What level of accuracy and resolution can industrial CT scanning achieve?

Industrial CT can use sub-micron voxels for a very small, low-density specimen, but no single accuracy or resolution applies to every scan. Usable feature resolution and dimensional uncertainty depend on part size, material, wall thickness, magnification, focal spot, detector, reconstruction, surface determination and calibration. Voxel size is not a machining tolerance or a guaranteed defect limit. The RFQ should state whether CT is needed for porosity detection, wall-thickness review, internal-geometry comparison, first-article evidence or metrology-grade dimensional acceptance.

Defining Accuracy and Resolution in Industrial CT

CT capability has three separate measures. Voxel size describes the sampling grid, spatial resolution describes the detail reproduced after blur and noise, and dimensional accuracy describes agreement with a traceable reference. Defect detectability is a fourth decision because feature contrast, orientation and reconstruction artifacts can matter more than nominal voxel size.

Spatial Resolution: The Finest Detail Distinguishable

Spatial resolution is the smallest separation that remains distinguishable in the reconstructed volume under the selected scan conditions. Geometric magnification can reduce voxel size when a small part sits close to the X-ray source, but focal-spot blur, detector response, motion and noise still limit real detail. A claim of reconstructed voxel size below 1 micron (0.001 mm) is credible only for a very small specimen, a restricted field of view and a qualified high-magnification setup. It does not prove that a 1 µm pore, wall or dimension can be accepted. A Precision Machining Service buyer should request the scan volume, voxel size, contrast-to-noise evidence, reference feature and smallest validated indication for the actual material.

Dimensional Measurement Accuracy: The Trueness of a Measurement

Dimensional accuracy is established by comparing CT measurements with calibrated artifacts or an accepted reference method and then stating uncertainty for the measurement task. A 5 micron (0.005 mm) dimensional result can be possible on a small, stable, well-calibrated part, but it is not a universal CT capability. Large envelopes, dense alloys, mixed materials, rough surfaces and long X-ray paths can move the usable uncertainty into a much wider range. The VDI/VDE 2630 series addresses CT dimensional metrology performance and task-specific qualification. Buyers using Multi-Axis Machining Service should define datum alignment, feature type, tolerance, surface state and the required uncertainty ratio before selecting CT over CMM or optical measurement.

Factors Influencing CT Scanning Performance

CT performance is limited by the complete imaging and measurement chain, not by one brochure number. Source geometry, detector sampling, material attenuation, reconstruction, edge selection and traceability must all support the required feature or defect.

1. X-Ray Source and Detector Capability

Detector pixel pitch divided by geometric magnification defines the nominal voxel sampling. Higher magnification can improve sampling, but the part must remain inside the field of view and clear of the source. A growing focal spot, vibration or insufficient projection count can erase the expected gain. Detector dynamic range and exposure also control noise and contrast. These limits matter for small CNC Machining Prototyping parts, microchannels, fine threads and selected Medical Device components. The quote should identify whether the claimed setting covers the whole part or only a region of interest.

2. Material Density and Penetration

X-ray attenuation increases with material composition, thickness and path length. Parts from Plastic CNC Machining or thin Aluminum CNC Machining stock often permit higher contrast and magnification. Dense sections from Stainless Steel CNC Machining or Superalloy CNC Machining may require more energy and exposure. Beam hardening, scatter and photon starvation can hide pores or bias an extracted edge. The RFQ should state grade, maximum X-ray path length, assembly stack and the minimum feature or discontinuity that matters.

3. Software and Calibration

Reconstruction software converts projections into voxels; metrology software then selects material boundaries and fits geometry. A threshold or surface-determination change can shift a bore, thin wall or rough edge without changing nominal voxel size. This partial-volume failure mode is especially important near metal-air boundaries and multi-material interfaces. Validation should use a calibrated artifact or representative feature, then compare selected CT dimensions with traceable CMM or reference measurements. The report should identify voxel size, filtering, artifact correction, surface method, alignment, calibration checks and task-specific uncertainty. Passing one length standard does not qualify every internal feature, material or scan position.

Practical Applications Across Critical Industries

CT is strongest when an internal feature cannot be reached by calipers, CMM or borescope and destructive sectioning would invalidate the part. The acceptance method still has to match the failure mode and governing requirement.

• Aerospace and Aviation: CT can evaluate cooling passages, trapped powder, casting porosity and wall-thickness variation. The buyer should define the material specification, critical zone, indication threshold, acceptance class and whether CT is approved for release or only engineering investigation.

• Automotive: CT can validate die-cast housings, connectors, battery assemblies and lightweight structures. Launch validation may use a detailed full-volume scan, while production monitoring may use sampling or a restricted region. The control plan must state which change triggers requalification.

• Nuclear Industry: CT may support diagnostic review of selected components, but density, shielding, radiation controls and construction-code acceptance can limit its role. A CT image should not be treated as code release evidence unless the governing procedure explicitly permits it.

Synergy with Manufacturing and Post-Processing

CT data can compare hidden geometry with CAD, locate trapped material, measure selected wall sections and support root-cause analysis. It can also verify whether internal passages remain open after cleaning or selected Electropolishing for Precision Parts. The manufacturing plan should place CT before the step where a rejected blank would create avoidable machining cost, then repeat it only when later processing can create or expose a new risk. CT does not replace surface roughness measurement, tactile datum verification, leak testing or functional testing.

A usable RFQ lists the scan purpose, drawing revision, material and condition, overall envelope, maximum path thickness, critical region, smallest relevant feature or defect, tolerance, datum alignment, required uncertainty, report fields and acceptance source. Ask the supplier to separate whole-part voxel size from region-of-interest voxel size. For tight GD&T or dense alloys, request a feasibility scan and a correlation plan before approving CT as the release method.

The release decision should use evidence from the actual part class: validated detectability for defect inspection, or traceable uncertainty for dimensional metrology. That distinction lets purchasing compare CT proposals without mistaking a sharp image for a qualified measurement.