ASTM A923 Deleterious Phase Detection for Duplex Stainless Steels

Overview and Scope of the Standard

ASTM A923 is the principal specification in the United States and a widely referenced document globally for verifying that the microstructure of duplex stainless steels such as UNS S31803 (2205) and UNS S32750 (2507) does not contain deleterious intermetallic phases. The standard defines three independent test methods — Method A (micrographic), Method B (impact), and Method C (corrosion) — and requires that all three be passed for a material lot to be considered acceptable. In engineering practice, particularly for hydrogenation reactors, sour gas separators, and offshore topside equipment, a single-method pass is never sufficient; the three-method combination is mandatory because each method probes a different manifestation of the same metallurgical problem.

Interpretation of the Three Methods

Method A is a micrographic evaluation performed on a transverse section of the material. The micrograph is examined at 200x and 500x magnification after appropriate etching. The critical phases of concern are sigma (σ), chi (χ), R-phase, and intermetallic chromium nitrides (Cr2N). The acceptance criterion is that the total area fraction of all deleterious phases must not exceed 2 percent when evaluated at 200x magnification. The engineer must pay close attention to the etchant selection — a standard 2 percent nital etch often fails to reveal sigma, whereas a specific reagent such as a 50 percent ammonium hydroxide and 50 percent hydrogen peroxide solution or a proprietary etchant is required. The micrograph is divided into a grid, and the area fraction is estimated by point counting or image analysis.

Method B is a Charpy V-notch impact test performed at a temperature that is sensitive to the presence of intermetallic phases. For 2205, the test temperature is typically 4 degrees Celsius (40 degrees Fahrenheit), while for 2507 the temperature is 16 degrees Celsius. The acceptance criterion is that the average absorbed energy must be at least 47 joules, with no individual specimen falling below 34 joules. The rationale is that sigma and chi phases embrittle the ferrite-austenite matrix and cause a sharp drop in absorbed energy at relatively mild temperatures. Method B is therefore a functional, macroscopic confirmation of microstructural soundness.

Method C is a corrosion test that exposes a coupon to a specific intergranular corrosion environment. For 2205, the test medium is 6 percent hydrochloric acid at 30 degrees Celsius for 24 hours, with the criterion being that the corrosion rate must not exceed 0.025 millimeters per year and the depth of attack at the grain boundary must not exceed 25 micrometers. For 2507, a more aggressive environment is used, reflecting the higher nickel and molybdenum content. The test is designed to reveal chromium nitride and sigma phases that form preferentially at grain boundaries.

Engineering Practice and Common Failure Modes

In my experience fabricating clad pressure vessels with 2205 or 2507 overlay layers, the most common cause of A923 failure is improper heat input control during welding. When the welding heat input exceeds approximately 2.5 kilojoules per millimeter on a 2507 cladding layer, the material passes through the 600 to 900 degrees Celsius temperature range slowly enough to precipitate sigma and chi phases. The countermeasure is to use a hot-wire TIG or pulsed GTAW process with tightly controlled heat input, to limit the number of passes, and to perform an interpass temperature control below 150 degrees Celsius.

Another frequent failure mode is the welding of a 2205 overlay onto a carbon steel base plate with a filler metal that is not properly balanced. If the filler metal is too austenite-rich, the resulting weld metal may shift toward a fully austenitic structure and lose the duplex character entirely; if it is too ferrite-rich, sigma precipitation risk increases dramatically. The recommended filler metals are ER31803 for 2205 and ER32750 or a higher-nickel variant for 2507, and the weld metal must be verified to contain between 35 and 65 percent ferrite.

Key Reflections and Implications

The A923 standard is not merely a quality gate; it is a diagnostic tool that tells the fabricator whether the entire thermal history of the component is acceptable. When a lot fails Method B but passes Method A, the engineer should suspect that the material has been subjected to a thermal cycle that did not produce a large area fraction of intermetallics but did produce enough embrittlement to drop the impact energy — often the signature of a slow-cool heat treatment or an over-tempered condition. When Method C fails while A and B pass, the issue is almost always localized grain boundary precipitation that is too fine to be seen at 200x but is aggressive enough to be attacked by the HCl solution.

For engineers working on export projects to the Middle East or Southeast Asia, where 2205 and 2507 are increasingly specified for sour gas and offshore applications, I recommend that the A923 testing be performed on the as-welded condition of a production weld, not merely on the base plate or a test coupon. This is because the welding process itself is the primary driver of deleterious phase formation, and a base plate that passes A923 may still produce a weld that fails. The cost of a single A923 test is small compared to the cost of a pressure vessel that is rejected at hydrostatic test or that suffers intergranular corrosion in service within a few years.

Understanding A923 deeply transforms the way one approaches duplex steel fabrication. It moves the engineer from a reactive stance — testing after welding and hoping — to a proactive stance — controlling the thermal history, selecting the correct filler metal, and verifying the microstructure at every critical step. This is the difference between a project that is accepted at first inspection and one that is rejected and reworked at great cost and schedule delay.