<?xml version='1.0' encoding='utf-8' ?>
<!DOCTYPE html>
<html xmlns="http://www.w3.org/1999/xhtml">
	<head>
		<meta http-equiv="Content-Type" content="text/html; charset=utf-8"/>
		<link type="text/css" rel="stylesheet" href="doxygen.css"/>
		<link type="text/css" rel="stylesheet" href="tabs.css"/>
	</head>
	<body>
	<div>
<!-- Generated by Doxygen 1.8.5 -->
</div><!-- top -->
<div class="header">
  <div class="headertitle">
<div class="title">Finite Element Formulation </div>  </div>
</div><!--header-->
<div class="contents">
<div class="toc"><h3>Table of Contents</h3>
<ul><li class="level1"><a href="#tm_finite_introduction">Introduction</a></li>
<li class="level1"><a href="#tm_finite_fundamental">Fundamental Continuum Mechanics Theory</a><ul><li class="level2"><a href="#tm_finite_fundamental_description">Description of Motion</a></li>
<li class="level2"><a href="#tm_finite_fundamental_definitions">Definitions of Strain and Stresses</a></li>
<li class="level2"><a href="#tm_finite_fundamental_virtual">Virtual Work Equations</a></li>
<li class="level2"><a href="#tm_finite_fundamental_dynamic">Dynamic Equilibrium Equations</a></li>
</ul>
</li>
<li class="level1"><a href="#tm_finite_implementation">Implementation of the Finite Element Method</a><ul><li class="level2"><a href="#tm_finite_implementation_large">Large Rotation in Space</a></li>
<li class="level2"><a href="#tm_finite_implementation_finite">Finite Rigid Bodies at Nodes</a></li>
<li class="level2"><a href="#tm_finite_implementation_updating">Updating of the Rotation Matrix</a></li>
<li class="level2"><a href="#tm_finite_implementation_bar">Bar Element</a></li>
<li class="level2"><a href="#tm_finite_implementation_beam">Beam Element</a><ul><li class="level3"><a href="#tm_internal_loads">Internal loads</a></li>
</ul>
</li>
<li class="level2"><a href="#tm_finite_implementation_connector">Connector Element</a></li>
</ul>
</li>
<li class="level1"><a href="#tm_finite_cross">Cross Sectional Modelling</a></li>
<li class="level1"><a href="#tm_finite_guideline">Guideline to System Modelling</a></li>
<li class="level1"><a href="#tm_finite_references">References</a></li>
</ul>
</div>
<div class="textblock"><p><a class="anchor" id="finite_element"></a></p>
<h1><a class="anchor" id="tm_finite_introduction"></a>
Introduction</h1>
<p>The structural analysis part of the RIFLEX program is based on finite element modelling. This chapter presents the theoretical basis for the finite element formulation. The current formulation is with few exceptions equal to the one used in FENRIS, which is a large scale, general purpose program for nonlinear finite element analysis. A detailed description of the theoretical basis for FENRIS has been given by Bergan et al. (1984, 1985), Mollestad (1983) and Engseth (1984).</p>
<p>The most important features included in the basic finite element formulation are:</p>
<ul>
<li>Unlimited rotations and translations in 3D space.</li>
<li>Beam and bar element based on small strain theory.</li>
<li>Use of beam as well as bar elements in system modelling.</li>
<li>Connector element for modelling of swivels, hinges, etc.</li>
<li>Description of nonlinear material properties.</li>
<li>Stiffness contribution from material properties as well as geometric stiffness, i.e. contribution of axial force to the transverse stiffness.</li>
<li>General element assembly allowing for description of arbitrary system topology, varying cross-sectional properties, arbitrary boundary conditions at supports, prescribed support displacements and specified external forces.</li>
</ul>
<p>The described features are crucial for adequate modelling and analysis of slender marine structures in general. Application of the basic finite element formulation in static and dynamic analysis are detailed in <a class="el" href="a00008.html#static_finite">Static Finite Element Analysis</a> and <a class="el" href="a00010.html#dynamic_time">Dynamic Time Domain Analysis</a>, respectively.</p>
