Why Do 2205 Duplex Fittings Crack After Heat Treatment? Understanding Sigma Phase and How to Verify Solution Annealing

Introduction

You ordered 2205 duplex fittings. They arrived with a mill certificate showing correct chemistry, passed dimensional inspection, and were installed. Then, during hydrotest or shortly after commissioning, a crack appeared—not at a weld, not at a mechanical joint, but in the fitting body itself. Or worse: the fitting held during installation but failed in service within months, with a brittle fracture that shows no sign of prior deformation.

The root cause is almost never the alloy composition. It is almost always the heat treatment.

2205 duplex stainless steel is uniquely sensitive to its thermal history in ways that 304, 316, or even super duplex 2507 are not. The grade’s two-phase microstructure—roughly equal parts ferrite and austenite—delivers its strength and corrosion resistance, but that same microstructure is thermodynamically unstable at intermediate temperatures. When a 2205 fitting passes slowly through the 600–900°C range during cooling after solution annealing, a brittle intermetallic phase called sigma (σ) precipitates. Even 1% sigma by volume is enough to destroy the impact toughness that makes duplex stainless steel suitable for pressure service.

This article explains what sigma phase is, why it forms during heat treatment, what cracking looks like when it does, and—most importantly—how to verify that your fittings were properly solution annealed before they leave the supplier’s facility.

What Is Sigma Phase and Why Is It So Damaging?

Sigma phase is a chromium- and molybdenum-rich intermetallic compound that forms in the ferrite phase of duplex stainless steel when the material is held or cooled through the temperature range of approximately 600°C to 1000°C (1,112°F to 1,832°F). The precipitation starts at around 920°C, with the maximum rate of formation occurring at approximately 850°C.

The formation mechanism is well understood. Sigma phase nucleates preferentially at ferrite-ferrite and ferrite-austenite grain boundaries and then grows into the adjacent ferrite grains. As it grows, it consumes chromium and molybdenum from the surrounding ferrite, causing the ferrite to transform into secondary austenite. Molybdenum is the main element controlling the precipitation kinetics, though chromium diffusion also contributes.

The damage is threefold:

Loss of toughness. Sigma phase is hard and brittle. Charpy impact tests on V-notched 2205 specimens show that the material has lost most of its impact toughness already at about 1% sigma content. This means a fitting that would normally absorb significant impact energy during installation or hydrotest can shatter under the same conditions.

Loss of corrosion resistance. Sigma phase is rich in chromium and molybdenum—the very elements that give 2205 its corrosion resistance. When these elements are tied up in the sigma phase, the surrounding matrix is depleted in them, reducing the alloy’s ability to form and maintain its protective passive film. Pitting and crevice corrosion resistance drop sharply.

Crack initiation sites. The interfaces between sigma phase and the surrounding ferrite or austenite are high-energy boundaries. Cracks are susceptible to propagate along or across these interfaces, and the sigma phase particles themselves can break off, severely deteriorating mechanical properties.

A failure analysis of a 2205 flange that cracked prematurely in offshore service found 17.0 vol% sigma phase in the heat-affected zone and 4.7 vol% in the base metal. The hardness in the affected zone reached 394 HV0.3, and tensile and impact fractures were brittle cleavage fractures. The root cause was improper aging during heat treatment, compounded by excessive welding heat input.

Why Does Sigma Phase Form During Heat Treatment?

The answer is simple: slow cooling through the critical temperature range.

The correct heat treatment for 2205 duplex is a full solution anneal followed by rapid cooling. Solution annealing dissolves any pre-existing sigma phase and restores the balanced ferrite-austenite microstructure. But if the cooling rate after annealing is too slow, the material spends too long in the sigma-forming temperature range, and the phase precipitates.

