Porosity is an internal defect, so the inspection method that images the bulk of the casting is the one that actually characterises it; for die-cast and investment-cast aluminium and steel parts, X-ray radiography is positioned as the primary volumetric NDT method, with digital radiography now replacing most film workflows [S3][S4].
Porosity comes in distinct morphologies, gas porosity, shrinkage porosity, and micro-shrinkage at grain boundaries, and each behaves differently under different NDT methods; ultrasonic testing, by contrast, reads flaws by sound reflection, and therefore responds best to larger or oriented reflectors rather than dispersed micro-voids [S2][S6].
What each method actually sees in a casting
Radiographic testing passes X-rays or gamma rays through the casting and records density variations on a detector; voids, inclusions, and porosity appear as contrast against denser metal, with thicker or denser sections requiring higher-energy sources and yielding lower contrast [S5]. The method is highly accurate for volumetric defects, including porosity, inclusions, voids, and lack of fusion in welds, and produces a permanent record that can be re-read for compliance or trend tracking [S5]. Acceptance is graded against reference radiographs such as ASTM E155 for aluminium and magnesium castings, ASTM E192 for investment-cast superalloys, and ASTM E1030 as the general radiographic test method for metallic castings [S3].
Ultrasonic testing sends high-frequency sound pulses into the part and reads the reflections from internal interfaces; it is positioned as the third major NDT method for castings after radiography and penetrant inspection, and is the preferred method where the operator needs depth information, where the casting is thick, or where access is limited to one side [S3][S1]. UT detects both surface and subsurface flaws and is deployable on robotic crawlers and drones for confined or hazardous environments, but it is sensitive to surface condition and coupling, and planar defects aligned parallel to the beam can be missed [S1].
Decision matrix: radiography vs UT for porosity
The right method depends on defect type, section thickness, and what the spec calls for, and neither method is universally "better" [S1]. The matrix below lines the two techniques up against the criteria that matter when a casting engineer is writing an inspection call-out.
Defect type: radiography excels at gas porosity, shrinkage porosity, inclusions, cold shuts, and hot tears because these are volumetric or produce measurable density contrast, and it is the documented best method for internal porosity in castings [S3][S4]; UT catches porosity, larger voids, and shrinkage, but tends to miss very small dispersed pores because the acoustic impedance contrast from a 0.5 mm pore is near the detection threshold of a standard transducer [S6].
Section thickness and geometry: radiography is less effective on very thick sections or complex geometries, while UT is more effective on thicker sections and where one-sided access is all that is available [S1]. For thin-walled investment castings, digital radiography delivers high contrast at low kV; for castings above roughly 50 mm in steel, high-energy X-ray sources or gamma sources are required, and UT phased-array becomes attractive because it can resolve depth [S5][S1].
Defect orientation: planar defects parallel to the X-ray beam are harder to detect with radiography, and planar defects parallel to the sound beam are harder to detect with UT, so a flat, tight crack in the through-thickness direction is a known weak spot for radiography, while a laminar shrinkage seam parallel to the probe face is a known weak spot for UT [S5][S1]. For cracks specifically, neither method replaces surface NDT: cracks are hard to detect with X-ray because they do not leave empty space in the casting, so magnetic particle or liquid penetrant inspection remains the surface method of choice [S2].
Safety, speed, and record: radiography involves strict safety protocols for radiation use and can be more time-consuming than other NDT methods, while UT gives real-time results with no radiation hazard; both methods depend on qualified operators, with radiographic interpretation typically certified to ASNT or EN ISO 9712 [S5][S1].
Where radiography wins, and where UT wins
For die-cast aluminium and investment-cast aerospace components, X-ray radiography is the workhorse because porosity is internal and radiography records it directly; an X-ray image lets the inspector describe the porosity in terms of size, shape, location, and frequency, which is what a spec reviewer needs to accept or reject a part [S4]. Standards such as ASTM E155 (aluminium and magnesium), ASTM E192 (investment-cast superalloys), and MIL-STD-2175 (defence castings) all reference radiographic reference images, so the call-out in a print is typically a radiograph severity level, not a UT dB threshold [S3].
