MODULE 1
From Prototype to Production

From Digital Model to Physical Part

Additive, subtractive & hybrid routes to a finished component
FramtidR | Advanced Manufacturing Technologies
The map
Three routes to the same part
Build it up, cut it down, or do both — no single method wins every time.
  • Additive (AM). Add material layer by layer from CAD — complex shapes, rough finish.
  • Subtractive (CNC). Cut material from a solid block — tight fit, but more waste.
  • Hybrid. Add and cut in one setup — great for repair.
  • The choice depends on shape, material, volume, tolerance, and cost.
Shape freedomSpeed / costPrecision / finishAMCNCHybrid
Figure: the trade-off triangle — where each method leans.
FramtidRFrom Digital Model to Physical Part2 / 29
Part 1

Additive

Building parts one layer at a time
FramtidRFrom Digital Model to Physical Part3 / 29
Layer by layer
Building parts from a digital model
AM stacks thin slices of material until the full shape exists.
  • Steps. CAD → slice → machine file → build → finish.
  • Materials. Plastics, metals, ceramics, and bio-materials.
  • Why engineers like it. Inner channels, fewer parts, graded structures.
  • Cheap for small batches — no costly tooling needed.
CADSliceBuildfilePrintPost-processstacked layers
Figure: the digital-to-physical pipeline.
FramtidRFrom Digital Model to Physical Part4 / 29
ISO/ASTM 52900
One standard, seven ways to lay down a layer
ISO/ASTM 52900 sorts all AM into seven groups by how each layer is made.
  • MEX / VPP / MJT. Squeezed paste, cured resin, jetted droplets.
  • PBF. Heat fuses a bed of powder — metals.
  • DED. Material is melted as it is laid down — metals.
  • BJT / SHL. Glued powder; bonded sheets — metals.
  • The metal workhorses: PBF, DED, BJT, and metal MEX.
AM (52900)MEXVPPPBFBJTMJTDEDSHLmetal-capable
Figure: the seven-category process tree (dark = metal).
FramtidRFrom Digital Model to Physical Part5 / 29
Powder Bed Fusion
Melting a bed of powder
A laser (LPBF) or electron beam (EBM) melts the layer; the bed drops; then repeat.
  • LPBF (SLM / DMLS). Fine detail, many alloys, inert argon gas.
  • EBM. Electron beam in vacuum; hotter bed → less leftover stress.
  • EBM is preferred for. Reactive titanium and set-porosity parts.
  • Great detail, but slow; you must manage stress and pores.
inert argon gasscan mirrorbeampowder bedpartsupportsbuild platerecoatermelt pool
Figure: LPBF chamber cross-section.
FramtidRFrom Digital Model to Physical Part6 / 29
Directed Energy Deposition
Depositing into a moving melt pool
Feed (powder or wire) is melted as it lands — bead by bead.
  • Laser/powder DED (LMD). Coatings and graded materials.
  • Wire DED / WAAM. Very fast build — kilograms per hour.
  • Near-net shape. Not for fine detail — cut to size after.
  • Can add material onto an existing part → the basis for repair.
headpowder /wiremelt pooldeposited beadsubstratetraverse
Figure: DED deposition head — powder or wire fed coaxially.
FramtidRFrom Digital Model to Physical Part7 / 29
Sinter-based routes
Print first, bake later
Both follow the same chain — green part → debind → sinter — and parts shrink as they get dense.
Binder Jetting (BJT)
no melting while printing
  • Inkjet glue onto a metal powder bed
  • Weak green part → debind → sinter
  • Fast and parallel — good for volume
Metal MEX (bound-metal)
metal-filled filament
  • Print metal-powder plastic filament
  • Green part → debind → sinter
  • Cheaper and safer than laser PBF
Parts shrink a lot, and evenly — so the design and fixtures must allow for it.
FramtidRFrom Digital Model to Physical Part8 / 29
Feedstock
Feedstock decides the outcome
Same process, different result — the form of the feed sets what the machine can do.
  • Atomized powder. Size, flow, and oxygen set density and defects.
  • Welding-grade wire. Cheap, traceable, fast build; coarser shape.
  • Bound-metal filament. Easy to handle, then debind + sinter.
  • Common alloys. Ti-6Al-4V, Inconel 625/718, 316L/17-4PH, AlSi10Mg.
  • Feed is a checked, qualified input — tracked by lot, with O₂ control.
powdercontrol O₂wire spoolhigh kg/hfilamentdebind+sinterCommon alloysTi-6Al-4V · IN625/718 · 316L17-4PH · AlSi10Mg · maraging
Figure: three feedstock forms and the common alloys.
FramtidRFrom Digital Model to Physical Part9 / 29
DfAM
Designing for a printer, not a mill
Design for how the machine really builds, not just for the final shape.
  • Build direction. Sets surface quality, strength by direction, and support load.
  • Overhangs. Parts that lean below ~45° usually need supports.
  • Topology & lattices. Put material only where the load goes — lighter parts.
  • Fewer parts. Replace a whole assembly with one printed part.
30° needssupport60° self-supporting45°solidoptimizedsame load, less mass
Figure: 45° overhang rule, and topology optimization.
FramtidRFrom Digital Model to Physical Part10 / 29
Post-processing
What happens after the build stops
A metal AM part is not finished when it leaves the machine.
1
Stress relief
heat treat before removal
▶
2
Cut & deburr
wire EDM / saw, remove supports
▶
3
HIP
close inner pores
▶
4
Finish-machine
key surfaces to size
▶
5
Inspect
CT, CMM, surface checks
After-work often costs more time and money than the print itself.
FramtidRFrom Digital Model to Physical Part11 / 29
HIP
Pressing the pores closed
Hot Isostatic Pressing (HIP) applies high heat and even gas pressure at the same time.
  • Pressure. Argon gas ~100 MPa (≈15,000 psi); range 50–310 MPa.
  • Heat. ~480 °C (aluminum) up to ~1,320 °C (nickel superalloys).
  • How it works. The hot metal flows and knits across each void → ~100% density.
  • Boosts fatigue life — often required for flight-critical AM parts.
before: poresafter: solidAr, ~100 MPa, T = alloy-dependent
Figure: isostatic pressure collapses voids while hot.
FramtidRFrom Digital Model to Physical Part12 / 29
Part 2

