A batch average can conceal a consistent difference between cavities in a wax injection tool. If one cavity produces a feature slightly larger than another, mixing all measurements may make the process look randomly variable. Keeping cavity identity allows a more useful investigation.

The objective is not to assume that every cavity needs a different correction. It is to determine whether the observed variation follows a cavity, a time period, a handling condition or the measurement method. Each pattern calls for a different response.

Preserve identity from the start of the study

Establish how patterns are identified without damaging functional geometry or confusing production markings. The study record should connect each measured pattern to its cavity and relevant process conditions. If identity is lost before measurement, later statistical analysis cannot reconstruct it reliably.

Where castings are included in the study, decide how the pattern identity will remain connected to the final sample. A traceable development set is more informative than a box of otherwise similar parts with no record of origin.

Compare within-cavity and between-cavity behavior

Look at repeated results from each cavity as well as differences among cavities. A stable offset from one cavity suggests a different investigation from a shared shift affecting all cavities at the same time. Plotting data by both cavity and sequence can make the distinction visible.

The table suggests questions rather than automatic causes. Confirm the mechanism before altering tooling.

Use the same measurement definition

The feature, support and measurement stage should remain consistent across cavities. Otherwise, a difference in how a pattern rests or is contacted can look like a cavity offset. Review the measurement method before treating every numerical difference as a tooling problem.

If the pattern geometry is compliant, avoid a fixture that forces it into the nominal shape unless that is the explicitly defined measurement condition. The study needs to report what it actually measures.

Correct only after the process is understood

A tooling change should address a demonstrated, repeatable condition. Changing a cavity to compensate for an uncontrolled storage effect may produce the wrong result once storage is improved. Stabilize the route and then evaluate any remaining tooling offset.

After a correction, repeat the relevant comparison and retain the original baseline. Confirm that the change did not improve one feature while creating a problem elsewhere. A cavity modification should be treated as a controlled manufacturing change.

Translate the findings into routine control

For TripleAce Castings projects using multi-cavity tooling, discuss cavity-related inspection needs during development. The final plan may use targeted checks based on the demonstrated process, but it should preserve enough information to investigate future changes.

The practical benefit is faster diagnosis and more credible dimensional evidence. A batch number remains useful for production records, while cavity identity can reveal variation that the batch average alone hides.

References for process studies

ICI: Atlas of Wax Pattern Defects provides pattern-process context. NIST: Process and Product Monitoring provides statistical background. The study design should reflect the actual tooling and measurement task.