Additive Manufacturing (3D Printing) — Engineering Reference

See also (Tier 3 family index): AM Taxonomy

1. At a glance

Additive manufacturing (AM) is the layer-by-layer fabrication of physical parts directly from 3D CAD data — no mold, no pattern, no subtractive blank. ISO/ASTM 52900:2021 codifies the field into seven process families (VAT photopolymerization, material jetting, binder jetting, powder-bed fusion, directed energy deposition, sheet lamination, material extrusion) and reserves the umbrella term additive manufacturing; 3D printing and rapid prototyping are colloquial.

The Wohlers Report 2024 puts the global AM industry at roughly USD 20 billion in 2024, with metal AM systems and services growing fastest (≈ 25 %/yr) while polymer systems dominate unit count. By tonnage, AM is still a rounding error against casting, forging, or machining — but in high-mix, low-volume, geometry-driven niches (aerospace brackets, medical implants, tool inserts with conformal cooling, on-demand spares) it is now production technology, not a prototype shop trick.

Where it sits in the design stack. First pick when any of the following dominate:

• Geometric freedom: internal channels, lattices, organic topology-optimised shapes that no mold could draw, no cutter could reach.
• Part consolidation: replace a 20-piece weldment or assembly with a single printed part (GE’s LEAP fuel nozzle famously consolidated 20 brazed pieces into 1 L-PBF CoCr nozzle, 25 % lighter and 5× more durable).
• Low-volume economics: tooling-free production below ~100–10 000 units depending on size and material.
• Distributed supply: print spares at the point of use rather than warehousing them (US Navy, Siemens Mobility, John Deere).
• Functional integration: conformal cooling in injection-mold tool steel inserts (cycle-time −30 to −50 %), embedded sensors, lightweight lattice cores.

Where it loses. High volumes of simple geometry (a sand-cast bracket is 1/10 the cost at 10 000 units), as-built surface finish where any mating contact is needed (PBF Ra 6–15 µm is the floor), parts requiring full isotropic mechanical properties without HIP, parts where the qualification burden (per ASTM F3122, AMS 7000 series, NASA-STD-6030) outweighs the geometric win. Hobby desktop printing and qualified production AM share the layer concept and almost nothing else — equating them is a category error.

---

2. First principles

2.1 The layer-wise build paradigm

Every AM process executes the same logical pipeline:

1. Slice. A watertight 3D model (STL, 3MF, AMF, STEP-AP242) is sectioned into 2D cross-sections at the chosen layer thickness Δz, typically 25–200 µm for polymers and 20–60 µm for metal PBF.
2. Deposit / cure / fuse. Each layer is built — extruded thermoplastic bead, jetted droplet, scanned laser melt of powder, arc-deposited wire — within the 2D slice contour and infill pattern.
3. Index. Build platform drops (or wire-DED nozzle rises) by Δz; recoater spreads fresh powder if applicable.
4. Repeat until top of part.
5. Post-process. Universally required: depowdering, support removal, stress-relief heat treatment, surface finish, inspection.

Resolution decomposes into two axes. Z-resolution equals layer thickness Δz. XY-resolution equals nozzle diameter (extrusion), laser/electron-beam spot diameter (PBF, DED), or projected pixel pitch (DLP/MSLA). For an EOS M290 L-PBF system: Δz = 30–60 µm, beam spot ~80 µm, achievable feature size ~200 µm, achievable dimensional tolerance ~±100 µm (about IT11–IT12 on small features) before any post-machining.

2.2 Anisotropy is intrinsic

Because layers bond after the underlying layer has partially or fully solidified, the Z-direction (build axis) is always the mechanically weakest. Typical strength ratios σ_z / σ_xy:

• FDM PLA/PETG: 0.40–0.65 (interlayer adhesion is the limit)
• FDM PEEK / PEI / CF-Nylon (heated chamber): 0.70–0.85
• SLS PA12: 0.85–0.95 (powder fusion is more isotropic)
• L-PBF Ti-6Al-4V as-built: 0.85–0.95 (columnar prior-β grains aligned in Z)
• L-PBF after HIP + solution-aged: 0.95–1.00 (recrystallised, near-isotropic)

Design corollary. Print orientation is a design variable, not a build-prep afterthought. Critical tensile loads must align with XY; fatigue-critical features should not place layer interfaces normal to the principal stress.

2.3 Support structures and overhang

Layer N is deposited onto layer N−1. Where the part geometry overhangs into empty space, sacrificial support structures are added during slicing. Rules of thumb:

• L-PBF metals: self-supporting up to 45° from vertical (the “45° rule”); steeper overhangs require supports or process-tuned tilt-and-rotate strategies (Velo3D Sapphire claims supportless up to 0°).
• FDM polymers: 45–60° self-supporting; soluble supports (HIPS, PVA, BVOH) for dual-extruder machines.
• SLA/DLP: ~30° rule; supports are pillars under overhangs and bridge the cured part to the build plate.
• SLS / Binder Jet polymer: no supports — surrounding powder bed supports the part. Major productivity advantage.

Supports cost time, material, post-processing labour, and surface quality at the contact face. Designing supports out is the dominant DfAM intuition.

2.4 Thermal physics drives metal AM

In L-PBF the laser delivers ~200–700 W into a melt pool ~100–200 µm wide moving at 500–2000 mm/s. Power density is on the order of 10⁶ W/cm² — comparable to laser welding. The local cooling rate is 10⁴–10⁶ K/s, producing very fine (often non-equilibrium) microstructures: cellular dendrites at sub-micron spacing, supersaturated solid solutions, martensitic phases not accessible from cast or wrought routes. This is why AM Ti-6Al-4V can hit wrought-equivalent strength as-built — but it is also why residual stresses are large (±200 to ±800 MPa locally), why distortion off the build plate is universal, and why post-build stress-relief is non-negotiable on any Ti or steel part.

2.5 Volumetric energy density (VED) — the master process parameter

For L-PBF a single scalar bundles laser power, scan speed, hatch spacing, and layer thickness:

VED = P / (v · h · t) [J/mm³]

where P = laser power (W), v = scan speed (mm/s), h = hatch spacing (mm), t = layer thickness (mm). Each alloy has a process window — typically 40–80 J/mm³ for Ti-6Al-4V, 50–90 J/mm³ for Inconel 718, 30–50 J/mm³ for AlSi10Mg (Al reflects 91 % of 1070 nm laser light, so the effective absorbed VED is lower — green/blue-laser machines like Trumpf TruPrint 5000 Green dramatically improve Cu and Al print quality).

