HIGH CYCLE THERMAL FATIGUE FATHER EXPERIMENT: NON DESTRUCTIVE AND METALLOGRAPHIC EXAMINATIONS

摘要

In May 1998, a leak (30 mVh) occurred in the reactor heat removal system (RHRS) of the Civaux 1 French power plant (PWR type N4, 1400 MWe) which was then in a hot shutdown situation [1,2]. A 180 mm through-wall crack was found in a 304L austenitic stainless steel elbow in a mixing area of high and low temperature fluids. The major root cause for cracking was identified as high cycle thermal fatigue (HCTF). The cracks were found in the mixing tees and at the roots of welds in mixing areas. The presence of ground surface finishes and of geometrical discontinuities like weld roots was associated to fatigue damage. For the new RHRS mixing zones of N4 plants, decision was then taken to suppress welds or locate them away from mixing area and to improve the surface condition. To reproduce the HCTF phenomenon occurring in mixing zones, a representative endurance thermal fatigue test named "FATHER" was performed by CEA under an EDF/CEA/AREVA agreement [3]. The test lasted 300 hours. It was performed on a 304L stainless steel mixing zone of 7 mm thick and 6" diameter with a temperature difference of 160°C between cold and hot fluids. Different internal surface finishes were introduced in the test mock-up: coarse and fine grinding, industrial polishing, as extruded surfaces and as welded or flushed joints. This paper proposes to present the numerous non destructive examinations (NDE) which were performed during and after the endurance fatigue test like ultrasonic examinations or dye liquid penetrant inspections. They led to the observation of many small thermal fatigue cracks located near as welded joints, on ground surfaces and on unpolished flushed welds. Cracks were not observed on industrially polished surfaces reproduced in straight piping sections or in flushed plus polished welds. This paper also focuses on the detailed metallographic examinations which were performed after the endurance test. These ones have been carried out after having axially cut the mock-up in two symmetric half parts and machined sampling plates containing HCTF cracks. In this way, more than 50 thermal fatigue cracks with depths of 100 to lOOOum were observed. Cracks initiate mainly on geometrical discontinuities like weld toes or grinding striations. Test results have allowed to improve and to validate methods and tools for predicting crack initiation in mixing zones. The FATHER experiment can be seen as a significant contribution for preventing the risk of HCTF in PWR equipment.