Addithive

Tag: medicine

  • Nondestructive Testing for Additive Manufacturing: Methods, Defects and Limitations

    Nondestructive Testing for Additive Manufacturing: Methods, Defects and Limitations

    Nondestructive testing (NDT) is essential for additive manufacturing, but no single method can guarantee that a complex AM part is defect-free. Inspection capability depends on the process, alloy, geometry, surface condition, defect type, defect orientation, required resolution and acceptance criteria.

    The correct question is not “Which NDT method is best for AM?” It is “Which combination of methods can detect the credible defects in this specific part with the required probability of detection?”

    Why AM inspection is different

    Metal additive manufacturing can produce internal channels, lattices, thin walls and highly integrated geometries that are difficult to inspect using methods developed for simple wrought or machined parts. AM defects can also be small, irregular, directionally oriented and distributed differently across the build.

    Inspection planning therefore begins with the complete manufacturing route: feedstock, machine, process parameters, build orientation, heat treatment, hot isostatic pressing, machining and surface finishing. Post-processing can close, reveal, reshape or remove indications, so the inspection stage matters.

    Common defect and imperfection classes

    ImperfectionTypical causeInspection challenge
    Lack of fusionInsufficient energy, poor overlap, contamination or unstable powder layerOften planar and orientation-sensitive
    Gas porosityEntrapped gas, powder condition or melt-pool behaviorSmall rounded pores may require high volumetric resolution
    Keyhole porosityExcessive energy density and unstable deep melt poolMay appear as irregular or elongated pores
    CracksResidual stress, hot cracking, thermal cycling or alloy sensitivityThin planar cracks can be difficult to detect when poorly oriented to the inspection beam
    Inclusions or contaminationForeign material, oxide, spatter or handling contaminationDetectability depends on density contrast and size
    Dimensional deviationShrinkage, distortion, support failure, thermal behavior or post-processingInternal geometry may be inaccessible to conventional metrology
    Surface-connected discontinuitiesSupport removal, machining damage, cracking or incomplete fusionRough as-built surfaces can create false or masked indications
    Trapped powderInsufficient escape paths or ineffective cleaningMay be hidden inside channels and cavities

    X-ray computed tomography

    Industrial X-ray computed tomography (CT) is one of the most powerful tools for AM because it can reconstruct internal and external geometry in three dimensions. It can detect porosity, lack-of-fusion regions, inclusions, dimensional deviations, trapped powder and inaccessible internal features.

    CT is not unlimited. Detectability depends on voxel size, focal spot, detector, material density, wall thickness, part diameter, scan geometry, reconstruction and analysis settings. A large dense nickel-alloy part cannot be inspected at the same resolution as a small aluminum coupon. CT resolution claims must be connected to the actual part and minimum defect size.

    • Strengths: volumetric data, internal geometry, pore distribution and dimensional comparison.
    • Limitations: cost, scan time, penetration, artifacts, resolution versus part size and interpretation complexity.
    • Best practice: validate the technique using representative artifacts or seeded flaws and document the scan and analysis parameters.

    Conventional radiography

    Two-dimensional radiography can detect volumetric discontinuities and density variations, but it compresses three-dimensional information into a projection. Overlapping features and complex geometry can mask defects. Planar flaws aligned unfavorably to the beam may be difficult to see.

    Radiographic inspection setup
    Radiographic testing

    Ultrasonic testing

    Ultrasonic testing can detect internal cracks, lack of fusion and other discontinuities in suitable geometries. Phased-array and advanced full-matrix techniques can improve coverage and imaging. However, rough surfaces, thin sections, complex curvature, internal channels and anisotropic microstructures can complicate coupling, wave propagation and signal interpretation.

    Machined inspection surfaces or purpose-designed access may be needed. Calibration blocks and reference reflectors should represent the alloy, heat treatment, geometry and expected defect orientation as closely as practical.

    Ultrasonic inspection of an aerospace component

    Liquid penetrant testing

    Liquid penetrant testing is effective for surface-breaking discontinuities on nonporous, clean surfaces. As-built AM roughness can retain penetrant and generate excessive background. The method is often more reliable after machining or surface finishing, when the inspection surface and cleaning process are controlled.

