Research Programs

Publications:

1. L.-W. Hu, M.S. Kazimi, and A. Sonin, Thermal Striping in LWR Piping Systems, MIT-NSP-PR-014 (August 2002).
2. L.-W. Hu, J. Lee, P. Saha and M.S. Kazimi, Thermal Striping in LWR Piping Systems, MIT-NSP-PR-016 (August 2003).
3. L-W. Hu and M. S. Kazimi, Large Eddy Simulation of Water Coolant Thermal Striping in a Mixing Tee Junction, NURETH-10, Seoul, Korea, October 5-9, 2003.

Investigators :

• Dr . Lin-wen Hu
• Dr. Pradip Saha
• Prof. Mujid S. Kazimi
• Mr. Joenik Lee

Funding :

• This project is supported by Tokyo Electric Power Company.

High Cycle Thermal Fatigue Due to Thermal Striping

High cycle fatigue at coolant mixing junctions has been observed in the secondary side of nuclear plants but is a challenging subject to predict and include in the life management of nuclear power reactor piping systems. Thermal striping at mixing tee junctions is one of the two phenomena that are identified to be the cause of such thermal fatigue failure. The evaluation of thermal striping can be performed by empirical laws extracted from mock-up experiments and/or computational fluid dynamics (CFD) simulations. In this project, we have simulated experiments conducted at TOSHIBA and at HITACHI as part of a joint Japanese project. The results have demonstrated recently that the Large Eddy Simulations (LES) approach in CFD codes is capable of predicting accurately the locations that are most susceptible to large temperature fluctuations at the mixing tee. However, it may not be practical to study thermal striping using an advanced turbulence model (e.g., LES or DNS) in a large system because the computer power needs normally require expensive parallel computers.

A simplified method has been proposed to identify potential areas that may be most susceptible to thermal striping. This method is based on the observations in our previous LES simulation study that the temperature fluctuation is proportional to the spatial temperature gradient. Since the time-averaged mean temperature profile obtained from LES should be close to those from a Reynolds averaged Navier-Stokes (RANS) turbulence model, e.g., k-ε or Reynolds-stress model (RSM), it may be possible to use a steady RANS simulation to locate the areas that are subject to large temperature fluctuations. Comparisons of the calculated root-mean-square (RMS) temperatures using LES, calculated time-averaged mean temperatures using LES, and the calculated steady-state temperatures using RSM have been performed in order to validate the analogy. The comparisons are made for two types of thermal striping flow configurations in the mixing tee-type A (collision type) and type B (co-current type) and demonstrate the promise in this approach.

Besides evaluating the fluid temperature fluctuations using a CFD code, simplified but practical methods have been developed for determination of solid temperature, thermal stress, stress intensity factor and finally the fatigue crack propagation rate in the pipe wall. Sensitivity studies have been performed with respect to the fluid-to-wall heat transfer coefficient, temperature difference between the hot and the cold fluid, system pressure, and the frequency of fluid temperature fluctuations. Attempts are being made to develop practical guidelines for avoidance of thermal fatigue problems at the coolant mixing regions.