CFD study of flow accelerated corrosion in 3D elbows

The objective of this paper is to examine the last step of the mechanistic model of wall thinning degradation mechanism i.e., convective mass transfer in feeder pipes under different environments of nuclear power plants (NPP). In the present study, the flow and mass transfer of demineralised water in carbon steel pipes such as single and double elbow was simulated under Indian NPP feeder water system conditions. The numerical simulations of mass transfer results are compared with the wall thickness measurement data of feeder pipes of Indian and CANDU NPP. The eddy structures and their interactions with the wall and the formation of vortex corelines were examined to analyze the flow changes in double elbow pipe. These vortex corelines appeared at the downstream of the bends. The intensity of these vortex corelines has been calculated by plotting the helicity along the vortex coreline. Due to the formation of vortex corelines and resulting flow changes, the mass transfer coefficient (MTC) varies circumferentially. MTC is the most important parameter to predict the highly susceptible FAC locations. For the MTC analysis, the Chilton-Colburn analogy in terms of wall shear stress was used. From this analogy, the effecting behavior of flow and geometrical parameters such as Reynolds number (Re) and the close proximity of bends, respectively, on MTC are studied. The locations of maximum MTC are calculated for both the single elbow of 73° and 90° and the double elbow of 90° and are shown in terms of contours. The flow singularity exists at the elbows; specifically in the double elbow, the downstream elbow is more vulnerable to FAC. The amplified Re results the increased MTC. The close proximity effect, calculated for the vertical limb heights from 0.01 m to 0.65 m, observed to be decrease with the increase of limb height. The changes in the angle of bend, varies the local maximum MTC locations. These reported results are useful for developing the targeted inspection plans in predicting the vulnerable FAC locations.