Adrià Vilanova Martínez | ce8597a | 2021-09-21 18:26:59 +0200 | [diff] [blame] | 1 | \documentclass[11pt,a4paper,dvipsnames]{article} |
| 2 | \usepackage[utf8]{inputenc} |
| 3 | |
| 4 | \input{../preamble.tex} |
| 5 | |
| 6 | \title{Homework 4\\Continuum Mechanics} |
| 7 | \author{Adrià Vilanova Martínez} |
| 8 | \date{April 11, 2021} |
| 9 | |
| 10 | \begin{document} |
| 11 | |
| 12 | \maketitle |
| 13 | |
| 14 | \begin{Problem} |
| 15 | Consider a solid sphere of constant density $\rho$ and undeformed radius $R$, subjected to its own gravitational force. It is made of an elastically homogeneous and isotropic material of Young's modulus $Y$ and Poisson's ratio $\nu$. |
| 16 | |
| 17 | \begin{enumerate} |
| 18 | \item Determine the displacement field of the material particles in the sphere. |
| 19 | \item Compute the corresponding strain and stress fields. |
| 20 | \end{enumerate} |
| 21 | |
| 22 | \underline{Note:} The gravitational force per unit volume experienced by a material particle is: $\vec{f} = \rho \vec{g} = - \rho g \frac{r}{R} \hat{e}_r$, where $g = \frac{4}{3} \pi G \rho R$, with $G$ the gravitational constant. |
| 23 | \end{Problem} |
| 24 | |
| 25 | \Solution |
| 26 | The displacement field in spherical coordinates will be $\vec{u} = (u_r(r), 0, 0)$ due to the symmetry of the problem. |
| 27 | |
| 28 | Let's use Navier-Cauchy's equation: |
| 29 | \[ \vec{f} + \mu \vec{\nabla}^2 \vec{u} + (\lambda + \mu) \vec{\nabla} (\vec{\nabla} \cdot \vec{u}) = 0 \] |
| 30 | which in spherical coordinates becomes: |
| 31 | \[ \begin{pmatrix} |
| 32 | - \rho g \frac{r}{R} \\ |
| 33 | 0 \\ |
| 34 | 0 |
| 35 | \end{pmatrix} + \mu \begin{pmatrix} |
| 36 | \partial^2_{rr} u_r + \frac{2}{r} \partial_r u_r - \frac{2}{r^2} u_r \\ |
| 37 | 0 \\ |
| 38 | 0 |
| 39 | \end{pmatrix} + (\lambda + \mu) \begin{pmatrix} |
| 40 | \partial^2_{rr} u_r + \frac{2}{r} \partial_r u_r - \frac{2}{r^2} u_r \\ |
| 41 | 0 \\ |
| 42 | 0 |
| 43 | \end{pmatrix} = 0 \iff \] |
| 44 | \[ \iff - \rho g \frac{r}{R} + (\lambda + 2 \mu) \left( \partial^2_{rr} u_r + \frac{2}{r} \partial_r u_r - \frac{2}{r^2} u_r \right) = 0 \iff \] |
| 45 | \[ \iff r^2 \partial^2_{rr} u_r + 2 r \partial_r u_r - u_r = \frac{\rho g}{(\lambda + 2 \mu) R} r^3 =: k r^3. \] |
| 46 | |
| 47 | If we take as an \textit{ansatz} the function |
| 48 | \[ u_r(r) := a \cdot r^b, \] |
| 49 | we can see that a particular solution to the ODE is |
| 50 | \[ u_r(r) = \frac{k}{10} r^3. \] |
| 51 | |
| 52 | Due to the fact that the homogeneous part of the ODE has the solution space |
| 53 | \[ \{ A r + B \frac{1}{r^2} : A, B \in \mathbb{R} \}, \] |
| 54 | and the fact that since $u_r(r = 0) \neq \infty \implies B = 0$, we conclude that \[ \vec{u}(r) = \left( \frac{k}{10} r^3 + A r, 0, 0 \right). \] |
| 55 | |
| 56 | Since we have |
