Document Type : Research Article
Authors
^{1} Faculty of Materials and Manufacturing Processes, Malek Ashtar University of Technology, Iran
^{2} Department of Aerospace Engineering, Ferdowsi University of Mashhad, Mashhad, Iran
Abstract
Keywords
Free Vibration Analysis of Combined CylindricalConical Composite Shells using FirstOrder Shear Deformation Theory
^{a} Faculty of Materials and Manufacturing Processes, Malek Ashtar University of Technology, Iran.
^{b }Department of Aerospace Engineering, Ferdowsi University of Mashhad, Mashhad, Iran
KEYWORDS 

ABSTRACT 
Free vibration, Cylindricalconical composite shells, Firstorder shear deformation theory 

In this paper, a solution procedure is presented for free vibration of combined cylindricalconical composite shells including the shear deformation effect of the shell. The solution presented in this study is obtained directly from the governing equations for five displacement components according to Hamilton’s principle. This solution is in the form of a power series in terms of a particularly convenient coordinate system. In this study, the effects of geometry and material parameters on the natural frequencies are investigated. Also, to illustrate the validity of the present solution procedure, analytical results are verified with many studies and compared with those of the present numerical ABAQUS analysis. The outcomes showed a good agreement between the obtained results. The novelty of the present study is incorporating the transverse shear deformation in calculating the natural frequencies of the joined cylindricalconical shells. In previous literature, this topic has not been studied in such a wide scope. 
Cylindrical and conical shells have widespread usage in the industry owing to their good mechanical and physical properties. Conical shells have not been as widely reported in the literature compared to cylindrical shells. This is due to the increased mathematical complexity associated with the effect of the variation of the radius along the length of the cone on the elastic waves. The approximate methods of calculating the natural frequencies for conical shells have been found by several authors [1], [2]. Many researchers, such as Love, DonnellMushtari, Timoshenko, Reissner, Flügge, and Sanders have presented various theories of shells with different assumptions and approximations for the sake of simplification. Many of them, such as DonnellMushtari, Timoshenko, Reissner, and Flügge have solved various shell problems based on Love’s postulates [3]. He et al. studied a unified power series method for vibration analysis of composite laminate conical, a cylindrical shell, and an annular plate. They investigated the effect of the geometric parameters and material constants (e.g., elastic restrained spring stiffness constants, the angle between the shell surface and axis, length to the radius ratios, and modulus ratios) on the free vibration characteristics of the composite laminated structure [4] Kouchakzadeh and Shakouri studied the vibrational behavior of the two joined crossply laminated conical shells. They reviewed the natural frequency and effects of semivertex angles, meridional lengths, and shell thicknesses by solving the problem using Donnell theory and Hamilton’s principle [5] Caresta and Kessissoglou offered a new recommendation based on DonnellMushtari and Flugge’s equations to describe the free vibration characteristics of coupled isotropic cylindricalconical shells. They also investigated the effects of the junction between the coupled shells and the boundary conditions [6]. Xianglong et al. studied the effects of the semivertex angle of the cone and the elastic restraint parameters on the free vibration behavior of the shell according to Reissner's thin shell theory [7]. Bagheri et al. evaluated the free vibration of joined conicalconical shells with higherorder shear deformation theory in various types of boundary conditions for the shell ends. The produced system of equations was discretized using the semianalytical generalized differential quadrature (GDQ) method [8]. Mohammadrezazadeh and Jafari studied the nonlinear vibration analysis of laminated composite angleply cylindrical and conical shells. Also, they investigated the effects of several parameters including the layers’ angle, the number of the layers, semivertex angle, length, and radius, as well as each layer's thickness on nonlinear frequency ratio, fundamental linear frequency [9]. Qin et al. studied a unified Fourier series solution for vibration analysis of FGCNTRC cylindrical, conical shells, and annular plates with arbitrary boundary conditions. They used a micromechanical model based on the developed rule of mixtures [10]. Previously, no study has been conducted on the composite joined conicalcylindrical structures using firstorder shear deformation theory (FSDT). In this research, the free vibration of joined conicalcylindrical composite shells is investigated based on higherorder shear deformation theory. After calculating the natural frequencies and verification of the results, the effect of geometric parameters material and layup is studied. The novelty of the present study is incorporating the transverse shear deformation in calculating the natural frequencies of the joined cylindricalconical shells. In previous literature, this topic has not been studied in this wide scope.
Consider a joined conicalcylindrical shell made of a composite material of a uniform thickness of h, small radius of , large (intersection) radius of , slanted length of , cylinder length of , and vertex half angle of . Meridional, circumferential, and normal directions of the conical and cylindrical shell are denoted by , , 0 ≤ θ ≤ 2π, and , respectively. The adopted coordinates system ( , , , , ), geometric characteristics, and sign convention of the joined shell are depicted in Fig. 1. Also, displacement components for the cone and cylinder in coordinate system directions are ( , , ) and ( , , ), respectively.


