This paper describes experimental analyses using the SIMMER-III computer code, which is a two-dimensional multi-component multi-phase Eulerian fluid-dynamics code. Two topics of key phenomena in core disruptive accidents were presented in this paper: debris-bed coolability and metallic fuel freezing behavior. Related experimental database were reviewed to choose suitable experiments. To analyze the debris-bed coolability, the ACRR-D10 in-pile experiments were selected. SIMMER-III well simulated the heat transfer mechanisms including conduction, boiling and channeling observed in the experiment. Metallic fuel may freeze onto the stainless steel (cladding or wrapper tube) together with eutectic formation during core disruption in a metallic-fueled reactor. The CAFÉ-UT2 experiment carried out using pure UO₂ melt to investigate such phenomena was selected for the experimental analysis. In spite of no eutectic formation model in the SIMMER-III code, the calculated fuel penetration behavior was in good agreement with the experimental data.

View full abstract

The present paper describes a new method for determining the target value of structural reliability in the framework of the System Based Code (SBC) by considering the safety point of view. This new method utilizes analysis models of a probabilistic safety assessment (PSA), and the reliability target is derived from (1) the proposal to a quantitative safety goal that was published by the nuclear safety commission (NSC) of Japan; (2) the quantitative safety design requirements on the core damage frequency (CDF) and the containment failure frequency (CFF) that were determined in the Fast Reactor Cycle Technology Development (FaCT) project by Japan Atomic Energy Agency (JAEA). This method was applied to determine the reliability target of the structures and components which constitute the reactor cooling system in the Japan sodium-cooled fast reactor (JSFR). The risk from the reactor is shown by the sum of combination of various elements in the PSA analysis model. Those elements include dynamic failures and static failures of the structures and components, and human errors. However, the present study focuses on the sequences including the static failure, and the probability of dynamic failures and human errors in those sequences is conservatively assumed as a unity. It was confirmed that the present method combined with the PSA analysis model for internal initiating events is applicable to determine the reliability target associated with a random failure of the structures and components, and also confirmed that the method related to seismic initiating events can derive the target value of the occurrence frequency at which any of the important structures and components fails due to an earthquake.

View full abstract

The present paper described the onset cavitation conditions, acoustic noise characteristics and void behaviors of flowing liquid sodium in the venturi made of 316 SS with an inner diameter of 6.5 mm and a length of 20 mm. An experiment was conducted in the range of sodium stagnant pressure in the expansion tank of 0.061-0.181 MPa-a and in the temperature range of 200°-400°C. As a result, it was found that the noise intensity increased greatly when the cavitation coefficient approached unity (onset of cavitation), and it became saturated in lower cavitation coefficient (developed cavitation). The cavitation coefficient at the onset of cavitation was nearly equal to unity. However, an increase in temperature shifted the cavitation coefficient to a value a little higher than unity. The results also showed that velocity in the venturi at the onset of cavitation also increased by increasing the stagnant pressure. Cavitation occurrence produced a noise in the range of 600-20,480 Hz. In developed cavitation condition, the noise intensity was nearly constant at around -50 dB which might be caused by the choking of the fluid. At 400°C, instabilities or fluctuations of the noise spectrums were observed. Simulated results by using CFD showed that voids were formed at the throat of the venturi and the void fraction increased as the static pressure decreased by the flow. The voids were conveyed and collapsed in the downstream of the venturi outlet because of an increase in the static pressure. These voids could cause choking and noise as well as erosion.

View full abstract

Supercritical carbon dioxide (S-CO₂) gas turbines can generate power at high cycle thermal efficiency, even at modest temperatures of 500-550°C, because of the markedly reduced compressor work near the critical point. Furthermore, the reaction between Na and CO₂ is milder than that between H₂O and Na. A more reliable and economically advantageous power generation system could be achieved by coupling with a sodium-cooled fast reactor. At Tokyo Institute of Technology, numerous development projects have been conducted for development of this system in cooperation with JAEA. Supercritical CO₂ compressor performance test results were obtained as described herein. Maximum design conditions of the supercritical CO₂ test apparatus are 11 MPa pressure, 150°C temperature, a 6 kg/s flow rate, and a rotation rate of 24,000 rpm. Different compressor design points are examined using impellers of three kinds. Then test data were obtained under steady-state conditions. The pressure ratio (compressor outlet pressure/inlet pressure) was obtained with the function of compressor rotation speed and the fluid flow rate. The data, reported herein for the first time, cover a broad region from sub-critical to supercritical pressures. No unstable phenomenon was observed in the area where the CO₂ properties change sharply. Results show that enthalpy rise needed to achieve the same pressure ratio near the critical point is smaller than in a sub-critical condition. Compressor test data were calculated using three-dimensional CFD code (CFX). Data of the pressure ratio vs. flow rate agreed with those implied by the fundamental compressor theory.

