- Non-destructive examination (NDE) of valve body castings is the engineering control that prevents sub-surface casting defects from reaching service — and the choice of NDE method, examination extent, and acceptance criteria must be driven by the service conditions and failure consequences, not by a default “do the minimum required” approach that may miss defects that matter.
- The three primary NDE methods for valve castings — radiographic testing (RT), ultrasonic testing (UT), and liquid penetrant testing (PT) — detect fundamentally different defect types and are not interchangeable: specifying RT when UT is required for thick sections, or omitting PT when surface-breaking defects are the primary concern, are engineering errors that leave real defects undetected.
- ASME B16.34 acceptance criteria for casting defects are the regulatory minimum — for safety-critical applications, more stringent acceptance criteria based on fitness-for-service analysis of the specific defect types and service loading should be specified in the purchase documents.
Why Casting Inspection Matters for Valve Reliability
Cast valve bodies are produced by pouring molten metal into a mould and allowing it to solidify. The solidification process is controlled but not perfect — shrinkage as the metal cools creates internal stresses, and the complex geometry of valve bodies creates conditions where solidification is uneven, generating internal voids, cracks, and inclusions that cannot be seen from the outside. These sub-surface defects are the time bombs in cast valve populations: they cause no problem during hydrostatic testing (where the test duration is too short and the test pressure too low to propagate a small crack to failure), but they propagate progressively under the cyclic loading of temperature and pressure cycling in service, eventually causing a leak or fracture.
The frequency of consequential casting defects in valve production varies with the casting process, alloy, complexity of the valve body geometry, and the foundry’s quality control practices. In a well-controlled foundry with systematic NDE, the defect rate in critical sizes and service classifications may be below 1%. In a poorly controlled foundry or in complex alloys prone to porosity (such as austenitic stainless steel castings), the defect rate can exceed 10% in thick-section areas. The economic case for systematic NDE is straightforward: the cost of NDE is small compared to the cost of a field failure, particularly in safety-critical or high-consequence service.
Radiographic Testing of Valve Castings
Radiographic testing (RT) — X-ray or gamma-ray imaging of the casting with film or digital detector on the opposite side — is the most widely specified NDE method for valve body castings because it provides a permanent photographic record of the internal condition that can be reviewed by independent inspection personnel. RT detects volumetric defects (porosity, shrinkage cavities, inclusions) that create contrast differences in the transmitted radiation. It is less sensitive to planar defects (cracks, cold shuts) that are oriented perpendicular to the beam direction because these provide little thickness variation and hence little radiation contrast.
ASTM E446 (reference radiographs for steel castings to 2 inches thick) and ASTM E186 (reference radiographs for steel castings 2–4.5 inches thick) provide the reference images against which the actual casting radiographs are compared. The acceptance criteria in ASME B16.34 reference these ASTM standards and define the maximum allowable severity level for each defect type. The interpretation of radiographs requires certified personnel — ASNT Level II or Level III — who compare the film images against the reference radiograph standards and judge whether defects observed are within or outside the acceptance criteria.
Ultrasonic Testing for Thick-Section Castings
Radiographic testing has a practical limitation for thick-section valve bodies (walls above approximately 75–100 mm) — the radiation source power required to achieve adequate film density through thick metal sections makes RT technically difficult and potentially dose-rate limited in populated work environments. Ultrasonic testing (UT) is better suited to thick sections because the ultrasound penetrates the material in a beam that can detect defects anywhere along the beam path without radiation dose considerations.
Phased array ultrasonic testing (PAUT) is the modern successor to conventional UT for valve body examination. PAUT uses an array of piezoelectric elements that can be electronically steered and focused to scan the volume of the casting, producing a cross-sectional image (B-scan or S-scan) that shows the location, orientation, and extent of any reflectors in the beam path. PAUT can detect both volumetric defects (like RT) and planar defects oriented at various angles to the examination surface (unlike conventional RT), making it a more comprehensive examination technique for thick valve bodies in safety-critical service.
Liquid Penetrant and Magnetic Particle Testing
Liquid penetrant testing (PT) is the standard method for detecting surface-breaking defects on non-magnetic materials — austenitic stainless steel, duplex stainless steel, nickel alloy, and aluminium valve bodies. The penetrant is applied to the clean metal surface, allowed to penetrate into any surface-opening cracks or pores by capillary action, the excess is removed, and a developer is applied that draws the penetrant out of the defects and makes them visible as coloured or fluorescent indications. PT does not detect sub-surface defects — only those that intersect the surface — but surface-breaking cracks (particularly at sharp-radius sections, internal corners, and weld-to-casting junctions) are a significant failure risk in high-pressure service and must be detected before the valve enters service.
Magnetic particle testing (MT) detects both surface-breaking and near-surface (up to approximately 3 mm depth) defects in ferromagnetic materials (carbon steel, low-alloy steel, and ferritic stainless steel). The casting is magnetised and iron particles applied — the particles accumulate at magnetic flux leakage locations caused by defects. MT is faster than PT for large numbers of ferromagnetic castings and is the standard method for surface examination of carbon and low-alloy steel valve bodies. MT cannot be used on non-magnetic materials (austenitic stainless, nickel alloys) and must be followed by demagnetisation to prevent residual magnetism from affecting downstream processing or measurement.