Fig. 1. Schematic of a thin joined cylindricalconical shell
To capture the thickness shear deformations and rotary inertia effects of the conicalcylindrical shell, FSDT was used to formulate the governing equations of the shell. Based on FSDT, components of the displacement on a generic point are represented according to the midsurface characteristics as follows:
For conical shell:

(1) 

(2) 

(3) 
For cylindrical shell:

(4) 

(5) 

(6) 
In the above equations, , , and are the meridional, circumferential, and normal displacements of the conical shell, respectively. Also, , , and are the radial, circumferential, and normal displacements of the cylindrical shell, respectively
A subscript 0 denotes the characteristics of the midsurface. Besides, , , , and are the transverse normal rotations about the x and θ axes for conical and cylindrical shells, respectively. According to FSDT, the components of the strain field on an arbitrary point of the conical and cylindrical are be obtained in terms of strains and curvatures of the midsurface of the shell as follows:
For conical shell:

(7) 
For cylindrical shell:

(8) 
According to Love’s theory, the kinematic relations between the displacement components and strain and curvatures are [13]:

(9) 

(10) 

(11) 

(12) 

(13) 

(14) 

(15) 

(16) 
In the above equations, is the angle of the semivertex of the conical shell. In the case of a cylindrical shell, is equal to 0 and .
In Eqs. (9) to (16), in case of a conical shell, , , , , and . Also, in the case of a cylindrical shell, , , , , and .
The dynamic version of the principle of virtual work (Hamilton’s principle) is expressed as follows [10]:

(17) 
where denotes the virtual strain energy, is the virtual potential energy due to the applied loads, and shows the virtual kinetic energy.

(18) 

(19) 
where [11]

(20) 
where is the density of shell material.
The relation between the stress resultants and stress fields is:
For the conical shell:

(21) 

(22) 
For the cylindrical shell:

(23) 

(24) 
where ( , , ) and ( , , ) are stress and moment resultants per unit length, respectively. Moreover, shows the elements of the stress vector, defined in [11].
After substituting the above relations in Eq. (17), and partial integration and collecting similar terms, governing equations for the equilibrium of the conical shell are obtained as follows:

(25) 

(26) 

(27) 

(28) 

(29) 
In the above equations, P is internal pressure and is axial force. For cylindrical shells, and is set to be equal to 0.
The constitutive relations for the laminated conicalcylindrical shell are expressed as follows:
(30) 


(31) 
where is the shear correction factor and is considered to be equal to according to Mindlin's assumptions [11]. Also, , , and are composite stiffness matrices defined in [12].
All types of boundary conditions can be used at both ends of the cone or the cylinder. The simply supported boundary conditions at are expressed as follows:
(32) 

The continuity conditions at the joint of the two shells can be represented as follows:


(33) 

(34) 

(35) 

(36) 

(37) 

(38) 

(39) 

(40) 

(41) 

(42) 

For solving the above equations, solutions of the following form are assumed such that they exactly satisfy both the boundary and the continuity conditions:
For the conical shell:
(39) 

(40) 

(41) 

(42) 

(43) 

For the cylindrical shell:
(44) 

(45) 

(46) 

(47) 

(48) 

where
(49) 

(50) 

(51) 

