[en] The new method allows to compute effective primary stress Peff with the help of an efficiency diagram. Distortion of the structure under cyclic loading is near the distortion obtained by the static application of this effective stress. At elevated temperature, the limitation of the value of this effective primary stress at St, insures the limitation of strain at 1%. Therefore it is recommended to use for Peff the same allowable value than for conventional primary stress. Such a recommendation is very different from the current practice at low temperature. It has been pointed out that this current practices tolerates very much larger elongation under cyclic load than under static load

[en] This paper presents a simplified method of analysis of buckling of thin structures like those of L.M.F.B.R.'s. The edification of the method is very similar to methods used for buckling of beams and columns having initial geometric imperfections and buckling in the plastic range. Particular attention is paid to the stress hardening of material involved (austenitic steel) and to possible unstable post buckling of thin structures. The analysis method is based on elastic calculation and diagrams that take into account various initial geometric defects

[en] Fatigue behaviour is essentially dependent on the real strain range, but the current practice is the use of elastic analysis. In low cycle fatigue conditions where inelastic strains predominate, elastic analysis never gives the real value of the strain range. In order to use these results some corrections are necessary. One of these corrections is due to the Poisson's effect (the Poisson ratio in inelastic behaviour is higher than in elastic behaviour). In this paper a method of correction of this effect is proposed. It consists in multiplying the results of the elastic analysis by a coefficient called Kν. A method to draw curves giving this coefficient Kν as a function of results of elastic analysis is developped. Only simple analytical computations using the unixial cyclic curve are needed to draw these curves. Examples are given. The proposed method is very convenient and low cost effective

[en] Many investigators tried to reproduce by inelastic calculations the experimentally observed behavior. Unfortunately, the material characteristics used in the computer code were established from monotonic tensile tests performed with specimens extracted from the plate product itself (1,45 mm thick) employed to construct the ratchetting specimen. It seems that, at the present time, cyclic behavior of the material is much more relevant to the observed phenomenon. So the need is expressed to make this of characterization. Nevertheless, the practical problem is to produce cyclic stresses, i.e. tensile and compressive stresses, with very thin specimen. The main difficulty is, in particular, to present buckling effect. It is described a new special device set up for this particular purpose. The adopted solution was to create uniformly distributed alternative pure bending stresses in the thin plate specimen. The bending moments are produced by two end-grips fixed to the specimen, and these grips are mounted on a conventional test machine which is displacement-controlled. In order to reduce tensile and compressive membrane stresses inside the specimen, the grips have two parallel axles of rotation. Forces produced by the machine and displacements of a few points of the specimen are continuously recorded during the test, so that cyclically stabilisized bending moment can be easily evaluated for each variation of curvature imposed. It is furnished the very first experimental cyclic data obtained for the material of the sodium test specimen, a 316 type stainless steel, at room temperature. Finally it is worth noticing that the specimen are of very simple fabrication and are for this reason not expansive

[en] Development of leak before break methodology in nuclear industry requires determination of conditions of crack stability in piping products. Due to tough and strain-hardening materials involved, a promising parameter to characterize such a behaviour seems to be the driving force J. An experimental program is carrying on in CEA at Saclay in order to establish and to verify analytical and experimental methods to predict conditions of crack stability. It concerns circumferentially cracked tubes. The through wall crack center angles range from 30⁰ to 150⁰. Blunt end notches are considered, as well as fatigue precracked notches. Loading is imposed monotonically in displacement controlled conditions until maximal load is reached. The paper will present experimental procedure and first results obtained. Instrumentation and typical recordings (load, rotations at different distances from crack section, crack opening displacement, electric potential drop, ovalization) will be described. Experimental results are interpreted in terms of limit analysis and J estimation to predict crack initiation and maximal loads as a function of crack length. Results obtained permit the adjustment of experimental scaling functions usually employed for J evaluation with one single specimen. These functions are compared with analytical ones based on theoretical considerations of a simple limit state of the pipe cracked section

[en] The paper provides a survey of the creep-fatigue design rules for the LMFBR in France. These rules are the ones currently implemented in French component manufacturing. The background of each item is discussed and the trends for improvements currently investigated are described. The creep-fatigue rules apply to elastic analysis only. (orig.)

[en] During the operation of nuclear plants, components experience low-cycle fatigue due to thermal transients. A good knowledge of the local strain range is therefore highly desirable for low-cycle fatigue analysis. There are two basic discrepancies between the plastic behaviour of the material and the linear elastic model. The first is the non-linear stress/strain relationship, which mainly affects the strain range in the vicinity of stress raisers. The second is that plastic deformation shows no change in volume. This paper covers the corrective action related to this absence of volume variation. Current practice is to require that stresses are evaluated on an elastic basis, although with a P.R. (Poisson's ratio) value different from the elastic P.R. This procedure is rather inconvenient, and seldom used. It appears preferable to use an alternative rule requiring only multiplying the computed equivalent elastic strain range by a corrective factor. Under plane stress conditions (e.g., close to a wall), the actual equivalent strain range can be computed using an equivalent P.R. This permits deriving the corrective factor Ksub(ν) and plotting curves showing Ksub(ν) as a function of the elastic strain range for a given material

[en] This paper highlights the main items emerging from a large bibliographical survey carried out on strain-induced buckling analysis methods applicable in the building of fast neutron reactor structures. The work is centred on the practical analysis methods used in construction codes to account for the strain-buckling of thin and slender structures. Methods proposed in the literature concerning past and present studies are rapidly described. Experimental, theoretical and numerical methods are considered. Methods applicable to design and their degree of validation are indicated

[en] This report is a State of the Art about practical methods to analyze buckling risks mainly due to thermal stresses in slender shell structures. A critical review of theoretical, numerical and experimental results available in open literature till 1986 is performed. They are particularly examined from the point of view of simplicity in the formulations and experimental validation. The final aim of this study is an attempt to propose analysis method of practical use for engineers. Most of used informations were obtained from aeronautic and nuclear (fast breeder reactors) domains

[en] The purpose of this paper is to describe and compare published methodologies available in Europe to perform crack assessment in nuclear components operating at elevated temperatures and where creep can play an important role. It concerns mainly the methods proposed in UK by the R5 rules and in France by Appendix A16 of RCC-MR, and developed mainly for fast breeders reactor applications. The paper is a summary presentation of the available rules, their domain of applications and gives references to the associated validation works (experimental and numerical). The organisation of the rules is very similar in both procedures: high and low temperature domains, evaluation of different phases in the crack behavior, and initiation and propagation under normal conditions and then stability analysis under accidental loading. Both procedures use the fracture mechanics parameters K_I, C^*, σ_ref. J parameter is explicitly calculated in A16 procedure whereas the R6 Failure Assessment Diagram approach, which is derived from a generalised J-estimation scheme in terms of the materials stress-strain curve, avoids J-calculation. The two methods described were considered in the writing of procedures proposed for the assessment of the design of EFR and discussed in the framework of AGT9B agreement. (J.S.). 20 refs., 7 figs