B31.3 vs. B31.1: Fatigue and Stress Range Changes in the 2025 Piping Codes

Historically, B31.1 has relied on the legacy Markl formulation, which uses a –0.20 slope (log stress vs. log cycles). B31.3 has already moved to the Hinnant Appendix W formulation, where the fatigue slope is –0.333. These differing slopes meaningfully change allowable stress ranges throughout most practical piping design cycles.

Low-Cycle Region (≈ 4,000 cycles)

At roughly 4×10³ cycles—typical for thermal start-ups, seismic qualification, or other displacement-controlled events—the difference is pronounced:

• B31.1 (Markl) allows a stress range about 28% lower than
• B31.3 (Hinnant) for the same number of cycles.

Where the Curves Converge—and Where They Don’t

As cycles increase toward the endurance range (10⁶–10⁷ cycles), the situation reverses: Markl’s flatter slope becomes non-conservative, producing stress ranges roughly a factor of two higher than Hinnant’s curve near 5×10⁶ cycles.

In real-world piping systems—which often accumulate somewhere between 4,000 and 100,000 cycles—this changing conservatism matters. Most industrial applications lie squarely in this mid-cycle band, where the two Code philosophies differ by 16–28% depending on the cycle count.

It is increasingly evident that B31.3 is aligned with the broader industry’s accumulated data set, while B31.1 has not yet harmonized its slope or methodology.

Expansion vs. Occasional: Why Stress Range Isn’t Always What It Seems

One of the recurring misunderstandings in design reviews involves stress range versus stress amplitude. For displacement-controlled loads (thermal, support movement, seismic reversal), the allowable is expressed as range; for occasional loads, it is typically expressed as amplitude.

This distinction leads to counterintuitive results:

• A 1.5 Sy amplitude allowable for seismic loads in older B31.1/B31.E language becomes a
  3 Sy stress range, simply because seismic loads reverse.
• This helps explain why seismic allowables can approach the collapse load of 4/3 Sy—
  the loads behave much more like thermal displacements than primary forces.

Understanding whether the Code intends “range” or “amplitude” is crucial for accurate stress reporting and for avoiding the common pitfall of comparing mixed terminology.

Is B31.1 Likely to Adopt the Hinnant Curve?

Speculation continues that B31.1 will ultimately adopt a slope aligned with B31.3 Appendix W. The technical rationale is clear:

• The Markl data set is limited in component diversity.
• The –0.20 slope produces unconservative values in the high-cycle range.
• Component behavior under displacement-controlled loads is better captured with updated fatigue formulations.

Until that change is formalized, engineers designing systems governed by both Codes will need to understand where their designs fall on the cycle spectrum and how much margin truly exists.

Design Implications for 2025 Projects

1. Low-Cycle Seismic and Thermal Transients
   Designs near 20–100 cycles, common in seismic qualification, are most sensitive to the slope difference. B31.1’s more restrictive allowables can drive reinforcement or flexibility changes that would not be required under B31.3.
2. High-Cycle Vibrations
   Reciprocating equipment, rotating machinery, and flow-induced vibration assessments should not rely on Markl’s high-cycle extrapolation without careful review. The Hinnant formulation provides a better match to modern fatigue data.
3. Lifecycle Assessments and Ratcheting Prediction
   Because ratcheting strain increments are more damaging than simple elastic shakedown, an incorrect fatigue basis at low cycles can yield misleading predictions in strain-based assessments.
4. Harmonization
   Once B31.1 aligns its curve with B31.3, designs previously constrained by Markl allowables will have additional margin. Conversely, designs relying on high-cycle Markl values may need reevaluation.
   
   

Where FEA Tools Fit

Modern FEA solutions—particularly those tuned for ASME fatigue and ratcheting assessments—greatly simplify comparison of:

• Markl vs. Hinnant allowables
• Low-cycle ratcheting predictions
• High-cycle elastic stress ranges
• Secondary displacement response under seismic loading

This is especially important in borderline cases where allowable stress range differences materially impact design decisions.

Takeaway

The 2025 cycle further widens the philosophical divide between B31.1 and B31.3 regarding fatigue behavior. B31.3 has already moved along with data-driven curve updates; B31.1 continues to use legacy Markl formulations. For engineers working across both codes, a clear understanding of cycle-dependent stress range differences is now essential to accurately predicting ratcheting, assessing seismic load combinations, and avoiding misplaced conservatism.