AM parts fail differently, so checking them the same way as a machined part misses the point
A 3D printed part doesn’t fail the way a machined or cast part does. There’s no tool wear, no core shift, no shrinkage in the way a mould shrinks. Instead you get layer-related porosity from incomplete fusion between passes, surface roughness on unsupported overhangs that never touched a support structure, and dimensional drift that changes depending on which axis the part was built along. Verifying an AM part with the same checklist you’d run on a machined equivalent misses most of what actually goes wrong with it.
Lattice structures and internal channels: why CT is the only option
Lattice geometry is often the entire reason a part gets printed instead of machined, weight reduction through an internal structure that would be physically impossible to cut into solid material. That same geometry is also physically impossible to probe or scan, since a lattice’s internal struts have no external access at all once the part is built. METROTOM captures the full lattice structure non-destructively, strut thickness, node connections, and any incomplete fusion sitting inside the structure, all in one scan. The same wall thickness principles that apply to a cast housing carry over here, except an AM lattice adds a layer of complexity a solid casting never has, hundreds of individual struts each needing their own thickness verified rather than a handful of walls on a simpler part.
Powder porosity isn’t the same defect as casting porosity
CT scanning finds voids in both a casting and a printed part, but the two porosity types come from different causes and need different acceptance thinking. Casting porosity forms from trapped gas or shrinkage during solidification, typically clustered near thick sections or gates. AM porosity forms from incomplete fusion between powder layers or gas trapped in the powder itself before it ever melts, and it tends to distribute differently through the part depending on laser or beam parameters rather than clustering near an obvious geometric feature. Treating AM porosity acceptance criteria as identical to a casting’s is a mistake worth avoiding, the root cause and the resulting risk profile aren’t the same, even though both show up as a void on a CT scan.
Build orientation changes the part’s own dimensional accuracy
A part printed standing upright measures differently to the same part printed lying flat, purely because of how thermal stress accumulates layer by layer during the build. This is genuinely unique to additive manufacturing, a machined or cast part doesn’t carry this kind of orientation-dependent drift, since it isn’t built up one layer at a time with each layer partially reheating the one beneath it. Dimensional verification on a new AM part should always capture which orientation it was printed in, since the same nominal design can come off the printer with a different accuracy profile purely from a build plate rotation nobody flagged.
Capturing the geometry: 3D scanner for complex surfaces, CMM for functional interfaces
AM parts frequently combine organic, freeform surfaces with a handful of precisely engineered functional features, a mounting hole, a bearing bore, a bolt pattern that has to mate with an existing assembly. That split determines the tool. ATOS Q or GOM Scan 1 handle the freeform surface work, capturing the full external geometry in one pass and comparing it against nominal CAD to catch warping or surface roughness issues across the whole part rather than at a handful of probed points. Where the part has genuine functional interfaces, a CMM measures those features directly, since a bore or a bolt pattern is a discrete geometric feature a probe handles more precisely than a scan.
Verifying the first part off a new build, not just the process
A new print job, whether it’s a new part design or a new machine parameter set, needs the same discipline as any other new production process. First article inspection principles apply directly here, verify the first part completely against the drawing before committing a full build to the same parameters, since a small deviation in laser power or layer thickness can produce a part that looks fine externally while carrying internal porosity that only shows up once someone actually checks for it. GD&T tolerances on functional features matter just as much on a printed part as a machined one, position and profile callouts don’t relax just because the part came off a printer instead of a mill.
Bring the part in, and tell us what it’s actually printed from
Whether it’s a lattice structure that needs its internal geometry verified, a batch that needs porosity checked against the right acceptance criteria, or a first article off a new build that needs clearing before full production starts, the inspection method has to match what actually needs checking. Global TechSolutions runs METROTOM for the internal geometry, ATOS Q and GOM Scan 1 for surface capture, and CMM for functional interfaces, as the ZEISS Partner Metrology Centre in Singapore. Our plastic manufacturing and metal processing pages both cover additive manufacturing as part of their wider application areas if you’re working across materials rather than a single build type.
Contact Us for a Free Demo
Whether you need a live demo with your own parts or expert measurement services, our specialists are ready to support your quality assurance needs.