<p>A short review of basic continuum mechanics theory is given in <a class="el" href="a00004.html#tm_finite_fundamental">Fundamental Continuum Mechanics Theory</a>. Using generalized strain/stress measures allows the description of large cross-sectional deformations. Implementation of the finite element method including description of translation and rotations in 3D space is described in <a class="el" href="a00004.html#tm_finite_implementation">Implementation of the finite element method</a>. All element formulations are based on the small strain approximation which is reasonable for application to slender marine structures.</p>
<h1><a class="anchor" id="tm_finite_fundamental"></a>
Fundamental Continuum Mechanics Theory</h1>
<h2><a class="anchor" id="tm_finite_fundamental_description"></a>
Description of Motion</h2>
<p>The theoretical foundation is based on well established principles of continuum mechanics. Details of the formulation may be found in the textbook by Malvern (1969).</p>
<p>The Lagrangian description is used to describe the motion of material particles. This motion is referred to a fixed global Cartesian coordinate frame defined by the base vectors <img class="formulaInl" alt="$\boldsymbol{I}_i$" src="form_0.png"/> (see <a class="el" href="a00004.html#Motion_of_a_material_particle">Motion of a material particle in space</a>). Each particle is identified by its position vector <img class="formulaInl" alt="$\boldsymbol{X}$" src="form_1.png"/> in an arbitrarily chosen reference configuration, which may be the initial configuration <img class="formulaInl" alt="$C_0$" src="form_2.png"/> as shown in <a class="el" href="a00004.html#Motion_of_a_material_particle">Motion of a material particle in space</a>. The motion of the particle can be expressed as </p>
<p class="formulaDsp">
<img class="formulaDsp" alt="\[ \boldsymbol{x} = \boldsymbol{x}(\boldsymbol{X} , t)\tag{2.1} \]" src="form_3.png"/>
</p>
<p> where <img class="formulaInl" alt="$\boldsymbol{x}$" src="form_4.png"/> denotes the position of the particle at the time <img class="formulaInl" alt="$t$" src="form_5.png"/>. The displacement vector <img class="formulaInl" alt="$\boldsymbol{u}$" src="form_6.png"/> is defined through </p>
<p class="formulaDsp">
<img class="formulaDsp" alt="\[ \boldsymbol{x} = \boldsymbol{X} + \boldsymbol{u}. \tag{2.2} \]" src="form_7.png"/>
</p>
<p> In a nonlinear analysis it is necessary to consider the initial configuration <img class="formulaInl" alt="$ C_{0} $" src="form_8.png"/> of the body, the deformed configuration <img class="formulaInl" alt="$C_n$" src="form_9.png"/> at a given time <img class="formulaInl" alt="$ t $" src="form_10.png"/> and a new incremental configuration <img class="formulaInl" alt="$C_{n + 1}$" src="form_11.png"/> for time <img class="formulaInl" alt="$ t + \Delta t $" src="form_12.png"/>. This is illustrated in <a class="el" href="a00004.html#Various_configurations_of_a_body">Various configurations of a body</a>. The new configuration <img class="formulaInl" alt="$C_{n + 1}$" src="form_11.png"/> is close to <img class="formulaInl" alt="$C_n$" src="form_9.png"/> since <img class="formulaInl" alt="$\Delta t$" src="form_13.png"/> is assumed to be small.</p>
<p>A formulation where the strains in <img class="formulaInl" alt="$ C_n $" src="form_14.png"/>, <img class="formulaInl" alt="$ C_{n+1} $" src="form_15.png"/>, and all other configurations are referred to the initial configuration <img class="formulaInl" alt="$ C_0 $" src="form_16.png"/>, is usually termed a "total