The standard heat treatment specification for 2205 requires:

• Solution annealing temperature: 1020°C to 1100°C

• Minimum temperature for most wrought product specifications: 1040°C (1,900°F)

• Cooling: rapid quench, typically water quenching or rapid cooling by other means

Allowing a fitting to cool in air from the annealing temperature, or furnace cooling, or even slow cooling in an insulated pack, is sufficient to precipitate sigma phase. The critical window is the 700°C to 980°C range, where sigma formation kinetics are fastest.

A chemistry certificate is not a heat treatment certificate. ASTM A923 specifically warns that "compliance with the chemical and mechanical requirements for the applicable product specification does not necessarily indicate the absence of detrimental phases in the product". The chemistry of a heat of 2205 can be perfect and still produce fittings full of sigma phase if the heat treater cooled them incorrectly.

The Common Heat Treatment Mistakes

Mistake 1: Air Cooling Instead of Water Quenching

Water quenching is the standard for 2205 fittings, flanges, and pipe. It removes heat from the part fast enough that the material passes through the sigma-forming range in seconds rather than minutes or hours. Air cooling, or even forced air cooling, is generally insufficient for anything thicker than thin sheet.

Mistake 2: Stacking Parts During Cooling

When multiple fittings are quenched in a basket, the parts in the center may not receive the same cooling rate as those on the outside. If the quench is not agitated or the parts are stacked densely, the interior parts can cool slowly enough to precipitate sigma. Heat treaters who are experienced with carbon steel or austenitic stainless may not appreciate how critical quench uniformity is for duplex.

Mistake 3: Insufficient Time at Solution Annealing Temperature

The part must reach the annealing temperature throughout its cross-section, not just at the surface. A thick flange or a heavy elbow requires sufficient soak time to homogenize. If the soak is too short, the core of the part may not fully dissolve prior sigma phase, and subsequent cooling will add more.

Mistake 4: Reheating for Stress Relief

Duplex stainless steel should not be stress-relief annealed in the same way as carbon steel or austenitic stainless. Any reheating into the 600–1000°C range for stress relief will precipitate sigma phase. If stress relief is required after welding or forming, it must be a full solution anneal at 1040°C minimum followed by water quenching, not a lower-temperature stress relief.

How to Verify Solution Annealing: The ASTM A923 Test Methods

The standard method for detecting detrimental intermetallic phases in duplex stainless steel is ASTM A923, which provides three complementary test methods:

Method A: Sodium Hydroxide Etch Test

This is a metallographic test. A polished cross-section of the fitting is etched in a sodium hydroxide solution, and the microstructure is examined under a microscope. Sigma phase etches differently from ferrite and austenite, appearing as a distinct third phase, typically in the form of islands or networks at the ferrite-austenite boundaries.

Method A is the most direct visual confirmation of sigma phase presence. It requires a skilled metallographer and a properly prepared sample, but it is relatively fast and inexpensive. The test is qualitative to semi-quantitative: it will tell you whether sigma phase is present, and in severe cases, the approximate volume fraction can be estimated.

Method B: Charpy Impact Test

This is the most sensitive indicator of embrittlement from sigma phase. Specimens are machined from the fitting and tested at a specified temperature (typically -46°C for 2205 per ASTM A923). The absorbed energy is measured and compared to the acceptance criterion.

Method B is the test that most directly correlates to the property that matters in service: the ability of the material to resist brittle fracture. A fitting that passes Charpy impact at the specified temperature is unlikely to fail by brittle cracking in service, even if small amounts of sigma are present.

Method C: Ferric Chloride Corrosion Test

This test measures the corrosion resistance of the material in a standard ferric chloride solution. Sigma phase reduces corrosion resistance by depleting chromium and molybdenum from the matrix, so a fitting with sigma will show higher corrosion rates than a properly annealed one.

Method C is particularly useful for verifying that the heat treatment has not degraded the corrosion performance of the fitting in the intended service environment. It is also the method most commonly specified by end users who are concerned about pitting and crevice corrosion in seawater or chloride-containing process streams.

