Computer Code QUARK-LP ver.4 for Simulation of Tube Rupture Propagation due to Overheating in LMFBR Steam Generator

摘要

Sodium-water reaction experiments performed in a variety of facilities in several countries have demonstrated that the reaction zone temperature varies in time and in space. These temperature variations arise from instability in the reaction jet with pre-reaction conditions varying with time and position caused by jet instabilities at the steam-sodium interface and by interactions of the jet with the tubes in the bundle. In modeling tube rupture due to overheating, it would therefore be very pessimistic to assume that the reaction zone temperature was uniform throughout at the maximum value and was constant with time, which are the assumptions adopted in the previous version of the LMFBR tube-rupture propagation analysis code QUARK-LP. In the present modification of the code the modeling of the reaction zone has been subdivided into three regions depending on the distance from the failed tube. It is assumed that the jet will expand with a solid angle 'omega', or in special case of adjacent to wall 0.5x omega. From the nearest location from the failed tube the three reaction zone regions are (1) the steam-rich zone, (2) the mixing zone in which the highest temperatures are generated and (3) the sodium rich zone. Once the mole ratio of sodium to steam, sodium temperature and pressure are given, it is possible to predict the post reaction temperature and the chemical/physical composition assuming thermodynamic equilibrium. The relation between mole ratio, the sodium temperature and the sodium pressure are included in this code in tabular forms. The recommended values of the mole ratio of each reaction region as well as the solid angle of jet expansion were obtained from temperature measurements from Super Noah experiments in UK and SWAT-3 experiments in Japan. The reaction zone to tube heat-transfer coefficient in the code depends on the local velocity and the post-reaction chemical and physical composition. The jet velocity is based on the distance from the origin of the jet, the angle of expansion omega and the free area of the tube bundle 'alpha'. Some additional modifications have been made to enable the specific geometry of the PFR Superheater to be modeled to allow data from the large leak event in 1987 to be used as a test case. The predicted number of failed tubes is 44 compared to 39 in the real event.