Lagrangian formulation". However, in some cases it is convenient to relate the motion of material particles within a body to a local, rectangular coordinate frame which translates and rotates along with the average motion of the body. The total motion is then found by combining the motion of the local position vector and the motion of the local reference system. An example of this is the concept of co-rotated ghost reference, introduced by Bergan (1985). This formulation assumes that the initial configuration translates and rotates as a rigid body so that it at any time is located close to the actual deformed configuration. In <a class="el" href="a00004.html#Various_configurations_of_a_body">Various configurations of a body</a> this is illustrated by <img class="formulaInl" alt="$ C_n $" src="form_14.png"/> and <img class="formulaInl" alt="$ C_{0n} $" src="form_17.png"/>, where the original base vectors <img class="formulaInl" alt="$ \boldsymbol {I}_i $" src="form_18.png"/> have rotated along with <img class="formulaInl" alt="$ C_{0n} $" src="form_17.png"/> and are denoted <img class="formulaInl" alt="$ \boldsymbol {\tilde{I}}_i $" src="form_19.png"/></p>
<p><a class="anchor" id="Motion_of_a_material_particle"></a></p>
<div class="image">
<img src="RIFLEX_TheoryManual_42_v0_23.jpg" alt="RIFLEX_TheoryManual_42_v0_23.jpg"/>
<div class="caption">
Motion of a material particle in space</div></div>
<p><a class="anchor" id="Various_configurations_of_a_body"></a></p>
<div class="image">
<img src="RIFLEX_TheoryManual_42_v0_24.jpg" alt="RIFLEX_TheoryManual_42_v0_24.jpg"/>
<div class="caption">
Various configurations of a body.</div></div>
<p>In RIFLEX the bar element is formulated using a total Lagrangian description, while the beam element formulation uses a co-rotated ghost reference description. This will be described in more detail in <a class="el" href="a00004.html#tm_finite_implementation_bar">Bar Element</a> and <a class="el" href="a00004.html#tm_finite_implementation_beam">Beam Element</a> in <a class="el" href="a00004.html#tm_finite_implementation">Implementation of the Finite Element Method</a>.</p>
<h2><a class="anchor" id="tm_finite_fundamental_definitions"></a>
Definitions of Strain and Stresses</h2>
<p>For the Lagrangian formulation the strains are measured in terms of the Green strain tensor <img class="formulaInl" alt="$ \boldsymbol {E} $" src="form_20.png"/>. If <img class="formulaInl" alt="$ {C}_0 $" src="form_21.png"/> is used as reference configuration, this strain tensor is defined by </p>
<p class="formulaDsp">
<img class="formulaDsp" alt="\[ \mathrm{d} S^2_n - \mathrm{d} S^2_0 = 2 \mathrm{d} \boldsymbol{X} \cdot \boldsymbol{E} \cdot \mathrm{d} \boldsymbol{X} \tag{2.3} \]" src="form_22.png"/>
</p>
<p>where <img class="formulaInl" alt="$ \mathrm{d} S_0 $" src="form_23.png"/> and <img class="formulaInl" alt="$ \mathrm{d} S_n $" src="form_24.png"/> denote the length of the line segment PQ before and after deformation, (see <a class="el" href="a00004.html#Various_configurations_of_a_body">Various configurations of a body</a>). If <img class="formulaInl" alt="$ C_{0n} $" src="form_17.png"/> is used as reference configuration, the definition of the Green strain tensor reads</p>
<p class="formulaDsp">
<img class="formulaDsp" alt="\[ \mathrm{d} S^2_n - \mathrm{d} S^2_{0n} = 2 \mathrm{d} \boldsymbol{\tilde{X}} \cdot \boldsymbol{\tilde{E}} \cdot \mathrm{d} \boldsymbol{\tilde{X}} \tag{2.4} \]" src="form_25.png"/>
</p>
<p>The strain tensors may also be expressed as</p>
<p class="formulaDsp">
<img class="formulaDsp" alt="\[ \boldsymbol{E} = E_{ij} \boldsymbol{I}_{i} \boldsymbol{I}_j \tag{2.5} \]" src="form_26.png"/>
</p>
<p class="formulaDsp">
<img class="formulaDsp" alt="\[ \boldsymbol{\tilde{E}} = \tilde{E}_{ij} \boldsymbol{I}_{i} \boldsymbol{I}_j = \overset{\approx}{E}_{ij} \boldsymbol{\tilde{I}}_{i} \boldsymbol{\tilde{I}}_j \tag{2.6} \]" src="form_27.png"/>