Below the window: lack-of-fusion pores (irregular, surface-connected, fatigue-killing). Above the window: keyhole porosity (deep, spherical vapor-collapse voids) and excessive spatter. The window is mapped by sweeping VED on coupon arrays and measuring density (Archimedes per ASTM B311) and tensile (ASTM E8).

2.6 Scan strategy

The 2D laser path inside each layer matters as much as the energy input:

• Stripes / chess — splits the layer into rectangles or chequerboard tiles scanned in alternating directions; reduces in-plane residual stress.
• Island / hexagonal — 5–10 mm tiles with randomised order; minimum stress, used on large Ti L-PBF plates.
• Contour + hatch — slow contour pass for surface quality, fast hatch for interior bulk. Standard on almost every L-PBF machine.
• Layer rotation (67° typical) — rotates the hatch by an irrational multiple per layer so that no two layers share scan lines; randomises anisotropy.
• Up-skin / down-skin / in-skin parameter sets — separate VED tuning for the top surface, bottom-facing overhangs, and bulk core.

---

3. The seven ISO/ASTM 52900 process families

Family (52900 abbrev.)	Mechanism	Materials	Layer Δz	Typical tolerance	Strength	Cost / kg
Material extrusion (MEX) — FDM/FFF	Heated thermoplastic extruded through nozzle	PLA, PETG, ABS, PA, PC, PEEK, PEI, CF/GF-reinforced	100–300 µm	±200 µm	Medium (anisotropic)	$20–$500 (polymer)
VAT photopolymerization (VPP) — SLA/DLP/MSLA	UV-cured liquid photopolymer	Standard, tough, flexible, castable, biocompatible resin	25–100 µm	±100 µm	Low–medium, brittle	$50–$500
Material jetting (MJT) — PolyJet/MultiJet	UV-cured photopolymer droplets from inkjet head, multi-material	Photopolymers (Stratasys VeroFamily, Agilus)	14–32 µm	±50 µm	Low	$200–$700
Binder jetting (BJT)	Powder bed + liquid binder droplet, then sinter (metal) or infiltrate	Sand, PA12 (HP MJF), 316L, 17-4 PH, Cu, W	50–150 µm	±300 µm green, ±100 µm sintered	Medium after sinter; HIP optional	$50–$1000
Powder-bed fusion (PBF) — L-PBF (SLM/DMLS), EB-PBF (EBM), SLS	Powder bed, scanning laser or electron beam fuses cross-section	Metals (Ti, Ni, Al, steel, Cu) for L-PBF/EBM; PA12, PA11, TPU for SLS	20–100 µm	±50–100 µm	High; near-wrought after HIP	Metal $200–$2000; polymer $50–$200
Directed energy deposition (DED) — L-DED, WAAM, EB-DED	Wire or powder fed into traveling melt pool from laser/EB/arc	Metals only; Ti, Ni, steel, Al, mixed-grade gradients	0.3–3 mm	±0.5–2 mm (machining stock required)	High; columnar	$50–$500
Sheet lamination (SHL) — LOM, UAM	Stack and bond sheets, then cut profile	Paper, polymer, Al, Cu, Ti (UAM solid-state)	50–200 µm	±100 µm	Low–medium (anisotropic)	Niche

3.1 Material extrusion (MEX / FDM / FFF)

Heated thermoplastic filament (1.75 mm or 2.85 mm) or pellet is extruded through a nozzle (0.25–1.20 mm; 0.40 mm typical) onto a heated build plate. The most accessible AM process — and the most variable in build quality.

• Cost band. $200desktop(BambuLabP1,PrusaMK4)to$200 000 industrial (Stratasys F900, Markforged FX20 with continuous CF).
• Materials. PLA (rapid prototyping), PETG (chemical resistance + ductility), ABS (engineering grade, warps without enclosure), PA6/PA12 nylon (hygroscopic — dry before printing), PC (impact + 130 °C T_use), PEEK and PEI Ultem 9085/1010 (aerospace-grade, requires 400 °C nozzle and 90–180 °C heated chamber), and fibre-reinforced grades (CF-PA, GF-PA, Markforged Onyx with chopped CF in PA). Continuous-fibre FDM (Markforged) lays continuous CF, fibreglass, or Kevlar tow alongside the matrix bead — quasi-composite performance.
• Use cases. Prototyping, jigs and fixtures, dunnage, low-volume end-use plastic parts, ESD-safe enclosures (PEKK-ESD), aerospace cabin parts under FAR 25.853 (Ultem 9085 is FAA-approved).
• Failure modes. Warping (large flat ABS), delamination (Z-axis interlayer adhesion), nozzle wear (CF/GF require hardened steel or ruby), oozing/stringing, hygroscopic filament shooting steam through the nozzle.

3.2 VAT photopolymerization (VPP)

A vat of UV-curable liquid resin is selectively cured layer by layer.

• SLA (stereolithography) — scanning UV laser, large build volume, accurate (Formlabs Form 4 35 µm XY, 25 µm Z).
• DLP (digital light processing) — entire layer cured at once by a projected image, fast (Carbon M3 DLS process).
• MSLA (masked SLA) — LCD mask over a UV-LED array, cheapest (Anycubic Photon, Elegoo Saturn).
• CLIP / DLS (Carbon) — continuous liquid interface; oxygen-permeable window suppresses cure at the bottom of the vat for vertical print speeds of 100 mm/h+.

Materials. Standard tough/rigid acrylates, flexible (Tango), castable wax-burnout (jewelry), high-temp (Formlabs High Temp 238 °C HDT), biocompatible Class I/IIa (Formlabs BioMed, NextDent dental), industrial DLS (Carbon EPU 40 elastomer, RPU 70 rigid).

Use cases. Dental aligners (Align Technology now prints 700 000+ aligners/day on Formlabs and SprintRay fleets), surgical guides, hearing-aid shells (the entire industry switched to SLA-printed shells by 2010), jewelry casting masters, dimensionally tight prototypes.

3.3 Material jetting (MJT — PolyJet / MultiJet)

UV-curable photopolymer dropletted from a piezo inkjet print head, multiple heads enabling multi-material and multi-colour prints in a single build.

• Stratasys J5/J55/J850/J35 Pro range; 3D Systems MultiJet (MJP).
• 14–32 µm layer, full-colour (J850), digital materials blended on-the-fly between rigid (VeroWhite) and elastomeric (Agilus30) — Shore 20A to glass-like in the same build.
• Use cases. Multi-material concept models, eye replicas, anatomical models for surgical planning (Stratasys Digital Anatomy), tactile colour prototypes.

3.4 Binder jetting (BJT)

A powder bed is selectively wetted with binder droplets, leaving a green part that is then sintered (metal/ceramic) or used directly (sand for foundry molds).

