# I. Introduction hell [1] is applied to bodies bounded by two curved surfaces, where, the distance between the surfaces is small in comparison with other body dimensions. The centre of points lying at equal distances from these two curved surfaces defines the middle surface of the shell. The lengths of the segment, which is perpendicular to the curved surfaces, is known as the thickness of the shell and is denoted by h. Shells have all the characteristics of plates, along with an additional one, which is curvature. Mindful of intrinsic, functional essence of shells, [2] presents shells as skin structures by virtue of their geometry and shell action, is essentially more towards transmitting the load by direct stresses with relatively small bending stresses. In line with this functional essence of the shell, [3], [15], shells are spatially curved surface structures which support applied external loads or forces. Shell structures [3], [5], [7] can be referred to as "form resistant structures". This implies a surface structure whose strength is derived from this shape, and which resists loads by developing stresses in its own plane [3], [5], [7]. An early form of shell construction [6] was the dome known to Romans thousands of years ago. With shell concrete construction, [6] it becomes quite possible to produce satisfactory domes which weigh only a fraction of the weights of the much earlier massive domes. Properties of shells [1], [5] which are of particular importance in structural usage and which also earn wide application of shell structures in engineering are the following: (i) Efficiency of Load carrying behavior; (ii) High degree of reserved strength and structural integrity; (iii) High strength versus weight ratio; (iv) Very high stiffness; (v) Containment of space. Areas where shell structures [1], [4] are used in Building and Civil Engineering are:-(i) Large-span roofs; (ii) Liquid retaining structures and water tanks; (iii) Containment shells of nuclear power plants; and (iv) Concrete arch domes. Shell forms in Mechanical Engineering [1], [4] are used in: (i) Piping systems; (ii) turbine disks; (iii) Pressure vessels technology. The use of shells [1], [4] in aeronautical and marine engineering are in the following forms:-(i) aircrafts; (ii) missiles; (iii) rockets; (iv) ships; and (v) submarines. Shells [8] found in various biological forms such as the eye, the skull and the egg, represent another application of shell engineering, this time, in the field of biomechanics. An egg, as a natural thin-walled structure, can be considered as one of the most beautiful structural shapes. It combines extreme fitness for its purpose with an economy of material and cleanliness of design. This account depicts only a small list of shell forms in engineering and nature. There are [1], [8] two different classes of shells: thin and thick shells. Shells are said to be thin when the ratio of their thickness, h, to the radius of curvature R of the middle surface is less than or equal to 1/20, i.e. h/R ? 1/20. For a large number of practical applications [1], the thickness of thin shells lies within the range: 1/100 ? h/R ? 1/20. Hence, shells for which these h/R stipulations do not lie within the stated range, belong to thick shells [1]. Thin plates and thin shells belong to a category of structures known as Thin-Walled Structures. Thinwalled structures [9] possess the following three characteristics: (i) two dimensions are much longer than the third, its thickness; (ii) They have a great strength as a result of their spatial character of working under the action of external loads; (iii) They make use of a minimal quantity of material. Depending on the curvature of the surface [1], [3], shells are divided into (i) cylindrical, comprising of noncircular and circular; (ii) conical; (iii) spherical (iv) ellipsoidal; (v) paraboloidal; (vi) toroidal; and (vii) hyperbolic paraboloidal shells. The usual theory of thin shells utilizes the main suppositions of the theory of thin plates. Nonetheless, thin plate and thin shell have a substantial difference in behaviour under external loadings. The static equilibrium of a plate element under lateral load [2], [8] is only possible by the action of bending and torsional or twisting moments, usually accompanied by shearing forces. Conversely, a