CT Scanning for 3D Printed Parts

Computed tomography (CT) scanning is one of the most powerful nondestructive evaluation (NDE) tools available for additive manufacturing (AM), but it is not automatically the “best” inspection choice for every build. For aerospace and defense programs—where internal defects, regulated documentation, and configuration control matter—CT can close inspection gaps that traditional metrology and surface NDE cannot. For many other situations, it can add cost and schedule with little added risk reduction.

This guide explains ct scanning for 3d printed parts from an engineering and procurement perspective: what CT actually reveals, when it is necessary versus optional, the resolution tradeoffs you must understand, what drives cost, what a usable CT report looks like, and a buyer checklist you can drop into an RFQ or supplier quality plan.

What CT reveals

CT scanning creates a 3D volumetric dataset (voxels) by reconstructing many X-ray projections. Unlike 2D radiography, CT allows you to “slice” the part in any plane and evaluate internal conditions without sectioning. For powder bed fusion (PBF) processes such as DMLS / SLM, CT is especially valuable because many critical risks are internal or sub-surface and cannot be verified with a CMM or visual inspection.

1) Internal porosity and void morphology
CT can detect gas porosity, lack-of-fusion voids, and shrinkage-type cavities and quantify them by size, location, and (with the right analysis) total porosity fraction. This matters for fatigue-driven hardware, thin-walled pressure boundaries, and parts with stress concentrations near internal features.

2) Lack-of-fusion and process signatures
For PBF, lack-of-fusion defects tend to be irregular and planar. When they align with layer-wise geometry or scan strategy artifacts, they can indicate a process control issue (powder condition, energy density, recoater events, or parameter drift) rather than random pores. CT can support corrective action by showing where the defect population concentrates (e.g., near down-skin surfaces, overhangs, or island scan boundaries).

3) Internal channels, lattices, and trapped powder
AM enables conformal cooling, fluid passages, and lattice energy absorbers. These are often impossible to inspect with conventional probes. CT can verify channel continuity, minimum wall thickness, obstructions, and powder entrapment risk. For heat exchangers and manifolds, it can also confirm that post-processing (e.g., powder removal, chemical cleaning) achieved the intended flow path.

4) Build-to-CAD comparison (internal and external)
CT datasets can be aligned to the CAD model to evaluate dimensional deviation, including internal features that a CMM cannot reach. This is often used for first article inspection (FAI) development and for validating the “as-built” condition prior to HIP and machining. CT-based metrology is not always a replacement for tactile CMM on critical datums, but it can augment CMM by providing insight into hidden geometry.

5) Assembly risk reduction and failure analysis feedback
On programs with tight integration windows, CT can prevent expensive downstream surprises by catching internal nonconformances before machining, coating, or joining. It is also valuable for root cause investigations when combined with process records, powder lot traceability, and post-processing history (stress relief, HIP, heat treat).

When CT is necessary

CT should be treated as a risk-based inspection method. The right question is not “Can CT find defects?” but “Does CT materially reduce risk that cannot be controlled through process qualification, in-process monitoring, and simpler NDE?” The following situations commonly justify CT scanning in defense and aerospace AM workflows.

1) Internal features are critical to function
If the part includes internal passages, impingement features, lattice cores, or enclosed volumes where blockage or geometry variation would cause performance loss, CT is often the only practical way to verify conformance. For example, a conformal cooling insert produced via DMLS may require CT to confirm channel diameter, minimum wall thickness, and powder removal effectiveness before it is placed into service.

2) You cannot inspect the critical region after machining
Many AM parts go through HIP (Hot Isostatic Pressing) and then 5-axis CNC machining to establish datums and meet final tolerances. If machining will remove access or change surfaces such that later inspection cannot confirm internal integrity, CT may be warranted earlier (often pre-HIP) to ensure you are not investing in a fundamentally nonconforming build.