    Penetrant testing cannot detect sealed internal defects and should not be treated as evidence of volumetric integrity.

    Liquid penetrant inspection

    Magnetic particle testing

    Magnetic particle testing can reveal surface and near-surface discontinuities in ferromagnetic alloys. It is not applicable to titanium, aluminum, austenitic stainless steels or most nickel alloys. Surface roughness, geometry and residual magnetism must be controlled.

    Magnetic particle inspection

    Eddy current testing

    Eddy current methods detect surface and near-surface discontinuities in electrically conductive materials. They can be sensitive to small cracks, but probe access, lift-off, curvature, roughness, conductivity variation and geometry affect performance. Eddy current inspection is generally local rather than a complete volumetric method.

    Eddy current induction principle

    Optical and dimensional inspection

    Coordinate measuring machines, structured-light scanners, laser scanners and optical microscopy verify dimensional and surface requirements. They do not replace volumetric NDT. Line-of-sight systems cannot measure hidden channels, and highly reflective or rough surfaces may require preparation or specialized scanning strategies.

    In-situ monitoring is not final NDT

    Melt-pool sensors, layer imaging, recoater monitoring, acoustic signals and machine logs can identify process anomalies. These data improve traceability and may support adaptive control. However, an anomaly signal is not automatically a verified defect, and the absence of an alarm does not prove that the part is acceptable.

    In-situ monitoring must be correlated with destructive testing, NDT and production outcomes before it can support acceptance decisions. It is best viewed as one layer in a broader process-control and inspection strategy.

    Probability of detection and validation

    A method may detect a large laboratory defect without reliably detecting the smallest critical defect in production. For safety-critical applications, inspection capability should be demonstrated using representative part thickness, geometry, alloy, surface condition and defect type.

    • Define the minimum relevant defect size and orientation.
    • Use representative reference standards, test artifacts or intentionally seeded flaws.
    • Control equipment, calibration, software, analysis thresholds and operator qualification.
    • Document false-call risk and inspection blind zones.
    • Revalidate the method when geometry, material, surface or equipment changes materially.

    How to build an AM inspection plan

    1. Identify critical functions and credible failure modes.
    2. Map likely imperfection types to the AM process and post-processing route.
    3. Define inspection zones and required detection capability.
    4. Select complementary surface, dimensional and volumetric methods.
    5. Design inspection access into the part where possible.
    6. Validate methods on representative artifacts or seeded flaws.
    7. Set acceptance criteria based on engineering significance, not merely visibility.
    8. Link results to build records, material genealogy and configuration control.

    Quick method comparison

    MethodBest atMain limitation in AM
    X-ray CTInternal geometry and volumetric defectsResolution, penetration, artifacts, cost and part-size trade-off
    RadiographyVolumetric density changes in suitable geometriesFeature overlap and limited 3D localization
    Ultrasonic testingInternal cracks and planar defects with suitable accessRough surfaces, complex geometry and anisotropic propagation
    Liquid penetrantSurface-breaking defectsAs-built roughness and no subsurface capability
    Magnetic particleSurface/near-surface defects in ferromagnetic materialsMaterial limitation and surface sensitivity
    Eddy currentSmall surface/near-surface cracks in conductive materialsLocal access, lift-off and geometry sensitivity
    Optical/CMMExternal dimensions and visible surface conditionNo internal volumetric capability
    In-situ monitoringProcess anomalies and traceabilityRequires correlation; not direct proof of final part integrity

    Conclusion

    Reliable AM inspection combines process knowledge, complementary NDT methods and validated detection capability. Complex geometry does not make a part uninspectable by definition, but it can create blind zones that must be understood during design. The inspection plan should be developed with the part and manufacturing route, not added after printing.