shell, in general, [10] is capable of transmitting the surface load by "membrane" stresses uniformly distributed over its thickness. This property of shells [8], [10] makes them to be not only more economical but also more rigid than plates and other types of construction under the same conditions. Apart from these obvious advantages over other systems [1], [8], shell structures are very well known and used for their performance, strength against accidental damage, resistance to fire and low upkeep cost as well as their aesthetic appearance. The economy and/or feasibility of many modern constructions necessitate light weight, a property which thin-walled structures are replete with. Strictly speaking, the aim in structural engineering has always been to lower, as much as possible, the cost and thus the quantity of material used without, compromising the structural integrity of the system. Thin-walled structures meet this requirement. The deliberate effort in the analysis and design of shells [3] is to make the shell as thin as practical requirements would permit, so the dead weight is reduced and the structure functions as a membrane free from the large bending stresses. Thin shell concrete structures [3] are pure compression structures formed from inverse catenary shapes. The inverse catenary is a pure compression scenario. Pure compression is ideal for concrete, as concrete has high compressive strength and very low tensile strength. These shapes maximize the effectiveness of concrete, allowing it to form thin light spans [3]. This paper presents the application of initial value model in analysis of underground circular cylindrical shell structure subject to axi-symmetrical loads of hydrostatic pressure but considered under two conditions: (i) when the tank is empty; (ii) when the tank is full. In addition, the paper seeks validity through the classical model. Aim in this study is to investigate stress effect on the underground cylindrical shell structure when full of water as well as when empty. The study intends to achieve the aim through the following objectives: 2) To evaluate the five internal stresses of the underground cylindrical shell structure when full of water, using the two models named above. 3) To determine the five internal stresses of the underground cylindrical shell when empty, also using the same two models. # II. Previous Works The analysis and design of shells attracted many researchers. Among them, perhaps the best known, are: Love, U. F., Pasternak, P. L. and Timoshenko, S. P. [11]. # a) The Membrane Theory Modern shell construction [11] has its origin in the work of Lame and Clopeyron who, in 1826, proposed the membrane analogy. This theory suggests that a shell is capable of resisting external loads by direct stresses called membrane stresses without bending. Hence, when membrane theory is applied for shell design, torsional or twisting, bending moments, and shear forces in the cross-section are neglected. This is only possible if torsion and bending stresses were small compared to stresses of normal or axial, and shear forces. Membrane theory fails to represent the true stresses in those portions close to the edges, since the edge conditions usually cannot be completely satisfied by considering only membrane stresses. It is expected [12] that membrane theory gives an approximate picture of stress distribution in the case of shells not long, say L ? 2R, where R is radius of the shell and L is its length. For longer shells a satisfactory solution can be obtained only by considering bending as well as membrane stresses [12]. # b) The Moment Theory Mathematical conceptions [6], developed during the 19 th century, made possible a more accurate analysis than could be achieved using any membrane theory. Aron, H., who derived an expression for potential energy of a shell as well as equations for shell equilibrium and strains, was the first to consider the new theory, evolved from those mathematical conceptions, which made room for both membrane and bending stresses [11]. Love [13] developed a detailed derivation of equilibrium equations and equations of motion of shells with correction to a number of slips in Aron's original treatment as analogous to the theory of plates of Kirchhoff [14], and was based on identical assumptions. In each particular