• HP Multi Jet Fusion (MJF) — variant of binder jetting using fusing/detailing agents and IR-lamp fusion of PA12 powder. Dominant industrial polymer AM platform by part throughput.
• ExOne / Desktop Metal / Digital Metal — metal binder jet, sintering at 1100–1380 °C, optional HIP.
• Voxeljet VX-series — sand molds for foundry casting, 20 mm/h build height, parts up to 4 m × 2 m × 1 m.

Use cases. High-throughput polymer parts (10 000+ parts/build on HP 5210), complex metal shapes at 1/3 the cost-per-part of L-PBF, sand cores and molds that replace pattern-tooling entirely for low-volume castings (Loramendi, Voxeljet supply core packs for Cadillac CT4-V and Cummins block castings).

3.5 Powder-bed fusion (PBF)

Powder bed selectively fused by a focused energy source.

Laser-PBF (L-PBF, also SLM, DMLS, LaserCUSING). Inert (Ar/N₂) atmosphere, 200–1000 W fibre laser, single or multi-laser (SLM NXG XII has 12 lasers building concurrently). Materials and machines:

Alloy	Spec	Powder size	Typical Δz	Machine families
316L SS	AMS 5648 / equiv.	15–45 µm	30 µm	EOS M290, GE M2, Renishaw RenAM 500, SLM NXG, 3D Systems DMP
17-4 PH SS	AMS 5643	15–45 µm	30 µm	All major L-PBF
AlSi10Mg	EN AC-43000	20–63 µm	30–60 µm	EOS M400, SLM 280/500
Ti-6Al-4V grade 23 (ELI)	AMS 4998 / ASTM F3001	15–45 µm	30–60 µm	EOS M290/M400, Velo3D Sapphire, GE Concept Laser
Inconel 625 / 718	AMS 5662/5663, AMS 5666	15–45 µm	30 µm	All; GE Additive most-qualified for aero
CoCr (F75)	ASTM F75, F1537	15–45 µm	20–30 µm	EOS M100/M290 medical
H13 tool steel	AISI H13	15–45 µm	30 µm	EOS, SLM
CuCr1Zr / pure Cu	C18150	15–45 µm	30 µm	TruPrint 5000 green, EOS M290 green laser

Electron-beam PBF (EB-PBF, EBM). Vacuum chamber, electron-beam gun, hot bed preheated to ~700 °C which dramatically reduces residual stress and allows nearly support-less builds in Ti-6Al-4V. Slightly higher deposition rate than L-PBF, coarser surface (Ra 25–35 µm), beam focus 100–200 µm. Machines: GE Additive Arcam Spectra L/H (formerly Arcam EBM), Wayland Additive Calibur3, Freemelt ONE. Dominant in orthopedic Ti implants and γ-TiAl turbine blades.

Polymer L-PBF (SLS). CO₂ or diode laser fuses PA12 / PA11 / PA6 / TPU powder in a heated bed at ~170 °C (just below sintering temperature). No supports required. Machines: EOS P-series (P396, P770, P810), Farsoon HT/Flight series, Sintratec S2, Formlabs Fuse 1+. End-use polymer parts at 0.05–0.5 kg.

3.6 Directed energy deposition (DED)

Wire or powder fed into a moving melt pool created by laser, electron beam, or arc.

• L-DED (powder) — Optomec LENS 860/1500, RPM Innovations 557Xtreme, BeAM (AddUp Modulo). 1–5 kW laser, 5–50 g/min deposition.
• L-DED (wire) / Hot-wire L-DED — preheated wire reduces porosity vs powder. Meltio M450.
• WAAM (Wire-Arc AM) — MIG/TIG/plasma + wire. Lincoln SCULPT, MX3D, Norsk Titanium Rapid Plasma Deposition (RPD, plasma + Ti wire in Ar). Deposition rates 2–10 kg/h — orders of magnitude above PBF. Norsk RPD produces certified Ti structural parts for Boeing 787.
• EB-DED — Sciaky EBAM. Vacuum chamber, EB + wire, up to 18 kg/h deposition. Largest AM parts in production (5 m+ Ti spars).

Use cases. Repair (turbine blade tip rebuild, mold cavity touch-up — AMS 7011), large near-net-shape preforms, multi-material functionally-graded parts (gradient from Inconel to steel along a turbine shaft). DED is not for small or fine-featured parts — minimum feature size 1–3 mm and machining stock is mandatory.

3.7 Sheet lamination (SHL)

Stack and bond sheets, cut profile per layer. Niche. LOM (laminated paper objects) is essentially obsolete. Ultrasonic AM (UAM, Fabrisonic) uses ultrasonic vibration to solid-state-bond Al / Cu / Ti / steel foils, allowing embedded electronics, dissimilar-metal joining, and very low residual stress — used for heat exchangers and embedded-sensor structures.

---

4. Reference data — materials, tolerances, finish

4.1 Polymer AM materials (common grades)

Material	Family	Process	σ_t (MPa, XY)	E (GPa)	T_use (°C)	Notes
PLA	Commodity TP	MEX	45–60	3.0	50	Easy, brittle, low T
PETG	Engineering TP	MEX	50	2.0	70	Chemical-resistant, food-safe
ABS	Engineering TP	MEX, MJF	35–45	2.0	90	Warps without enclosure
PA12	Engineering TP	SLS, MJF	48	1.7	90	Workhorse SLS material
PA12 + 40 % CF	Composite	MEX (Markforged Onyx FR-A)	80–100	7.0	145	Anisotropic, ESD-safe variant
PC (polycarbonate)	Engineering TP	MEX	60	2.3	130	Tough, impact
Ultem 9085 (PEI)	High-perf TP	MEX	70	2.2	170	FAA-approved cabin interiors
PEEK	High-perf TP	MEX, SLS	95	4.0	250	$500/kg; 400 °C nozzle, heated chamber
Formlabs Tough 2000 resin	Photopolymer	VPP	46	2.2	63 (HDT)	ABS-like
Formlabs High Temp v2	Photopolymer	VPP	58	3.6	238 (HDT)	Mold inserts, lighting
Formlabs BioMed Amber	Photopolymer	VPP	73	2.4	105	USP Class VI, autoclavable

