Physical Address
304 North Cardinal St.
Dorchester Center, MA 02124
ASME STP-PT-031-2010 — Pressure Induced Fatigue
📈 ASME STP-PT-031-2010 is a Pressure Technology research report prepared by Joseph A. Kapp, PhD, PE (Elmhurst Research, Inc.) under project 07-06 (B31#3). In response to several fatigue failures of welding tees and branch connections in B31.1 piping caused by pressure cycling, the report systematically evaluates cyclic-pressure fatigue test data from 41 piping intersections (welded tees/branch connections), validates the classic Markl power-law relationship i·S·N^0.2 = C, compares the data against the ASME Section VIII Division 2 and Division 3 fatigue design curves, and examines experimental determination of the pressure-induced stress intensification factor ip. It provides the basis for the B31 Technical Committee on Mechanical Design (MDC) to develop fatigue evaluation rules for pressure cycling of piping components.
💡 Piping-component fatigue performance has long been understood in terms of thermal-expansion (secondary) stresses, but fatigue from pressure cycling is less well understood. With higher-strength materials (thinner components) and higher pressures (e.g., hydrogen fuel technology), pressure cycling effects are expected to become much more pronounced — this report lays the data and method foundation for that shift.
1. Background and Study Objectives 📌
- The B31.1 committee received reports of multiple fatigue failures of welded tees and branch connections due to pressure cycling; a long-standing MDC agenda item sought stress multipliers of some sort for fatigue evaluation of piping components under pressure cycling.
- In the traditional Markl approach, fatigue life follows a power-law relationship between applied stress and life: i·S·N^0.2 = C, where N is life, C a material constant, i the stress intensification factor (SIF), and S the nominal stress; Markl suggested C = 245 ksi for welded carbon steel.
- For pressure-induced fatigue, the nominal stress is the tangential stress in the pipe from internal pressure, and ip is determined experimentally by actual fatigue testing — experimental development of ip is considered superior to analytical methods because actual welded structures are not well described by continuous homogeneous models.
2. Test Data and Markl Analysis 🔧
| Item | Content / Result |
|---|---|
| Test scale | 41 piping intersections (welded tees/branch connections) cyclic-pressure tested; data from Decock’s 119-vessel program with intersections used as supplementary reference |
| Data fit | Fatigue results follow the Markl-type power law (linear Log S – Log N, exponent 0.2) with considerable scatter |
| Scatter source | Inherent variability of fatigue failure in large welded structures |
| Design-curve comparison | Compared with Section VIII Div 3 fatigue analysis, Div 2 analysis method (calculated ip values), and safety factors for various design curves (Tables 4/5) |
ip·S·N0.2 = C (C ≈ 245 ksi, welded carbon steel)
⚠️ The report emphasizes the “equivalent-life principle” in defining ip: for the same fatigue life, take the ratio of the nominal stress in a welded pipe without the intersection to the nominal stress in a pipe with the intersection. Tests confirm this definition applies consistently across the 41 intersections, but the scatter warns that a conservative design curve must be used in engineering.
3. Life-Prediction Results and Design-Curve Selection 🧪
- Figures 1/2 compare measured fatigue lives with Section VIII Div 3 and Div 2 design curves: the data lie broadly within the Markl power-law band, and the relationship to code design lines depends on stress level and joint detail.
- Tables 2/3 give the design-life analysis results by the Div 3 and Div 2 methods; Table 4 compares measured results with Equation (7) as the design curve; Table 5 summarizes safety factors for various design curves.
- The conclusion is that several design curves are appropriate as a design rule for pressure-induced fatigue; engineering should select by material, joint type, and desired margin. The Markl power-law form (ip·S·N^0.2 = C) remains the simplest and most data-consistent basis.
4. Engineering Practice Essentials 🛠️
- 🔍 Distinguish stress types: pressure-cycling fatigue is dominated by primary stress (tangential membrane), unlike thermal-expansion secondary stress; it cannot simply reuse the same SIF — the pressure-specific ip is required.
- 🔧 Joint details: tee/branch welded joints are the weak points for pressure-cycling fatigue; stress concentration, weld residual stress, and defects act together. Optimize branch-connection geometry and control weld quality in design.
- 📊 Data margins: the scatter across 41 intersections shows single-point estimates are unreliable; use a conservative design curve covering the data and verify its safety factor (Table 5).
- 📋 Iterative approach: the report concludes that existing data and analysis already describe the fatigue performance of pressure-loaded intersections well, so a dedicated test program to determine ip in the traditional manner may not be necessary — engineering can adopt the power-law-based design curves directly.
🚨 For high D/t power and process piping, pressure-cycling fatigue is traditionally weaker than thermal-expansion fatigue and easily overlooked — but when high-strength materials thin the wall, or pressure rises significantly (hydrogen transport and hydrotreating), pressure-cycling fatigue can become the controlling factor. Any service with frequent start-stop and pressure fluctuation must be checked for pressure-cycling fatigue separately.
5. Frequently Asked Questions ❓
❓ Q: What is ip in the Markl formula?
A: It is the pressure-induced stress intensification factor (SIF), defined as the ratio of nominal stress in a straight pipe without the joint to that in a jointed specimen at the same fatigue life, determined by paired fatigue tests; it reflects the stress-concentration effect of the welded joint.
A: It is the pressure-induced stress intensification factor (SIF), defined as the ratio of nominal stress in a straight pipe without the joint to that in a jointed specimen at the same fatigue life, determined by paired fatigue tests; it reflects the stress-concentration effect of the welded joint.
❓ Q: Why use a dedicated SIF for pressure fatigue instead of the thermal i?
A: The two involve different stress natures: thermal expansion is dominated by secondary stress, pressure by primary tangential stress; the stress distributions and failure mechanisms differ, so a unified i may underestimate or overestimate the actual damage.
A: The two involve different stress natures: thermal expansion is dominated by secondary stress, pressure by primary tangential stress; the stress distributions and failure mechanisms differ, so a unified i may underestimate or overestimate the actual damage.
❓ Q: The test data scatter is large — how should engineering use it?
A: The report recommends a conservative design curve covering the data (Tables 4/5 give comparisons and safety factors) and explicit statement of the chosen curve in design documents; a sensitivity study with reliable ip values is advisable where needed.
A: The report recommends a conservative design curve covering the data (Tables 4/5 give comparisons and safety factors) and explicit statement of the chosen curve in design documents; a sensitivity study with reliable ip values is advisable where needed.
🔍 Q: How does this report relate to the B31 codes?
A: It provides the data and method basis for the B31 MDC committee to develop stress-multiplier rules for pressure-cycling fatigue; until formal rules are published, engineering may check pressure fatigue with the Markl power law and Section VIII curves.
A: It provides the data and method basis for the B31 MDC committee to develop stress-multiplier rules for pressure-cycling fatigue; until formal rules are published, engineering may check pressure fatigue with the Markl power law and Section VIII curves.