case, the moment theory involves the solution of a system of three differential equations, which is very complicated. # c) The Semi-Moment Theory In the 20 th Century, [15] suggested a simplified method of analyzing and designing cylindrical shells using the theory of semi-moment. This theory, on the basis of experimental data concerning medium length cylindrical shells, length/diameter = 2 -8, neglects the effect of longitudinal bending moment, shear forces and torques, and introduces geometrical hypothesis. This method has the advantage to be simpler than the moment theory and gives accurate solution when a distributed load is applied. This paper admires this method for the simple reason it appears to be the most appropriate for circular cylindrical shell loaded symmetrically with respect to its axis -the axi-symmetric loading scenario. Detailed information on general thin-walled structure theory were given in many general treatises such as: [9], [8] and [16], as well as monographs: [10], [17], [18], and [19]. These consider, in a great extent, the mathematical theory of thin-walled structures and the derivation of differential equations pertaining to them. Furthermore, the possibility of solving various shell-theory problems using analytic methods, has also been discussed in: [25], [26], [27], [28], [29], [30], [31], [32], and [33]. # d) Methods of Solution As far as techniques of solving the derived differential equations are concerned, difficulties involved in realizing mathematically rigorous methods, led to approximate techniques of integrating equilibrium equations. One of such methods, known as asymptotic integration, consists in replacing a given differential equation by another with specially selected coefficients different from those in the exact equation, and whose solution can be obtained by strict method and expressed in elementary functions. Blumenthal [20] suggested the method in 1912. In 1913, Timoshenko [21] applied it to shell equations. Shtayerman, Novozhilov [8], and Gol'denveiser [22] perfected the same asymptotic integration method. Another widely accepted method was a version of the membrane theory which takes boundary effects into cognizance. Geckeler's equations [23] replaced exact equations owing to the presence of boundary effect. Generally speaking, methods for obtaining solutions for the shell differential equations [8] can be classified as follows: (i) Exact analytical methods, also known as classical solution; (ii) methods using variational calculus; (iii) numerical methods such as: finite difference, finite element, finite strip, to mention but a few; (iv) approximate methods based upon exact equilibrium of the problem. Among others [1], [8], the classical solution, the finite difference, the finite element, the finite strip, the boundary element, the boundary collocation, and the boundary value methods are some of the well-known models used for solution of shell problems. The classical solution gives accurate results but fails to capture the boundary conditions. The finite difference model considers the boundary conditions but does not give room to further optimization for the simple reason solutions are obtained only for some selected nodes. Finite element generates very large matrices for considerable accuracy, handling of those large matrices being only suitable for use of computer. Besides, they are numerical models. Numerical models are more inclined to approximate solutions than exact solutions. Those models connoting boundary in their names, sound though they may be, involve a level of mathematics beyond the scope of an average engineer! The semi-moment theory is as suitable for circular cylindrical shells under hydrostatic pressure and uniform gas pressure as it is amenable to application of initial value model. This study adopts the initial value model for among the suggested analytical solutions. Initial value model gives the least number of unknowns enabling such manual handling as to imbue the analyst with the ability and privilege to solve a problem in so systemic a manner that represents a mandatorily profound understanding in the-every-step-of-the-way of the problem for the analyst. ( ) 0 3 2 2 0 2 0 1 h 4 (Z)W y (Z) W ? ? + = Z y Eh R Z y ? ? ? - ( ) 0 4 2 4 Q Z y Eh R ?????????????.. (2) ( ) ( ) ( ) ( ) 0 4 2 3 0 1 4 4 4 M Z y Eh R Z y Wh Z y Z h ? ? ? ? ? + ? = - ( ) 0 4 2 2 4 Q Z y Eh R ? ?????????????.. (3) ( ) ( ) ( ) ( ) 0 1 0 4 3 3 0 3 2 2 M Z y Z Y R Eh W Z y R Eh Z M h + + = ? ? ? ??????????????.. (4) . ( ) ( ) ( ) 0 3 2 2 0 2 2 ? ? ? Z Y R Eh W Z y R Eh Z Q h + = ( ) ( ) 0 1 0 4 4 Q Z y M Z y + ? ?????????????.. (5) ( ) ( ) ( ) 0 2 0 1 ? ? ? ? ? ? ? ? + ? ? ? ? ? ? = Z y R Eh W Z y R Eh Z N h ( ) ( ) 0 4 0 3 2 4 4 Q Z y M Z Ry ? ? ? ?????????????.. (6) ii. The Initial Value Particular Solution Consider the following: d Q x -x dx x 1 1 1 x Q 0 A B C ? Z ? 