Nuclear engineering and design/fusion : an international journal devoted to the thermal, mechanical, materials, structural, and design problems of fusion energy最新文献

This paper has been prepared in the framework of the safety analysis of a breeding blanket proposed for NET (Next European Torus). The basic features of the system are the following:

• -

  Li₁₇Pb₈₃ as breeder;

• -

  pressurized (5 MPa) water as coolant;

• -

  AISI 316 SS as structural material.

The breeding blanket consists of 24 segments with an angular opening of 15° placed side by side in the toroidal direction and arranged in the inboard and outboard part of the plasma chamber. The outboard part of the segment is presently under development, and two different design options are proposed:

• -

  a modular concept in which the breeding units (arranged in five rows and four columns), named modules, look like boxes;

• -

  a tubular concept in which the breeding units are tubes bent in the poloidal direction.

In both concepts the vessel of the breeding unit must operate as the first barrier against the accident propagation in case of a pipe break in the unit's cooling system.

The mechanical behaviour of the modular concept, loaded by the pressure transient due to such a pipe break, has been investigated and is presented in detail. The analysis of the results, taking into account material non-linearities, fluid-structure interactions and dynamic effects, shows that the structural reliability of the module vessel cannot be guaranteed, and suggests to continue the development of the tubular concept for which a much better mechanical behaviour is expected.

A major goal of designing a tokamak is to minimize the size of the device and achieve lowest cost. Two key factors influencing the size of the device employing superconducting magnets are toroidal field (TF) winding current density and its nuclear heat load withstand capability. Lower winding current density requires larger radial build of the winding pack. Likewise, lower allowable nuclear heating in the winding requires larger shield thickness between the plasma and TF coils. In order to achieve a low-cost device, it is essential to maximize the winding's current density and nuclear heating withstand capability. A methodology for determining optimum current density is developed by using the Tokamak Fusion Core Experiment (TFCX) as an example. A winding current density of 3500 A/cm² is determined to be optimal at a peak field of 10 T and peak nuclear heat load limit of 1 mW/cm³. This study is based on employment of Nb₃Sn cable-in-conduit conductors cooled with forced-flow helium.

Beryllium is an attractive alternative to graphite for use as armor material for plasma interactive components in fusion devices because of its low atomic number, high strength, and good compatibility with hydrogen. However, beryllium is susceptible to damage from cyclic thermal stresses because of its high elastic modulus and thermal expansion coefficient. We have performed 2-D elastic-plastic finite element stress analyses of prototype beryllium limiter tiles for the JET project that are exposed to a surface heat flux of 250 W/cm² for 15 second pulses. Plastic deformation was predicted to occur at the heated surface during both the heating and cooling phases of the cycle, thereby causing cyclic plastic strain. We also performed thermal fatigue tests using a rastered electron beam to apply the heat load to prototype limiter specimens. After 10 000 thermal fatigue cycles, the only damage of the beryllium tile was microcracking of the heated surface. The depth of this microcracking, 4 mm, corresponds closely to the calculated depth of cyclic plastic strain. These favorable results show that the operating conditions for the JET limiter design can be extended into the regime of cyclic plastic deformation without causing overall structural failure, despite the formation of thermal fatigue cracks.

In this paper an analysis of stress and strain for orthotropic toroidal shells on the basis of the linear theory of thin elastic shell is presented. The asymptotic solution has been obtained.

The results are suitable for λ = E₁ / E₂ > 0.3, where E₁, E₂ are reduced modulus of elasticity in the direction of the meridian and the parallel circle, respectively.

A cross-section sensitivity and uncertainty analysis on four types of fusion reactor blankets has been performed, based on cross-section covariance matrices. The design parameters investigated in the analysis include the tritium breeding ratio, the neutron heating and the fast neutron leakage flux from the inboard shield. Uncertainties in Secondary Angular Distribution (SAD) and Secondary Energy Distribution (SED) of scattered neutrons have been considered for lithium. The collective standard deviation, due to uncertainties in the evaluated cross-section data presently available, is 2–4% in the tritium breeding ratio, 2–3% in the neutron heating, and 10–20% in the fast neutron leakage flux. Contributions from SAD/SED uncertainties are significant for some parameters, such as those investigated in the present study. SAD/SED uncertainties should be considered in the sensitivity and uncertainty analysis on nuclear design of fusion reactors.

An analysis is made of the engineering problems related to the structures facing the plasma in experimental tokamak-type reactors. Attention is focused on the so-called “current first wall”, i.e. the wall side of the blanket segments facing the plasma, and on the collector plates of the impurity control system. The design of a first wall, developed at the JRC-Ispra for INTOR/NET and based on the idea of conceiving it as one of the sides of a box which envelopes a blanket segment, is described. The progress in the structural analysis of the first wall box under operating and abnormal (plasma disruption) conditions is presented and discussed. The design of the collector plates of the single-null divertor of INTOR/NET, as developed at the JRC-Ispra, is described. This design is based on a W-Re protective layer and a water-cooled heat sink, including cooling channels in Cu-alloys and a Cu-matrix for bonding. The results of the elastic and elasto-plastic evaluations are discussed, together with a layout of the experimental activity in progress. It is concluded that, even if the uncertainties related to the plasma-wall interaction are still relevant, the engineering solutions identified look manageable, although they require a large research and development effort.