Subtractive

Cutting a part out of solid stock
FramtidRFrom Digital Model to Physical Part13 / 29
Orthogonal cutting
Cutting a part out of solid stock
CNC cuts metal by bending and shearing it — a chip forms and breaks away.
  • Cutting speed. How fast the edge moves over the metal.
  • Feed. How far the tool moves per turn or tooth.
  • Depth of cut. Together these set removal rate, heat, force, and finish.
  • A mature process: very tight fits and a fine finish right away.
workpiecespeed, feed, depthtoolrake faceflankchipshear planeheat zone
Figure: orthogonal cutting — shear plane and chip.
FramtidRFrom Digital Model to Physical Part14 / 29
Tool wear modes
Why cutting edges wear out
Heat and sticking speed up all three — tool life, finish, and cost rise and fall together.
◢
Flank wear
the flank rubs and wears down; this steady wear sets tool life
◯
Crater wear
the chip wears a dip in the rake face when it gets very hot
▲
Built-up edge
metal sticks to the edge, then breaks off and spoils the finish
FramtidRFrom Digital Model to Physical Part15 / 29
Difficult-to-machine
The alloys that fight back
Titanium and nickel superalloys stay strong when hot — and that same strength wrecks cutters.
  • Poor heat flow. Heat stays at the edge → fast wear.
  • Strong when hot. High cutting forces even at high heat.
  • Reactive and hardens. Metal sticks to the edge (built-up edge) and notches it.
  • Titanium is usually cut slowly (carbide roughly below ~60 m/min).
ordinary steelheat spreads outTi / Inconelheat stays at the edge
Figure: heat spreads in steel but stays trapped at the edge in superalloys.
FramtidRFrom Digital Model to Physical Part16 / 29
Countermeasures
Winning the cut
Hold tight, probe, reach, and cool — to control force, access, and heat.
◫
Rigid workholding
Stop shake and chatter; superalloys punish any flex
◎
In-machine probing
Set reference points and check sizes without unclamping
⟳
5-axis toolpaths
Reach hard surfaces in one setup, with the best tool angle
❄
Strong cooling
High-pressure coolant and very cold LN₂ (~−196 °C)
FramtidRFrom Digital Model to Physical Part17 / 29
Part 3

Hybrid

One machine that both builds and cuts
FramtidRFrom Digital Model to Physical Part18 / 29
Hybrid loop
One machine that both builds and cuts
One 5-axis machine holds both a DED head and a milling spindle — the part never leaves the table.
1
Deposit (DED)
add near-net metal
▶
2
Inspect
probe in-machine
▶
3
Mill to tol.
cut key surfaces
▶
4
Reorient
no re-clamping
↺ Repeats each cycle — finish inner surfaces before the next layer buries them.
FramtidRFrom Digital Model to Physical Part19 / 29
Wire vs powder
The case for feeding wire
Wire DED (including wire-arc / WAAM) uses normal welding wire.
  • Cheap feed. Standard welding wire, easy to buy.
  • Almost no waste. Nearly all the wire ends up in the part.
  • Safer. No fine, reactive powder cloud to contain.
  • Fast build. Big near-net shapes, quickly.
  • The cost: a rougher finish → lean on the built-in mill.
Wire-fed DED~100% usedno powder cloudhigh kg/hcoarser finishPowder Bed Fusion!recover/re-qualifyfine detailslowpowder hazard
Figure: wire-vs-powder scorecard — a trade, not a winner.
FramtidRFrom Digital Model to Physical Part20 / 29
Remanufacturing
Rebuilding worn parts, not scrapping
Costly parts (e.g. turbine blades with worn tips) hold huge material and energy value — rebuild, don't scrap.
1
3D scan
map the worn part
▶
2
DED rebuild
add metal to tip
▶
3
5-axis mill
machine to size
▶
4
Inspect
prove it meets spec
↺ Circular making — extend a part's life in one setup: less waste, less energy, shorter lead time.
FramtidRFrom Digital Model to Physical Part21 / 29
Part 4

