In-depth

Visual inspection of die-cast parts: defects and technologies

Porosity, scratches, burrs and unmachined areas: how to detect die-casting defects with machine vision


Close-up of a die-cast engine block in a workshop
Die-cast product. Photo by Jakub Zerdzicki

Background

Die casting makes it possible to mass-produce aluminium, magnesium and zinc components with complex geometries. The process, however, always leaves traces.
Molten metal injected at high speed can trap gas, and it shrinks as it cools. Subsequent machining can then bring to light what the raw casting was hiding. In the automotive sector, where sealing, strength and safety matter, 100% inspection of parts is increasingly the rule.

Defects: what to look for and acceptance criteria

On die-cast parts, visual inspection focuses on five families of defects.

  • Porosity and inclusions: These are voids caused by gas or shrinkage, which after machining surface on faces and inside holes and threads. Pore size, concentration and the distance between pores all matter.
  • Unmachined areas: These are zones of residual raw casting surface left untouched by machining.
  • Scratches, dents and surface breaks: These occur during machining or handling.
  • Burrs: These are material residues along the edges.
  • Non-conforming machined edges: These are chamfers or fillets that are out of specification.

Acceptance criteria

There is no single criterion that suits everyone. Each customer defines, in their own drawing or specification, which defects are allowed, in which area of the part and within which limits. A pore that is tolerated on a non-functional surface may be unacceptable on a sealing seat, a thread or a mating surface. For this reason, every inspection project starts from a shared definition of what is conforming and what is not, area by area.

From specification to vision station

The parameters used to describe porosity are size, concentration and distance. These quantities can be measured by a machine vision system, provided that technology and lighting are chosen carefully. The system must also take into account the other categories of defects, which have different characteristics and require different approaches.

The technologies and features to consider when building a vision system to check for porosity are:

  • Lighting. The light must be designed to highlight pores and scratches. Structured light helps where contrast is low.
  • 3D technologies. Laser scanners and profilometers provide depth information, which is useful for telling a defect from a simple stain or a machining residue.
  • Holes and threads. Liquid lenses change the focal distance in a few milliseconds, so internal surfaces can be inspected without lengthening the cycle time.
  • Configuration. Stations can be in-line, with devices mounted on robots or axes that inspect the part through 360°. Alternatively, inspection bays with fixed cameras are used, where parts are picked up and presented to the station.
  • Criteria and traceability. Configurable software such as ZeroD, the suite developed by Sistemi 2 Visione for machine management and image processing, makes it possible to set non-conformity criteria and acceptance tolerances for each area of the part. It also records the outcome of every inspection, so each component remains traceable.

What vision can and cannot do

Visual inspection sees what is exposed at the surface: porosity uncovered by machining, aesthetic and functional defects, burrs, unmachined areas. It does not replace radiography or tomography for internal porosity. In practice the two approaches complement each other. Section analysis or CT is used to validate the process and to sample parts, while machine vision checks every part in-line, with objective and repeatable criteria.

Conclusions

An effective visual inspection of die-cast parts starts from the customer's acceptance criteria, which are translated into measurements and thresholds set in the vision station. Good design of lighting, optics and software makes the difference between a system that finds defects and one that finds them reliably, with few false rejects.

Sistemi 2 Visione designs machine vision stations for die-cast parts and other automotive components. Contact us to assess your case with a test on your own parts.