3) Fatigue or fracture performance is program-critical
For highly loaded brackets, hinges, rotating components, or pressure-containing hardware, internal defect size and location can dominate fatigue life. Even when HIP reduces porosity, CT can be used to (a) establish baseline defect populations during qualification, (b) screen parts when risk is high, and (c) support damage-tolerance arguments when permitted by the design allowables approach.

4) New build parameter set, new machine, or new material lot
During process qualification and supplier onboarding, CT can serve as a “truth dataset” to validate that the PBF process window and powder controls are producing acceptable internal quality. This is especially relevant when moving between machines, adopting new scan strategies, or introducing a new powder supplier. CT is not a substitute for a robust qualification plan, but it can accelerate learning and reduce the chance that you qualify a process that hides unacceptable internal discontinuities.

5) Contracts, specs, or customer flowdowns require volumetric NDE
Some programs explicitly require CT or volumetric NDE on specific geometries. In regulated environments (AS9100, NADCAP-controlled special processes, and defense flowdowns such as DFARS and ITAR), the governing requirement is the contract, drawing notes, and inspection plan. If CT is required, align early on acceptance criteria, data deliverables, and how CT fits into the certificate of conformance (CoC) and traceability pack.

6) You need to validate HIP effectiveness or PM-HIP densification
For some components, especially in development or first-time qualification, CT can be used before and after HIP to demonstrate porosity reduction. For PM-HIP (powder metallurgy + HIP) parts, CT can help evaluate consolidation quality in complex geometries. Note: HIP can close voids, but it may not heal every defect type equally; acceptance criteria should consider the defect morphology and the service load case.

When CT is usually not the right primary tool
If the part is simple, fully accessible for conventional inspection, and the risk of internal discontinuities is low (or already controlled by process qualification and periodic witness coupons), CT may be unnecessary. Many production programs rely on a combination of: process control, build monitoring, coupon testing, surface NDE (penetrant), and dimensional inspection by CMM—reserving CT for first articles, engineering investigations, or targeted screening.

Resolution and limitations

CT is sometimes treated as a “microscope for the inside,” but it has real limits. Misunderstanding these limits is a common source of disputes between engineering, quality, and suppliers.

Voxel size is not the same as minimum detectable defect
A scan might be reported as a 25 µm voxel size, but that does not guarantee reliable detection of 25 µm pores. Practical detectability depends on signal-to-noise ratio, part material and thickness, beam energy, focal spot size, reconstruction settings, and the contrast between defect and base metal. A good rule for planning is that reliable detection typically requires multiple voxels across the feature and stable segmentation thresholds.

Material and geometry drive X-ray attenuation
High-density alloys (e.g., nickel superalloys, some cobalt alloys) and thick sections can require higher energy, which can reduce contrast and increase artifacts. Titanium is often CT-friendly, but thick titanium sections still challenge resolution. Long path lengths through metal create beam hardening and scatter, which can hide small defects or distort dimensional measurements if not corrected.

Artifacts can mimic or mask defects
Common CT artifacts include beam hardening, ring artifacts, metal streaking, and partial volume effects. AM-specific surface roughness and near-surface porosity can also make segmentation difficult. This is why acceptance criteria should be tied to a validated scan plan and analysis method rather than a single “porosity %” number generated with unknown thresholds.

Near-surface defect detection is tricky
CT is excellent for internal voids, but detecting discontinuities very near the surface can be less reliable because the boundary between air and metal dominates the signal. For surface-breaking cracks or laps, penetrant inspection (or other surface NDE methods) may be more appropriate, especially after machining when surfaces are accessible and smoother.

CT metrology is powerful but needs controls
Using CT for dimensional inspection requires careful fixturing, calibration, and alignment to datums. In aerospace workflows, CT dimensional results are often treated as supplementary unless the CT system and method are qualified for metrology. When tolerances are tight, many teams use CT to understand deformation or internal feature position, then still rely on CMM for final acceptance of critical external datums.

HIP and machining change what CT “means”
If you CT scan as-built, then HIP, heat treat, and machine, you should not assume the same defect population or geometry remains. HIP can reduce internal porosity; machining removes surface-connected flaws and can also expose subsurface conditions. Align CT timing with the decision you’re trying to make: screen as-built builds, validate HIP closure, or confirm final internal geometry after machining (when feasible).