UT earns its place on heavy steel castings, large valve bodies, and thick-walled castings where depth sizing matters, and on welds where one-sided access is the only access; phased-array UT in particular gives the inspector a cross-sectional image of the casting wall in real time, which is useful for locating shrinkage at the centre of a thick section [S3][S1]. On porosity alone, however, UT is not the first call, because dispersed micro-porosity returns weak echoes that can fall below the noise floor, especially in coarse-grained cast metals where attenuation and scattering dominate [S2][S6].
Failure modes and limitations you have to spec around
Radiography false negatives concentrate in three places: very tight cracks running parallel to the beam, very small pores below the resolution of the source-detector combination, and thick sections where contrast collapses because high-energy photons scatter through the metal [S5]. The published weaknesses are operator qualification, geometric setup (source-to-object and object-to-detector distance), and the radiation-safety overhead of running a sealed source or X-ray tube on the shop floor [S5][S1].
UT false negatives concentrate in coarse-grained austenitic stainless and high-nickel castings, where grain scattering attenuates the beam and the noise floor rises; the published weaknesses are coupling quality, surface roughness, and the need for a Level II or Level III interpreter, plus the planar-defect blind spot when a flaw is parallel to the sound beam [S1][S3]. For castings where porosity is the controlling defect rather than cracks, radiography has the more direct physical signal, and many foundries run radiography as the primary method with UT as a complement for depth verification of specific indications [S1][S4].
Spec language and standards to anchor the call-out
A defensible casting porosity spec names the method, the reference standard, and the acceptance level, because the inspector needs all three to make a binary accept/reject decision [S2]. For radiography, anchor the call-out to ASTM E1030 for the test method, ASTM E155 for aluminium and magnesium severity grades, and ASTM E192 for investment-cast aerospace superalloys; for defence work, MIL-STD-2175 sets the radiographic acceptance levels; and the operator certification line typically reads ASNT or EN ISO 9712 [S3][S5].
For ultrasonic testing on castings, the typical anchor is ASTM A609 for heavy steel castings or ASME Section V Article 4 for the UT procedure, with acceptance levels written in dB or as a reference-block response; this matters because UT on castings requires lower frequencies (commonly 1-2.25 MHz) and damping to push through coarse grain, and the spec must allow for that or the inspector will be reading grain noise rather than porosity [S3][S1].
Selection guide by casting type
Aluminium die castings: specify radiography to ASTM E1030 with ASTM E155 severity grading, because porosity is dispersed and fine, and the customer typically wants a permanent image record for traceability [S4][S3]. Investment-cast aerospace superalloys: specify radiography to ASTM E1030 with ASTM E192 reference radiographs, and add UT only for sections thicker than about 25 mm where depth sizing adds value [S3]. Steel and iron castings over 50 mm wall thickness: specify UT as the primary volumetric method, with radiography reserved for areas of known shrink or for weld repairs on the casting [S1][S6].
Welded fabrications attached to castings: combine radiography for volumetric defects with phased-array UT for depth and one-sided access, because weld geometry mixes planar and volumetric indications and no single method covers both cleanly [S1][S5]. On high-stakes fracture-critical parts, the aerospace procurement chain typically requires both methods, with radiography as the porosity screen and UT as the crack and lamination screen; this dual-method approach is also why a comparison of phased array and X-ray testing for butt welds keeps coming up in weld-procedure qualification records [S1][S9].
The next decision node to track is whether your foundry is moving from film radiography to digital radiography (DR) or computed radiography (CR), because the image library changes, the severity grading reference (ASTM E155 / E192) is unchanged, but the file format and the inspector's review workflow do, and that will change how porosity indications are archived for the next ten years [S1]. A second trackable signal is the NADCAP or ASNT Level III auditor findings on UT calibration blocks for coarse-grained castings, which is where most porosity-vs-UT false negatives actually originate [S3].
Detailed specification references: industrial x ray, tensile testing machine, and ultrasonic flowmeter.