No abstract

Four carefully machined cylinder-to-cylinder shell models were tested, and the experimentally determined stresses were compared with theoretical predictions obtained from a thin-shell finite-element analysis. The models were idealized structures consisting of two circular cylindrical shells intersecting at right angles. The first model tested had a nozzle-to-cylinder diameter ratio of 0.5 and a diameter-to-thickness of 100 for both nozzle and cylinder. The second model had a nozzle-to-cylinder diameter ratio of 1.0 with a diameter-to-thickness ratio of 100. The third and fourth models had a nozzle-to-cylinder ratio of 0.129. For these models the diameter-to-thickness ratio was 50 for the cylinders and 7.68 for the nozzle of model 3, while it was 20.2 for the nozzle of model 4. All models were strain gaged and subjected to 13 separate loading cases. Comparisons of measured and predicted stress distributions are presented for three of these loadings—internal pressure and in-plane and out-of-plane moments applied to the nozzle. The analytical predictions were obtained using a finite-element program that used flat-plate elements and which considered five degrees of freedom per node in the final assembled equations. The agreement between these particular finite-element predictions and the experimental results is shown to be reasonably good for the four models.

The effect of varying the radii of circumferential fillets in the transition region of cylindrical shells with a step change in outside diameter and loaded with internal pressure has been examined by means of finite element analyses for 27 models. Parameters studied were ratios of the thickness of the thicker section to the thickness of the thinner section, ratios of the fillet radius to thickness of the thinner section, and ratios of the inside radius of the cylindrical shell to the thickness of the thinner section. Comparisons are presented between experimental and theoretical results for two experimental models, and duplicate analyses for two cases were made with different computer programs for comparison purposes.

This report collects and uses available data to mation. Therefore, it is recommended that this cornreexamine the criteria for controlling primary loads in bined loading evaluation procedure be replaced for nuclear piping branch connections as expressed in primary loads by the separate leg evaluation procedure Section III of the ASME Boiler and Pressure Vessel specified in NC/ND-3653.3(c) and NC/ND-3653.3(d). Code. In particular, the primary load stress indices No recommendation is made for fatigue or secondary given in NB-3650 and NB-3683 are reexamined. The load evaluations for Class 1 piping. Further work report concludes that the present usage of the stress should be done on ti_e development of better criteria indices in the criteria equations should be continued, for treatment of combined branch and run moment However, the complex treatment of combined branch effects.

1. INTRODUCTION 2 2. EXPERIMENTAL ANALYSIS 5 2.1 Model Construction 6 2.2 Strain-Gage layout 6 2.3 Test Description 8 2.If Data Acquisition and Reduction 12 3. FIHITS-EIiEkEBT AHAIYSIS 12 3.1 Background 12 3.2 Finite-Element Method 14 3.3 Finite-Element Idealization of Model 19 k.