How to Inspect Cavity Volume and Flatness on Multi-Cavity Silicone Molds

Checking one center cavity is not enough for a 20-, 50- or 100-cavity production mold.

A useful inspection plan looks for position patterns:

• center;
• edges;
• corners;
• long unsupported spans.

The goal is to separate true cavity-volume variation, tray deflection, filling/depositor variation and measurement error.

Start with a sampling map

For a rectangular tray, identify:

• four corners;
• center;
• midpoint of each long edge;
• midpoint of each short edge.

For higher-risk applications, expand the map.

Use consistent cavity IDs such as R1C1, R1C10, R5C5. This makes repeat-order comparison possible.

CAD volume is the design reference, not the final production proof

CAD can calculate nominal cavity volume, but the physical mold can differ because of master error, molding distortion, support state, surface finish or trimming.

For gummy, chocolate, soap or wax programs, physical validation may use:

• controlled liquid fill;
• actual casting weight;
• dimensional measurement.

The best method depends on the process.

Water fill is useful-but define the method

If water volume is used:

• level the mold;
• define support;
• remove trapped bubbles;
• use the same fill plane;
• use calibrated measuring equipment;
• control temperature if precision requires it.

A flexible tray that sags under water load can change the measurement. That is exactly why support condition must be defined.

Actual product weight can be the more useful CTQ

For gummy, chocolate, soap or wax, final piece weight may matter more than mathematical cavity volume.

But piece weight also includes process variation:

• density;
• depositor shot;
• drying;
• scraping;
• cooling.

If finished weights vary, compare the physical cavity and process before reworking tooling.

Cavity-to-cavity variation should be mapped, not averaged away

An average can hide a bad edge row.

Example:

• center pieces near target;
• left edge consistently heavy;
• right edge consistently light.

That pattern points toward tray tilt, support, scraper plane or depositor alignment.

If only one cavity is wrong, inspect the local cavity/master.

Relaxed flatness and loaded flatness are different measurements

A tray can be flat while empty and sag when filled.

Measure:

Relaxed

On a defined flat reference with no production load.

Supported

Seated in the production tray or nest.

Loaded

Filled to actual working mass or representative load.

For depositor applications, the loaded state is often the most important.

Flatness can directly change piece weight

If the top plane tilts:

• flood-and-scrape fill height changes;
• nozzle Z-height changes;
• edge cavities may underfill or overfill.

A "weight variation" complaint can therefore be a tray-structure complaint.

Use position data to connect the two.

Inspect repeat batches using the same map

For a reorder, compare:

• same cavity positions;
• same support;
• same method;
• same approved reference.

Do not change the sampling method between lots.

The comparison should answer:

Did the new batch reproduce the approved production behavior?

Acceptance limits should come from the product

There is no universal cavity-volume or flatness tolerance for all silicone molds.

The acceptable variation depends on:

• product label weight;
• machine tolerance;
• casting fit;
• customer specification;
• support method.

A hand-poured craft mold may tolerate more variation than a depositor tray controlling a supplement dose.

Relationship to existing RUUIPON content

Keep:

• /multi-cavity-production-silicone-molds/ = commercial category;
• P33 = spacing/border design;
• P47 = gummy cavity count/flatness;
• P85 = inspection/metrology.

That search intent is distinct enough to justify a dedicated QC guide.