[en] Beryllium is a favoured candidate for a neutron multiplier in solid breeder blankets of fusion reactors. This is mainly due to its low (n, 2n)-reaction threshold and because of its good thermal and mechanical properties. Its behaviour under intense neutron irradiation, however, is a crucial issue for its use in future fusion reactors. Displacement damage in beryllium so far has been calculated both with data related and methodological deficiencies. First of all, there is a need to have accurate cross-section data in order to obtain reliable spectra of primary knock-on atoms (PKA's). Furthermore, there are principal restrictions of the NRT-model in general used to calculate secondary displacements initiated by PKA's. The underlying theory of damage-energy (part of kinetic energy of PKA transferred elastically to matrix atoms) according to Lindhard is strictly valid only for medium and heavy mass ions with moderate energies in targets of the same element. In this work improved damage cross-sections and displacement rates (dpa/s) in beryllium have been calculated based on cross-section data from ENDF/B-VI (with a significantly improved (n, 2n)-evaluation) and on an appropriate treatment of damage-energy that is suitable for fusion relevant damage of light mass materials. ''This work has been performed in the framework of the Nuclear Fusion Project of the Forschungszentrum Karlsruhe and is supported by the European Communities within the European Fusion Technology Program''. (orig.)

[en] For performance assessment of high-level radioactive waste disposal in salt formations, corrosion tests were carried out, using high active R7T7-type glass containing reprocessing waste, produced by CEA Marcoule. The objective of this investigation was to describe the extent to which Np, Pu, Am and Tc are mobilized from vitrified high-level waste into the near field, when a repository-relevant (Gorleben) concentrated salt solution intrudes the emplacement locations. Furthermore, it should be assessed, if the large data base generated on the reaction behavior of inactive R7T7-type glass under saline conditions is applicable to the high active glass. Glass powder with an average grain size of 86 μm (surface area to solution volume ratio S/V=9370 m^-1) was corroded in a halite saturated concentrated Mg(Ca)Cl₂- salt solution for periods of time up to 720 days at 110 , 150 and 190 C. In order to assess, whether the results from previous inactive tests are applicable to those of the active experiments, the results of both type of experiments are compared. pH changes of the radioactive leachate are the same as in inactive leachates. Acidification of the leachate during glass dissolution results from the formation of solid alteration products, in particular of the clay mineral saponite. The time and temperature dependence of the reaction for the radioactive glass are in excellent agreement with that of the inactive glass suggesting that the corrosion mechanism remains the same. The data for Pu and Am release indicate that fractions of these glass constituents are retained to various degrees during glass corrosion, controlled by the temperature, whereas Np and Tc are released congruently with the soluble elements from the glass. In the presence of corroding container material concentrations of Np and Tc were found to be about 100 times lower whereas Pu concentrations were higher than in the parallel experiment without iron. (orig./DG)

[en] Attention has recently been drawn to the importance of the melt properties, among others the viscosity of oxidic melts, on the heat transfer and on the melt behaviour during MCCI. In the first step the viscosity model in WECHSL was changed in order to achieve a better agreement between the available viscosity measurements and the viscosities predicted by WECHSL. In order to account for the variation of the viscosity with temperature an iterative scheme was used in WECHSL to determine the temperature gradient of the pool bulk temperature to the interface temperature. In addition, the influence of porosity of the melt on the thermal conductivity was modeled. Finally, the so-called property ratio method was implemented for correcting the Nusselt number. A validation matrix which consists of the ACE and SURC experiments was used to assess the improvements of the code. (orig.)

[en] A series of so-called PREMIX experiments is being performed in which the mixing behaviour is investigated of a hot alumina melt which is discharged into water. The tests are part of a multi-lateral programme carried out in support of the licencing procedure of LWR. The programme, which aims at the safety of the reactor, includes relevant experiments and the numerical simulation by computer codes. The parameters of the 1st experimental series were: Melt masses of 10 and 20 kg, released through nozzles of 40 and 56 mm in diameter, resp., the type of melt discharge, a slender size of the water pool, and the degree of sub-cooling. The phenomena of mixing can well be described by means of both high speed and video films and a variety of measurements. The results show that it is the period up to about 0.4 to 0.5 s in which the decisive processes of premixing occur. The very first material release generally occurs as single droplets. In 3 of the first 6 tests, a more or less compact stream of melt formed soon after that. In the other tests, the spray-type discharge of melt continued. A funnel-shaped interaction region is formed in the water pool. Its radial and axial growth rates are determined by the discharge mode and mass flux of the melt. A vapour explosion did not occur. The test results are well suited to be used in the verification and validation of computer models. The results of the measurements and the post-test examination of the debris give an indication of a possible inherent limitation of the masses involved in premixing. Consequently, the probability of a steam explosion to occur would also be limited. This limitation is anticipated to occur in 2 stages: 1. The bulk portion of the water is displaced far enough by the growing interaction region and is prevented from taking part immediately in the thermal interaction. Vapour is the continuous phase in the interaction region. 2. In case of a compact melt stream, the major part of the melt quickly flows down through the interaction region. (orig.)

[en] The neutrino flux produced by the spallation neutron source ISIS at the position of the KARMEN neutrino experiment is calculated to an accuracy of 6.7%. Contributions from the spallation and μSR targets to the ν_μ, ν_e, and anti ν_μ fluxes, due to π⁺ and μ⁺ decay at rest, are evaluated. Results are presented in terms of neutrinos per proton for incident proton beams of 750 and 800 MeV. The contamination of anti ν_e, from the π^- decay-in-flight and μ^- decay-at-rest chain, is found to be 0.06% with an accuracy of 12%. (orig.)