4.2 Metal AM materials — typical as-built and post-treated properties

Alloy	Process	Condition	σ_y (MPa)	σ_u (MPa)	Elong. (%)	Reference
316L SS	L-PBF	As-built, XY	480–540	600–680	40–55	AMS 7003
316L SS	L-PBF	Stress-relieved	450	590	50
17-4 PH SS	L-PBF	H900 aged	1170	1310	10	AMS 5643 wrought baseline
Ti-6Al-4V (gr 23)	L-PBF	As-built	1100	1230	7	ASTM F3001
Ti-6Al-4V (gr 23)	L-PBF	Stress-relieved + HIP	870	970	14	ASTM F3001
Ti-6Al-4V (gr 5)	EB-PBF	As-built (hot bed)	925	1010	14	ASTM F2924 (EB variant)
AlSi10Mg	L-PBF	As-built	230	380	6	EN AC-43000 cast 220 MPa baseline
AlSi10Mg	L-PBF	T6 (530 °C SHT + age)	245	330	11
Inconel 718	L-PBF	Solution + aged AMS 5664 cycle	1100	1320	16	AMS 5664 wrought 1030 baseline
Inconel 625	L-PBF	Solution annealed	540	880	39	AMS 5666
CoCr (F75)	L-PBF	Stress-relieved	700	1100	14	ASTM F75 / F3213
H13 tool steel	L-PBF	Hardened + tempered 50 HRC	—	1700–1900	4	AISI H13 wrought baseline
CuCrZr (C18150)	L-PBF	Aged	380	450	14	High thermal conductivity
Ti-6Al-4V	WAAM (Norsk RPD)	Stress-relieved	850	950	12	AMS 4999 wire baseline
Inconel 718	L-DED	HIP + age	1050	1280	19	AMS 5662

Values typical. Always require lot-by-lot witness coupons (per ASTM F3122) for qualified production.

4.3 Achievable tolerance and surface finish by family

Family	Δz	Min feature	Tolerance (small parts)	Ra as-built	After polish
MEX (FDM)	100–300 µm	0.8 mm	±300 µm or ±0.2 %	10–25 µm	1–3 µm (vapour smooth, sand)
VPP (SLA/DLP)	25–100 µm	150 µm	±100 µm	1–5 µm	< 0.5 µm (UV cure + polish)
MJT (PolyJet)	14–32 µm	200 µm	±50 µm	1–3 µm	0.5 µm
BJT polymer (MJF)	80 µm	500 µm	±300 µm	6–10 µm	1 µm
BJT metal	50–100 µm	500 µm	±300 µm green / ±100 µm sintered	6–8 µm	0.4 µm
L-PBF metal	20–60 µm	200 µm	±100 µm	6–15 µm	< 0.4 µm (EP, Hirtisation)
EB-PBF metal	50–100 µm	400 µm	±200 µm	25–35 µm	0.8 µm (machine)
SLS polymer	80–120 µm	500 µm	±300 µm or ±0.3 %	5–10 µm	1–2 µm
L-DED metal	0.3–1 mm	1 mm	±500 µm + 1 mm machining stock	25–50 µm	0.4 µm (machine)
WAAM metal	1–3 mm	3 mm	±1.5 mm + 3 mm machining stock	100+ µm	0.4 µm (machine)

---

5p. Theory — Design for Additive Manufacturing (DfAM)

5p.1 The DfAM design space

Subtractive design starts from a billet and removes; AM design starts from nothing and adds. Five principles drive the change in mindset:

1. Complexity is free — within reason. A lattice-filled bracket costs the same per layer as a solid bracket; the metallurgist does not care.
2. Topology optimisation pays. Run a load-case-driven optimiser (Altair Inspire, ANSYS Discovery, nTopology, Fusion 360 Generative Design) to find the minimum-mass material distribution under stated constraints, then smooth and verify. The GE bracket challenge winner (Atkins) cut a Ti-6Al-4V bracket from 2.066 kg to 0.327 kg — 84 % lighter — with no loss of stiffness.
3. Consolidate. Replace assemblies with monoliths. Fewer fasteners, fewer welds, fewer quality inspections.
4. Embed function. Conformal cooling channels in mold inserts; internal lattices for impact absorption; integrated flow paths in manifolds (Liebherr printed a Ti-6Al-4V hydraulic block for Airbus A380 with 35 % fewer parts and 30 % less mass).
5. Design supports out, or design to remove them. Self-supporting angles (≤ 45° from vertical in L-PBF), tear-drop overhangs for circular holes printed in horizontal orientation, sacrificial keep-out for support-tool access.

5p.2 Lattice and infill structures

Lattices replace solid bulk with periodic unit cells (BCC, octet truss, Kelvin, gyroid, Schwarz primitive). Engineered for:

• Stiffness-to-mass — octet truss approaches the Hashin-Shtrikman bound.
• Energy absorption — bending-dominated BCC for crash structures.
• Heat transfer — TPMS (triply periodic minimal surface) gyroid for compact heat exchangers; one print produces a 200-tube equivalent at 1/3 the volume.
• Bone ingrowth — 500–700 µm pore trabecular for orthopedic implants (Stryker Tritanium, Renovis Tesera).

Specialist tools (nTopology, Materialise 3-matic) generate implicit-modeled lattices that survive Boolean operations without exploding the polygon count.

5p.3 Build orientation as design variable

Build orientation simultaneously affects:

• Mechanical anisotropy (load along XY > Z)
• Surface finish (down-facing surfaces are roughest, ~2× as-built Ra of up-facing)
• Dimensional accuracy (circular features in the XZ plane print as stair-stepped ovals; circular features in XY plane print true)
• Support volume and access
• Build time (taller builds = more layers = longer)

In practice, orientation is chosen as a multi-objective optimisation, often by hand, sometimes by build-prep software heuristics (Materialise Magics, Autodesk Netfabb).

5p.4 Pre-machining stock

Where AM surface finish and tolerance fall short, leave 0.5–1.0 mm of machining stock on mating, sealing, or bearing surfaces. The workflow is AM + machining hybrid: build near-net-shape, then 5-axis CNC the critical features. Machines that combine both in one envelope (Mazak Integrex i-400 AM, DMG Mori Lasertec 65 3D Hybrid) automate this for high-value parts.

