No single inspection method is best for every thread or bearing seat. Visual or borescope inspection maps accessible openings, CT can map detectable three-dimensional distributions, radiography can screen suitable projections, microsections reveal local morphology and functional tests determine whether the finished interface performs. Selection depends on feature size, aluminum thickness, orientation, required resolution and whether the question is surface acceptance, internal extent or function.
Buyers should define the decision first, then qualify the method on representative good and reject conditions.
Map indications on loaded flanks, thread root, entry and engaged length. A pore outside the engagement may be less critical than one removing material across several loaded threads. Inspect burrs and torn metal separately because tapping or thread milling damage can resemble a void.
Go/no-go gauges confirm assembly envelope but cannot prove local flank strength. Torque or pull-out validation can support qualification when specified, yet one destructive result does not screen every production part. Use material continuity rules and traceable process evidence.
Thread Question | Method Direction | Limitation |
|---|---|---|
Is an opening visible? | Magnification/borescope | Access and no internal extent |
Is engagement geometry acceptable? | Gauge/profile measurement | Does not prove material continuity |
Does a detectable void extend behind flank? | Qualified CT or section | Resolution/artifact or destructive locality |
Can the joint carry required load? | Qualified torque/pull-out test | Sampling and assembly dependence |
A bearing seat needs size, roundness, cylindricity or other drawing geometry, surface finish and continuous support appropriate to the fit. An indication can create local loss of contact, edge breakout or a path for lubricant. Position relative to peak bearing load matters, but the actual load case must come from product engineering.
Map the complete circumference and depth. A bore view can miss an opening at the back edge or under a press-fit zone. Repeated insertion should not be used to burnish away evidence.
Radiography compresses thickness into a projection; geometry can superimpose bosses and walls over the region of interest. Multiple views can help. CT separates volume but its voxel size is not equal to guaranteed defect detectability. Contrast, noise, beam hardening, reconstruction and thresholding affect results, especially near dense geometry and machined edges.
Use reference samples or known indications where practical, record scan parameters and establish what sizes, orientations and locations are reliably found. The industrial CT resource provides context but cannot establish capability for an untested part.
Method | Best Role | Release Boundary |
|---|---|---|
Visual/borescope | Production surface-opening map | Defined lighting, magnification and access |
Radiography | Projection screening | Qualified view and sensitivity |
CT | Development 3D distribution | Validated detectability and analysis settings |
Microsection | Mechanism/morphology investigation | Mapped destructive plane |
Functional test | Finished interface performance | Named load, assembly and acceptance |
Assign each part a common coordinate and feature identity. Mark surface openings, retain image slices and section selected locations. Compare method findings rather than assuming a CT indication and visible pore are separate defects. Track false positives, misses and ambiguous results during qualification.
A negative scan does not override a visible out-of-specification opening, and a visual pass does not override a functional failure. The inspection hierarchy should state which evidence controls each decision.
Request part, cavity, lot, machining operation, feature zone, instrument or scan setup, calibration/reference, raw images or results, analyst acceptance and disposition. The post-machining process should preserve the indication before local blend or additional cut.
A method should be challenged with representative known conditions near the actual thread or bore. Natural indications confirmed by sectioning can be valuable, while manufactured holes or inserts may not reproduce contrast and orientation perfectly. Use references to establish a practical probability of detection and analyst consistency, not to claim that every smaller void will always be found.
Scan or inspect the same reference in different orientations and positions within the equipment field. Threads and curved bearing walls create artifacts and changing path length. If detectability drops at one side of the bore, add a view, adjust the method or retain another inspection for that zone.
Capability Element | Evidence | Risk if Omitted |
|---|---|---|
Known location | Sectioned or otherwise confirmed indication | False confidence from image noise |
Orientation challenge | Multiple views or part rotations | Planar or hidden feature missed |
Repeated analysis | Same data reviewed over time/operators | Threshold depends on analyst |
Good controls | Acceptable parts without reported feature | High false-positive rate |
Revalidate after equipment, software, reconstruction, source-to-part geometry or acceptance threshold changes. An old capability claim cannot automatically release a new larger casting or deeper feature.
The best inspection strategy combines accessible surface mapping with a qualified internal or destructive method when extent matters and a functional test when performance matters. Method choice follows the question; no image technology replaces the product requirement.