</p>
<p>where <img class="formulaInl" alt="$E_{ij}$" src="form_28.png"/> and <img class="formulaInl" alt="$ \tilde{E}_{ij}$" src="form_29.png"/> denote the components of <img class="formulaInl" alt="$ \boldsymbol{E}$" src="form_30.png"/> and <img class="formulaInl" alt="$\boldsymbol{\tilde{E}} $" src="form_31.png"/> in the original coordinate frame <img class="formulaInl" alt="$\boldsymbol{I}_j $" src="form_32.png"/>. <img class="formulaInl" alt="$ \tilde{E}_{ij}$" src="form_29.png"/> denotes the components of <img class="formulaInl" alt="$\boldsymbol{\tilde{E}} $" src="form_31.png"/> in the co-rotated coordinate frame <img class="formulaInl" alt="$ \boldsymbol{\tilde{I}}_j $" src="form_33.png"/>. The strain tensor <img class="formulaInl" alt="$\boldsymbol{\tilde{E}} $" src="form_31.png"/> is generally different from <img class="formulaInl" alt="$ \boldsymbol{E}$" src="form_30.png"/>. However, it can be shown that the following identity holds between the components</p>
<p><a class="anchor" id="tm_eq_2_7"></a> </p>
<p class="formulaDsp">
<img class="formulaDsp" alt="\[ E_{ij} = \overset{\approx}{E}_{ij} \tag{2.7} \]" src="form_34.png"/>
</p>
<p><a class="el" href="a00004.html#tm_eq_2_7">Eq. (2.7)</a> signifies that the strain components remain unchanged when both the reference body and the base vectors are rotated in the same way. No tensor transformations of the components are necessary. This result is very useful and makes this formulation especially attractive.</p>
<p>The rectangular components of <img class="formulaInl" alt="$ E $" src="form_35.png"/> referred to <img class="formulaInl" alt="$ \boldsymbol{I_i} $" src="form_36.png"/> and <img class="formulaInl" alt="$ \boldsymbol{I_j} $" src="form_37.png"/>, may be expressed as</p>
<p><a class="anchor" id="tm_eq_2_8"></a> </p>
<p class="formulaDsp">
<img class="formulaDsp" alt="\[ E_{ij} = \frac{1}{2}\left( \frac{\partial u_i}{\partial X_j} + \frac{\partial u_j}{\partial X_i} + \frac{\partial u_m}{\partial X_i} \frac{\partial u_m}{\partial X_j}\right) \tag{2.8} \]" src="form_38.png"/>
</p>
<p>where the components of the displacement vector <img class="formulaInl" alt="$ \boldsymbol{u} $" src="form_39.png"/> have been introduced. Note that the differentiation in <a class="el" href="a00004.html#tm_eq_2_8">Eq. (2.8)</a> is performed with respect to the material coordinates <img class="formulaInl" alt="$\boldsymbol{\mathrm{X_i}}$" src="form_40.png"/>. It may also be seen that <img class="formulaInl" alt="$\boldsymbol E $" src="form_41.png"/> is a symmetric tensor consisting of both linear and quadratic terms. The quadratic contributions are only of importance for large displacements and stability problems.</p>
<p>An appropriate stress measure is also needed in the analysis. The symmetric Piola-Kirchhoff stress tensor <img class="formulaInl" alt="$ \boldsymbol{S} $" src="form_42.png"/> is always used in conjunction with Green strain. A detailed description of this stress measure is omitted here, but may be found in Malvern (1969).</p>
<p>The symmetric Piola-Kirchhoff stress tensor referred to <img class="formulaInl" alt="$ C_0 $" src="form_16.png"/> may be expressed as</p>
<p class="formulaDsp">
<img class="formulaDsp" alt="\[ \boldsymbol{S} = S_{ij} \boldsymbol{I}_{i} \boldsymbol{I}_j \tag{2.9} \]" src="form_43.png"/>
</p>
<p>Similar to what was done for the strains, this stress tensor may also be referred to a co-rotated ghost reference <img class="formulaInl" alt="$ C_{0n} $" src="form_17.png"/>. In this case the stress tensor may be written as</p>
<p class="formulaDsp">
<img class="formulaDsp" alt="\[ \boldsymbol{\tilde {S}} = \tilde{S}_{ij} \boldsymbol{I}_{i} \boldsymbol{I}_j = \overset{\approx}{S}_{ij} \boldsymbol{\tilde{I}}_{i} \boldsymbol{\tilde{I}}_j \tag{2.10} \]" src="form_44.png"/>
