| name | computed-tomography-ndt |
| description | CT scanning for NDT — industrial CT, voxel resolution, flaw detection threshold, porosity analysis, GD&T verification, ASTM E1570/E1814, aerospace and AM applications. |
| metadata | {"priority":7,"promptSignals":{"phrases":["CT scanning NDT","industrial CT","computed tomography NDT","CT scan inspection","X-ray CT","CT porosity analysis"],"minScore":3}} |
Computed Tomography (CT) for NDT — Complete Skill
CT Scanning Principles
X-ray CT: rotate part through X-ray beam; collect 2D projections at many angles; reconstruct 3D volume via filtered backprojection or iterative algorithms
Main components:
- X-ray source: microfocus (focal spot < 20 μm) for high resolution; conventional (> 100 μm) for large parts
- Detector: flat panel detector; typical 1024×1024 to 4096×4096 pixels
- Rotary table: part rotates through 360°; 1000–3000 projections
- Reconstruction: Feldkamp-Davis-Kress (FDK) cone beam or iterative (SART/TV)
Voxel: 3D volume pixel; minimum detectable feature ≈ 2–3× voxel size
Resolution vs. part size:
Voxel size = (detector pixel size × magnification)⁻¹ × SDD / SOD
SOD = source-to-object distance; SDD = source-to-detector distance
Practical voxel sizes:
| Part size | Achievable voxel [μm] |
|---|
| < 5 mm | 1–5 |
| 5–50 mm | 5–50 |
| 50–200 mm | 20–100 |
| 200–500 mm | 50–200 |
| > 500 mm (steel) | Limited by penetration (100–500 μm or CT not feasible) |
X-Ray Penetration and kV Selection
Lambert-Beer law:
I = I₀ × exp(-μ × t) [μ = linear attenuation coefficient [m⁻¹]; t = material thickness [m]]
μ = function of material and X-ray energy (keV)
kV selection by material and thickness:
| Material | Thickness [mm] | kV recommended |
|---|
| Aluminum | 10–50 | 60–150 kV |
| Aluminum | 50–200 | 150–250 kV |
| Steel | 5–20 | 200–350 kV |
| Steel | 20–100 | 350–450 kV |
| Inconel | 5–30 | 250–450 kV |
| Titanium | 10–50 | 200–350 kV |
Beam hardening: polychromatic X-ray → soft X-rays absorbed preferentially → cupping artifact → correction filter or iterative reconstruction
Flaw Detection
Detectability limit:
Detectable flaw size ≈ 2–3 × voxel size
For 50 μm voxel: detect pores ≥ 100–150 μm
ASTM E1570 (aerospace CT acceptance):
Compares CT image to reference CT of known defects
Sensitivity classification by ability to detect standard reference flaws
Acceptance: define class of defects not acceptable (e.g., no pore > 0.5 mm, no crack > 5 mm)
Porosity analysis:
Threshold scan volume → identify void voxels
Report: pore volume fraction [%], max pore size [mm], pore distribution (histogram)
ASTM E1814: CT for AM porosity characterization
Crack detection:
Cracks < 1 voxel wide → not detectable as open crack; detectable only as density artifact
Effective minimum crack width detection: ~3 voxels
Dimensional Verification
GD&T from CT:
Extract surface (isosurface at threshold density) → point cloud → align to CAD → compare
Dimension deviation: typical ± 2–5 voxel accuracy for dimensional measurements
Accuracy qualification:
Calibrate with NIST-traceable artifact; measure gauge ball or step wedge at various orientations
VDI/VDE 2630: metrological CT calibration guideline
Tolerance capability:
For 50 μm voxel CT: dimensional accuracy ≈ ±0.1 mm (3σ); better with calibration
Not suitable for ISO IT5 or finer tolerances without extreme care
CT for Additive Manufacturing
Key inspections:
- Internal porosity (LOF pores, keyhole, gas pores): 3D map by size and location
- Internal channels: verify flow passage geometry (no partial blockage from un-melted powder)
- Wall thickness measurement: verify compliance to design
- AM lattice verification: check strut diameter and connectivity
Acceptance criteria (aerospace AM example):
No LOF pore > 0.3 mm in critical stress zones
No interconnected porosity chains
Total porosity < 0.1% in critical zones
CT for Assemblies (Non-Destructive)
Assembled CT: scan complete assembly without disassembly
Observe: joint gaps, interference fits, internal component alignment, trapped voids
Multi-material challenges:
High-density metals + low-density polymer: large dynamic range → ring artifacts, beam hardening
Reduce by: dual kV, filtered reconstruction, iterative algorithms
Industry Applications
| Industry | CT application |
|---|
| Aerospace | Turbine blade internal cooling channels; casting porosity; composite delamination |
| Medical devices | Implant geometry verification; coating thickness |
| Automotive | Cylinder head casting porosity; engine assemblies |
| Electronics | PCB solder joint inspection; BGA void analysis |
| Nuclear | Fuel rod pellet density; cladding thickness |
Limitations
Penetration: steel > 100 mm → requires high energy (> 400 kV or 1–9 MeV linac); long scan times; poor resolution
Scan time: 30 min – 8 hr (depends on resolution, projections, averaging)
Cost: high (capital $250K–$5M+; scan $500–$5000+ per part)
Artifacts: beam hardening; ring artifacts; scatter (reduce with scatter correction hardware/software)
Standards
| Standard | Scope |
|---|
| ASTM E1570 | CT examination in aerospace |
| ASTM E1814 | CT examination of AM parts |
| ASTM E2339 | Digital detector array radiography |
| VDI/VDE 2630 | CT in dimensional metrology |
| EN 13068 | Radioscopic examination |
Output
Provide: X-ray voltage kV, tube current [mA], focal spot size [μm], SOD and SDD [mm], voxel size [μm], number of projections, minimum detectable flaw size [mm] (≈ 3× voxel), dimensional accuracy [mm] (3σ), porosity report (total [%], max pore [mm], distribution), specific flaw type detected (pore/crack/inclusion/channel deviation), comparison to accept/reject criteria, artifacts present (beam hardening/ring), and applicable standard (ASTM E1570, ASTM E1814, VDI/VDE 2630).
