[en] The following problems are discussed within the framework of a project for the purpose of studying the metallurgical and chemical interaction between corium and reactor pressure vessel during a hypothetical accident: definition of corium composition, experimental methods for examining the properties of a corium (analytical methods, determination of melting point), isothermal crucible experiments (reaction behaviour of the binary mixtures steel-zircaloy, zircaloy-UO₂, steel-UO₂, and of the corium at high temperatures, thermodynamic analysis of experimental results), interaction between the corium and the reactor pressure vessel wall (kg-scale), experiments for producing a corium. (GSCH)

[en] When the calculated results of the concrete decomposition model are compared with experiments the concrete decomposition enthalpy is found to be in good agreement for silicaceous concrete. For lime-stone concrete, however, a considerable discrepancy exists which needs further investigation. The evaluation of thermite experiments (topic 06.01.12) showed that the transient erosion of a horizontal concrete surface and the time dependent surface temperature of the melt agrees in tendency with the WECHSL calculations. The starting time of solidification processes as well as the maximum erosion depth also agree in order of magnitude with the experimental observations. The results of the extended film model for the heat transfer to inclined walls was checked with model experiments (subliming dry ice slabs in water). (orig.)

[en] Short communication

[en] A volumetrically-heated pool with gas injection at the boundaries is used to simulate the heat transfer processes taking place in molten core debris-concrete systems. Measurements of the upward, downward, and sideward heat transfer rates at the pool boundaries have been made over wide ranges of power density, superficial gas velocity, and pool aspect ratio. Pools with either a free upper surface or an isothermal solid upper boundary to simulate an overlaying solid crust of oxides have been examined. Detailed measurements of the temperature distribution within the pool have also been made. A total of nearly 400 experiments, generalized correlations for the downward and sideward Nusselt numbers have been developed. The results indicate that the Nusselt numbers are nearly independent of the internal Rayleigh number and depend only on the nondimensional quantity (pV³/μg) where V is the superficial gas velocity, p is the pool density, μ is the viscosity, and g is the gravitational acceleration. The results also indicate that the downward and sideward Nusselt numbers in pools with gas release at the boundaries are comparable. This result is in contrast to the natural convection case where the sideward Nusselt numbers are significantly higher than those in the downward direction. The correlations developed in this investigation cover a wide range of the relevant nondimensional quantity (5 x 10^-10 <= (pV³/μg) <= 5 x 10^-3). They can be used to determine the heat transfer rates at the pool boundaries, and hence, the pool growth rate into the concrete. Such information is necessary since it constitutes the last line of defence against violation of the containment. (orig.)

[en] TROI experiments on the interaction of molten corium and water are performed to quantify the risk of a potential steam explosion in a nuclear reactor. In the experiments, a cold crucible technique was used to generate the molten corium. The temperature of a molten corium is one of key parameters which have a strong effect on the behavior of the molten corium and water interaction. As the molten corium has a very high temperature of 2830-3200K, its temperature measurement is very difficult. Generally thermocouples are only used up to 2600K. So an optical pyrometry should be applied to measure the temperature of the molten corium. In the TROI experiment, optical two-color pyrometers are used to measure the corium temperatures. One of the pyrometer is to measure the melt temperature during the melt delivery to the water pool. The melt temperature by the pyrometer has uncertainties due to a spectral transmittance of the view port, and a transient response of the two-color pyrometer. In this paper, uncertainty factors of the optical pyrometry in the TROI experiment, such as a spectral transmittance of a view port, and a time response were discussed

[en] Worldwide on a steam explosion, there are only a few tests using real reactor materials. So the TROI steam explosion experiments have been carried out to evaluate the explosivity of corium. In the previous TROI tests, the effects of the water depth, corium composition and an external trigger were studied on the occurrence of a steam explosion in a 60cm wide interaction vessel (3-D geometry). It turned out that a shallow water pool, eutectic or pure material and the use of an external trigger easily led to steam explosions. However, the use of an external trigger did not cause a steam explosion in a 30cm wide interaction vessel (1-D geometry) with non-eutectic corium of which the composition was 80 : 20 at a weight percent of UO₂ : ZrO₂. In this paper, the results of two TROI tests(test No. 40 and 41) are presented using spontaneously explosive eutectic corium (70 : 30 at weight percent of UO₂ : ZrO₂) in a 30cm wide interaction vessel by applying an external trigger