(52) 

(53) 

After substituting the above functions in the governing Eqs. (25) to (29), the following recurrence relations are obtained:
(54) 

(55) 

(56) 

(57) 

(58) 

where parameters are introduced in the Appendix.
After applying the continuity and boundary conditions and global assembling, a general relation is obtained as follows:

(59) 
where M is the generalized mass matrix, K is the generalized stiffness matrix, and X is the unknown displacement vector (mode shape). The natural frequencies of the structure are obtained by solving the above equation using numerical techniques within a MATLAB code.
Finite element modeling (FEM) was performed in ABAQUS 13.6 software. The geometric model made in the software is a shell model. A total of 13,460 S4R shell elements were used for meshing. The procedure described in the previous sections is used herein to study the free vibration of joined conicalcylindrical shell system made of composite material. In this section, first, some comparison studies are conducted. Next, parametric studies are performed to examine the influences of involved parameters. Numerical calculations show that including 50 terms (m=50) in the series solution provides satisfactorily converged results for the natural frequencies. The element type is quadratic 8node shell element (S8R). The number of elements increases from 4500 up to 520000 elements. When the number of elements reaches 94,000, the results started to converge (Fig. 2).
Before presenting the results, the nondimensional frequency parameter is introduced as follows:

(60) 
Fig. 2. The convergence of the frequency parameter
for a simply supported  simply supported joined isotropic conical–cylindrical shell based on element number
The first comparison study calculates the first six frequencies of a joined cylindricalconical shell for each circumferential mode number n. The numerical results of this study are compared with those reported in several references as shown in Table 1.
For an isotropic combined conicalcylindrical shell, the material and geometric parameters are:
where and are referred to as the frequency and its parameter of the combined shell.
The frequency parameters were compared for a combined shell with the above materials, geometric parameters, and simplysupported boundary conditions at both ends with ABAQUS. The model is shown in Fig. 3.
The results are shown in Table 2.
(a)

(b) 
Fig. 3. Finite element model in ABAQUS: a) Elements mesh pattern and b) Lowest asymmetric shape (n=5)
Table 1. Comparison of dimensionless frequency parameter ω_c for a freeclamped joined isotropic conicalcylindrical shell 


n 

Irie et al. [4] 
Efraim and Eisenberger [11] 
Caresta and Kessissoglou [6] 
Kouchakzadeh and Shakouri [5] 
Present 







0.5047 
0.5038 
0.5038 
0.5038 
0.5035 
0 
0.2930 
0.2929 
0.2929 
0.2929 
0.2928 
1 
0.1010 
0.1000 
0.1020 
0.1000 
0.1103 
2 
0.9076 
0.0876 
0.09377 
0.0876 
0.0952 
3 
0.1477 
0.1446 
0.1506 
0.1446 
0.1512 
4 
0.2021 
0.1995 
0.2039 
0.1995 
0.1986 
5 
Table 2. Comparison of the dimensionless frequency parameter for a simply supported joined isotropic conicalcylindrical shell

n 


Error Percents 
Present (ABAQUS) 
Present (Analytical) 


2.91% 
0.067 
0.0651 
0 


0.27% 
0.4898 
0.4925 
1 


0.03% 
0.3512 
0.3509 
2 


0.12% 
0.2609 
0.2597 
3 


0.04% 
0.2121 
0.2117 
4 


0.19% 
0.2102 
0.2121 
5 


0.1% 
0.2458 
0.2448 
6 


0.19% 
0.2963 
0.2944 
7 


0.68% 
0.3580 
0.3512 
8 


0.14% 
0.4150 
0.4136 
9 

According to Table 2, the fundamental frequency occurs at n=4 (for analytical method) and n=5 (for the ABAQUS method), indicating a close agreement.
After validating the proposed solution method, we analyzed the results of the natural frequencies for the free vibration of a shear deformable conicalcylindrical shell system made of a linearly elastic composite material. In this section, the composite cylindricalconical shell with Eglass epoxy properties (Table 3) and geometrical properties (Table 4) is investigated. The considered layup is . Unless stated otherwise, the properties mentioned in Tables 3 and 4 were applied. The boundary condition is simply supported at two ends.
The effect of the semivertex angle of the composite cylindricalconical shell on the natural frequencies is investigated in Figs. 4 to 6.
Table 3. Properties of the material of composite shell [14], [15]