5p.5 DfAM rule-of-thumb checklist (L-PBF metal)

Feature	Minimum / rule	Why
Wall thickness	≥ 0.4 mm; 0.6–0.8 mm preferred	Below 0.4 mm: incomplete melt + warp
Hole diameter (vertical axis)	≥ 0.5 mm printed; circular	Smaller drills better than printing
Hole diameter (horizontal axis)	≥ 1 mm; teardrop or diamond profile preferred	Round horizontal holes sag at top
Overhang angle	≥ 45° from vertical (Velo3D supportless)	Lower angles need supports
Overhang bridge	≤ 0.5 mm un-supported span	Surface tension + recoater drag
Internal channel	≥ 0.8 mm diameter; teardrop preferred	Powder removal feasibility
Pre-machining stock	0.5–1.0 mm on critical surfaces	As-built Ra 6–15 µm insufficient
Aspect ratio (height : base)	≤ 8 : 1 unsupported	Heat accumulation + curl
Engraved text	≥ 0.8 mm stroke, 0.5 mm depth	Smaller blurs out
Embossed text	≥ 0.5 mm stroke, 0.3 mm height	Smaller doesn’t resolve
Threaded hole	Print under-size, tap post-build	Printed threads fail load
Press-fit hole	Print under-size, ream post-build	Tolerance > printed capability

---

6p. Application — build prep, slicing, and post-processing

6p.1 File formats

Format	Geometry representation	Color / multi-material	Metadata	Standard
STL	Watertight triangle mesh	No	Filename only	De facto, no formal owner
OBJ	Triangle mesh + texture	Texture, not material assignment	UV map	Wavefront
AMF	Triangle mesh, curved patches	Yes, per-material	Limited	ISO/ASTM 52915:2020
3MF	Triangle mesh + properties	Yes	Rich (lattice, supports, build)	3MF Consortium (Microsoft) — emerging dominant standard
STEP-AP242	CAD B-rep	Yes	Full PMI	ISO 10303-242
Build file (CLI, SLM, slc)	Per-machine sliced path data	n/a	Process parameters	Vendor-specific

STL remains entrenched but is lossy: every curved surface becomes faceted, and there is no provenance, no units (though widely assumed mm), no scale check. 3MF is the modern replacement and should be preferred whenever the slicer and machine support it.

6p.2 Slicers and build prep

• Polymer FDM: PrusaSlicer, Bambu Studio, UltiMaker Cura, Simplify3D, Markforged Eiger.
• Polymer VPP: Formlabs PreForm, Chitubox, Lychee, Carbon Cloud.
• Polymer SLS / MJF: EOS RP-Tools / EOSPRINT, HP SmartStream 3D Build Manager.
• Metal L-PBF: Materialise Magics + Build Processors, Autodesk Netfabb, EOSPRINT 2, 3DXpert, GE Concept Print, Velo3D Flow.
• Process simulation (distortion + thermal): ANSYS Additive Print/Suite, Autodesk Netfabb Simulation, Materialise Process Simulation, Velo3D Flow. Predicts in-build distortion and pre-compensates the input geometry by inverse warp.

6p.3 Post-processing pipeline (metal L-PBF reference)

1. Powder removal / depowdering. Brush, vibration, blast cabinet under inert atmosphere if reactive (Ti, Al).
2. Stress-relief heat treatment, before removing the part from the build plate. Ti-6Al-4V: 650–800 °C / 2–4 h in Ar. 316L: 650 °C / 2 h. Skipping this on Ti will produce visible curl and possible cracking when the plate is cut.
3. Plate removal. Wire EDM or band saw.
4. Support removal. Mechanical (cutter, grinder), wire EDM for inaccessible supports.
5. HIP (Hot Isostatic Pressing) if fatigue or pressure-containment critical. 100–200 MPa Ar + sintering-range temperature (Ti-6Al-4V: 920 °C / 100 MPa / 2 h). Closes internal porosity > 99.95 % theoretical density. Surface-connected porosity is not closed by HIP — drives the need for clean surface finish before HIP for vacuum-tight parts.
6. Solution + aging as per alloy spec (17-4 PH H900, Inconel 718 AMS 5664).
7. Machining of critical surfaces.
8. Surface finishing. Bead blast (light cosmetic + cleaning); tumbling/vibratory; electropolish (Ra 0.1–0.4 µm on Ti, stainless); Hirtisation (chemical-electrochemical) for internal surfaces of complex geometries; abrasive flow machining (AFM) for internal channels.
9. Inspection. Industrial CT (Nikon XT H, Zeiss Metrotom, Yxlon, GE Phoenix) for porosity and dimensional; CMM for external features; surface profilometer (Mitutoyo Surftest SJ-410) for Ra; fluorescent dye penetrant (ASTM E1417) for surface defects; ultrasonic for sub-surface; lot-by-lot witness coupons tensile-tested per ASTM E8 and density per Archimedes (ASTM B311).

6p.4 Typical process parameters (L-PBF, reference values)

Alloy	Laser P (W)	Scan v (mm/s)	Hatch h (µm)	Layer t (µm)	VED (J/mm³)	Atmosphere
316L SS	200–280	800–1200	100–110	30	60–90	N₂ or Ar, O₂ < 100 ppm
17-4 PH SS	200–280	700–1000	100	30	65–90	N₂ or Ar
Ti-6Al-4V	250–340	1000–1400	100–140	30–60	40–75	Ar, O₂ < 100 ppm
AlSi10Mg	350–400	1300–1800	170–190	30–60	35–55	Ar (Al detonation hazard)
Inconel 718	250–300	800–1100	100–110	30–40	65–95	Ar
Inconel 625	200–300	800–1100	100	30	60–90	Ar
CoCr (F75)	180–220	700–1000	80–110	20–30	80–130	Ar
H13 tool steel	200–280	600–900	100	30	75–110	N₂ or Ar
Pure Cu (green laser)	350–500 (515 nm)	600–800	80–100	30	145–260	Ar
Pure W	300–400	200–400	60–80	20	470–830	Ar, hot bed

Values illustrative; every machine + powder lot needs its own characterised window.

6p.5 Qualification flow for production AM parts

Production AM parts in aerospace, medical, and pressure-containment service require a structured qualification pipeline:

1. Material qualification — powder lot characterisation: PSD per ISO 13320, morphology per SEM, flow per ISO 4490 (Hall flowmeter), apparent density per ISO 3923, chemistry per ASTM E1019 (O, N, H) + ICP for trace elements. Specs reference SAE AMS / ASTM F3001 / etc.
2. Machine qualification — IQ / OQ / PQ per ISO 13485 conventions; calibration to ASTM F3303; laser power meter, scan accuracy (Renishaw QC20 ballbar or equivalent), atmosphere O₂ + dew point check.
3. Process qualification — VED window mapped, coupons built at corners and centre of envelope, tested per ASTM F3122 (tensile + density + hardness + metallography).
4. Part qualification — first-article inspection (FAI) per AS9102; statistical process capability (Cpk ≥ 1.33) on critical-to-quality dimensions; destructive testing per applicable spec.
5. Production control — every build: powder lot record, machine log, in-process monitoring (melt-pool camera, layer photo, O₂ trace), witness coupon tested per build, dimensional sample per build.
6. Release — Certificate of Conformance (CoC) with full traceability chain: powder lot → build job → heat treatment lot → inspection records → operator.