| 57 | \[ \vec{\nabla} \vec{u} = \begin{pmatrix} |
| 58 | \frac{3}{10} k r^2 + A & 0 & 0 \\ |
| 59 | 0 & \frac{1}{10} k r^2 + A & 0 \\ |
| 60 | 0 & 0 & \frac{1}{10} k r^2 + A |
| 61 | \end{pmatrix}, \] |
| 62 | Cauchy's strain tensor is: |
| 63 | \[ \TT{u} = \frac{1}{2} \left( \vec{\nabla} \vec{u} + (\vec{\nabla} \vec{u})^T \right) = \vec{\nabla} \vec{u}. \] |
| 64 | |
| 65 | Via Hooke's law, the only non-zero components of the stress vector are the ones in the diagonal: |
| 66 | \[ \sigma_{ii} = 2 \mu u_{ii} + \lambda \Tr \TT{u} \implies \] |
| 67 | \[ \TT{\sigma} = \begin{pmatrix} |
| 68 | \frac{6 \mu + 5 \lambda}{10} k r^2 + (3 \lambda + 2 \mu) A & 0 & 0 \\ |
| 69 | 0 & \frac{2 \mu + 5 \lambda}{10} k r^2 + (3 \lambda + 2 \mu) A & 0 \\ |
| 70 | 0 & 0 & \frac{2 \mu + 5 \lambda}{10} k r^2 + (3 \lambda + 2 \mu) A \\ |
| 71 | \end{pmatrix}. \] |
| 72 | |
| 73 | If we impose the boundary condition $\sigma_{rr} (r = R) = 0$ (free surface), we get: |
| 74 | \[ (3 \lambda + 2 \mu) A = \frac{6 \mu + 5 \lambda}{10} k R^2. \] |
| 75 | |
| 76 | We can transform $Y$ and $\nu$ into the Lamé coefficients by applying the following transformation: |
| 77 | \[ \begin{cases} |
| 78 | \lambda = \frac{Y \nu}{(1 - 2 \nu) (1 + \nu)}, \\ |
| 79 | \mu = \frac{Y}{2 (1 + \nu)}. |
| 80 | \end{cases} \] |
| 81 | |
| 82 | \newpage |
| 83 | |
| 84 | \begin{Problem} |
| 85 | A cylindrical pipe of inner radius $R_0$, outer radius $R_1$, and Lame coefficients $\mu$ and $\lambda$, is subjected to an internal pressure $p_0$, an external pressure $p_1$, and a uniform tensile force per unit area $F_z/A$ along the symmetry axis of the pipe $z$. Consider the end of the pipe at $z = 0$ is clamped to a rigid wall. |
| 86 | |
| 87 | \begin{enumerate} |
| 88 | \item Determine the displacement field corresponding to the elastic deformation of the pipe. Neglect gravity and corrections due to the finite length of the pipe. |
| 89 | |
| 90 | \item Compute the stress and strain fields corresponding to such deformation. |
| 91 | \end{enumerate} |
| 92 | \end{Problem} |
| 93 | |
| 94 | We'll consider the displacement field resulting from the hydrostatic pressure, and the one resulting from the tensile force. |
| 95 | |
| 96 | \textbf{Displacement field resulting from the hydrostatic pressure:} |
| 97 | |
| 98 | Due to the symmetry of the problem in this case, we have that $\vec{u} = (u_r(r), 0, 0)$ in cylindrical coordinates. |
| 99 | |
| 100 | By applying the Navier-Cauchy equation, we get: |
| 101 | \[ (\lambda + 2 \mu) \frac{d}{dr} \left( \frac{1}{r} \frac{d}{dr} (r u_r) \right) \implies u_r = \alpha r + \beta \frac{1}{r}. \] |
| 102 | |
| 103 | Therefore, we have that the only non-zero components of the strain tensor are: |
| 104 | \[ u_{rr} = \frac{d u_r}{dr} = \alpha - \beta \frac{1}{r^2}, \] |
