NUI Galway – UL Alliance First Annual ENGINEERING AND - ARAN ...

Thermomechanical characterisation of P91 power plant components
T.P. Farragher 1 , N.P. O'Dowd 2 , S. Scully 3 , S.B. Leen 1
1 Mechanical and Biomedical Engineering, College of Engineering and Informatics, NUI Galway, Galway
2 Dept of Mechanical and Aeronautical Engineering, Materials & Surface Science Institute, University of Limerick, Limerick
3 ESB Energy International, ESB Head Office, 27 FitzWilliam St., Dublin 2
Abstract
A time-dependent, elastic-plastic-creep, thermomechanical
methodology has been developed to
characterize P91 power plant components subjected to
realistic loading conditions.
1. Introduction
9-12%Cr ferritic steels are commonly used in fossil fuel
power plant. The P91 alloy (9Cr-1Mo) is the focus of this
investigation. This is primarily due to its high creep
resistance at high temperatures. Increasing operational
steam pressures and temperatures will improve plant
efficiency. Therefore to achieve more efficient plant there
is a need to understand the performance of candidate
materials at higher temperatures and pressures [1].
Concomitantly, changes in the operational schedule of
current plant from 'base load' operation to ‘loadfollowing’
(in order to facilitate electricity supply due to
the unpredictable output of wind and other such
renewable energy sources) exposes plant to an increased
frequency of thermal and pressure cycles, thus increasing
the risk of thermal fatigue.
2. Material model
An anisothermal cyclic viscoplasticity material model has
been implemented for P91 material. Sample isothermal
results are shown in Figure 1. The material parameter
identification process is based on published experimental
material data from [2-4]. This material model
encompasses cyclic isotropic hardening/softening effects,
non-linear kinematic hardening effects, strain rate
(viscous) effects, as shown in Figure 1 (a to d).
(a)
(c) (d)
Figure 1 Isothermal hysteresis loops for P91 material at
different temperatures; softening effects shown in a - c
and strain rate sensitivity effect in d.
(b)
3. Methodology
A methodology was then developed to incorporate the
cyclic viscoplasticity model into time-dependent, thermomechanical
analysis of P91 power plant components, as
depicted in Figure 2.
181
3.1 Power plant operating cycle
Figure 3 shows measured steam temperature and pressure,
and outside surface temperature histories of a power plant
pipe steam header. This history represents a routine startup
of plant from ambient conditions, attaining an
intermediate fluctuating temperature, associated with
attemperation, before temperature and pressure increase
to fully operational conditions of 500 o C and 17 MPa, for
a period before cooling to ambient conditions (the process
is repeated for three consecutive cycles). The measured
data of Figure 3 was used as input to the
thermomechanical analysis.
Temperature ( o C)
Material
Model
Model
500
400
300
200
100
0
Power Plant Operating Cycle
Transient Thermal Analysis Model
Thermomechanical Model
Stream temperature
Measured outside pipe temperature
Steam pressure
Heat Transfer
Properties
Heat Transfer
Thermo
mechanical
properties Thermo -
Figure 2. Material characterisation methodology.
10000 30000 50000 70000
Time (secs)
90000 110000
Figure 3. ‘Cold-start’ temperature and pressure history.
3.2 Thermal model
The thermal model employed for the plant cycle
incorporates: (i) forced steam convection on pipe inside
surface, based on measured time-dependent steam
temperature data, e.g. Figure 3, (ii) transient conduction
through pipe wall (with temperature-dependent properties
[4]) and (iii) natural convection on pipe outside surface.
4. Results and Conclusion
A cyclic viscoplasticity material model has been
developed for multiaxial thermomechanical analysis of
P91 material. A transient thermo-mechanical analysis
methodology has been developed for P91 power plant
components, incorporating a transient thermal model and
a multiaxial viscoplasticity model. Figure 3 (inset) shows
typical predicted thermomechanical loops for power plant
pipes. Stress ranges of more than 300 MPa with tensile
mean stresses were predicted, with significant cyclic
inelastic strains of at least 0.2 %.
5. References
[1] Viswanathan V. et al. Advanced Materials and
Processes, 166(8), 2008
[2] Saad A. A. et al. ASME PVP Conference 2010
[3] Okrajni J. et al., Int. J. Fatigue 2007
[4] ASME B31.1, ASME, NY, 1995
20
15
10
5
0
Pressure (MPa)