Material 
0.28 
3.7 
8.6 
39.2 
2100 
EGlass/Epoxy 
Table 4. Geometric parameters of the conical shell
Composite layup angles 





0.01 
0.4226 

200 
Fig. 4. Natural frequency parameter versus semivertex angle of the cone for circumferential wave numbers (n=0,1)
Fig. 5. Natural frequency parameter versus semivertex angle of the cone for circumferential wave numbers (n=2:5)
Fig. 6. Natural frequency parameter versus semi vertex angle of the cone for circumferential wave numbers (n=6:9)
As can be seen from Figs. 4 to 6, by increasing the semivertex angle of the cone, for higher values of the circumferential wave number (n=6:9), the natural frequency is converged to an almost constant value. Also, the minimum natural frequency for the values of n greater than 1 occurs at while the maximum natural frequency occurs at for n=0 and n=1. For n=0, by increasing the semivertex angle, the natural frequency is first increased and then decreased. However, for other values of n, no special trend exists. Just for higher values on n (n=6:9), the natural frequency is converged to a special value for higher values of.
The effect of the thickness of the shell on the natural frequency is represented in Table 5
As can be seen in Table 5, by increasing the value of from 0.01 to 0.03, the fundamental mode number (n) is decreased. However, it is unchanged by increasing the value of from 0.03 to 0.05 (n=3).
The effect of the length ratio of the combined shell on the natural frequency is studied in Fig. 7. The radius of the cylindrical shell is assumed to be constant. Also, for a specific value of the ratio, the ratio of is changed by changing the value of . The layup of the composite shell is considered to be .
Table 5. Effect of thickness of the shell on the frequency parameter of a conicalcylindrical composite shell

n 

0 
1 
2 
3 
4 
5 

0.01 
0.663 
0.534 
0.371 
0.261 
0.220 
0.191* 
0.02 
0.715 
0.562 
0.399 
0.310 
0.270* 
0.295 
0.03 
0.756 
0.584 
0.410 
0.321* 
0.333 
0.368 
0.04 
0.772 
0.606 
0.424 
0.359* 
0.391 
0.436 
0.05 
0.781 
0.611 
0.421 
0.378* 
0.438 
0.491 
*: Lowest (fundamental) mode number
(a)
(b)
(c)
(d)
Fig. 7. Effect of length ratio of shell on natural frequency for n=5, a) b) c) , and d)
As can be seen, regardless of the value of ratio, the natural frequency is increased by increasing the value of .
Fig. 8 presents the variation of the frequency parameter versus the orthotropic ratio ( ) studied for different layups.
Based on the obtained results, the natural frequency of the shell is decreased by increasing the ratio of . For higher ratios of , the rate of decreasing natural frequency is decreased. Moreover, it is observed that the layups and have the greatest and smallest natural frequencies, respectively.
In this section, the effect of the number of layers ( ) on the natural frequency is investigated. Figs. 9 to 12 show the effect of changing the number of layers for a constant thickness and layup and its effect on the natural frequency.
Fig. 8. Effect of elasticity modulus on the natural frequency (n=5)
Fig. 9. Effect of number of layers for a constant thickness and layup on the natural frequency at circumferential wave numbers n=0 to n=3
Fig. 10. Effect of number of layers for a constant thickness and layup on the natural frequency at n = 4 to 5
Fig. 11. Effect of number of layers for a constant thickness and layup on the natural frequency at n = 6
Fig. 12. Effect of number of layers for a constant thickness and layup on the natural frequency at n=7
As can be seen in these figures, the natural frequency of the shell is increased by increasing the number of layers in a constant thickness. The rate of increase is very much greater at lower values of the number of layers. The maximum percentage discrepancy between the numerical and analytical results is 4%, which occurs in Fig. 9 at and n=1.
The effect of internal pressure on the natural frequency is studied as shown in Table 6.
Table 6. Effect of internal pressure on the frequency parameter of a conicalcylindrical composite shell