    Related Addithive research: Aerospace AM Qualification Guide · Design for LPBF

    References and further reading

  • Surface Finishing for Metal Additive Manufacturing: Methods, Selection and Design Rules

    Surface Finishing for Metal Additive Manufacturing: Methods, Selection and Design Rules

    Metal additive manufacturing can produce geometries that conventional processes cannot, but the as-built surface is rarely the final engineering surface. Roughness, partially fused particles, stair-stepping, support scars, down-facing irregularities and near-surface defects can affect fatigue, sealing, friction, flow, cleaning, coating and dimensional performance.

    Surface finishing should not be treated as a cosmetic operation added after printing. It is part of the manufacturing route and should be planned during design.

    Why AM surfaces are difficult

    Surface condition depends on more than layer thickness. Process category, alloy, powder size, energy input, orientation, local heat flow, support strategy, recoating, scan parameters and post-build handling all contribute. The same part can contain several different surface states:

    • Up-facing surfaces: often smoother and more dimensionally stable than unsupported down-facing areas.
    • Down-facing surfaces: vulnerable to dross, partially fused powder and geometric sag.
    • Vertical walls: affected by layer stair-stepping, contour strategy and powder adhesion.
    • Support-contact regions: retain witness marks and local damage after removal.
    • Internal channels: may be inaccessible to conventional tools and difficult to inspect after finishing.
    • Machined stock surfaces: intentionally printed oversized so final geometry is created later.

    Because surface requirements vary by function, a single roughness target for the complete part is usually inefficient. A better approach classifies surfaces by engineering need.

    Start with functional surface requirements

    Surface functionWhat mattersTypical finishing route
    Structural fatigue surfaceNotches, attached particles, near-surface defects and residual stressMachining or controlled material removal, followed by polishing or peening where justified
    Sealing faceFlatness, waviness, roughness and dimensional controlMachining, grinding, lapping or honing
    Bearing or sliding interfaceGeometry, roughness direction, hardness and wearMachining, grinding, honing, superfinishing or coating
    Fluid channelPressure loss, debris retention, cleanability and geometry preservationAbrasive-flow, chemical/electrochemical or specialized internal finishing
    Biological interfaceCleanliness, controlled texture and biocompatibilitySelective machining, blasting, chemical treatment and validated cleaning
    Coating substrateAdhesion, contamination, oxide state and target profileBlasting, machining, cleaning or chemical preparation
    Cosmetic surfaceVisual uniformity and touchBlasting, tumbling, polishing or coating

    Mechanical finishing methods

    Machining

    CNC milling, turning, drilling, reaming, grinding and EDM remain the most reliable methods for creating tight tolerances and functional interfaces. Machining is commonly used for datums, bores, sealing faces, threads and mounting surfaces.

    Designers should define machining stock, workholding, tool access and the datum strategy before printing. AM parts can contain residual stress, thin walls and interrupted surfaces that make machining less stable than a wrought blank. Heat treatment and build-plate removal sequence can materially change the final geometry.

    Abrasive blasting

    Bead, grit or shot blasting can remove loose particles and create a more uniform appearance. It is fast and scalable for accessible surfaces, but it does not create precision geometry. Media type, pressure, angle and contamination control must suit the alloy and final application.

    Mass finishing and vibratory finishing

    Tumbling and vibratory systems use media and relative motion to smooth exposed edges and surfaces. They work well for smaller robust parts, but can round sharp features, damage thin sections and provide uneven access to recessed geometry.

    Abrasive-flow machining

    Abrasive media is forced through a passage to remove high points and smooth internal channels. It can improve flow performance in manifolds and heat exchangers, but material removal is geometry-dependent. Process development should verify that critical wall thickness and channel shape remain within limits.

    Peening and surface mechanical treatments

    Shot peening, laser shock peening and related methods can modify near-surface residual stress and improve fatigue performance in suitable applications. Peening does not automatically remove surface-connected defects and should not be used as a substitute for required machining or inspection.

    Chemical and electrochemical finishing

    Chemical polishing

    Chemical polishing removes material through a controlled reaction across exposed surfaces. It can reach complex areas without direct tool contact, but the rate depends on alloy chemistry, solution condition, temperature, flow and local geometry. Entrapped chemicals, selective attack and dimensional loss must be controlled.

