Thermal Striping in LWR Piping Systems

Report Date: December 2003
Appendices: No

Abstract

The current study addresses high-cycle thermal fatigue that can result at tee junctions of the Light Water Reactor piping systems. Our study concentrates on numerical analyses of the magnitude of temperature fluctuations and structural response of the pipes at a tee junction. The three key aspects of this study are: i) benchmark studies of small-scale thermal striping' experiments using the Large Eddy Simulation (LES) model of a commercial CFD code - FLUENT, ii) development of a simplified methodology that uses the temperature gradient predicted by RANS turbulent modeling to identify the locations that are most susceptible to thermal striping fatigue cracking and to correlate the magnitude of temperature fluctuations with steady-state temperature gradient, and iii) numerical evaluation of the pipe wall temperature response, thermal stress, and Stress Intensity Factor (SIF) for thermal stress fatigue cracking analysis.

The Large eddy simulation (LES) model is currently the only practical alternative to Direct Numerical Simulation (DNS) in modeling the time-dependent fluctuating quantities of turbulence. These quantities directly affect the coolant temperature fluctuations that are the major concern of thermal striping. Benchmark studies were performed for two types of tee junction flow configurations, type A (collison type) and type B (co-current type). The simulation results are in good agreement with measurements. In particular, the predicted locations of the maximum temperature fluctuations are normally within the range of ID of measurements. The calculated maximum normalized fluctuating temperatures are somewhat higher than measurements. This may be due to the high-frequency temperature fluctuation components that are not resolved by the temperature measurements. It is therefore concluded that LES is a promising tool to accurately analyze the coolant temperature fluctuations associated with thermal striping.

As a result of our FLUENTILES simulation study, a novel method is proposed to identify potential locations that may be most susceptible to thermal striping and to correlate the temperature fluctuation with the mean temperature gradient. This method is based on the observations throughout our LES simulation study that the temperature fluctuation is proportional to the spatial temperature gradient. It is concluded that it may be possible to use a steady RANS calculation to locate the areas that are subject to large temperature fluctuations. Furthermore, a correlation could in the future be developed between the temperature gradients and temperature fluctuations. The attenuation of the temperature fluctuation magnitude in the near-wall region can then be calculated using such a correlation.

A numerical scheme is developed and evaluated for the thermal stress fatigue analysis. This model uses the calculated instantaneous coolant temperatures from FLUENTLES as input, and then calculates the pipe wall temperature response, stress and SIF separately. The fatigue crack propagation correlation is then applied to calculate the number of cycle numbers of crack propagations to reach outside the safe zone.. The model is a useful tool in identifying the relative importance of various parameters affecting fatigue cracking failure. These parameters include the magnitude and frequency of temperature oscillations, operating conditions of the system (pressure and temperature), heat transfer coefficient, and initial crack size. Our sensitivity analysis shows that the heat transfer coefficient and the temperature difference between the hot and cold fluids have the most significant effect on the stress intensity due to thermal gradients in the pipe, and that these two are the dominant factors that govern the thermal fatigue phenomenon.

Program:     NSP Nuclear Systems Enhanced Performance

Type:     PR 

RPT. No.: 17