Exafuse Public Proof Map
This page connects public Exafuse stories to the technical themes I write about here: large-part LMD, repairability, build-and-coat workflows, material breadth, process monitoring, AI image processing, and inspection-aware decision support.
Claim boundary
These are public company proof signals. They support technical context and buyer education; they are not engineering approval, certification, guaranteed outcomes, or confidential project details.
Open proof-map JSON3-axis Titan LMD build space
4 m3
Public Exafuse capability signal for controlled large-part LMD build-up, repair and cladding routes.
Titan component window
2 x 1 x 2 m
Part-envelope language that makes large-component LMD concrete for buyers, engineers, and technical readers.
rotary-table support
1,000 kg
Robotic LMD positioning context for heavy components, repair access and contour-following deposition.
SLM / LPBF window
400 x 400 mm
Powder-bed reference point for comparing LMD scale logic with compact fine-detail metal AM.
Case proof
The strongest proof is not just a number. It is the chain from damaged or designed part, through route selection, deposition, finishing, monitoring, inspection, and release planning.
CS15
Large structural LMD is a CAD-to-production system problem: manufacturability review, path planning, parameter development, monitoring, independent validation and final inspection.
This is the strongest public anchor for my interest in AI-assisted process understanding without replacing inspection evidence.
CS01
A credible hammer repair is not one hardness number. It requires surface preparation, crack context, layer strategy, finishing, bond quality and release evidence.
This maps directly to the LMD Repairability Index and the Quality Evidence Ladder.
CS10
Repair value often comes from a local failure with a large downtime risk. The damaged material must be removed before rebuilding, not hidden below new deposition.
This is a clean RFQ-intelligence example: damage boundary, lead time, machining route and inspection context decide the recommendation.
CS13
LMD can combine geometry creation and functional surface strategy when material compatibility, coating duty, finishing and validation are planned together.
This supports the site's build-and-coat logic: geometry and surface function should be evaluated as one workflow.
Knowledge signals
The articles are useful because they are practical and bounded: they explain fit, limits, inputs, evidence and what not to claim.
A03
LMD is framed around local melt-pool deposition, repair, modification, cladding and hybrid routes. Public examples include a 130 mm drill and a 750 mm multi-material water-cooled nozzle with 1.8 mm thin-wall context, about 50 hours of uninterrupted printing and more than 1,070 layers.
A04
LMD is positioned for large parts, repair, local feature addition and cladding; SLM / PBF is positioned for compact parts, fine geometry and internal channels.
A06
Large LMD depends on bead width, overlap, heat management, machining allowance, fixturing and inspection planning. Public capability context includes 1.8-3.7 mm Titan wall structures and 1.5-4.5 mm robotic zoom-optic wall adjustment.
A12
Monitoring tracks process consistency through signals such as melt-pool behavior, path execution, deposition continuity and thermal history; it does not replace final inspection or qualification.
A21
The wide-bead research target is public: conventional LMD tracks are often 1-4 mm, while the project explores roughly 5 mm and potentially 10 mm tracks, 30-50% time-saving potential and >95% powder-utilization target as project goals.
A25
Exafuse publicly reports more than 1,850 kg of LMD material in 2024, including more than 1,600 kg of 316L and around 250 kg across nickel, wear-resistant, copper and specialty steel routes.
A29
The hammer repair guide frames fit around local accessible damage, viable base material, metallurgical bond, machinability, inspection and explicit release criteria rather than generic hardfacing.
A37
Pix2Pix-style image models are framed as research tools for segmentation, normalization and visual interpretation; model outputs must be validated against process context and inspection evidence.
Interpretation
The site should read like technical interpretation over real industrial work, not a generic portfolio.
Use bridge, hammer, screw and drill examples to explain decisions, not just define acronyms.
Every monitoring or AI statement should route back to inspection, release evidence and claim boundaries.
The best content asks for material, geometry, damage, finish, inspection and timeline before recommending a process.
Material names are useful only when tied to substrate, duty, dilution, heat input, machining and validation.
Sources
Established domain
