The FAC (Flow Accelerated Corrosion) issue is studied through measuring local velocities in the curved pipe in order to validate a CFD (Computational Fluid Dynamics) code. The ultrasound technique that has no fluid temperature dependency is selected to measure local velocities because it is non-intrusive to a flow and an accurate velocity measurement is possible. FLUENT is selected for the ultrasound velocimetry calibration in the straight pipe, for a data comparison in the curved pipe and for an analysis of the FAC problem.
A local velocity measurement using the ultrasound velocimetry is done not only in the 89.75mm diameter with 3m length straight pipe but also in the same diameter, 0.784m curve radius and 90° curved pipe. In the same flow condition and pipe geometry, the predictions by FLUENT are compared to the experimental data and the maximum error is 7.8%.
FLUENT is applied to a flow analysis for the FAC problem. A flow analysis includes a local velocity and shear stress measurement near the wall. A RNG κ-ε turbulence model is selected for the analysis. In the curved pipe with 0.482m/s uniform inlet velocity, this model shows only 0.0025 RMS error of the velocity, compared to the velocity by the ultrasound velocimetry. Various wall roughness heights are applied to boundary conditions of the pipe wall for the simulation that the pipe wall becomes rough as the FAC proceeds. A CANDU feeder pipe is selected for the real scale application of the FAC problem. The selected CANDU feeder pipe has 2.5in (0.0635m) inner diameter, 3.75in (0.09525m) curve radius and 73.1° curved pipe.
As a result, it is predicted that the severe FAC could proceed at the inner side of the pipe wall just after the inlet of the curved section and at the outer side of the pipe wall just after the curved section. Moreover, from the roughness heights variation result, it is predicted that the region at the outer side of the pipe wall just after the curved section becomes the most severe re...