Understanding Your 3MF Analysis

3MF is a package-based manufacturing format. A file may contain one or more meshes together with objects, component references, transforms, units, materials, colours, textures, thumbnails and application-specific metadata.

This means that analysing a 3MF file is not always equivalent to analysing one STL mesh. The visible model may be assembled from several objects or component references.

Geometry findings describe the mesh itself. Typical measurements include vertex and triangle counts, dimensions and conditions such as open boundaries or non-manifold geometry.

A geometry warning should be investigated, but the presence of a warning alone does not determine the correct repair. The affected geometry and the intended use of the model must also be considered.

A 3MF model can reference components rather than storing the entire visible model as one direct mesh. Components may also carry transforms that control their position, orientation and scale within the build.

For this reason, a tool counting one source object may report different geometry totals from a tool that resolves and measures the complete assembled model.

3MF can carry appearance information that STL does not normally preserve. Depending on the file, this can include material resources, colour groups, textures and preview images.

Moksha3D reports evidence that is actually present in the file. If a material is not explicitly specified, the analysis should not invent one from appearance alone.

Units are essential because the same numerical coordinates can represent very different physical dimensions depending on the unit declared by the file.

Some 3MF packages also contain application or manufacturer metadata such as layer height, nozzle information, filament settings, support configuration or build-plate information. These settings describe manufacturing intent and should not be confused with the underlying mesh geometry.

Different tools can analyse different representations of the same model. One tool may inspect an individual mesh, another may resolve an assembly, and another may apply its own definitions for boundaries, manifold conditions or repairable geometry.

Different numbers therefore do not automatically prove that one tool is wrong. The important question is what each tool measured and how it defined the finding.

Independent tools are valuable because they can provide additional evidence. Their results should be compared and interpreted rather than treated as competing scores.

Repair should be considered when a finding represents a genuine defect that could affect geometry integrity, slicing, manufacturing or the intended use of the model.

Automatic repair is not always the correct response. A repair operation can close intentional openings, remove geometry, change topology or otherwise alter the design.

If the fault is severe or the intended geometry is uncertain, manual inspection is preferable to forcing the model to obtain a PASS result.