1 Z Z ? ? 1 dZ ? 1 Z IV. Analysis a) The Governing Equation of Equilibrium [1] and [24] derived the governing differential equation of equilibrium as: D W dx W d x ? ? = + 4 4 4 4 ????? (1) Using, for the derivation, the following: i. Forces acting on the shell; ii. Moments acting on the shell; iii. Equilibrium equations of stress; iv. Equations of forces and moments displacements. # b) Initial Value Solution of The Governing Differential Equation of Equilibrium for Circular Cylindrical Shell Structure i. In Summary, initial value homogenous solution becomes: 1) is applicable only to circular cylindrical shell structure subjected to axi-symmetric loading. This means the loading or forces and moments do not vary along the circumferential section. In other words, the loads are radially symmetrical loads. 0 2 Q Z y ? + Equation ( For radially symmetrical loads, the governing differential equation of circular cylindrical shell structure, the equation of equilibrium, is equivalent to that of Beam on an elastic Winkler Foundation [34]. Understanding an origin transformation leads to finding the initial value particular solution. The origin that was previously at A is shifted to B, while introducing, at the same time, a new variable x 1 . ? ? 1 1 1 1 Z Let dZ dx x = ? = He distributed load at the new origin, B, is given by: ( ) ? ? 1 1 Z Z Z q w ? = ? The elemental force can be expressed as: ( ) ( ) 1 2 1 1 1 dZ Z Z dZ Z q dQ ? ? ? ? = = ???? (7) In summary, the initial value particular solution would be: ( ) [ ] Z y Z Eh R W p 2 2 ? = ? ? ?????? (8) ( ) [ ] Z y Eh R p 1 2 1 ? = ? ? ???.. (9) ( ) Z y M p 4 3 ? ? ? = ???.. (10) ( ) Z y p 3 2 ? ? ? ?????. (11) ?????. (11) ( ) [ ] Z y Z R N p 2 ? = ? ? ???.. (12) Obviously, general solution = homogeneous solution + particular solution. # c) Classical Model Homogeneous Solution Basically, in summary, the homogenous solution would be as follows: ( ) 17) Equation [13] can be expressed as: x i x x i x i x x x i x h( ) ( )( ) ( )( ) ( )( ) ( )( ) x iSin x Cos x Sinh x Cosh C x iSin x Cos x Sinh x Cosh C x iSin x Cos x Sinh x Cosh C x iSin x Cos x Sinh x Cosh C x W h ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? + ? + ? ? + ? + + + + = 4 3 2 1 Expanding the above equation gives: ( ) x xSin Cosh A x xCos Cosh A x W h ? ? ? ? 2 1 + = x xSin Sinh A x xCos Sinh A ? ? ? ? 4 3 # + + ???????????? (18) Where: A 1 = C 1 + C 2 + C 3 + C 4 A 2 = (C 1 -C 2 -C 3 + C 4 ) i A 3 = C 1 + C 2 -C 3 -C 4 A 4 = (C 1 -C 2 + C 3 -C 4 ) i d) # Classical Model Particular Solution Recall the static equation of equilibrium: x D W W IV ? ? = + 4 4 ????????????????????????..(19) Assuming the particular solution is of the form: W p (x) = ax ???????????????????????? (20) # Global Journal of Researches in Engineering ( ) Volume XVI Issue III Version I It follows that: ( ) 0 = x W IV p ???????????????????????.. (21) Making use of equations [20] and [21] in eqn [19] gives: Therefore, the classical model particular solution would be: A comparison of co-efficient in equation [22] yields: ( ) x D x W p 4 4 ? ? = ??????????????????????? (24) # b) Internal Stresses The deflection curve, for the hydrostatic loading, fig. 2, has a parabolic like shape with a speak of 0.854 mm at x = 3.2m from the top. Zero deflection is obtained at both top and bottom ends of the reservoir, justifying zero deflection at the supports. Also, tables 3 and 4 refer. Under hydrostatic pressure, fig. 2, tables 3 and 4, the reservoir presented a maximum rotation of 0.078 x 10 -3 radians at its top end. For the hydrostatic loading, fig. 3, tables 3 and 4, the bending moment varies from zero, at the top, to 20.5KMN at the bottom, describing a concave parabolic-like curve due to the cantilever action. In the case of hydrostatic loading, the shearing