Cost considerations

CT cost is not just “machine time.” It is a combination of scan time, setup, analysis, reporting, and—often overlooked—data management. For procurement teams, the goal is to scope CT so it answers the acceptance question without creating unnecessary cost or turning CT into an open-ended research project.

What drives CT price and lead time

1) Part size and maximum wall thickness
Larger parts require larger field-of-view and longer source-to-detector distances, which tends to increase voxel size (lower resolution). Thick sections require higher energy and longer exposures. Both increase cycle time and may reduce defect detectability, sometimes forcing multiple scans (sub-volume scanning) to hit required resolution in critical areas.

2) Alloy selection
Dense or high-Z alloys increase attenuation and artifacts. Scanning Inconel-class materials at meaningful resolution can be more expensive than scanning aluminum or thin titanium components. If the program can accept it, scanning representative coupons alongside the part can sometimes provide cost-effective process insight without scanning every production unit at maximum settings.

3) Acceptance criteria definition
“CT scan and provide images” is ambiguous. Cost rises sharply when the supplier must perform advanced segmentation, pore sizing distributions, CAD comparisons, and engineering interpretation. If your requirement is “no indications larger than X in region Y,” define X, Y, and the measurement method. If your requirement is “compare to CAD within tolerance,” define the datums, tolerances, and whether this is informational or acceptance.

4) Number of regions of interest (ROI) and scan strategy
A single full-volume scan is not always optimal. Many programs use a two-tier approach: a lower-resolution full scan to locate anomalies and verify gross internal features, plus high-resolution ROI scans on critical sections. This can cut cost compared with scanning the entire part at the highest possible resolution.

5) Data handling and export requirements
Volumetric datasets are large. If you require DICONDE (Digital Imaging and Communication in Nondestructive Evaluation) formatting, encrypted transfer, long-term retention, or ITAR-controlled handling, factor those controls into cost and schedule. For ITAR programs, confirm where scanning occurs and how data is stored and accessed.

How CT fits into a practical AM production workflow

Step 1: Define inspection intent during design and RFQ
Identify critical-to-function internal features and critical regions for defect control. Decide whether CT is for (a) process qualification, (b) first article, (c) periodic audit, or (d) 100% screening.

Step 2: Align CT with post-processing sequence
Typical PBF workflow: build → stress relief → support removal → (optional) CT → HIP → heat treat → rough machine → (optional) CT of ROI → finish machine → surface NDE (e.g., penetrant) → CMM → final documentation pack. Not every program uses every step, but planning CT insertion points prevents rework and surprises.

Step 3: Tie CT results to disposition rules
If CT finds an indication, what happens? Can the part be reworked (e.g., additional machining), or is it scrap? Is engineering review required? Define thresholds and escalation paths so CT does not become a schedule risk.

Reporting

A CT scan is only as useful as the report package and traceability that accompanies it. In regulated aerospace and defense manufacturing, CT results need to integrate with the broader quality system: material traceability, calibrated equipment, controlled procedures, and a clear link to the part serial number and configuration.

What a procurement-ready CT report should include

1) Part identification and configuration control
Part number, revision, serial/lot number, build ID, and drawing/spec references. If multiple builds share a scan method, the report should state that the method is controlled and revision-managed.

2) CT equipment and calibration status
System identification, calibration/verification references, and confirmation that the equipment is within calibration. For high-reliability programs, buyers often request evidence that the CT system is maintained under an accredited quality system and that the procedure is controlled.

3) Scan parameters
Source energy and current, filtration (if used), voxel size, number of projections, exposure time, reconstruction method, and any artifact correction applied. These parameters matter for repeatability and for interpreting detectability limits.

4) Acceptance criteria and evaluation method
Explicit criteria for indications: maximum allowable pore size, density of indications per volume, exclusion zones, or “no indications in region” statements. The report should state how defects were segmented and measured (thresholding approach, connectivity rules, minimum reportable size).

