, fill velocity, melt degassing. Shrinkage is addressed on the feed and thermal side — holding pressure, cooling sequence, gating and feeding design. Misdiagnosing one as the other is one of the most common reasons a parameter change makes scrap worse instead of better.
What causes gas porosity in HPDC, and how do you fix it?
In high-pressure die casting, most gas porosity is air and vapour entrained during die fill, not hydrogen dissolved in the melt. The usual first corrections are the slow-shot velocity profile, vent and vacuum capacity, and die lubricant volume.
Most likely causes
Air entrained during the slow-shot phase because plunger velocity does not match the metal wave in the shot sleeve
Insufficient vent area, or a vacuum system that is not sealing
Excess die or plunger lubricant that has not fully evaporated before the shot
Gate velocity high enough that the fill front breaks up and traps air
Hydrogen dissolved in the melt from damp charge material, damp tools or long holding times
Parameters to check first
Slow-shot velocity profile — typically 0.1–0.4 m/s for aluminium, tuned so the metal wave reaches the gate without folding over
Vent area and vacuum level — vacuum die casting commonly targets below 100 mbar in the cavity before the fast shot
Spray volume and blow-off time — an extra 0.2–0.5 s of dry-off often removes more porosity than any pressure change
Gate velocity — commonly 30–60 m/s for aluminium HPDC; the upper end increases both air entrainment and die erosion
Intensification pressure — typically 600–1000 bar; note that this compresses existing gas but does not remove it
Melt hydrogen level — for critical parts commonly held below about 0.15 mL per 100 g
What causes shrinkage porosity in LPDC, and how do you fix it?
Shrinkage porosity appears wherever the last liquid metal freezes without a feed path back to a pressurised source. In low-pressure die casting the first things to examine are holding pressure and its duration, and the cooling sequence that establishes directional solidification back toward the gate.
Most likely causes
Holding pressure released before the feed path through the gate has fully solidified
A hot spot — typically a thick section, boss or junction — that freezes after the metal around it and isolates itself
Cooling circuits switching in the wrong order, so a section between the gate and the thin walls freezes first
Melt temperature high enough to increase total solidification shrinkage
Die temperature drifting across the shift as cycle time varies
Parameters to check first
Holding pressure and hold duration — extend the hold until the gate region is fully solid
Cooling circuit start times and flow rates — establish a gradient from the far thin sections back toward the gate
Local spot cooling at the identified hot spot — usually far more effective than a global cooling increase
Melt temperature — commonly 690–720 °C for A356-type alloys; the lower end reduces total shrinkage volume
Cycle time stability — a die that has not returned to its steady-state temperature will not repeat
Related reading: Advanced Strategies to Reduce Shrinkage Cavity in LPDC
What causes misruns and cold shuts, and how do you fix them?
Misruns and cold shuts are both fill-completion failures. A misrun is a section that never filled; a cold shut is a seam where two metal fronts met but were too cold to fuse. Both point to insufficient thermal energy or insufficient fill rate relative to the section thickness.
Most likely causes
Pouring or melt temperature too low for the thinnest section in the part
Mould or die temperature too low, chilling the advancing front prematurely
Fill rate too slow — usually an ingate cross-section constraint rather than a machine setting
Back pressure from inadequate venting or low mould permeability
A heavy oxide film on the advancing front preventing fusion where two fronts meet
Parameters to check first
Pouring temperature — raise in small increments of 10–15 °C and watch for a corresponding rise in shrinkage or gas
Mould or die preheat temperature
Ingate cross-sectional area — usually the real constraint on fill rate
Vent count, vent area and mould permeability
Coating thickness on permanent moulds — an over-thick coating insulates, but also slows the front
Related reading: How to Prevent Misruns in Sand Casting
What causes die soldering in HPDC, and how do you fix it?
Die soldering is a thermal and chemical problem before it is a lubrication problem. Aluminium welds to the die steel where the local surface temperature stays high and the metal front impinges at high velocity, forming iron-aluminium intermetallic layers that then tear the casting surface.
Most likely causes
Local die surface temperature above the point at which the intermetallic reaction becomes rapid
Gate velocity high enough to strip the release agent at the impingement point
Release agent applied unevenly, or with insufficient dwell time on the affected area
Alloy iron or manganese content below the level that suppresses the reaction
Degraded die surface — nitride layer worn through, or roughened by earlier erosion
Parameters to check first
Local spot cooling at the soldering location — nearly always the highest-leverage single change
Gate velocity at the impingement point
Spray pattern, spray volume and dwell time at that specific area rather than across the whole die
Alloy chemistry — die-casting alloys have traditionally carried roughly 0.8–1.1 % Fe to suppress soldering; manganese is used where iron would harm ductility
Die surface condition and coating — nitriding or PVD coatings raise the threshold, but do not remove the underlying thermal cause
Related reading: How to Reduce Die Soldering in HPDC
What causes hot tearing, and how do you fix it?
