Views: 0 Author: J-VALVES Publish Time: 2026-11-06 Origin: Site
A forged body starts as a billet that has been worked, so its flaws are few and mostly flat. A cast body solidifies in a mould, and every defect it carries forms in the minutes between pouring and freezing. Shrinkage, gas, inclusions and hot tears all appear where the metal froze last or could not contract, so inspecting a casting means checking a known list at known places.
Shrinkage porosity forms in the hot spot, the section that freezes last because thinner metal surrounds it. In a cast steel gate valve body that is usually the junction between barrel and port, or the metal under a flange hub; it looks like a sponge on a fracture face and is nearly always internal. Gas porosity is rounder, from hydrogen or nitrogen released during freezing or from moisture in the mould. Sand and slag inclusions are non-metallic and settle where the flow slowed. Hot tears are intergranular cracks opened while the casting was still weak, starting at a change of section.
Radiography images density, so it is the natural method for porosity, shrinkage and inclusions. MSS SP-54 is the valve industry reference for steel castings, with acceptance levels written against reference radiographs; ASTM E446 covers sections up to 2 inch, E186 from 2 to 4 1/2 inch and E280 heavy walls above that. Two limits decide the result. Sensitivity falls as the wall thickens, so the same indication grades differently in a 1 inch and a 4 inch wall. And a tight crack standing parallel to the beam may not appear at all.
Ultrasonic testing finds internal planar defects, cracks, cold shuts and laminar shrinkage, and still works on thick sections where radiography runs out of energy. On cast steel it is harder than on wrought product: coarse grains scatter the beam, the screen fills with grass and small indications drown in it. Austenitic and duplex castings are worse, because columnar grains are anisotropic and bend or split the beam. ASTM A609 describes the practice, and a result only means something against a reference block of the same material and section.
MT and PT examine the surface, and they catch what matters in service, because fatigue and corrosion both start there. The rule most often forgotten is that magnetic particle inspection needs a ferromagnetic material. Carbon steel and 13 percent chrome bodies can be magnetised; CF8, CF8M, CF3M and duplex castings cannot, so a drawing that asks for MT on a CF8M body is unenforceable and penetrant testing is what actually gets done. Penetrant needs a clean, opened surface, which is why the machined gasket face is examined and the rough boss beside it is not.
Acceptance depends on location as much as on size. A shrinkage cavity in a boss that carries no pressure is cosmetic; the same cavity under a seat ring, or in the wall between bore and bonnet joint, is a leak waiting for a thermal cycle. Three zones deserve their own line in the specification: the seating area and the body to bonnet gasket face, the stuffing box wall, and the flange raised faces. A workable split sets radiography or ultrasonics on the pressure boundary, penetrant on the machined faces, and visual examination under MSS SP-55 on the rest.
Most standards permit weld repair and most foundries rely on it. What makes a repair acceptable is the paper behind it: a procedure qualified to ASME BPVC Section IX, a welder qualified to that procedure, preheat matched to the carbon equivalent, post-weld heat treatment where required, and re-examination of the repaired area by the same method and acceptance level as the original. Two things go wrong commercially: the repair is made and not recorded, so the certificate describes a casting that was never shipped; or it is made on a machined sealing face, where distortion costs more than the original defect.
| Defect | Where it forms | Best method | What it does in service |
|---|---|---|---|
| Shrinkage porosity | Hot spot, port to barrel junction | Radiography, ultrasonic on thick walls | Weeps under the seat ring or through the wall |
| Gas porosity | Last metal to freeze, near cores | Radiography | Cosmetic unless it breaks a machined face |
| Sand and slag | Where the flow slowed | Radiography, penetrant after machining | Laps open on the seating face |
| Hot tear | Change of section, core print | Penetrant, ultrasonic | Crack grows under thermal cycling |
| Cold shut | Long thin sections, cold metal | Radiography, ultrasonic | Laminar weakness, leaks when stressed |
On an A352 LCB or LCC body the impact test is the property being bought, and welding changes it. Charpy specimens come from a test block cast from the same heat and heat treated in the same furnace load, so a repair that needs a second heat treatment separates the casting from its block unless the block goes through the furnace too. That is the failure which passes every dimensional check and then fails at minus 46 degrees C.
Five lines settle it: which zones are examined and by which method, the acceptance level against MSS SP-54 or ASTM E446 and the reference block for ultrasonics, the surface method and whether it runs after machining, the weld repair rule including who approves a repair and what is re-examined, and the documents that ship with the valve. J-VALVES supplies API 600 gate valves from 1/2 to 56 inch, ASME Class 150 to 2500, with shell tests at 1.5 times and seat tests at 1.1 times rated pressure, documented through quality control. The other pressure boundary on the same valve is the body joint, covered in bolted and welded bonnet joints.
Can radiography find a crack in a cast steel valve body? Only sometimes. Radiography is sensitive to volumetric defects such as porosity, shrinkage and inclusions, but a tight crack is sensitive to orientation: one parallel to the beam can be invisible, and one at a shallow angle shows as a faint line. Cracks are better found by penetrant or magnetic particle at the surface, and by ultrasonic testing inside the wall.
Can a cast steel valve body be weld repaired? Yes, and most standards allow it, but only as a controlled procedure: qualified welding procedure and welder, preheat matched to the carbon equivalent, post-weld heat treatment where required, re-examination to the original acceptance level, and a record on the certificate. Repairs inside gasket faces and seating areas are commonly prohibited.
Why is penetrant used instead of magnetic particle on a CF8M body? Because magnetic particle inspection requires a ferromagnetic material, and austenitic stainless castings such as CF8, CF8M and CF3M are not magnetic. A magnetic particle requirement on that material cannot be executed as written, so surface examination is done with penetrant after machining.