Feasibility Study of a Spectral Density Modeling of Thermal Fluctuations in Mixing Tees

In mixing tees, hot and cold fluids meet at a nozzle with different flow rates. The resulting temperature fluctuations create mechanical and thermal stresses on the pipes. Predicting the subsequent fatigue is the subject of on-going research. A new approach was developed in order to obtain a qualitative and quantitative estimate of the damage caused by a flow without necessarily acquiring measures, which is often difficult when studying flows in nuclear plant piping. Several time-series of fluid temperature from mock-up tests were studied. Four specific profiles were recognized. Internal stress was computed and fatigue was calculated according to an algorithm. The influences of the profiles, and of the temperature difference between hot and cold legs, were studied. Finally, an approach was proposed, whereby thermal fluctuations are modeled by an envelope spectral density which is transposed to the characteristics of the studied flow (standard deviation, mean temperature, cold and hot legs temperature difference, flow rate). The use of the envelope signal gives a good estimate of the thermal fluctuations but the damage is over-estimated. On-going calculations are attempting to modify the envelope spectral density to attain more realistic values.