Magnetic Parametric Composition: From Snaps to Explicit Relationships

Dragging a mounting lug toward a housing looks like a simple positioning task. In engineering terms, several different decisions may be hidden inside that gesture.

The engineer might want to align two axes, mate two faces, attach a separate component, or unite both objects into one solid. Similar screen positions can correspond to very different models.

Magnetic Parametric Composition is our research concept for helping an engineer discover these possibilities without allowing proximity to decide the design.

Propose candidates without committing them

As the lug approaches the housing, a future interaction could identify plausible references: a mounting plane, a bore axis, or a named attachment interface.

A candidate would need a clear explanation. “Align this axis with the selected bore” is more useful than an unexplained jump in position.

Proximity could help rank suggestions. It would not prove that a surface can carry a load, that two parts should be fused, or that a suggested axis is the intended datum.

A snap is not a permanent constraint

A temporary snap can help place an object during a drag. A persistent relation must describe what should happen after either object changes.

For example, the lug might remain flush with a housing face with additional constraints leaving one permitted sliding direction. Alternatively, the engineer might lock its placement completely. Those choices need different constraints.

The interaction should show which freedoms remain and require confirmation before creating a persistent dependency.

Joining changes the engineering meaning

Positioning the lug against the housing does not determine whether they become one body.

A Boolean union, an assembly mate, and a contact relationship are distinct actions. In the proposed workflow, the engineer would explicitly choose between them. Material assignments and component identities should not disappear merely because two objects touch.

Subtraction would likewise need an explicit target and operand. Reversing them can produce an entirely different result.

Design for refusal and ambiguity

Consider two nearby bores with similar diameters. The system should expose the ambiguity instead of silently selecting one. An incompatible connection should remain incompatible even if the cursor is close.

A useful prototype would measure unintended commits, successful undo, reference survival after edits, and the engineer’s ability to explain the resulting relationship. Attractive snapping animations would not answer these questions.

The concept is promising only if it reduces interaction effort while preserving deliberate control. NeuroCAD has not established that this complete interaction is available or qualified; the research task is to specify and test those conditions.

Research note: These proposed workflows do not establish current NeuroCAD feature availability, performance, or engineering qualification.