Combined Thermal & Presssure Fatigue 1

[Bridgy]

Does anyone have any views on this?

Just trying to get a feel of how others tackle the problem of fatigue due to combined pressure and thermal cycling.
The approach we have taken is to use ASME VIII Div 2 App 5 for the pressure fatigue. Calculated the thermal transient stresses at nozzles in acc with PD5500 Annex G4.
We have added the membrane stresses from both results and calculated the stress intensity for the stress range and input this into the ASME fatigue curve to give us the allowable cycles.
Is this a feasible approach? How do others do this?
Is there an FE package that can do this?

Looking forward to some wisdom from my mentors!

Thanks in advance
Roy

 

[davefitz]

This is handled in the european codes via the TRD 301 annex 1 and the new EU PED code related to boilers. The simplified version assumes the pressure cycle is coodinated with the thermal stress cycle in the case of initial design calcs. In the design case, this is the approximation I usually use. It is useful for estimating the operating limitations of the equipment, for input to the O+M manulas or plant monitoring alarms.

In terms of monitoring actual operations, the pressure stress is monitoried by press Xmitter and the thermal stress by a near-ID thermocouple- these process data points are evaluated in real time via a series of algorithms , sometimes called a "boiler stress analyzer". The life consumption is estimated using the rainflow cycle counting algorithm, also defined in the EU code . But these approximations are based on unnotched specimens, while the acutal most likely source of fatigue failure is a welded joint , with weld associated pre-exisitng cracks and material properties of the HAZ that are inferior to the parent material.

The new EU PED inlcudes the PD5500 fracture mechanics approach to the life of welded joints, if the part in question includes a weld interface. This is claimed to also be used in the new 2007 re-write of sect VIII div 2.

It is often considered to be more sophisticated to use a 3D finite element model, with elastic-plastic formulations, to determine the stress-strain response of the most stresed region of the vessel, but it soemtimes seems to a PR gimmick to impress the client, if no account is taken of the weld cracks, weld HAZ, and inferior HAZ properties. Not to mention the complete disconnect between assumed operating schenarios vs the wild operating gyrations imposed by neandrathal operators.