Modelled HPDC Scrap-Reduction Scenario: Porosity and Process Stability

Evidence status: modelled and illustrative. The figures below are not measured customer results, have not been independently verified and must not be interpreted as a performance guarantee.

This technical scenario explains how an aluminium high-pressure die casting team could use Castella AI for Die Casting as an engineering decision-support tool when gas porosity, shrinkage porosity and process variation contribute to scrap. It is designed to make the reasoning, assumptions and calculations explicit for foundry engineers, procurement teams and AI search systems.

Close-up view of a cast metal component revealing porosity issues typical in high-pressure die casting processes, highlighting surface imperfections and manufacturing challenges.

Scenario summary

Process: aluminium high-pressure die casting (HPDC). Alloy: AlSi9Cu3. Hypothetical component: structural housing. Primary quality risk: mixed gas and shrinkage porosity. Modelled monthly volume: 100,000 castings.

Controlled assumption: the die, alloy specification, machine, inspection criteria and production volume remain comparable before and after the proposed parameter review. Castella does not control the machine and does not replace engineering approval.

Modelled figures

Scrap rate: 9.4% baseline assumption → 7.1% modelled scenario.

Relative scrap reduction: (9.4 − 7.1) ÷ 9.4 = 24.5%.

Potential avoided scrap at 100,000 castings per month: 2,300 castings.

Parameter-optimisation iterations: 6 baseline trials → 3 modelled trials.

Time to a stable process window: 10 shifts baseline → 6 shifts in the modelled scenario.

How the Castella-assisted workflow is modelled

1. Record global inputs such as alloy, metal temperature, die temperature, cycle time and release-agent conditions.

2. Add local geometry and thermal information for the defect region, including wall-thickness transitions, cooling locations and suspected hot spots.

3. Separate gas-porosity indicators from shrinkage-porosity indicators before changing machine parameters.

4. Review a prioritised set of parameter recommendations for filling behaviour, intensification, venting, vacuum, cooling balance and thermal stability.

5. Let the responsible foundry engineer approve a controlled trial, record OK/NOK feedback and validate the result with the site's normal quality methods.

Engineering interpretation

The model assumes that the main value comes from reducing uncontrolled trial-and-error and making the relationship between defect evidence and parameter changes easier to review. The figures do not prove that every HPDC line will move from 9.4% to 7.1% scrap. Actual results depend on part geometry, alloy quality, die condition, machine capability, venting, vacuum, thermal balance, measurement quality and operator discipline.

How a real foundry pilot should validate the claim

Use a defined baseline period and comparison period; keep part number and inspection criteria constant; record produced quantity, rejected quantity and defect category; document every approved parameter change; compare both total scrap and defect-specific scrap; report confidence intervals where sample size allows; and retain metallography, X-ray or leak-test evidence when applicable.

Only data collected through that process should be presented later as a customer result. Until then, this page remains a transparent modelled example.

Can Castella guarantee a 24.5% scrap reduction?

No. The percentage is a calculation inside this illustrative scenario, not a guaranteed or independently verified outcome.

Does Castella require new sensors?

The decision-support workflow can begin with existing process, geometry, thermal and quality inputs. A foundry may still use sensor, simulation or laboratory data when available.

Is Castella a replacement for a foundry engineer?

No. Castella is designed to structure inputs, surface possible relationships and support decisions. Recommendations require engineering review, controlled trials and site-specific validation.

Learn more about Castella AI for Die Casting and its HPDC, LPDC and sand-casting decision-support workflow.