force varies from 0.472KN, at the top, through 42.503KN, at the bottom, fig. 5, tables 3 and 4. The graphs, fig. 6, have the same shape as those obtained for direct deflection. For real, deflection and hoop tension are directly proportional. For the hydrostatic pressure, a peak of 180.729 KN is reached at x = 2.67m, fig. 6, tables 3 and 4. The tables 3 and 4 show values for deflection and rotation at the bottom, x = 4m, not exactly equal to zero, but very close to zero in the extent they can be taken as zero with sufficient accuracy. Table 5 reveals: (i) percentage difference for applied pressure on tank wall, when the tank is empty versus when the tank is full of water, is equal to: 36.572; (ii) Overall average percentage difference value for results of internal stresses for, when the tank is empty versus when the tank is full of water, is equal to: 36.572. # c) Conclusion and Recommendation The half-moment theory, otherwise known as semi-moment theory appears to have proven to be one of the most accurate theories in shell analysis. It has the advantage to be more realistic than the membrane theory which does not consider the bending effect. The semi-moment theory is simpler than the moment theory which generates heavy equations. In terms of capturing the boundary conditions of systems, the initial value model is more equipped than the classical model. Sequel to results revealing that: percentage difference for applied pressure on tank wall: when the tank is empty versus when the tank is full of water is equal to 36.572%; again the overall average percentage difference value for results of internal stresses with respect to: when the tank is empty versus when the tank is full of water came to 36.572% as well, stress effect on the water reservoir which resulted from the two conditions would not diminish the fundamental structural integrity of the underground circular cylindrical shell structure, if all other sources of stress such as from weather elements are under control. ![Journal of Researches in Engineering ( ) Volume XVI Issue III Version I 32 Year 2016 E © 2016 Global Journals Inc. (US) Application of Initial Value Model in Circular Cylindrical Shell Analysis 1) To solve the governing differential equation of equilibrium of cylindrical shells subjected to radially symmetrical loads using both initial value and classical models.](image-2.png "Global") 1X(m)ZDeflectionSlope ?( z)BendingShear ForceHoop TensionW( z )(mm)radiansMomentQ( z ) (KN)N( z )(KN)M( z ) (KNM)0000.00012253600.74385975500.81.036008026 0.0000527880.000123922-0.4857128716-1.18317430991.039381511.62.072016051 0.0001659170.000129602-1.516548554-3.019258682170.08852452.43.108024077 0.0004734290.000145863-4.247252695-8.416752018226.05944082.673.453360086 0.0006704750.000156135-5.962376315-11.88682999284.93527863.24.144032102 0.0001346480.000190643-11.73091094-23.73505572211.33603984.05.1800401283.55 x 10 -11-4.4 x 10 -11-32.31896214-67.010174432.7 x 10 -7 2X(m)ZDeflection W(Z)(mm)Slope ( ) Z ? radiansBending M(Z) (KNM) MomentShear Force Q(Z) (KN)Hoop Tension N(Z)(KN)0000.0001224860-0.74385973100.81.036008026 0.0000527640.000123834-0.48571286i8-1.18317434291.039381841.62.072016051 0.0001658550.000129585-1.516548484-3.019258654170.08852582.43.108024077 0.0004733960.000145792-4.247252615-8.416752001226.05944632.673.453360086 0.0006704930.000156193-5.962376293-11.88682675284.93527653.24.1440321020.001345920.000190589-11.73091124-23.73505498211.333603584.05.1800401284.27 x 10 -123.84 x 10 -13-32.31896187-67.010174132.3 x 10 -8 3Year 201636Global Journal of Researches in Engineering ( ) Volume XVI Issue III Version I EX(m) 0 0.8 1.6 2.4 2.67Z 0 1.036008026 2.072016051 3.108024077 3.453360086Deflection W(Z)(m) 0 0.000033482 0.000105238 0.000300286 0.000425269Slope ( ) Z ? (radians) 0.000077722 0.000078601 0.000082204 0.000092518 0.000099033Bending Moment M(Z) (KNM) 0 -0.3085154309 -0.9619133775 -2.693938928 -3.781804103Shear Force Q(Z) (KN) -0.471813873 -0.750461426 -1.915049339 -5.338560581 -7.539554672Hoop Tension N(Z)(KN) 0 57.7444589 107.8828632 143.3830003 180.72874853.24.1440321020.0008540470.000012092-7.44065868-15.05462351134.04622334.05.180040128-2.53 x 10 -11-3.4 x 10 -11-20.49920654-42.50307896-1.14 x 10 -7© 2016 Global Journals Inc. (US) 4X(m)ZDeflection W(Z)(m)Slope ( ) Z ? (radians)Bending