[en] The European Union has been engaged since 1989 in a programme to develop tritium breeding blankets for application in a fusion power reactor. There are four blanket concepts under development. Two of them use lithium ceramics, the other two concepts employ an eutectic lead-lithium alloy (Pb-17Li) as breeder material. The two most promising concepts were to select in 1995 for further development. In order to prepare the selection, a Blanket Concept Selection Exercise (BCSE) has been inititated by the participating associations under the auspices of the European Commission. This BCSE has been performed in 14 working groups which, in a comparative evaluation of the four blanket concepts, addressed specific fields. The working group safety addressed the safety implications. This report describes the methodology adopted, the safety issues identified, their comparative evaluation for the four concepts, and the results and conclusions of the working group to be entered into the overall evaluation. There, the results from all 14 working groups have been combined to yield a final ranking as a basis for the selection. In summary, the safety assessment showed that the four European blanket concepts can be considered as equivalent in terms of the safety rating adopted, each concept, however, rendering safety concerns of different quality in different areas which are substantiated in this report. (orig.)

[en] FLUTAN is a highly vectorized computer code for 3D fluiddynamic and thermal-hydraulic analyses in Cartesian or cylinder coordinates. It is related to the family of COMMIX codes originally developed at Argonne National Laboratory, USA, and particularly to COMMIX-1A and COMMIX-1B, which were made available to FZK in the frame of cooperation contracts within the fast reactor safety field. FLUTAN 2.0 is an improved version of the FLUTAN code released in 1992. It offers some additional innovations, e.g. the QUICK-LECUSSO-FRAM techniques for reducing numerical diffusion in the k-ε turbulence model equations; a higher sophisticated wall model for specifying a mass flow outside the surface walls together with its flow path and its associated inlet and outlet flow temperatures; and a revised and upgraded pressure boundary condition to fully include the outlet cells in the solution process of the conservation equations. Last but not least, a so-called visualization option based on VISART standards has been provided. This report contains detailed input instructions, presents formulations of the various model options, and explains how to use the code by means of comprehensive sample input. (orig.)

[en] This publication deals with results of a series of containment thermalhydraulic and aerosol studies made with the multicompartment code system CONTAIN 1.12 for the VANAM experiment M3. This work was done as a participation to the open ISP37. VANAM M3 was planned to give experimental data for verification of aerosol computer codes and test their thermohydraulic and aerosol results on such a large integral multicompartment containment experiment. Five different calculations are compared in this paper. The influence of simplifications was studied for example by - taking only oxygen instead of air as atmospheric gas - modeling the outer structures only by concrete - using flow coefficients according to Idel'chick or improved values -including radiation transport effects. CONTAIN thermohydraulic results are within the accuracy that can be obtained regarding the uncertainties of the experiment and the the boundary conditions used as code input. CONTAIN is able to simulate atmospheric stratifications as observed in the experiment. VANAM M3 used water soluble NaOH aerosol to study a 'dry' aerosol depletion phase and later after a second aerosol injection the behaviour in a 'wet' surrounding. The aerosol densities differ by more than an order of magnitude in the different compartments. To calculate this behaviour a multicell code like CONTAIN with its powerful MAEROS tool for the hygroscopic aerosol treatment was needed. The aerosol transport to different cells and the aerosol densities in these rooms are calculated qualitatively correct with CONTAIN 1.12. (orig.)

[en] SIMMER-III is a computer code to investigate core disruptive accidents in liquid metal fast reactors but should also be used to investigate safety related problems in other types of advanced reactors. The code is developed by PNC with cooperation of the European partners FZK, CEA and AEA-T. SIMMER-III is a two-dimensional, three-velocity-field, multiphase, multicomponent, Eulerian, fluid-dynamics code coupled with a space-, time-, and energy-dependent neutron dynamics model. In order to model complex flow situations in a postulated disrupting core, mass and energy conservation equations are solved for 27 density components and 16 energy components, respectively. Three velocity fields (two liquid and one vapor) are modeled to simulate the relative motion of different fluid components. An additional static field takes into account the structures available in a reactor (pins, hexans, vessel structures, internal structures etc.). The neutronics is based on the discrete ordinate method (S_N method) coupled into a quasistatic dynamic model. The code assessment and verification of the fluid dynamic/thermohydraulic parts of the code is performed in several steps in a joint effort of all partners. The results of the FZK contributions to the first assessment and verification phase is reported. (orig.)

[en] Within the framework of the European Fusion Technology Programme, a primary vacuum pump for the ITER reactor is being developed. As the tritium accumulated by the pumps must be limited, short pumping cycles are necessary and as a consequence to that regeneration times of about 4 min only are required by the intermittently working cryopumps; approximately 60 s are available for the heating process from LHe temperature (4.2 K) to LN₂ temperature (77 K). Methods for fast heating were tested in component tests. The heating tests were performed at the TITAN test facility. According to the basic planning, the LHe-cooled panel consisted of seven flow channels in quilted design (500 x 350 mm²); the detailed planning meanwhile showed that a smaller number of channels per panel will be sufficient. The panel was mounted in a LN₂-cooled rig, which worked as first pumping stage. After having worked out a screening study comprehensive test series with three different heating methods were performed. (orig.)