Physical AI Lab

The same 6 mm offset magnitude consumed different perpendicular interfaces, while a 22 mm X shift withdrew one tongue and left the other centered.

All three cone pairs had approximately perpendicular axes, but only the matched pair shared an actual apex and corresponding finite generator.

Equal cross-sectional area did not preserve the selected-axis area moments when rectangular link sections changed shape or orientation.

The same recovered drum turn produced about 75.648 mm or 150.876 mm of modeled cable centerline length, depending on winding radius.

Four sampled wheel-and-pawl poses were separated while two intersected, showing why wheel phase and pawl stop need independent finite-solid checks.

The same nominal bore clearance produced shoulder contact, finite overlap, or an unsupported relieved gap depending on shaft fillet and axial position.

The same rigid dovetail insert fits when seated and remains clear during sampled end entry, yet intersects both retaining rails at the sampled top throat.

An actual Blender comparison finds lip overlap in sampled edge-axis openings, while an offset axis keeps the same closed cover clear in all nine tested poses.

Two actuator base pivots map the same pin-length range to different arm angles. Both sampled layouts still fail the finite barrel-clearance check.

Two static correction layouts reach the same tool-face endpoint. Moving only the carriage preserves upstream clearance; moving the whole assembly intersects a fixed guard.