Thermal Cycling and Fatigue: Why Temperature Swings Destroy Nozzles

The Silent Killer: Thermal Cycling

Your vessel starts up at ambient temperature. Within an hour, the process fluid is pumped through at 450°F. Inside the nozzle, metal is expanding. Stresses are building.

Twelve hours later, the reactor goes through a planned cooldown. Back to ambient. The metal contracts. Stresses reverse direction. The metal wants to shrink, but it is bolted to neighboring components that are also shrinking at slightly different rates.

This cycle repeats 365 days a year. For decades.

Welcome to the world of thermal fatigue-one of the most underestimated failure modes in pressure vessel design.

Fatigue Is Not Like Static Loading

Most engineers understand static stress. You load a beam, it bends. You exceed the yield strength, and it deforms permanently. Simple cause-and-effect.

Fatigue is different. Even if the stresses are far below the yield strength, repeating cycles of stress will accumulate microcracks in the metal grain structure. Each cycle advances the crack a tiny bit. After millions of cycles, the crack reaches critical size and-ping-failure.

The Goodman diagram is your friend here. It shows how the allowable stress range depends on the mean stress and the number of cycles. The higher the mean stress, the lower the allowable stress range. And with 365 cycles per year for 20 years, you are looking at 7,300 cycles minimum.

Now add in stress concentrations-and guess what creates the worst stress concentration? Welds. The Heat Affected Zone of a weld is a prime fatigue crack initiation site.

This is one of the many reasons nozzle flanges (single forgings with no welds) outperform DIY nozzles (flange + pipe + welds) in cyclic service. You have eliminated the weakest link.

Differential Expansion: When Materials Grow at Different Rates

Every material has a Coefficient of Thermal Expansion (CTE). Steel expands about 6.5 microinches per inch per °F. Hastelloy expands about 7.8 microinches per inch per °F.

Now imagine you have a carbon steel vessel with a Hastelloy nozzle welded to it. When you go from 70°F to 450°F (a 380°F rise), the Hastelloy expands 27% more than the carbon steel. The weld literally gets pulled from both sides.

Over 20 years of thermal cycling, this difference in expansion creates a stress pattern that is mathematically certain to fatigue. The weld gets work-hardened in compression on one side and tension on the other. Microcracks initiate. They propagate. Suddenly you have a leak.

The Stress Concentration Problem Revisited

If you have a corner, a transition, or even a slight radius change in your nozzle design, that is where fatigue cracks will initiate. Stress concentrations (often denoted by a stress concentration factor, Kt) can multiply the nominal stress by 3 or 4 times locally.

You might design the nozzle with nominal stresses that are acceptably low. But if there is a sharp corner in the transition from the nozzle body to the weld, the local stresses spike 3x higher. That concentrated stress sees the full brunt of thermal cycling, and cracks start propagating.

This is why good design practice calls for generous radii in all transitions, especially on nozzles subject to cyclic loading. A 1/4-inch radius might cost an extra hundred bucks on a forging, but it prevents a $100,000 catastrophic failure.

Frequency Response and Resonance

If your thermal cycling frequency happens to match the natural vibration frequency of the nozzle and piping system, you have a serious problem. This is called resonance, and it can amplify stresses by an order of magnitude.

A process that cycles slowly (once per day) might be fine. A process that cycles rapidly (multiple times per hour) needs careful seismic and dynamic analysis to avoid hitting a resonant frequency. This is why large vessels in earthquake-prone regions require extensive FEA (Finite Element Analysis) of the nozzle designs.

Material Matters: Some Alloys Fatigue Better Than Others

Carbon steel has a clear fatigue limit-stresses below about 30-40% of the yield strength will not cause fatigue failure, no matter how many cycles.

Stainless steels and Hastelloy do not have a clear fatigue limit. They will continue accumulating microcracks at indefinitely low stress levels. For these materials, you must use Goodman diagrams or S-N curves (stress vs. number of cycles) to establish the allowable stress for your specific cycle count.

Chrome-Moly actually has decent fatigue resistance at elevated temperature, which is one reason it is popular for high-temperature low-alloy (HTLA) applications.

Mitigating Fatigue: Design Strategies

If you know thermal cycling is going to be a problem, build it into your nozzle design from the start:

Use a nozzle flange. The solid forging has no internal welds, dramatically improving fatigue life. Optimize the radius of all transitions. Add stress relief grooves if necessary. Specify post-weld heat treatment (PWHT) to relieve residual welding stresses. Use flexible piping connections outside the vessel to accommodate differential expansion. Oversizing the nozzle increases the cross-sectional area experiencing stress, spreading the load over more material.

The Bottom Line

Fatigue is insidious because it does not announce itself. The nozzle looks fine right up until the moment it fails. This is why you must design for it explicitly. Understand your thermal cycling profile. Calculate the stress ranges. Use appropriate materials. Eliminate stress concentrations. And when in doubt, upgrade to a nozzle flange design. Your vessel's longevity depends on it.