View full abstract

The concept of System Based Code (SBC) has been proposed to achieve compatibility in matters of reliability, safety, and cost of fast breeder reactors (FBRs). This code extends the present structural design standard to include the broad areas such as load setting, fabrication, inspection, and maintenance related to FBR design and operation. Therefore, a quantitative index that can correlate these different areas is required. One such index is the probability of failure. The determination of a target value is also one of the key points to implement the SBC concept. We have proposed a new method to determine reliability targets for structures and components in FBR plants from the viewpoint of safety. Analysis models based on probabilistic safety assessments are used in the method for determining reliability targets. In this study, the effectiveness of the probability of failure as an index and the reliability targets produced using the new method are investigated through a trial setting of an in-service inspection (ISI) requirement for a reactor vessel. The probability of failure due to fatigue-creep interaction is calculated using a Monte Carlo simulation. The reliability targets for failure related to the risks arising from internal initiating events are derived using the new method. The summation of the probability of failure and of the reliability targets up to the end of the in-service period enables us to compare them directly. The results show that the reactor vessel has sufficient reliability even without an ISI. Through this example, we demonstrate that the probability of failure is a promising index and that reliability targets derived using the new method are compatible with SBC.

View full abstract

Ferritic-martensitic steel with several welding methods in LBE were observed. The material was welded HCM12A. The specification of TIG, YAG laser and electron beam welding were 40A/12V of electrical conductance, 278W/14.3J of heat input and 5.0 mA of beam electric current, respectively. The stagnant corrosion test was conducted in liquid LBE at 600°C for 500 h. Oxygen concentration was controlled to be ∼4.7x10^-6wt% by using Ar-H₂-H₂O mixed gas. After corrosion test, cross sections of the specimens were analyzed by using optical microscope and SEM/EDX. In order to know the hardness profiles of fusion zone and base metal were observed by using Vickers hardness tool. The results show coarse grain structure around fusion zone and typical ferritic-martensitic grain structure in base metal. It was found that Cr-rich spinel oxide layer and diffusion zone formed on the surface. However, it appears much thicker in the fusion zone, about 18-30µm, with respect to the base metal (10-15µm). The hardness number increase in fusion zone.

View full abstract

The present paper describes the damage behavior caused by cavitation in flowing liquid sodium using venturi test section. The venturi test section was made from 316 SS with an inner diameter of 6.5 mm and a length of 20 mm. The experiment was conducted with pressure of argon cover gas in the expansion tank ranging from 105-110 kPa. The liquid sodium temperature was kept at 200°C and the sodium flow rate was 27-28 L/min, which corresponded to developed cavitation condition with the cavitation coefficient of 0.59-0.51. This condition was kept for 600 hours during the experiment. The results showed that cavitation bubbles collapse created damage on the inner surface of the venturi test section in the form of pits due to eroded surfaces. Under SEM, most of the pits on the inner surface of the venturi test section were in the sizes of around 25 µm. Some pits with sizes larger than 25 µm were also observed with the largest diameter around 110 µm which indicated intense collapse of cavitation bubbles. Observation on the inner surface of the test section also indicated some micro cracks, which might be caused by the cyclic process of the formation and collapse of the bubbles during cavitation process.

View full abstract

Dissolved hydrogen is one of the key factors of PWSCC (primary water stress corrosion cracking) in the primary systems of pressurized water reactors. The authors consider that application of alternative reductant for hydrogen may mitigate PWSCC. The corrosion tests of 304 stainless steel specimens in methanol solution and hydrazine solution at 320 °C were carried out under γ-ray irradiation, and the corrosion environment was evaluated. Electrochemical corrosion potential (ECP) of the stainless steels was -605 mV and -643 mV vs. SHE in 2 mmol dm^-3 hydrazine solution and 2 mmol dm^-3 methanol solution at 320 °C, respectively. After the immersion tests, oxide films formed on the stainless steel specimens were analyzed. It is concluded from the ECP measurement and comparison with previous results that corrosion environment under following conditions is similar: (1) 1.5 ppm DH without irradiation, (2) 2.9 ppm methanol without irradiation, (3) 2 mmol dm^-3 hydrazine under γ-ray irradiation and (4) 2 mmol dm^-3 methanol under γ-ray irradiation. During the immersion tests, the test waters were sampled, and analyzed, in order to understand radiolysis of the methanol solution and the hydrazine solution at 320 °C. Hydrazine is decomposed predominantly through thermal decomposition, and an ammonia molecule is formed from a hydrazine molecule. Formaldehyde and ethylene glycol were detected in the methanol solution. But, carboxylates were not detected. The present results do not demonstrate oxidation of methanol to CO or CO₂. Irradiation experiments of higher absorbed dose are necessary.

View full abstract

Radioactive corrosion products (CP) are primary cause of personal radiation exposure during maintenance work at FBR plants with no breached fuel. The PSYCHE code has been developed based on the Solution-Precipitation model for analysis of CP transfer behavior. We predicted and analyzed the CP solution and precipitation behavior of MONJU to evaluate the applicability of the PSYCHE code to MONJU, using the parameters verified in the calculations for JOYO. From the calculation result pertaining to the MONJU system, distribution of ⁵⁴Mn deposited in the primary cooling system over 20 years of operation is predicted to be approximately 7 times larger than that of ⁶⁰Co. In particular, predictions show a notable tendency for ⁵⁴Mn precipitation to be distributed in the primary pump and cold-leg. The calculated distribution of ⁵⁴Mn and ⁶⁰Co in the primary cooling system of MONJU agreed with tendencies of measured distribution of JOYO.

View full abstract