Physical Address
304 North Cardinal St.
Dorchester Center, MA 02124
ASME STP-PT-088-2017 — Effect of Cooling Rate on the Microstructure and Properties of Duplex Stainless Steel Welds
🌡️ ASME STP-PT-088-2017 is a technical report funded by ASME Standards Technology, LLC (ASME ST-LLC), coordinated by Kiefner & Associates, Inc., with metallurgical testing and analysis contracted to The Ohio State University. The study systematically evaluates the effect of cooling rate on the impact toughness of the heat-affected zone (HAZ) of welds in six duplex stainless steel alloys. The austenite/ferrite phase balance of the HAZ is governed by weld cooling rate; cooling too fast can leave a HAZ microstructure with lower toughness than the unaffected base material — exactly the effect this report quantifies.
💡 Engineering value: the report supplies systematic “cooling rate – ferrite content – impact toughness” data for duplex stainless steels, especially high-alloy grades such as 2205 and 2507, directly supporting control of welding PQR variables (heat input, interpass temperature) required by ASME and industry codes.
1. Background and Literature Review 📌
Duplex stainless steels combine an approximately 50:50 ferrite (α) / austenite (γ) two-phase structure, offering austenitic toughness with ferritic strength and resistance to stress-corrosion cracking. Under a weld thermal cycle, however, the HAZ cools rapidly; if cooling is too fast the austenite cannot re-precipitate sufficiently, leaving excess ferrite and lower toughness, and in severe cases precipitating harmful phases. Chapter 2 of the report systematically reviews previous work on the “relationship of cooling rate to toughness in duplex stainless steels” and “duplex stainless steel welding industry practice,” listing the welding procedure qualification variables recommended by API TR 938-C as a baseline against which the study’s own matrix is compared.
| Alloy | UNS Number | Grade Type | Typical Applications |
|---|---|---|---|
| 2304 | S32304 | Lean duplex | Food, chemical tanks, medium-strength structures |
| 2003 | S32003 | Lean duplex | Pulp and paper, seawater heat exchangers |
| 2101 | S32101 | Lean duplex (LDX) | Bridges, pressure vessels, seawater piping |
| 2205 | S32205 | Standard duplex | Oil and gas, chemical, marine engineering |
| 2507 | S32750 | Super duplex | Sour gas, deep-water risers, high-chloride media |
| Zeron 100 | S32760 | Super duplex | Offshore platforms, pumps and valves, corrosive duty |
⚠️ The six alloys span the full spectrum from lean duplex (2304/2003/2101) through standard duplex (2205) to super duplex (2507/Zeron 100); their corrosion resistance and cost differ enormously, and filler metals and process parameters must not be interchanged.
2. Test Approach and Methods 🔬
The program combined three parallel lines of evidence — thermal simulation, computer modeling, and actual weld testing:
- Gleeble 3800® thermal simulation: imposing different cooling curves on the six alloys to reproduce the weld HAZ thermal history, then measuring ferrite number (FN) and performing Charpy V-notch (CVN) impact testing on the simulated specimens.
- Computer model simulation: using t12-8 (the time to cool from 1200 °C to 800 °C) as the cooling-rate parameter to predict cooling curves for different welding processes (GMAW/GTAW) and heat inputs, guiding welding-parameter selection.
- Weld testing: preparing standard butt-joint configurations, welding with ER2209 filler (with 2003/2304/2205/2101) and ER2594 filler (with 2507), covering different heat-input levels, and measuring weld-metal and fusion-boundary microstructures plus CVN toughness at -40 °C.
🚨 The biggest difference between Gleeble simulation and real welding is stress state and multi-pass thermal cycling: simulations reliably establish the cooling–microstructure–toughness trend, but procedure qualification must still rely on actual weld tensile, impact, and bend testing.
3. Cooling Rate and Ferrite Number 📊
Table 4-1 of the report gives ferrite-number (FN) results for the Gleeble thermal-simulation samples. As cooling rate increases (t12-8 shortens), austenite re-precipitation is incomplete and ferrite content rises. The alloys differ in sensitivity: lean and standard grades tolerate moderate cooling rates, while super-duplex grades (2507, Zeron 100) require much tighter cooling control to retain sufficient austenite content and toughness margin.
| Alloy | UNS | Fast-cooling Microstructure Character | CVN Toughness (-40 °C) Trend |
|---|---|---|---|
| 2304 | S32304 | Ferrite enrichment, austenite as grain-boundary films | Drops markedly at fast cooling |
| 2003 | S32003 | Austenite precipitation sensitive | Moderately sensitive |
| 2101 | S32101 | Austenite precipitation aided by nitrogen | Moderately sensitive |
| 2205 | S32205 | Standard balanced (ferrite/austenite) structure | Good toughness when properly controlled |
| 2507 | S32750 | High alloy, narrow austenite window | Cooling must be tightly controlled |
| Zeron 100 | S32760 | Similar high-alloy sensitivity to 2507 | Cooling must be tightly controlled |
Figures 4-1 through 4-6 show the microstructure evolution of the six alloys as a function of cooling rate; Figures 4-7 through 4-12 give the impact-energy transition curves and shear-area transition curves, illustrating the toughness margin at -40 °C and below.