[en] In-vessel corium retention (IVR) by external reactor vessel cooling (ERVC) shown in Fig. 1(a) has been regarded as a favorable severe accident management for decades. In order for the IVR to be successful, the heat flux at the outer surface of the vessel should be less than critical heat flux (CHF) mostly at the metallic region where heat flux may sharply concentrate. This is called as 'focusing effect', which is a situation where metallic layer above heavy oxide molten pools, when it is thin, may focus heat load to the reactor vessel as shown in Fig. 1a (red-colored arrows). This has been generally regarded as a critical issue for successful IVR since the reactor vessel remaining thickness pertaining its original strength at this region is much smaller or even the vessel would melt through. For illustration of the focusing effect issue, a simple conduction-only computation for a bare slab considering heat loads from two depths of metallic layer with same heat flux is performed as shown in Fig. 2. Important result is that the temperature of the vessel goes higher for the thick heat load rather than for the thin case. For the same heat flux, thick load is thus more threatening the vessel and this is because thick heat load retards heat diffusion in vertical directions. Therefore, 'thin or thick' is not a major factor for estimating the reactor vessel remaining thickness. Validation of computational method of solid melting is performed for an existing experiment for gallium melting in a two-dimensional cavity. The result shows that present computation of melt front advancement is in good agreement with experimental data and existing computations. This method can be and will be later applied to accurate estimation of the reactor vessel ablation behavior under external vessel cooling conditions for diverse inner corium heat load conditions

[en] A gap between a molten material and a lower head vessel is formed in the LAVA experiment, a phase 1 study of SONATA-IV program. In this paper, the quantitative results of the gap measurement using an off-line ultrasonic pulse echo method by digital signal processing are presented. However, the gap measurement signal using an ordinary ultrasonic test would be lack of reliability due to the structural complexity of the specimen. The structural complexity may result from the external reason from the shape and the internal reason from the material characteristics. This paper aims at the development of an appropriate ultrasonic test method, by analyzing the problems from the internal characteristic reason. As the ultrasonic test is affected by internal structures and configurations of test materials, back-scattering, that is, time invariant noise is generated in large grain size materials. Due to the above reason, the received signal results in low signal to noise ratio. Split spectrum processing(SSP) technique is effective to suppress the grain noise in the multi-gap signals. As the result of experimental test, the signal to noise ratio was improved by applying the new SSP technique with multi-gap signal. In this test, the signal of the propagational direction and reflectional direction through solid-liquid-solid specimen was analyzed to understand the behavior of the reflectional signal in a multi-layered structure by filling the gap with water between the melt and the lower head vessel.

[en] Electrothermal complex Vulcan intended to heavy reactor accidents final stages out-reactor research is worked out. Complex consists of 3 experimental facilities: Slava, Lava, Lava M and systems ensuring its operating. Main part of all these facilities is electro-melting furnace of induction type. Complex allows to conduct 3 groups of researches: - on study of melt fall in dry trap; - on study of interaction of melt with water; - on study of interaction of melt with water and concrete under cooling water feed on melt surface situating in concrete trap, and residual power imitation in melt. With help of experiments conducted in complex has been created contrivance for melt confinement with temperature 3000 deg C and above. Time of melt with reactor constructive materials interaction down to the vessel melting that is final stage of catastrophic accident is determined

[en] Although many studies have been performed worldwide on a steam explosion, there are only a few tests using real reactor materials. So the TROI steam explosion experiments have been carried out to evaluate the explosivity of corium. In the previous TROI tests, the effects of the water depth, corium composition and an external trigger were studied on the occurrence of a steam explosion in a 60 cm wide interaction vessel. It turned out that a shallow water pool, eutectic or pure material and the use of an external trigger easily led to steam explosions. However, tests with metal added corium are needed since the real core material contains iron and zirconium. In this paper, the results of two TROI tests (test No. 43 and 44) are presented using a mixture of spontaneously explosive eutectic corium (70 : 30 at weight percent of UO₂ : ZrO₂) and iron with or without applying an external trigger