NASA-STD-6030 codifies this for spaceflight; FAA Advisory Circular AC 33.15-3 for engine parts; FDA 21 CFR 820 + ISO 13485 for medical.

6p.6 In-process monitoring

Modern L-PBF machines instrument the build to give a real-time digital twin:

• Melt-pool monitoring (MPM) — co-axial photodiode + high-speed camera looking down the laser path; samples melt-pool intensity at 100 kHz, flags hot/cold anomalies layer-by-layer. EOS EOSTATE MeltPool, GE Concept Laser QM Meltpool, SLM Solutions MPM.
• Layer imaging — each freshly recoated powder layer photographed; ML compares to expected pattern to catch recoater streaks, short-feeds, part curl. EOSTATE PowderBed, SLM LCS, Velo3D Assure.
• Acoustic emission — microphones on the build chamber catch porosity-formation cavitation events; research-grade today, productionising.
• Atmosphere monitoring — O₂ and dew-point sensors interlock the laser if oxygen exceeds setpoint (typically 100–500 ppm).

These data streams feed Velo3D Assure, EOSTATE, Materialise Streamics, and AMFG MES platforms — establishing a digital build record that ships with the Certificate of Conformance for every part.

---

7p.4 Example D — Sand-mold binder jet for cast aluminum manifold

Problem. A 50-unit prototype run of an aluminum intake manifold, 320 × 220 × 180 mm, complex internal runners. Conventional pattern + corebox tooling: ~USD 80 000 lead time 10 weeks. Use a sand-binder-jet mold instead.

Step 1. Mold design. CAD the manifold solid; offset −5 mm for machining stock on critical surfaces; Boolean-subtract from a parting-line mold block; design runner + riser system in the same model. Export 3MF directly to the foundry shop.

Step 2. Print. Voxeljet VX1000, silica sand + furan binder, 280 µm layer, 100 mm/h build height. Mold prints in ~10 h; total of 5 mold packs in one build (50 castings ÷ 1 pour per pack ÷ 10 pours-per-mold-life = 5 packs).

Step 3. Cast. A356.2 Al at 720 °C, gravity-pour into the printed mold; modify with 100 ppm Sr; grain-refine with Al-5Ti-1B. Shake-out, T6 heat treat (solution 540 °C / 8 h, water quench, age 160 °C / 6 h).

Step 4. Machine + inspect. 5-axis CNC the mating flanges; X-ray inspect runners; CMM check. Yield ~88 % (typical for sand cast).

Step 5. Economics. Mold cost: USD 800/pack × 5 = USD 4 000. Casting + post-process: USD 350/unit × 50 = USD 17 500. Total USD 21 500, 4-week lead time. Conventional tooled route: USD 80 000 + USD 12 000 casting = USD 92 000, 14-week lead time. 77 % cost save, 71 % lead-time save — and the design can change between builds at zero tool cost. This is the dominant production use of binder-jet AM today.

---

7p. Worked examples

7p.1 Example A — FDM topology-optimised bracket, CF-Nylon

Problem. Replace a 6061-T6 milled L-bracket (250 g, 100 × 80 × 40 mm) used as a sensor mount on a mobile robot arm. Static load case: 500 N tip load at the end of the 80 mm arm. Target: minimum mass with ≥ 90 % of the original stiffness, fabricated in-house on a Markforged X7.

Step 1. Topology optimisation. In nTopology, fix the four mounting holes, apply 500 N at the load point, solve for minimum-compliance subject to a 40 % volume fraction. The optimiser produces a tree-like web.

Step 2. Smooth and finalise. Use the implicit-body smoothing to remove jagged edges; add 1.5 mm minimum wall, 3 mm fillets at corners. Export 3MF.

Step 3. Material + slicer. Markforged Onyx (PA6 + 20 % chopped CF), 0.20 mm Z, 4 perimeters, 50 % gyroid infill. Add continuous CF reinforcement layers along the principal stress direction — Markforged Eiger places 8 layers of continuous CF along the lattice flanges.

Step 4. Estimate.

• Mass: 105 g (vs 250 g aluminum → −58 %).
• Build time on X7: 6 h 10 min.
• Material cost: ~USD 35 (Onyx + CF tow).
• Machine amortised cost: ~USD 80 (typical 100 k machine over 5 yr at 50 % duty).
• Per-part cost ≈ USD 115, vs ~USD 90 milled (commodity 6061 bracket). Cost parity at the part, but with elimination of CNC programming, fixturing, deburr, and a 10-day lead time. Order-of-one mounting tools at 1 day lead time.

Step 5. Verify. FEA on the printed part with anisotropic Onyx-CF properties (E_xy ≈ 24 GPa with continuous CF flanges, E_z ≈ 7 GPa interlayer) shows 0.32 mm deflection at 500 N, vs 0.30 mm for the aluminum original. 93 % of original stiffness — meets the criterion.

7p.2 Example B — L-PBF Ti-6Al-4V orthopedic spinal cage

Problem. Produce 12 intervertebral spinal fusion cages, 24 × 14 × 11 mm with a 700 µm trabecular lattice core for bone ingrowth. Material Ti-6Al-4V grade 23 ELI (ASTM F3001). Build on an EOS M290.

Step 1. Design. Solid endplates (1 mm thick) on each face for instrument grip and load transfer; gyroid lattice infill at 75 % porosity, 700 µm strut spacing optimised to mimic cancellous-bone modulus (3–8 GPa) and reduce stress shielding vs solid Ti (110 GPa).

Step 2. Build setup. EOS M290, 30 µm layer, 280 W laser, 1200 mm/s scan speed, 100 µm hatch, 67° rotation per layer. 12 cages oriented vertically on the build plate, total volume ~80 cm³, build time ~24 h. Powder: virgin Ti-6Al-4V Grade 23 ELI, 15–45 µm spherical, plasma-atomised, O < 0.13 %, N < 0.05 % (per AMS 4998 Class 1).

Step 3. Post-process.

• Stress-relief 730 °C / 2 h in Ar, on plate.
• Wire EDM off plate.
• Manual support removal under stereo microscope.
• HIP 920 °C / 100 MPa / 2 h in Ar — closes the typical 0.2 % residual porosity to < 0.01 %.
• Chemical etch (HF–HNO₃, very mild) to remove partially-sintered powder from the lattice — critical for medical use, residual powder is unacceptable.
• Passivate per ASTM F86.
• Clean to ISO 13485 specification, package, sterilise by gamma per ISO 11137 (25 kGy).

Step 4. Test and qualify.

