Troubleshooting bambu lab slicer CLI for H2D two-material prints (PLA & TPU)

[Copilot]

The issue describes a slicer CLI bug: re-loading profiles via --load-settings/--load-filaments causes per-part extruder assignment to receive a bogus extruder id (e.g. 21842) for TPU under manual multi-extruder mode. Those flags and per-part extruder semantics belong to the OrcaSlicer / Bambu Studio / PrusaSlicer family, not to this repository.

Repository scope check

• This repo is a LaTeX/academic project (proposal.tex, idetc-abstract.tex, references.bib, figures/, reviews/, sterling-cv/). No source code, CLI, or slicer logic.
• grep -i for extruder, load-settings, load-filaments across the tree → 0 matches.
• Matches for PLA/TPU are exclusively prose in .tex / .bib / .md.

Result

No code changes made. There is no surface here to patch without fabricating unrelated edits to the proposal text, which would violate the minimal-change rule.

Suggested next step (pick one)

• Re-file upstream at the relevant slicer project (likely SoftFever/OrcaSlicer or bambulab/BambuStudio) — most probable correct destination.
• Reframe as a manufacturing-risk paragraph in proposal.tex / sections/ if the intent is to document this as a known PLA+TPU multi-material print risk — point me to the target section.
• Share the companion code repo / print-prep script if one exists outside this clone and the bug lives there.

[sgbaird]

@copilot you were meant to clone bambu slicer repo and try to figure out how to sort this out

[Copilot]

@copilot you were meant to clone bambu slicer repo and try to figure out how to sort this out

Cloned bambulab/BambuStudio @ e150b502 and traced it. The bogus 21842 is an out-of-bounds read on filament_maps in src/BambuStudio.cpp around L6775–6796: after --load-filaments shrinks the active filament set, per-volume extruder configs on the loaded 3MF still reference original (higher) slot numbers, so filament_maps[plate_filaments[f_index] - 1] reads past the end of the vector and returns uninitialised memory. Two nearby compounding defects: (1) shadowed inner loop with the same f_index (L6775–6776), and (2) no range check before status >> (filament_extruder - 1) (UB shift) at L6784.

Full diagnosis + reproduction + proposed upstream diff + interim workarounds for our print pipeline are in manufacturing-notes/bambustudio-cli-multimaterial-bug.md (commit in this PR). Since I can't push to bambulab/BambuStudio from here, the upstream issue/PR will need to be filed from your account — the doc includes a ready-to-paste issue title and body.