Two-Dimensional Evidence, Three-Dimensional Truth
What each method actually produces
Radiography produces a two-dimensional projection: everything along the beam path is superimposed. It is fast, cheap per part, and excellent for finding voids, cold shuts and inclusions in a known region of interest. Its weakness is depth — a defect's position along the beam is not resolved, so a small void at the surface and a large one deep inside can look similar.
Computed tomography reconstructs a three-dimensional volume from many projections. It resolves position and size, allows arbitrary slice views, and can compare a casting against its CAD nominal to show wall-thickness deviation. The cost is time and money per part, which is why it is used for first article, failure analysis and process development rather than for routine 100 percent inspection.
Choosing by the decision you need to make
If the question is "is this part acceptable today", radiography on a sampled basis with a defined acceptance standard answers it. If the question is "where exactly is the porosity and why", CT answers it, and radiography will only start the argument. For new tooling, CT on the first article frequently pays for itself by identifying the wall-thickness and shrinkage locations before they become a tooling change.
For either method, the acceptance criteria must be written before inspection. A porosity level that is acceptable in a non-structural boss and unacceptable in a pressure-tight wall cannot be judged from an image without a drawing that says which is which.
FAQ
When is CT better than X-ray for castings?
When you need defect position and size in three dimensions, wall-thickness comparison to CAD, or failure analysis. For routine accept/reject on a known region, radiography is faster and cheaper.
Can X-ray tell how deep a void is?
Not directly. A radiograph superimposes everything along the beam, so depth is unresolved unless you take multiple views at different angles.
Is CT practical for production inspection?
Rarely for 100 percent inspection because of cycle time. It fits first article, process development and failure analysis.
What must be defined before inspection?
Acceptance criteria per region, on the drawing. The same porosity level can be acceptable in one boss and rejectable in a pressure-tight wall.
All figures above refer to standard product data sheets; final confirmation is subject to written quotation.