• ASTM F2077 (intervertebral body fusion device static and fatigue compression).
• ASTM F1264 sub-clauses as applicable.
• Tensile witness coupons built alongside, tested per ASTM E8: σ_y 870 MPa, σ_u 970 MPa, elong 14 % — within wrought-equivalent envelope.
• CT inspection per ASTM E1570: density > 99.95 %, no defects > 100 µm.
• 510(k) submission to FDA referencing predicate device.

Step 5. Economics. Build cost per cage ≈ USD 350 (powder, machine time, post-process). Conventional machined Ti cage with trabecular surface plasma spray: USD 600–900. L-PBF wins on cost and on lattice fidelity, which is why every major spinal vendor (Stryker, NuVasive, Medtronic, K2M) now ships L-PBF cages.

7p.3 Example C — L-DED repair of aero turbine blade tip

Problem. Service-worn Inconel 718 HPC turbine blade, tip eroded by 5 mm and 0.5 mm out-of-round. Original blade cost USD 25 000. Repair scheme certified under FAA-PMA (Parts Manufacturer Approval), AMS 5663 material spec.

Step 1. Prep. Strip damaged tip back to clean parent metal by 5-axis grind (5.0 mm material removed, perpendicular tip face). Fluorescent penetrant inspection per ASTM E1417 to confirm no cracks remain. Mount in fixture in Optomec LENS 860 build chamber.

Step 2. Deposit. 5-axis L-DED, 1 kW Yb-fibre laser, IN718 powder 45–105 µm, 6 g/min feed rate, Ar shield + carrier. Deposit 5 mm of new tip plus 1 mm machining stock. Total deposition time: ~12 min. Continuous in-process pyrometry of melt pool to maintain 1380 ± 30 °C.

Step 3. Heat treatment. AMS 5663 solution + age cycle: solution 980 °C / 1 h / air cool, double age 720 °C / 8 h / FC to 620 °C / 8 h / air cool. Restores γ′ / γ″ precipitate microstructure to wrought-equivalent.

Step 4. Finish. 5-axis CNC machine to final tip profile, ±50 µm dimensional. Tip clearance gauge check, root-to-tip length within drawing tolerance. CMM full inspection. Fluorescent penetrant on finished part.

Step 5. Economics. Repair labour + DED + heat treat + machine ≈ USD 6 500. Replacement blade ≈ USD 25 000. 74 % cost save; 4-week shop turnaround vs 12-week OEM new-blade lead time. Several engine MROs (StandardAero, MTU Aero Engines) run hundreds of LENS tip repairs per year on commercial CF6, CFM56, V2500 fleets.

---

8p. Edge cases and gotchas

• Anisotropy underestimated. Z-strength 60–95 % of XY depending on process; static-load designs and fatigue lives both suffer if a critical principal stress aligns with Z.
• As-built surface is rarely usable. Plan for surface finishing on any sealing, bearing, sliding, or mating surface. PBF Ra 6–15 µm is the floor, not the ceiling.
• Internal porosity. L-PBF parts contain 0.05–0.5 % residual porosity as-built; lack-of-fusion pores are the worst. HIP closes pores only if not surface-connected — vacuum-tight parts must be electropolished or sealed before HIP if surface-connected porosity exists.
• Residual stress. Ti and steel L-PBF: ±200 to ±800 MPa locally. Causes plate-curl, edge lift (recoater crash), distortion after plate removal. Stress-relief is mandatory before removing supports.
• Recoater crash. Curl + recoater blade contact mid-build will tear off the part and (worst case) damage the recoater. Mitigations: stress-balanced support layouts, simulation-driven part orientation, soft-blade recoaters.
• Witness lines / staircasing. Layer boundaries are stress concentrators and fatigue-crack initiators. Surface polish (electropolish, AFM, Hirtisation) addresses external; internal channels are harder.
• Powder hygiene. Reactive metal powders (Ti, Al, Mg) absorb oxygen and nitrogen over recycle cycles. Specifications limit O (Ti AMS 4998 Class 4: O ≤ 0.13 %, N ≤ 0.05 %). Powder lot traceability per AS9100 / ISO 13485 is non-negotiable in qualified production.
• Build-plate adhesion failure (FDM): warping pulls the part off the bed. Mitigations: heated bed, brim/raft, enclosure for ABS/PC/PEEK, PEI/garolite/glass bed surfaces.
• Calibration drift. Laser power, scan accuracy, layer thickness all drift. ASTM F3303 is the calibration baseline. Periodic test coupons + machine logs are part of any qualified facility’s quality system.
• Powder bed cross-contamination. Switching alloys requires full machine purge — Ti powder traces in an Al-Si build will detonate; CoCr traces in a Ti build will ruin the implant’s biocompatibility certification. Dedicated machines per alloy is the medical-device norm.
• Qualification & traceability. ASTM F3122 (mechanical test of AM metals), ASTM F3303 (process qual), ISO/ASTM 52920 (qualification principles), NASA-STD-6030 / NASA-HDBK-6030 (spaceflight). Per-build witness coupons, full lot traceability of powder → build → part → heat treat → inspection → release.
• ITAR / export control. Build files, scan parameters, and qualified material specs for military or aerospace parts are export-controlled (ITAR/EAR). Counterfeit AM parts entering MRO supply chains is a growing serious concern (DoD reports increasing instances).
• IP / file distribution. STL files leak easily. Encrypted-on-machine workflow (Identify3D, GrabCAD Print encrypted print packages) is the emerging norm for distributed manufacturing.
• Cost trap. AM is cheaper per part than CNC only in low-volume + complex geometry. For high-volume, simple geometry, casting + machining wins. Always do a unit-economics break-even (typically AM wins below 100–1000 units depending on size and material).
• Volume scaling. AM cycle time scales roughly with part height (in layers) and bed coverage; doubling part count doubles build time unless nest density rises. Casting/molding cycle time is largely independent of part count per cycle. The crossover quantity is the key business decision.

