# Revision 3 engineering record

## Geometry change

The former upper channel contained an 8 mm-high horizontal rib whose underside spanned a 57 mm gap, over a 7 mm width: 57 × 7 = 399 mm². The revised sleeve replaces the void and both ribs with a continuous 7 mm side wall. Its C-section is open at both vertical ends. Longitudinal edge chamfers are 1.0 mm; entrance bevels are 0.6 mm laterally and 1.2 mm axially. The two fastener bores retain tangent 60-degree roof faces.

The lower triangular weight-relief window has been removed. The outer profile, wall attachment and case envelope remain unchanged, with 1.2 mm longitudinal chamfers and 0.6 × 1.2 mm end-face bevels. Nominal lower bearing width is 14 mm; edge bevels mean the full nominal strip cannot be credited everywhere as contact area.

## Applied load

With the assumed loaded mass of 3.2 kg:

    W = 3.2 × 9.80665 = 31.38128 N
    F = 2 W = 62.76256 N

The entire F is allocated independently to either lower shoe. Equal sharing is not required by the stated load case. Cable pull, upward movement, impact, seismic loads and attachment strength are not qualified.

The prior open-frame chord-force model and its 1.78 MPa estimate are not reused as a calculation for this solid geometry. A separate elementary screen of the solid section uses an effective width of 12.8 mm and, for 25 ≤ z ≤ 78 mm, depth d(z) = 54 - (34/78)(z-8) mm. Under a point load at z=78:

    nominal bending stress(z) = 6 F (78-z) / (12.8 d(z)^2)

Its maximum over that interval is approximately 0.718 MPa at z=25 mm. This is a nominal section screen only. It does not resolve torsion from the 29.7 mm maximum lateral eccentricity, root/fastener stress concentrations, enclosure compliance, interlayer adhesion or creep. It is not a bound on the actual maximum stress or a material allowable. The actual physical safety factor remains unknown.

## Geometric verification

All five exported parts are valid single solids, watertight and consistently oriented. They fit within 256 mm dimensions and have positive flat bed contact. A continuous swept recorder envelope clears all four mounted brackets during vertical removal. The actual enclosure, feet, cables, tools and ventilation are not represented by that envelope.

The non-bed downward STL faces are at least 45 degrees above the bed; no non-bed horizontal underside remains. An independent section-by-section check at 0.16, 0.20 and 0.24 mm checks whether each cross-section lies within the previous cross-section expanded horizontally by the layer rise, plus 0.0001 mm numerical tolerance. All 15 probes pass. Those are geometric criteria, not native slicer/toolpath results.

The 3MF geometry is derived from the same canonical STL triangles. The four-part plate retains at least 5 mm brim allowance within the 256 × 256 mm bed. Native Bambu off/normal/tree-support probes and a review of bridge toolpaths remain unperformed.

## Physical acceptance still pending

Verify actual loaded mass and contact on sound chassis corners. Use properly selected stud fasteners and metal compression tubes. Proof-test a dummy with the real bearing geometry and centre of gravity at twice the measured mass, checking each lower shoe's allocation without assuming equal sharing. Measure bracket temperature under the actual PoE/drive load and observe time-dependent movement and permanent set. The existing 0.5 mm displacement stop criterion is project-specific, not an established ABS lifetime threshold. Do not infer a long-term capacity from short-term geometry or strength calculations.

## Design reference

Prusa's manufacturer guidance, "Modeling with 3D printing in mind", distinguishes bridging from supported overhangs and recommends chamfers rather than bed-facing fillets where surface finish matters: https://help.prusa3d.com/article/modeling-with-3d-printing-in-mind_164135 . It is general design guidance, not qualification of these brackets.