مستخلصات رسائل السعوديين لدرجة الدكتوراة — الصفحة ٥٦١

Technology 407 council. Taylor, Lind, and Schresults of three-dimensional weiker," and Taylor and Lind" reported photoelastic studied of reinforced shell interections. Un- fortunately, as Mershon's pointed out, the results of these studies can not be used to check the validity of theoretical solutions due to scatter of the photo- elastic data. Results of experiments on «as supplied» steel models were reported by Har- denbergh and Zamrik.19 Riley20 reported results for spherical and cylindrical shells subjected to a variety of loadings. At present a program is under way at Oak Ridge National Laboratory to test several machined steel models under a variety of loadings. Corum, Bolt, Greenstreet, and Gwalthy41,22, reported results of tests on tow of these models. Since the emergence of the finite element as a powerful tool of stress analysis a number of solutions of the problem of stress distribution near thin-shell clynd- er-to-cylinder intersections have appeared bending elements for shell analysis. As an exampio to show the capability of their program, they analyzed the ONRL model of Ref 22. Their results for the internal pressure loading were in reasonable agreement with experimental values. The same model was analysed by Corum et al. 21,22 using a program originally developed by Greste2 to analyse tubular joints. The element used in Greste's program is a rectangular elements that included bending and membrance behavior. Numerical results obtained using this program showed generally good agreement with experimental values. Finite element analysis of the model tested by Riley20 was reported by peterson and Kulunk. Analysis of inclined intersections along with normally intersecting cylinders was given by Ando, Yagawa, and Kikuchi26. The element used in their study was based on Novozhilov and Mushtari-Viasov shell theory. 1.2 Objective and Scope It is clear from the above review of previous work in the subject area that a reliable general analytical method of stress analysis of cylinder-to-cylinder in- tersections is not yet available. With regard to the fininite element solutions, the following is noted: 1 - Whenever flat elements are used, the smooth shell surface is represented by a faceted surface. This gometric approximation introduces an error in the analysis, however, the magnitude of the error dimicishes with refinement of the mesh; and. 2- In thin shell theory, rotation about the normal to the shell middle surface is usually neglected. This introduces a problem at shell intersections because
ص ٥٦١

النص مستخرج آليًّا من صور الكتاب، وقد يقع فيه خطأ.