| 105 | \[ u_{\phi\phi} = \frac{u_r}{r} = \alpha + \beta \frac{1}{r^2}. \] |
| 106 | |
| 107 | By using Hooke's law, we get: |
| 108 | \[ \sigma_{rr} = 2 \alpha (\lambda + \mu) - 2 \beta \mu \frac{1}{r^2}, \] |
| 109 | \[ \sigma_{\theta \theta} = 2 \alpha (\lambda + \mu) + 2 \beta \mu \frac{1}{r^2}, \] |
| 110 | \[ \sigma_{zz} = 2 \alpha \lambda. \] |
| 111 | |
| 112 | We can now impose the boundary conditions $\sigma_{rr}(z = R_0) = - p_0$, $\sigma_{rr}(z = R_1) = - p_1$, and we get: |
| 113 | \[ \alpha = \frac{p_0 R_0^2 - p_1 R_1^2}{2 (\lambda + \mu) (R_1^2 - R_0^2)}, \] |
| 114 | \[ \beta = \frac{(p_0 - p_1) R_0^2 R_1^2}{2 \mu (R_1^2 - R_0^2)}. \] |
| 115 | |
| 116 | \textbf{Displacement field resulting from the tensible force:} |
| 117 | |
| 118 | Due to the uniform force, we have that: |
| 119 | \[ \sigma_{zz} = \frac{F_z}{A}, \] |
| 120 | which is the only non-zero component of the stress tensor. Therefore, from Hooke's law we have that the only non-zero components of the strain tensor are: |
| 121 | \[ u_{zz} = \frac{1}{Y} \sigma_{zz} = \frac{F_z}{A Y}, \] |
| 122 | \[ u_{rr} = - \nu u_{zz} = - \frac{F_z \nu}{A Y}, \] |
| 123 | \[ u_{\theta \theta} = - \nu u_{zz} = - \frac{F_z \nu}{A Y}. \] |
| 124 | |
| 125 | Now we can integrate (in cylindrical coordinates) the previous expressions to find the displacement field: |
| 126 | \[ u_{zz} = \frac{du_z}{dz} \implies u_z(z) = \frac{F_z}{AY} z + c, \] |
| 127 | and since $u_z(0) = 0$, we have: |
| 128 | \[ u_z(z) = \frac{F_z}{AY} z. \] |
| 129 | |
| 130 | Also: |
| 131 | \[ u_{rr} = \frac{du_r}{dr} \implies u_r(r) = - \frac{F_z \nu}{A Y} r + c', \] |
| 132 | and since $u_{\theta \theta} = \frac{u_r(r)}{r} \implies c' = 0$, then: |
| 133 | \[ u_r(r) = - \frac{F_z \nu}{A Y} r \] |
| 134 | |
| 135 | \textbf{Conclusion:} |
| 136 | |
| 137 | Thanks to the superposition principle, we can sum the previous displacement fields to calculate the total displacement field: |
| 138 | \[ \vec{u}(r) = \begin{pmatrix} |
| 139 | \left( \alpha - \frac{F_z \nu}{A Y} \right) r + \beta \frac{1}{r} \\ |
| 140 | 0 \\ |
| 141 | \frac{F_z}{A Y} z |
| 142 | \end{pmatrix}. \] |
| 143 | |
| 144 | As for the strain and stress tensors, we can do the same in order to get their total expressions: |
| 145 | \[ \TT{\sigma} = \begin{pmatrix} |
| 146 | 2 \alpha (\lambda + \mu) - 2 \beta \mu \frac{1}{r^2} & 0 & 0 \\ |
| 147 | 0 & 2 \alpha (\lambda + \mu) + 2 \beta \mu \frac{1}{r^2} & 0 \\ |
| 148 | 0 & 0 & 2 \alpha \lambda + \frac{F_z}{A} |
| 149 | \end{pmatrix}, \] |
| 150 | \[ \TT{u} = \begin{pmatrix} |
| 151 | \alpha - \beta \frac{1}{r^2} - \frac{F_z \nu}{A Y} & 0 & 0 \\ |
| 152 | 0 & \alpha + \beta \frac{1}{r^2} - \frac{F_z \nu}{A Y} & 0 \\ |
| 153 | 0 & 0 & \frac{F_z}{A Y} |
| 154 | \end{pmatrix}. \] |
| 155 | |
| 156 | \end{document} |