M(Z) (KNM) MomentShear Force Q(Z) (KN)Hoop Tension N(Z)(KN)0000.0001776940-0.47181379200.81.036008026 0.0000334710.000078595-0.3085154293-0.75046138657.74445361.62.072016051 0.0001051680.000082193-0.9619133724-1.915049295107.88285842.43.108024077 0.0003002450.000092512-2.693938892-5.338560528143.38300012.673.453360086 0.0004252180.000099021-3.781804084-7.539554622180.72874233.24.144032102 0.0008540140.000012024-7.44065818-15.05462316134.04621944.05.180040128-3.62 x 10 -122.94 x 10 -13-20.49920612-42.503078541.21 x 10 -8 5tank is full = 36.572Fig. 2 : Deflection Diagram Year 2016 E © 2016 Global Journals Inc. (US) © 2016 Global Journals Inc. (US) Application of Initial Value Model in Circular Cylindrical Shell Analysis © 2016 Global Journals Inc. (US)This page is intentionally left blank * EduardVentsel TheodorKrauthammer Thin Plates and Shells -Theory, Analysis, and Applications Marcel Dekker, Inc. New York 2001 * Design aids for Fixed Support Reinforced Concrete Cylindrical Shells under Uniformly Distributed Loads, Int'l J of Engineering SChandrasekaran SKGupta Carannante F Science and Technology 1 1 2009 * A Consistent First Approximation in the General Theory of Thin Elastic Shells WTJoiter Proc. Symp. on Theory of Thin Elastic Shells Symp. on Theory of Thin Elastic ShellsDelft; North -Holland Amsterdam August, 1959. 1960 * Shell Structures in Civil Engineering, Theory and Closed-Form Analytical Solution AlphoseZingoni 1997 Thomas Telford, London * CRCalladine Theory of Shell Structures London Cambridge University Press 1983 * A new Classification for Shell Problems ABlouza FBrezzi CLovadina Publicazioni IAN -CNR n. 1128 1999 * Design and Analysis of Reinforced Concrete Shells VSJoythi Int'l J. for Scientific Research and Development 3 09 2015 * Theory of Thin Elastic Shells, 2 nd edn VVNovozhilov P. Noordnoff 1964 * VZVlasov General Theory of Shells and its application in Engineering, NASA TTF -99 1964 * Flexible Plates and Shells in [Russian ASMir 1956 Gostekhizdat, Moscow * UFinsterwalder FDischenger BauingenierIn 1928 9 Berlin * A New Derivation of the Equations of the Deformation of Elastic Shells EReissner Am J. Math 63 1 1941 * A Treatise on the Mathematical Theory of Elasticity, 1 st edn AE HLove 1982. 1944 Dover New York 4 th edn * GRKirchhoff Sungenuber Mathematishe Physik, Mechanik 1877 * A General Theory of Shells VZVlasov 1949 Fizmatgiz, Moscow * SPTimoshenko SWoinowsky-Krieger theory of Plates and Shells New York McGraw-Hill 1959 * Theory of Thin Shells Monograph VVNovozhilov 1962 Sudostroerue, Leningrad Russian * Statics of Thin-Walled Elastic Shells ALLur'e 1947 Gostekhizdat Moscow -Leningrad in Russian * Theory of Elastic Thin Shells in [Russian ALGol'denveize 1953 Gostekhizdat, Moscow * OBlumenthal Proc of the Fift Int'l Conf. on Math of the Fift Int'l Conf. on MathCambridge 1912 * SPTimoshenko Bull.Eng.Tech 1913 * AGol'denveizer Theory of Thin Shells New York Pergamon press 1961 * Nonlinear Theory of Thin Shells, The Israel Program for Scientific translations KhMMushtari KZGalimov 1961 Jerusalem * Theory of Plates and Shells, 2 nd edn., Tata McGraw-Hill Edition SPTimoshenko SWoinowsky-Krieger 2010. 2010 New Delhi; New York * Theory of Anisotropic Shells SAAmbartaumyan 1961 Nauka, Moscow in Russian * Theory of Thin Shells Weakened by Holes ANGuz ISChernyshenko VNChekhov Methods of Calculating Shells Kiev 1980 1 in Russian * Lectures on the Analytical Mechanics of Shells NAKil'chevskii GHIzdebskaya LMKiselevskaya Vishcha Shkola Kiev 1974 in Russian * YaSPodstrigach RNShvets Thermoelasticity of Thin Shells Naukova Dumka, Kiev 1983 in Russian * Plates and Shells SPTimoshenko SVoinovskii-Krieger 1963 Fizmatgiz, Moscow in Russian * Fundamentals of the Calculation of Elastic Shells NVKolkunov Vishcha Shkola 1972 in Russian * Statics of Thin-Walled Shells of Revolution VSShernina 1968 Nauka Moscow in Russian * Linear Theory of Shells KFChernykh Izd. LGU. Leningrad, Part 1 1962. 1964 in Russian * Development of Applied Methods in Problems in the Static Design of Thin-Walled Spatial Systems -Shells and Folds IEMileikovskii VDRaizer proceedings of the seventh All-Union Conference on the Theory of Shells and Plates the seventh All-Union Conference on the Theory of Shells and PlatesMoscow 1970 in Russian * Die Lahre Vonder Elastizit and Festigkeit, Praque EWinkler 1867