5) Regions of interest and annotated evidence
Screenshots/slices with coordinates, orientation references, and callouts. If using CAD comparison, include alignment approach and a color map or deviation summary with tolerances clearly stated.

6) Results summary and disposition
A clear pass/fail statement when CT is an acceptance inspection, or an “informational only” statement when used for engineering evaluation. Avoid ambiguous language. If indications exist but are acceptable, the report should state why they meet criteria.

7) Data deliverables and retention
If you need the volumetric dataset, specify the format (often DICONDE or agreed internal format), compression, and retention period. For ITAR-controlled programs, specify access controls and whether data must remain on U.S.-based servers and be handled by authorized persons.

How CT ties into certification packs
For aerospace/defense buyers, CT is typically one element in a larger deliverable set: material certifications (chemistry and mechanicals as applicable), powder lot traceability, heat treat/HIP records, in-process traveler, dimensional inspection (CMM), surface NDE results, and a final certificate of conformance (CoC). Ensure the CT report is clearly cross-referenced in the pack so auditors can trace acceptance decisions back to objective evidence.

Buyer checklist

Use this checklist to scope CT requirements in an RFQ, supplier quality clause, or inspection plan. The goal is to specify CT in a way that is measurable, auditable, and aligned to actual risk.

1) Define the purpose
State whether CT is for first article inspection (FAI), process qualification, periodic audit, or 100% production screening. Overuse drives cost; underuse can leave critical risks unaddressed.

2) Identify regions of interest (ROI)
Call out critical internal features (channels, manifolds, lattice cores, thin walls, high-stress regions). If only certain areas matter, request ROI scanning rather than full-volume maximum-resolution scans.

3) Specify acceptance criteria in engineering terms
Examples: maximum indication size in a defined volume; exclusion zones near critical surfaces; blockage limits for internal channels; minimum wall thickness for pressure boundaries. Avoid vague requirements like “no porosity” unless it is truly achievable and measurable for the alloy and process.

4) Require a documented scan plan and method control
Ask for a written procedure that defines setup, parameters, reconstruction, and analysis approach. In AS9100 environments, method control and revision history matter for repeatability and audits.

5) Clarify required resolution and how it will be demonstrated
Request the intended voxel size and a statement of practical detectability for your indication size threshold. If the program is critical, ask how the supplier validates sensitivity (e.g., using reference standards, representative coupons, or agreed verification artifacts).

6) Define deliverables
At minimum: a signed CT report with annotated slices and results summary. If you need raw volumetric data, specify format, naming convention, and transfer method. Also specify whether CAD comparison outputs are required.

7) Address regulatory and controlled data handling
If the part, drawing, or dataset is export-controlled, include ITAR/DFARS requirements and confirm that scanning, data storage, and personnel access comply. Define whether data must remain within a controlled network and whether subcontracted CT is permitted.

8) Integrate CT into the overall manufacturing plan
Ensure CT timing matches the additive + post-processing route: PBF build, stress relief, support removal, HIP/PM-HIP densification (if used), heat treat, and machining. Specify whether CT is pre-HIP, post-HIP, post-rough machine, or post-finish machine.

9) Define nonconformance and disposition workflow
State what happens if CT finds indications: MRB process, engineering review thresholds, rework options, and who owns the decision. This prevents CT from becoming a late-stage schedule shock.

10) Qualify the supplier, not just the scan
Confirm the supplier’s quality system (e.g., AS9100), calibration program, and experience with AM geometries. If special processes are involved (HIP, heat treat, NDE), ensure the workflow is compatible with your program requirements and that any NADCAP-controlled operations are addressed where applicable.

Bottom line: CT scanning is a high-leverage tool when internal quality and internal geometry matter, when failure consequences are high, or when conventional inspection cannot see the risk. When specified with clear regions of interest, measurable acceptance criteria, and controlled reporting, CT becomes an engineering and procurement asset rather than a cost driver.