---

9p. Tools and software ecosystem

9p.1 Hardware — major machine families

Process	Vendor / family	Notes
Polymer MEX (industrial)	Stratasys F123/F900/Fortus 450mc, Markforged X7/X10/FX20, Roboze Argo, Essentium HSE	Ultem/PEEK and continuous-CF specialty
Polymer MEX (prosumer)	UltiMaker S5/S7, Raise3D Pro 3, Bambu Lab X1C, Prusa MK4S / XL	$1–10 k, engineering thermoplastics
Polymer VPP	Formlabs Form 4 / 4L, EnvisionTEC P/Vector, Carbon M2/M3 (DLS), 3D Systems Figure 4, Anycubic, Phrozen	Carbon DLS dominates digital footwear (Adidas 4D, New Balance)
Polymer SLS	EOS P396 / P770 / P810, Farsoon HT/Flight, Sintratec S2, Formlabs Fuse 1+ / Fuse Blast	Workhorse end-use polymer
Polymer BJT	HP Multi Jet Fusion (5210, 5420W), Voxeljet	HP MJF dominant production polymer AM
Metal L-PBF	EOS M290 / M400, SLM Solutions NXG XII (12 lasers), GE Additive Concept Laser M2 / M Line / X-Line 2000R, Trumpf TruPrint 3000/5000, Renishaw RenAM 500S/Q, Velo3D Sapphire / XC, 3D Systems DMP Flex/Factory, AddUp FormUp, Aconity3D, Nikon SLM	Velo3D unique for supportless deep-cavity Ti
Metal EB-PBF	GE Additive Arcam Spectra L/H, Wayland Additive Calibur3, Freemelt ONE	Orthopedic Ti, γ-TiAl turbine
Metal DED (laser-powder)	Optomec LENS 860/1500, RPM Innovations 557Xtreme, BeAM (AddUp Modulo), Trumpf TruLaser Cell, DMG Mori Lasertec 65	Repair, cladding, gradient
Metal DED (wire-arc/WAAM)	Lincoln SCULPT, MX3D M1, Gefertec arc603, WAAM3D	Large near-net-shape preforms
Metal DED (EB-wire)	Sciaky EBAM	Largest Ti AM parts (5 m+)
Metal BJT	Desktop Metal Production System P-50, ExOne S-Max + InnoventX, Digital Metal	High-throughput metal alternative to PBF
Sand BJT	ExOne S-Max Pro, Voxeljet VX1000/VX4000	Foundry mold/core packs
CT inspection	Nikon XT H 225/450, Zeiss Metrotom 1500, Yxlon CT Compact, GE Phoenix v	tome

9p.2 Software

• Design / topology / lattice: nTopology, Altair Inspire, ANSYS Discovery, PTC Creo Generative, Autodesk Fusion 360 Generative, Materialise 3-matic.
• CAD (general): SOLIDWORKS, NX, Creo, CATIA, Fusion 360, Onshape.
• Slicers / build prep (metal): Materialise Magics + Build Processors, Autodesk Netfabb, EOSPRINT 2, 3DXpert, GE Concept Print, Velo3D Flow.
• Slicers (polymer): PrusaSlicer, Bambu Studio, UltiMaker Cura, Formlabs PreForm, Markforged Eiger, HP SmartStream.
• Process simulation: ANSYS Additive Print/Suite, Autodesk Netfabb Simulation, Materialise Process Simulation, Velo3D Flow, Atlas3D Sunata.
• MES / digital thread: Materialise Streamics, Identify3D, 3YOURMIND, AMFG, Link3D (Markforged).

---

10. Cross-references

• [[Engineering/materials-polymers]] — base polymer mechanical, thermal, and creep behaviour relevant to FDM/SLS/SLA grades.
• [[Engineering/materials-steel]] — wrought-steel baselines that L-PBF 316L / 17-4 PH / H13 / tool steels must equal or surpass.
• [[Engineering/materials-aluminum]] — wrought 6061/7075 baseline vs L-PBF AlSi10Mg.
• [[Engineering/materials-composites]] — continuous-fibre FDM, prepreg layup as a competing process.
• [[Engineering/materials-selection]] — Ashby-chart positioning of AM-feasible alloys.
• [[Engineering/mechanics-of-materials]] — anisotropic stress analysis fundamentals for layer-built parts.
• [[Engineering/heat-transfer]] — conformal cooling channel design, lattice heat exchangers.
• [[Engineering/fasteners-bolts]] — printed-thread reliability and printed insert reinforcement.
• [[Languages/Tier3/3d-scene]] (planned) — STL, 3MF, AMF, build-file (CLI, SLM, SLI) syntax.
• [[Languages/Tier3/construction-bim]] (planned) — STEP-AP242 and the CAD-to-AM data thread.
• [[Robotics/end-effectors]] (planned) — printable soft and rigid grippers, mass-customised tooling.

---

11. Citations

• Gibson, I.; Rosen, D.; Stucker, B.; Khorasani, M. Additive Manufacturing Technologies: 3D Printing, Rapid Prototyping, and Direct Digital Manufacturing, 3rd ed., Springer, 2021. The canonical text.
• Milewski, J. O. Additive Manufacturing of Metals: From Fundamental Technology to Rocket Nozzles, Medical Implants, and Custom Jewelry, Springer, 2017.
• Yang, L.; Hsu, K.; Baughman, B.; Godfrey, D.; Medina, F.; Menon, M.; Wiener, S. Additive Manufacturing of Metals: The Technology, Materials, Design and Production, Springer, 2017.
• ASM Handbook Vol. 24 Additive Manufacturing Processes, ASM International, 2020. ASM Handbook Vol. 24A Additive Manufacturing Design and Applications, 2023.
• Wohlers Associates, Wohlers Report 2024: 3D Printing and Additive Manufacturing — Global State of the Industry. Annual industry-state survey.
• ISO/ASTM 52900:2021 — Additive manufacturing — General principles — Terminology.
• ISO/ASTM 52910:2022 — Additive manufacturing — Design — Requirements, guidelines and recommendations.
• ISO/ASTM 52915:2020 — Specification for additive manufacturing file format (AMF) Version 1.2.
• ISO/ASTM 52920:2023 — Additive manufacturing — Qualification principles — Requirements for industrial additive manufacturing processes and production sites.
• AWS D20.1/D20.1M:2024 — Specification for fabrication of metal components using additive manufacturing.
• SAE AMS 7000-series — PBF metals specifications (AMS 7003 316L L-PBF, AMS 7008 Ti-6Al-4V L-PBF, etc.).
• ASTM F3122 — Standard guide for evaluating mechanical properties of metal materials made via additive manufacturing.
• ASTM F3001 — Standard specification for additive manufacturing Ti-6Al-4V ELI with powder bed fusion.
• ASTM F2924 — Standard specification for additive manufacturing Ti-6Al-4V with powder bed fusion.
• ASTM F2077 — Test methods for intervertebral body fusion devices.
• ASTM F3303 — Standard for additive manufacturing — Process characteristics and performance — Practice for metal powder bed fusion process to meet critical applications.
• NASA-STD-6030 — Additive manufacturing requirements for spaceflight systems.
• GE Aviation LEAP fuel-nozzle case study — multiple papers; Kellner, T. (GE Reports, 2017).
• Norsk Titanium Rapid Plasma Deposition for Boeing 787 — first FAA-certified structural Ti AM part (2017).
• Frazier, W. E. Metal Additive Manufacturing: A Review, J. Materials Engineering and Performance 23 (2014) 1917–1928.