    Electropolishing

    Electropolishing removes material electrochemically and preferentially smooths peaks. It is widely used for stainless steels and selected nickel, cobalt-chromium, titanium and other alloys with appropriate electrolytes. Electrical contact, current distribution, shielding and access determine uniformity. Deep blind passages may remain difficult.

    Both chemical and electrochemical routes require validated cleaning and waste-control procedures. For medical, aerospace or fluid-service components, residual chemicals and altered surface chemistry can be as important as roughness.

    Thermal and energy-based finishing

    Laser polishing

    Laser polishing locally remelts a thin surface layer so surface tension redistributes material. It can smooth selected accessible regions without abrasive media, but the thermal cycle can change microstructure, residual stress, oxide condition and dimensions. Line-of-sight access and parameter development are required.

    Plasma and electrochemical plasma processes

    Plasma-based finishing methods can reduce roughness on suitable conductive materials and complex shapes. Their effectiveness depends strongly on chemistry, current density, edge effects and process access. They are specialized processes rather than universal solutions.

    Internal-channel finishing

    Internal channels are one of AM’s strongest design advantages and one of its hardest post-processing problems. A channel can be printable but impossible to depowder, finish, clean or inspect. The finishing strategy should answer:

    • Can media or fluid reach every surface and exit completely?
    • Will bends, branches or changes in section create uneven removal?
    • Can the process preserve minimum wall thickness and calibrated flow area?
    • How will residual abrasive, chemical or powder be verified?
    • Can CT, flow testing or other methods confirm the finished geometry?

    For critical channels, coupons and representative flow artifacts are usually more useful than generic roughness samples.

    Surface finishing and fatigue

    As-built metal AM surfaces can reduce fatigue performance because attached particles, valleys, support scars and near-surface imperfections act as local stress concentrators. The improvement obtained from finishing depends on how much material is removed and whether the controlling defect is actually eliminated.

    A polished appearance does not prove that subsurface lack of fusion or porosity has been removed. Fatigue-critical routes should connect surface treatment with material allowables, residual stress, heat treatment, inspection and representative testing.

    How to select a finishing process

    1. Define function. Specify why the surface needs treatment: tolerance, sealing, fatigue, flow, wear, coating, cleaning or appearance.
    2. Map accessibility. Separate open, recessed, internal and support-contact surfaces.
    3. Set material-removal limits. Protect thin walls, sharp features, lattices and calibrated flow areas.
    4. Choose the sequence. Heat treatment, HIP, support removal, machining, polishing, peening, coating and cleaning can interact.
    5. Validate on representative geometry. Flat coupons rarely reproduce internal channels, downskin or support scars.
    6. Measure more than Ra. Use the parameters and spatial scales that correlate with function.
    7. Verify cleanliness and integrity. Finishing must not introduce contamination, cracks, embedded media or unacceptable dimensional loss.

    Common mistakes

    • Specifying one roughness value for all surfaces regardless of function
    • Adding finishing only after the geometry and orientation are frozen
    • Assuming every internal passage can be polished uniformly
    • Using polishing to hide process instability rather than correcting the build route
    • Ignoring fixture, datum and tool-access requirements
    • Failing to account for removed material in tolerance and wall-thickness analysis
    • Evaluating visual appearance without checking fatigue, cleanliness or dimensional performance

    Conclusion

    Surface finishing is one of the main bottlenecks between a successful AM build and an accepted production part. No method is best for every surface. The correct route combines design, orientation, machining access, controlled material removal, cleaning and inspection around the actual functional requirement.

    Related Addithive resources: Design for LPBF · NDT for Additive Manufacturing

    References and further reading

  • Medical Additive Manufacturing: Devices, Patient-Specific Workflows and Regulatory Controls

    Medical Additive Manufacturing: Devices, Patient-Specific Workflows and Regulatory Controls

    Medical additive manufacturing includes several very different activities: anatomical models, surgical guides, standard implants, patient-matched devices, prosthetics, instruments and early-stage bioprinting research. Each application has a different risk profile, evidence burden and manufacturing route.

