NX Clearance/Interference Analysis: Handling False Positives in Routed Assemblies

Routed systems (piping, tubing, harness bundles) are dense, contact-rich assemblies, so NX's general Assembly Clearance Analysis tooling is heavily relied on by routing designers to catch real interferences without drowning in noise. This note captures how experienced users on Eng-Tips actually configure and interpret clearance analysis to avoid false-positive fatigue — a real gap in official docs, which explain the mechanics but not the "how to actually use this without crying" workflow.

Why routed assemblies generate so many "false" hits

A clearance zone in NX means "report any parts closer together than this distance" — not just literal interpenetration. If you specify a 1 mm clearance requirement, NX will flag every pair of components within 1 mm of each other even if they aren't touching at all. In a harness or piping assembly where clamps, clips, grommets, and stock deliberately sit close to structure by design, a nonzero clearance zone can generate a large volume of "violations" that are actually normal, intended proximity — not defects. (Source: Assembly Clearance, NX Help mirror)

"False hard interferences" — a distinct failure mode

Separately from clearance-zone noise, Eng-Tips users describe genuine false hard interferences: cases reported as interpenetration with zero penetration distance, where the interference geometry is perfectly flat with 0 distance from one face to the opposing face. These are frequently attributed to faceted (tessellated) representations used for the interference check rather than exact analytic geometry — i.e., the check is comparing simplified mesh approximations of two surfaces that are nominally coincident (e.g., a mating face), and the faceting introduces spurious "penetration" at the boundary. (Source: Clearance Analysis False Hard Interferences, Eng-Tips)

Practical filtering options practitioners actually use

To cut down on noise before ever looking at results:

• Ignore pairs within the same subassembly and ignore pairs within the same group — options built into the Clearance Analysis setup, useful for routing assemblies where you don't care about self-interference within a single harness bundle subassembly, only interference between the bundle and the rest of the product.
• Ignore pairs within the same part — relevant for multi-body routing stock parts.
• Manually ignore specific reported pairs — right-click a listed interference in the Clearance Browser and mark it ignored; NX moves it to an exclusions folder so it stops appearing in future runs (until the model changes enough to need re-review). This is the standard way to permanently suppress a known-intentional touching condition (e.g., a clip designed to grip a tube) rather than re-triaging it every analysis run.

Practical recommendation for routing-heavy assemblies

Because routed systems are full of deliberate near-contact by design (clips, straps, grommets, stock resting against brackets), teams that rely on a single global clearance-zone value tend to get buried in noise. The pattern reported by experienced users is to run interference-only checks first (0 clearance, true hard interferences) to find real problems, then apply a small nonzero clearance zone as a secondary pass and expect to triage/ignore a batch of intentional-proximity results rather than treating every flagged pair as a defect.

Sources:

• NX: How to ignore touching pairs in Clearance Analysis and only review for hard interferences?, Eng-Tips
• Clearance Analysis False Hard Interferences, Eng-Tips
• Assembly Clearance (NX Help mirror)