P(Pa) 
0.1910 
0 
0.1910 
1 
0.1910 

0.1910 

0.1912 

0.1930 

0.2125 

0.3381 

0.9265 

2.2111 

As shown in Table 6, by increasing the internal pressure, the natural frequency is increased as well. Although the natural frequency is not affected considerably by lower values of P, for higher values of P, it increases dramatically.
The effect of axial compressive force on natural frequency is studied in Table 7.
Based on the obtained results, by increasing the axial compressive force, the natural frequency is decreased and reaches 0 at a specific load (buckling load).
To study the effect of fiber angle on the natural frequencies, a composite shell with layup was considered. The results are shown in Fig. 13.
Table 7. Effect of axial compressive force on the frequency parameter of a conicalcylindrical composite shell

Na*2*pi*R(N) 
0.1910 
0 
0.1908 
10 
0.1898 
100 
0.1789 
1000 
0.1623 
2000 
0.1426 
3000 
0.1013 
4000 
0 
5000 
Fig. 13. Effect of fiber angle of the composite shell with layup on the natural frequency
As can be seen, in each circumferential wave number (n), the behavior of natural frequency is changed by changing the fiber angle.
In axisymmetric mode (n=0), by increasing the angle of the fiber, the natural frequency is increased up to the fiber angle . Beyond , the natural frequency is decreased. The minimum frequency occurs at . For n=1, by increasing the angle of the fiber, the natural frequency is increased up to the fiber angle of ∅=30°. Beyond ∅=30°, the natural frequency is decreased. Naturally, the slope of decreasing frequency is increased for fiber angles between and . The minimum frequency occurs at and .
At n=2, by increasing the angle of the fiber, the natural frequency is decreased strongly between and . From ∅=10° to ∅=20°, the natural frequency is increased and then decreased. The minimum frequency occurs at and the maximum value of it occurs at .
In this section, the effect of orthotropic ratio and fiber angle on the natural frequency of the singlelayer composite is studied. Fig. 13 presents the variation of the natural frequency correspondent to the axisymmetric mode (n=0) versus fiber angle for a singlelayered composite combined shell for different orthotropic ratios.
As shown in Fig. 14, for lower orthotropic ratios, the change in the natural frequency caused by varying the angle of the fiber is negligible. For , the natural frequency is first increased up to 45° at which the maximum occurs. For , the maximum occurs at . For orthotropic ratios equal to 3:10, this maximum occurs at a fiber angle of . As the orthotropic ratio increases, the natural frequency is decreased. Regardless of the value of the orthotropic ratio, by increasing the fiber angle, the natural frequency is first increased and then decreased.
Fig. 14. The natural frequency of combined shell with single layer versus angle of fiber for various orthotropic ratios (n=0)
In the present study, a new approach is introduced to obtain the free vibrational characteristics of combined composite cylindricalconical shells. In the previous studies, the present problem was solved using classical shell theories but the effect of shear deformation was not taken into account. This approach contains the effect of firstorder shear deformation in the shell theory. Moreover, the effect of geometric and material parameters on the natural frequency is investigated and the effect of preload of internal pressure and an axial compressive force is studied. The results showed that the fiber angle and orthotropic ratio affect the frequency behavior of joined conicalcylindrical shells. It shows that for lower orthotropic ratios, the change ub the natural frequency by varying the angle of the fiber is negligible. Furthermore, the natural frequency of the shell is increased by increasing the number of layers in a constant thickness. The natural frequency of the shell is decreased by increasing the ratio of . For higher ratios of , the rate of decreasing natural frequency is decreased. Finally, by increasing the semivertex angle of the cone, the natural frequency is converged to an almost constant value for higher values of the circumferential wave number.
Results are verified by making a comparison with the literature and the results obtain using ABAQUS. Good agreement is observed.
[1] Irie, "Natural Frequencies of Truncated Conical Shells," Journal of Sound and Vibration, vol. 92, pp. 447453, 1984.
[2] L. Tong, "Free Vibration of Orthotropic Conical Shells," International Journal Engng Science, vol. 31, no. Pergamon Press Ltd, pp. 719733, 1993.
[3] L. Tong, "Free Vibration of Laminated Conical Shells Including Transverse Shear Deformation," International Journal of Solid Mechanics, vol. 31, pp. 443456, 1994.
[4] D. He, D. Shi, Q. Wang, Ch. Ma, "A unified power series method for vibration analysis of composite laminate conical, cylindrical shell and annular plate", Structures, Vol. 29, pp. 305327, 2021.
[5] M. Kouchakzadeh and M. Shakouri, "Free vibration analysis of joined crossply laminated conical shells," International Journal of Mechanical Sciences, vol. 78, pp. 118125, 2014.
[6] Mauro Caresta and Nicole J. Kessissoglou, "Free vibrational characteristics of isotropic coupled cylindricalconical shells," Journal of Sound and Vibration, vol. 329, pp. 733751, 2010.
[7] M. Xianglong, J. Guoyong, X. Yeping and L. Zhigang, "Free and forced vibration analysis of coupled conical–cylindrical shells with arbitrary boundary conditions," International Journal of Mechanical Sciences, vol. 88, pp. 122137, 2014.
[8] H. Bagheria, Y. Kiani and M. Eslami, "Free vibration of joined conicalconical shells," ThinWalled Structures, vol. 120, p. 446–457, 2017.
[9] Sh. Mohammadrezazadeh and A. Jafari, Nonlinear vibration analysis of laminated composite angleply cylindrical and conical shells, Composite Structures, vol. 255, 2021,112867.
[10] B. Qin, R. Zhong, T. Wang, Q. Wang, Y. Xu, Z. Hu, A unified Fourier series solution for vibration analysis of FGCNTRC cylindrical, conical shells and annular plates with arbitrary boundary conditions, Composite Structures, vol. 232, 2020, 111549.
[11] E. Efraim and M. Eisenberger, "Exact vibration frequencies of segmented axisymmetric shells," ThinWalled Structure, vol. 44, pp. 281289, 2006.
[12] H. Garnet and J. Kempner, "Axisymmetric Free Vibrations of Conical Shells," Journal of Applied Mechanics, no. ASME, 1964.
[13] J. Reddy, Mechanics of Laminated Composite Plate and Shell, Theory and Analysis, United States of America: CRC PRESS, 2003.
[14] ASM HANDBOOK  Composites, vol. 21, ASM International, Dec.2001.
[15] G. Lubin, Handbook of composites, New York: Van Nostrand Reinhold, 1982.
A1 