    A material described as biocompatible is not automatically a safe medical device. The intended use, finished geometry, manufacturing process, cleaning, sterilization and clinical performance must be evaluated together.

    Medical AM application map

    ApplicationTypical valueMain controls
    Anatomical modelsVisualization, education and surgical planningImage segmentation, dimensional accuracy and intended-use labeling
    Surgical guidesTransfer a digital plan to the patient during a procedureFit, guide geometry, sterilization and procedural validation
    Patient-matched implantsGeometry adapted to a patient’s anatomyImaging data, design boundaries, mechanical performance and traceability
    Standard porous implantsRepeatable lattice or porous structures for fixationProcess validation, fatigue, residue removal and biological evaluation
    External prostheses and orthosesCustomization, comfort and rapid iterationFit, load capacity, skin contact and durability
    Instruments and fixturesComplex geometry and low-volume productionCleaning, sterilization, strength and lifecycle testing
    Drug and biologic researchNovel dosage forms, scaffolds and laboratory modelsApplication-specific drug or biologic regulation
    Bioprinting researchCellular constructs and research modelsCell viability, biology, vascularization and translational evidence

    Standard designs vs patient-matched devices

    Not every medical AM part is patient specific. Many commercially manufactured spine cages, orthopedic implants and instruments are produced repeatedly from standard designs. Patient-matched devices use imaging or anatomical data to create a defined variation within an approved design envelope.

    Personalization increases the importance of software validation, segmentation review, design rules and data traceability. A unique geometry does not remove the requirement for controlled manufacturing or device testing.

    The patient-specific digital workflow

    1. Acquire patient data: CT, MRI, optical scan or another validated source is collected at appropriate resolution.
    2. Segment the anatomy: Relevant structures are separated from the medical image and reviewed for artifacts or missing information.
    3. Create the device or model: The design is generated within approved rules, offsets and anatomical interfaces.
    4. Clinical and engineering review: Qualified personnel confirm intended use, fit, orientation and critical features.
    5. Prepare the build: Orientation, supports, nesting and process parameters are released.
    6. Manufacture and post-process: Printing, cleaning, heat treatment, machining, finishing or curing are performed under controlled instructions.
    7. Inspect and test: Identity, geometry, material and functional requirements are verified.
    8. Clean, package and sterilize: The validated route is completed where required.
    9. Release and retain records: The part is linked to patient, design, software, machine, material and processing history.

    The FDA describes a similar chain covering device design, software workflow, material controls, printing, post-processing, validation and testing.

    Process and material selection

    ProcessCommon medical roleImportant limitations
    Metal laser powder bed fusionTitanium, cobalt-chromium and selected implant or instrument applicationsResidue, surface condition, fatigue, heat treatment and NDT
    Electron-beam powder bed fusionSelected titanium implants and porous structuresPowder-cake removal, feature resolution and platform-specific material routes
    Polymer powder bed fusionModels, instruments, orthoses and selected devicesPowder reuse, moisture, porosity and sterilization compatibility
    Vat photopolymerizationModels, guides, dental devices and selected patient-contact productsResin identification, washing, post-cure, extractables and aging
    Material extrusionModels, prosthetics, fixtures and research devicesAnisotropy, porosity, dimensional accuracy and cleaning
    Material jettingMulti-color or multi-material anatomical modelsMaterial aging, support removal and limited structural performance
    Bioprinting platformsResearch constructs, tissues and disease modelsBiological complexity and early translational maturity

    Design and manufacturing controls

    The FDA’s final guidance on additively manufactured medical devices organizes technical considerations around design/manufacturing and device testing. A practical control plan should address:

    • Device orientation and build location
    • Minimum feature capability and dimensional compensation
    • Support removal and inaccessible surfaces
    • Feedstock or resin specifications and lot control
    • Machine, software and parameter validation
    • Heat treatment, HIP, washing, post-curing and machining
    • Process monitoring and change control
    • Worst-case geometry and build configuration
    • Part identification and patient-data linkage
    • Supplier and production-site controls

    Cleaning and residue removal

    Complex AM geometry can trap powder, uncured resin, support material, solvent or process debris. Cleaning validation should consider the actual device geometry rather than an easily accessible test coupon.

