رفتار سازه در طول PTS گرفتن گذرا به حساب اثر WPS

The Reactor Pressure Vessel (RPV) is an essential component liable to limit the life duration of PWR's. Its behavior in service is limited in time due to the embrittling effects of irradiation. The structural integrity of the RPV is assessed by conventional fracture mechanical studies, where it is assumed that the failure of a flawed structure occurs when the stress intensity factor at the crack tip reaches the toughness value of the structure material. Toughness curves of materials are obtained from monotonously increasing and isothermal loading. On the other hand, RPV integrity assessment involves loading conditions with coupled cooling, heating, increasing and decreasing load. The safety analyses made at the European level study, the behavior of defects in the vessel subjected to loading resulting from thermal transients. These analyses usually do not take into account the effect of load history/warm pre-stressing (WPS) of the defects, which is observed in a wide range of experimental studies. The non-consideration of the beneficial effect of this physical phenomenon has two major consequences: • a poor knowledge of the real margins associated with the transients to which the vessel is subjected, • an economical penalty due to large under-estimation of the life duration of the vessel. This paper presents the results of two independent programs. The first dealing with four WPS tests performed at CEA in France on CT specimens manufactured of ferritic 18MND5 steel undergoing different types of loading during the cooling phase as follow: • Load Cool Fracture (LCF) • Loading Maintained CMOD Cooling Fracture (LM2CF) WPS effect is observed in the first case, while curve in the second case the failure occurs during the crossing of the transition because of the monotonous increase of the force due to constant CMOD (Crack Mouth Opening Displacement). The second program deals with tests performed at MPA in Germany in collaboration with EDF on CT25 and CT50 specimens using the same material with five types of WPS cycles: • LCF • Load Unload Cool Fracture (LUCF) • Load Transient Fracture (LTF) • Load Oscillation Cool Fracture (LOCF) • Load Oscillation Transient Fracture (LOTF) In all these cases, WPS effect is demonstrated. For both the programs, numerical analyses were performed at MPA using Weibull [Metall Trans A, 14A (1987) 2277] and Chell [Fourth Int Conf Pressure Vessel Technol, Inst Mech Engng, 1980, Paper C22/80, London, U.K., 117] models to predict the WPS effect, details of which are presented here.

The warm pre-stress effect means the application of a load to a structure at high temperature prior to a later event or load application at lower temperature. The warm pre-stress effect has been largely reported in the literature [1]. On one hand, authors generally say that the WPS effect elevates the effective fracture toughness of the material, so that crack propagation does not occur when the stress intensity reaches the KIC value of the virgin material identified with an isothermal loading. On the other hand they speak about the conservative principle of the Warm Pre-Stressing effect, which means that after a pre-load of the structure at high temperature, fracture does not occur if the stress intensity factor decreases or holds constant while the crack-tip temperature decreases, even if the virgin material toughness is attained. Up to now the WPS effect is not considered in the French mechanical assessment of flawed vessels subjected to events such as Loss Of Coolant Accident (LOCA) or Pressurized Thermal Shock (PTS). The integration of this effect into analyses is full of promise since it can prove their present over-pessimism. It is the reason why a certain number of programmes are underway in France similar to those performed in Germany [2], [3] and [4] to evaluate the beneficial effects.

The experiments conducted in the framework of this study show that fracture occurs only for a monotonous increase in loading in the transition regime, as for example imposed during a maintained crack mouth opening displacement experiment (LM2CF). This conclusion is consistent with the definition itself of the values defining the toughness in the transition regime, which have been obtained by testing specimens with increasing loading. In all cases of WPS cycles (LCF, LUCF, LTF, LOCF, LOTF), fracture load of the tested specimens are higher than the upper bound of the virgin material and prove that the ferritic steel 18 MND 5 is subjected to the WPS effect. Tests performed with LCF, LTF, LOCF and LOTF WPS cycles proved that the conservative principle exists and can be extended. After a pre-load of the structure at high temperature, fracture does not occur if the stress intensity factor decreases or remains constant with oscillations in the range of 20% while the crack-tip temperature decreases, even if the virgin material toughness is attained. We have presented some experimental evidences (Fig. 9) showing that oscillations during a transient could act as many WPS cycles preventing fracture. Further two tests have been modeled with CASTEM2000:LCF and LM2CF. The finite element results are close to the test results in terms of global behavior. Moreover the LCF and LUCF tests performed at MPA have been simulated with the local and Chell approaches. As for the CASTEM2000 simulations, the finite element results obtained with Adina are close to the test results in terms of global behavior. The application of the modified Beremin model shows that this model has a good potential to predict the test results. Further developments are underway at EDF to improve this model to take into account all cases of loading paths including mechanical and thermal unloading. The analytical Chell's model gives results extremely close to these obtained with the Beremin model. The LCF test results are very well predicted, whereas an important conservatism is observed for the simulations of LUCF tests.