4. Principal Conclusions and Engineering Significance 🛠️
- ✅ Cooling rate is the dominant HAZ toughness variable: too-fast t12-8 cooling leaves insufficient austenite re-precipitation and high ferrite numbers, so HAZ toughness falls below the base metal; managing heat input and interpass temperature controls the cooling rate effectively.
- ✅ Sensitivity is graded by alloy: lean and standard duplex grades are relatively tolerant of moderate cooling rates, while super-duplex grades demand stricter cooling control and better-matched filler metals.
- ✅ Cooling curves can be predicted: computer models based on t12-8 (Appendix A gives simulation results for 2205/2507 under GMAW and GTAW) allow a proposed welding-parameter combination to be judged in advance for hazardous cooling rates.
- ✅ Filler matching and weld verification: actual-weld tests with ER2209/ER2594 fillers (Table 4-3 gives -40 °C CVN results) confirm that the thermal-simulation trends hold in real welding.
5. Engineering Practice Recommendations 🏭
- 🔍 Manage heat input: duplex welding should observe both a minimum heat input (to avoid overcooling and brittleness) and a maximum interpass temperature (to avoid harmful-phase precipitation) — two extremes to avoid.
- 🔧 Filler and shielding gas: higher-nickel fillers (e.g., ER2209/ER2594) promote austenite re-precipitation; nitrogen-bearing shielding gas improves the austenite balance.
- 📊 Ferrite-content inspection: verify ferrite content in weld metal and HAZ by magnetic (FN) or metallographic methods per the specification, typically within about 30%–70% depending on project requirements.
- 📋 PQR variables: list heat input, interpass temperature, and pass thickness as essential variables; qualify within the ranges recommended by documents such as API TR 938-C, with impact testing commonly at -40 °C.
6. Frequently Asked Questions ❓
❓ Q: Why must cooling rate be controlled so carefully when welding duplex stainless steel?
A: The HAZ re-establishes its two-phase balance through austenite re-precipitation during cooling; too fast a cool leaves insufficient austenite, excess ferrite, and possible σ-phase precipitation, collapsing toughness and corrosion resistance. Controlling cooling rate means controlling phase balance.
A: The HAZ re-establishes its two-phase balance through austenite re-precipitation during cooling; too fast a cool leaves insufficient austenite, excess ferrite, and possible σ-phase precipitation, collapsing toughness and corrosion resistance. Controlling cooling rate means controlling phase balance.
❓ Q: What is t12-8 and why use it to characterize cooling?
A: t12-8 is the time (in seconds) to cool from 1200 °C to 800 °C — the window in which austenite re-precipitation and harmful-phase precipitation compete. Larger values mean slower cooling. The report uses it as a unified cooling-rate variable for both computer and thermal simulation.
A: t12-8 is the time (in seconds) to cool from 1200 °C to 800 °C — the window in which austenite re-precipitation and harmful-phase precipitation compete. Larger values mean slower cooling. The report uses it as a unified cooling-rate variable for both computer and thermal simulation.
❓ Q: How does super-duplex (2507, Zeron 100) welding differ from 2205?
A: Super-duplex grades are more highly alloyed, have a narrower austenite-precipitation window, and are more sensitive to cooling rate. They require higher-nickel fillers (e.g., ER2594), stricter heat-input and interpass control, and often a lower ferrite-content ceiling.
A: Super-duplex grades are more highly alloyed, have a narrower austenite-precipitation window, and are more sensitive to cooling rate. They require higher-nickel fillers (e.g., ER2594), stricter heat-input and interpass control, and often a lower ferrite-content ceiling.
🔍 Q: What is this report’s significance for the ASME code system?
A: As an STP-series technology publication, it provides the data foundation for revisions of duplex stainless steel welding and inspection clauses in codes such as ASME Section VIII and B31, and is commonly cited as background support for welding procedure qualification.
A: As an STP-series technology publication, it provides the data foundation for revisions of duplex stainless steel welding and inspection clauses in codes such as ASME Section VIII and B31, and is commonly cited as background support for welding procedure qualification.