    • Define internal passages, pores and lattices that can retain residue.
    • Validate the cleaning process at worst-case locations.
    • Measure extractable or recoverable residue using appropriate methods.
    • Control cross-contamination between materials and patient-specific jobs.
    • Confirm that cleaning does not damage surface, dimensions or material properties.
    • Retain evidence linking cleaning parameters to the released device.

    ASTM F3335 provides guidance for assessing residue removal from powder-bed-fusion medical devices and is recognized by the FDA.

    Biocompatibility and material claims

    Biocompatibility is assessed for the finished device in its intended contact condition. Relevant variables can include:

    • Base chemistry and additives
    • Powder reuse or resin aging
    • Build orientation and surface area
    • Heat treatment, post-cure and cleaning
    • Residual particles, monomers, solvent or support material
    • Coatings, polishing and machining
    • Contact type, location and duration

    A resin or alloy cleared or used for one application should not be assumed acceptable for another indication or contact type.

    Sterilization and dimensional stability

    Sterilization can alter polymer dimensions, mechanical properties, surface condition and residual chemistry. Metal devices can also be affected by packaging, cleaning or repeated processing. The selected method—such as steam, radiation or gas—must be compatible with the finished device and its intended lifecycle.

    • Measure dimensions and function after the complete sterilization route.
    • Evaluate repeated cycles for reusable devices.
    • Confirm that packaging permits effective sterilization and protects the device.
    • Control time between manufacturing, cleaning, sterilization and use.
    • Use the same post-processing condition for performance and biological testing.

    Mechanical and functional testing

    Test methods should represent the actual device, loading and manufacturing route. Depending on the application, evidence can include:

    • Static strength and fatigue
    • Wear, corrosion and fretting
    • Porous-structure characterization
    • Dimensional and fit verification
    • Pressure, flow or leak performance
    • Sterilization and shelf-life effects
    • Usability and procedural performance
    • Biological evaluation and clinical evidence as required

    ASTM F3604-23 provides a framework for validating laser-powder-bed-fusion production systems used for medical devices, including machine qualification, software, raw materials and IQ/OQ/PQ concepts.

    Point-of-care manufacturing

    Hospitals and clinical centers may use 3D printing for models, guides or other devices. Point-of-care location does not remove manufacturing responsibility. Governance should define:

    • Who is the legal manufacturer?
    • Who approves segmentation and design?
    • Which software, printer and material combinations are validated?
    • How are maintenance, environmental conditions and operators controlled?
    • How are patient data, cybersecurity and privacy protected?
    • How are nonconformances and adverse events handled?
    • Which records are retained and for how long?

    Bioprinting: promising research, different maturity

    Bioprinting uses cells, biomaterials or biologically active materials to create research constructs. It should not be grouped with commercially mature metal implants or surgical guides. Major challenges include cell viability, vascularization, tissue maturation, reproducibility, immune response and long-term function. Claims about printing complete transplantable organs remain research-stage rather than routine clinical manufacturing.

    Medical AM readiness checklist

    1. The intended use and regulatory classification are defined.
    2. Design inputs and patient-matching boundaries are controlled.
    3. The exact material, machine and post-process route is validated.
    4. Worst-case geometry can be cleaned and inspected.
    5. Finished-device biocompatibility and sterilization are addressed.
    6. Mechanical and functional testing represents final condition.
    7. Software, imaging and segmentation are verified.
    8. Traceability links patient, design, build, material and processing records.
    9. Changes to site, machine, software or material trigger defined review.
    10. Clinical claims match the available evidence.

    Conclusion

    Medical additive manufacturing is already mature in selected areas such as hearing aids, dental workflows and porous metal implants, while other areas remain developmental. Success depends on matching the process to a defined medical need and validating the complete finished-device route—from patient data and design through cleaning, sterilization, testing and traceability.

    Related Addithive resources: Dental Additive Manufacturing · Straumann Dental AM Profile · NDT for AM

    References and further reading