A2 

A3 

A4 

A5 

A6 

A7 

A8 

A9 

A10 

A11 

A12 

A13 

A14 

A15 

A16 

A17 

A18 

A19 

A20 

A21 

A22 

A23 

A24 

A25 

A26 

A27 

A28 

A29 

A30 

A31 

A32 

A33 

A34 

A35 

A36 

A37 

A38 

A39 

A40 

A41 

A42 

A43 

A44 

A45 

A46 

A47 

A48 

A49 

A50 

A51 

A52 

A53 

A54 

A55 

A56 

A57 

A58 

A59 

where:
A60 

A61 

A62 

A63 

A64 

A65 

A66 

A67 

A68 

A69 

A70 

A71 

A72 

A73 

A74 

A75 

A76 

A77 

A78 

A79 

A80 

A81 

A82 

A83 

A84 

A85 

A86 

A87 

A88 

A89 

A90 

A91 

A92 

A93 

A94 

A95 

A96 

A97 

A98 

A99 

A100 

A101 

A102 

A103 

A104 

A105 

A106 

A107 

A108 

A109 

A110 

A111 

A112 

A113 

A114 

A115 

A116 

A117 

A118 

A119 

A120 

A121 

A122 

A123 

A124 

A125 

A126 

A127 

A128 

A129 

A130 

A131 

A132 

A133 

A134 

A135 

A136 

A137 

where:
A138 

A139 

A140 

A141 

A142 

A143 A144 

A145
A146
A147
A148