Prototype → Production

Proving, costing & choosing the route
FramtidRFrom Digital Model to Physical Part22 / 29
Metrology
Proving the part
A part is finished only when you can prove it meets spec — measured, not assumed.
◎
CMM & probing
Sizes, GD&T, and feature tolerances from reference points
◉
X-ray CT
Inner pores, unfused spots, and trapped powder
∿
Surface (Ra)
Surface-roughness checks on critical faces
▯
Steady process
Repeated good results, not just one good part
FramtidRFrom Digital Model to Physical Part23 / 29
Economics
Counting the cost: break-even by volume
Cost per part is part fixed and part variable, and the winner flips with volume.
  • AM. No tooling, slow build → cheap at low volume, costly at high volume.
  • CNC. Small setup, fast cycle → strong for small-to-mid volumes.
  • Casting/molding. High tooling up front → cheapest only at high volume.
  • AM's hidden cost: after-work often costs more than the print itself.
cost / partquantityAMCNCcasting/moldingAM bestCNC besttooled bestexample only, not to scale
Figure: cost per part vs. quantity — two break-even points.
FramtidRFrom Digital Model to Physical Part24 / 29
Decision guide
Choosing a route
Work the five questions in order — most real parts are a sequence, not one box.
Need ametal partGeometry?Tolerance/finish?Volume?Material/repair?New orrestore?AdditiveSubtractiveTooledDED/HybridHybrid-repairUsually a sequence, not one box: build near-net → finish-machine → inspect.
Figure: process-selection decision tree.
FramtidRFrom Digital Model to Physical Part25 / 29
The gap
A lab part is not yet a production part
A part that prints in the lab is not a part you can ship — each gap below is real engineering work.
  • ✓Tolerance as-built parts drift (shrink, warp); production needs microns→ finish-machine
  • ✓Surface finish as-built metal AM can be tens of microns Raoften not good enough as-is
  • ✓Pores / defects unfused spots and gas pores cut fatigue life→ HIP + inspection
  • ✓Qualification one good part is not a steady, repeatable process
  • ✓Speed & cost AM is slow and machine-time is costly at volume
  • ✓Design trade-offs the design must suit the whole chain, not just the printer
FramtidRFrom Digital Model to Physical Part26 / 29
Recap
Pulling it together & looking ahead
  • Seven AM groups; four make metal. PBF, DED, BJT, metal MEX.
  • Subtractive = shear. Superalloys keep their heat and their strength — so they fight the cutter.
  • Hybrid. AM shape freedom + cutting precision; wire-fed repair drives reuse.
  • The real work. Prototype → production: tolerance, finish, pores, qualifying, cost.
  • Next. Measuring and proving it — metrology, inspection, and linked digital data.
Making a part is needed but not enough — production is a checked, repeatable, low-cost system.
FramtidRFrom Digital Model to Physical Part27 / 29
Further Reading
Additive
  • The Seven AM Processes — Wohlers Associates
  • Additive manufacturing, explained — MIT Sloan
  • EBM vs L-PBF — 3Dnatives
  • Wire Arc Additive Manufacturing overview — ScienceDirect
  • Topology Optimization for Metal AM — IntechOpen
Subtractive
  • Machining Additively Manufactured Ti-6Al-4V — PMC / NIH
  • Cryogenic Cooling of Difficult-to-Machine Alloys — PMC / NIH
Hybrid & remanufacturing
  • Additive/subtractive hybrid manufacturing of DED — ScienceDirect
  • Turbine Blade Failures Repaired by Laser Remanufacturing — Materials (MDPI)
  • Circular economy principles — Ellen MacArthur Foundation
Standards, post-process & open
  • ISO/ASTM 52900:2021 — ISO/ASTM
  • Hot Isostatic Pressing — Wikipedia + Metal AM (Inovar)
  • MIT OpenCourseWare — CC BY-NC-SA
FramtidRFrom Digital Model to Physical Part28 / 29
⚠

Before you rely on this

This series was made by FramtidR with the help of AI. AI can make mistakes, including about people. Please check this information before you use it.

Made with AI help • checked by FramtidR
FramtidRFrom Digital Model to Physical Part29 / 29