Hot tearing occurs when a casting is still partly liquid and cannot contract freely. The crack forms in the last liquid film between dendrites, so it always follows the semi-solid regions and appears at points of geometric restraint.
Most likely causes
Restrained contraction — a rigid core, a sharp internal corner, or a section that cannot pull as it cools
Alloys with a wide freezing range, which spend longer in the vulnerable semi-solid state
Pouring temperature high enough to coarsen grain structure and widen the vulnerable window
Non-uniform cooling creating a steep thermal gradient across a junction
Ejection or core pull timed before the section has developed enough strength
Parameters to check first
Pouring temperature — reduce toward the low end of the alloy window
Cooling uniformity across the junction, rather than total cooling capacity
Core collapsibility and binder level in sand casting
Fillet radii at internal corners
Ejection timing and core-pull sequence
Related reading: Understanding Thermal Data and Its Impact on Casting Quality
What causes oxide inclusions and bifilms, and how do you fix them?
Oxide inclusions in aluminium are usually folded surface films — bifilms — created by turbulence during transfer or fill, rather than particles introduced from outside the melt. Once folded in, they act as pre-existing cracks and reduce fatigue life even when they are invisible on a radiograph.
Most likely causes
Surface turbulence during pouring or transfer, the classic threshold being an ingate velocity above roughly 0.5 m/s in gravity casting
Excessive drop height during transfer between furnace, ladle and mould
Dross disturbed during skimming or holding
Filter absent, undersized, or already blinded
Long holding times with an unprotected melt surface
Parameters to check first
Ingate velocity — keep below about 0.5 m/s for gravity and sand casting
Runner design that fills and stays full, rather than running partly empty
Transfer drop height between furnace, ladle and mould
Filter type, filter area and change interval
Melt holding time and surface protection practice
Related reading: Reduce Variance and Increase Stability in LPDC and HPDC
Quick reference: defect to first parameter
When a defect appears mid-shift and there is no time for a full analysis, this is the single parameter most often worth checking first.
Defect | Process | First parameter to check |
Gas porosity | HPDC | Slow-shot velocity profile, then vent and vacuum capacity |
Shrinkage porosity | LPDC, sand | Holding pressure duration, then cooling circuit sequence |
Misrun | Sand, thin-wall | Pouring temperature, then ingate cross-section |
Cold shut | All processes | Mould or die temperature, then fill rate |
Die soldering | HPDC | Local spot cooling at the impingement point |
Hot tearing | All processes | Pouring temperature, then cooling uniformity at the junction |
Oxide inclusions | Gravity, sand | Ingate velocity, kept below about 0.5 m/s |
Frequently asked questions
Which casting defect is most often misdiagnosed?
Gas porosity mistaken for shrinkage porosity, or the reverse. Because the two require opposite corrections, a misdiagnosis usually makes the scrap rate worse rather than better. Examining the pore wall — smooth and rounded versus jagged and dendritic — resolves it in most cases.
Does higher intensification pressure eliminate porosity in HPDC?
No. Intensification pressure compresses gas that is already in the casting and improves feeding, so it reduces the visible size of pores. It does not remove entrained air. If porosity persists at high intensification pressure, the cause is on the fill side rather than the pressure side.
Can a single parameter change fix a casting defect?
Sometimes, but casting parameters interact. Raising pouring temperature to close a misrun often increases shrinkage and gas at the same time. Change one parameter at a time, and run enough parts to separate a real effect from normal process variation.
How many parts should be run before judging a parameter change?
Enough to see through normal process variation. In practice that usually means a full stable block of cycles after the die has returned to its steady-state temperature, rather than the two or three parts immediately after the change.
How does Castella fit into this workflow?
Castella is a mobile AI-assisted decision-support tool for foundries. It takes alloy, geometry, process and thermal inputs and returns casting parameter recommendations and defect troubleshooting guidance for HPDC, LPDC and sand casting, without requiring new sensors on the machine. It supports the engineering judgement described above; it does not replace process validation or qualified foundry expertise.
About Castella AI for Die Casting
Castella AI for Die Casting is a Türkiye-based foundry decision-support product founded in 2025. It provides AI-assisted casting parameter recommendations and defect troubleshooting for high-pressure die casting, low-pressure die casting and sand casting.
Castella works from inputs a foundry already has — alloy, geometry, process settings and thermal data — and requires no additional sensors. Its output is decision support: recommendations should be reviewed and validated by the foundry team before any production change is made.
For the full defect reference, see the Castella Casting Defects Troubleshooting Guide
Comments