Support strategy economics: when supports cost more than the part

First, build time. Supports add scan time proportional to their footprint; a heavily supported part can spend 20–40% of its build hours printing scaffolding it will throw away.

Second, removal labour. Manual support removal on a complex internal lattice is the single most expensive post-processing step in many cells — hand-finishing rates run $80–$200 per hour and the work doesn't parallelise well.

Third, surface damage. Witness marks from support contact require additional machining or AFM passes to clean up. Designs that pack supports against critical surfaces inherit those passes whether the surfaces need them functionally or not.

The classic guideline — overhangs steeper than 45° self-support — is right on average and wrong on the edges. On L-PBF Ti-6Al-4V with a 30 μm layer and a moderate hatch, the practical self-support angle is closer to 40°. On AlSi10Mg with its higher thermal conductivity, 50–55°. On Inconel 718, 45° holds for downskins up to ~3 mm wide; wider downskins droop and require support regardless of angle.

Replace flat downward-facing surfaces with chamfered or arched profiles. Move parting lines so supported faces become non-critical. Add sacrificial print-in-place fillets that re-enter the part envelope after machining. Re-orient the build so the worst overhangs face the build direction.

• Chamfer or arch every downward-facing surface > 5 mm wide
• Re-orient to put critical faces upskin, not downskin
• Convert flat overhangs into self-supporting truss patterns
• Consolidate parts to share build plate area; supports amortise over more parts

The wizard's cost ranking includes a per-process support model: estimated support volume from the candidate orientation, an estimated removal-labour bin based on accessibility, and a machining allowance per supported face. When the support cost dominates total cost, the wizard surfaces a redesign hint with the highest-payback geometry change.