@vbogdanYLC3R
In a fully automated clash test where you are not able to define specific rules on how treat for instance secondary insulation layers in each test direction, the results will always slightly vary. It is a bounding box test after all, if you want a more detailed clash test you need to use Navisworks Manage or a similar application from a 3rd party.
The inconsistency you are observing primarily arise from how the clash detection settings are set up, object geometry and bounding box of the primary elements set and the elements of the test set it is being clashed against. Subsequently, depending on the order of the sets, clashes between the same objects are in certain cases reported twice (once for each direction) depending on how tolerances and clash rules are applied. Hence, the additional count which you have witnessed.
Other reasons for the inconsisty in reports in bi-directional tests are:
- Clash tolerances: If tolerances for one direction allow for small overlaps or bounding box conflicts but the reverse direction uses stricter settings, the results will differ.
- Object Geometry: Considering the fact that clash test in ACC is entirely automated and does not allow for custom rule sets, the Object Geometry and type of category it belongs to becomes a crucial factor that influence the accuracy and consistency of the report.
- Detail Levels and Secondary Elements/Layers: Structural and mechanical elements have different geometrical representations and may as well have different levels of detail, which affect how clashes are detected in ACC.
PS: You are taking about 6 clashes though in a total of what may be 62. Did you cross check to see which ones are those 6 missing ones or if they might actually be duplicates (one test may sometimes detect clash occurances twice in both direction.) Using filters you might be able to exclude these from the report if they are duplicates