Practical Verification Checklist for Buyers

If you are purchasing 2205 duplex fittings for a critical application, here is what to require from your supplier:

1. Specify the Correct Product Standard

2205 fittings are covered by different ASTM standards depending on the product form:

• ASTM A182 F51 covers forged flanges and fittings

• ASTM A815 covers wrought butt weld fittings

• ASTM A995 covers cast fittings (Grades 4A/CD3MN)

The product standard specifies the required solution annealing temperature and the required testing. Make sure your purchase order references the correct standard.

2. Require ASTM A923 Testing on the Actual Fittings

Do not accept a mill certificate that simply states “solution annealed.” Require ASTM A923 test results for the actual production fittings, not just the heat of raw material. The heat treatment is performed after the fitting is formed, and the cooling rate depends on the fitting’s geometry and mass. A thin pipe that cooled rapidly may pass, while a thick flange from the same heat may fail.

For critical applications, specify ASTM A923 Method B (Charpy impact) as the acceptance test. It is the most direct measure of embrittlement and the most difficult to falsify with paperwork alone.

3. Require Test Specimens from the Thickest Section

Sigma phase precipitation is a function of cooling rate. The slowest-cooling section of a fitting—typically the thickest wall or the center of a heavy flange—is the most likely location for sigma to form. Require that Charpy specimens be taken from the thickest section, not from a convenient thin section that cools faster.

4. Consider Supplementing with ASTM A1084 for Lean Duplex

If your project includes lean duplex grades (such as 2304 or LDX 2101), ASTM A1084 provides the corresponding test methods for detecting detrimental phases in those alloys. The principles are the same: verify that the heat treatment has not created brittle intermetallic phases.

5. Require Ferrite Content Verification

The target ferrite content for solution-annealed 2205 is 40% to 60% ferrite in an austenite matrix. A ferrite content significantly outside this range indicates either improper solution annealing or sigma phase precipitation (which consumes ferrite). Ferrite measurement can be performed by magnetic methods or by metallographic point counting.

Advanced Detection Methods for Field Verification

For fittings already in service or in inventory where destructive testing is not possible, several non-destructive methods are available:

Low-Field Magnetic Analysis (LFMA). Sigma phase is non-magnetic, while ferrite is ferromagnetic. As sigma precipitates, the magnetic response of the material changes. LFMA uses a Hall effect sensor to measure magnetic flux density and can detect sigma phase without cutting the part.

Magnetic noise and first harmonic analysis. This technique applies a sinusoidal magnetic excitation to the component and analyzes the magnetic noise and harmonic content of the response. It has been demonstrated to detect sigma phase in SAF 2205 steel.

Heat tinting metallography. For laboratory confirmation, heat tinting is a thermal etching technique that reveals the phases in duplex stainless steel through interference colors, making sigma phase clearly distinguishable from ferrite and austenite.

Electron backscatter diffraction (EBSD). This is the most precise method for phase identification and quantification. EBSD maps the crystal structure of the sample and can distinguish ferrite, austenite, and sigma phase with high spatial resolution. It is typically used for failure analysis rather than routine quality control.

A Summary of Acceptance Criteria

Key Takeaways

Sigma phase forms during slow cooling, not during the anneal itself. The solution annealing temperature is critical, but the cooling rate after annealing is equally critical. Water quenching is not optional for 2205.

ASTM A923 is the verification tool. Method A (etch), Method B (Charpy), and Method C (ferric chloride) each verify a different aspect of heat treatment quality. For critical service, specify all three.

A chemistry certificate proves the heat, not the heat treatment. Do not accept a mill certificate as evidence that the fitting was properly solution annealed. Require ASTM A923 results on the actual product.

Test the thickest section. Sigma phase forms where cooling is slowest. A fitting that passes Charpy from a thin section may still contain sigma in the core.

1% sigma is already too much. The impact toughness of 2205 drops dramatically at 1% sigma. There is no safe threshold for sigma phase in pressure-containing components.