</p>
<p>As for the strains, it may be shown that</p>
<p class="formulaDsp">
<img class="formulaDsp" alt="\[ S_{ij} = \overset{\approx}{S}_{ij} \tag{2.11} \]" src="form_45.png"/>
</p>
<p>which means that no transformation of the stress components is necessary when the co-rotated ghost reference formulation is used.</p>
<h2><a class="anchor" id="tm_finite_fundamental_virtual"></a>
Virtual Work Equations</h2>
<p>The virtual work equation may be used to express equilibrium for a finite body. Using Green strains and Piola-Kirchhoff stresses, this equation may be written</p>
<p><a class="anchor" id="tm_eq_2_12"></a> </p>
<p class="formulaDsp">
<img class="formulaDsp" alt="\[ \int_{V_0} \boldsymbol {S} : \delta \boldsymbol{E} \mathrm{d} V_0 = \int_{A_0} \boldsymbol {t}_0 \cdot \delta \boldsymbol{u} \mathrm{d} A_0+ \int_{V_0} \boldsymbol {f}_0 \cdot \delta \boldsymbol{u} \mathrm{d} V_0\tag{2.12} \]" src="form_46.png"/>
</p>
<p>where <img class="formulaInl" alt="$ \delta $" src="form_47.png"/> indicates virtual quantities, and <img class="formulaInl" alt="$ A_0 $" src="form_48.png"/> and <img class="formulaInl" alt="$ V_0 $" src="form_49.png"/> express the surface and volume of the initial reference configuration. The surface traction <img class="formulaInl" alt="$ \boldsymbol {t}_0$" src="form_50.png"/> and body forces <img class="formulaInl" alt="$ \boldsymbol {f}_0$" src="form_51.png"/> are referred to a unit surface and a unit volume in the initial reference state.</p>
<p>Most solution procedures for nonlinear problems employ a linearized incremental form of the equilibrium equations based on two neighbouring equilibrium configurations. <img class="formulaInl" alt="$ C_n $" src="form_14.png"/> and <img class="formulaInl" alt="$ C_{n+1} $" src="form_15.png"/> in <a class="el" href="a00004.html#Various_configurations_of_a_body">Various configurations of a body</a> are examples of this. The incremental equations are best expressed using an incremental form of the virtual work principle.</p>
<p><a class="anchor" id="tm_eq_2_13"></a> </p>
<p class="formulaDsp">
<img class="formulaDsp" alt="\[ \int_{V_0} (\boldsymbol {S} : \delta \Delta \boldsymbol{E} + \Delta \boldsymbol{S} : \delta \boldsymbol{E})\mathrm{d} V_0 = \int_{A_0} \Delta \boldsymbol {t}_0 \cdot \delta \boldsymbol{u} \mathrm{d} A_0+ \int_{V_0} \Delta \boldsymbol {f}_0 \cdot \delta \boldsymbol{u} \mathrm{d} V_0\tag{2.13} \]" src="form_52.png"/>
</p>
<p>where the <img class="formulaInl" alt="$ \Delta $" src="form_53.png"/> is used to denote finite but small increments between <img class="formulaInl" alt="$ C_n $" src="form_14.png"/> and <img class="formulaInl" alt="$ C_{n+1} $" src="form_15.png"/>.</p>
<p>The first term in <a class="el" href="a00004.html#tm_eq_2_13">Eq. (2.13)</a> which depends on the current state of stress <img class="formulaInl" alt="$ \boldsymbol{S} $" src="form_42.png"/>, gives rise to the geometric stiffness matrix. The second term which depends on the incremental material law, is the basis for the material stiffness matrix. Note that both of these stiffness matrices are symmetric.</p>
<p>Equations <a class="el" href="a00004.html#tm_eq_2_12">Eq. (2.12)</a> and <a class="el" href="a00004.html#tm_eq_2_13">Eq. (2.13)</a> are also valid when a co-rotated ghost reference system is used. Merely replace <img class="formulaInl" alt="${E}_{ij} $" src="form_54.png"/> and <img class="formulaInl" alt="${S}_{ij} $" src="form_55.png"/> by <img class="formulaInl" alt="$\overset{\approx}{E}_{ij} $" src="form_56.png"/> and <img class="formulaInl" alt="$\overset{\approx}{S}_{ij} $" src="form_57.png"/>.</p>
<h2><a class="anchor" id="tm_finite_fundamental_dynamic"></a>
Dynamic Equilibrium Equations</h2>
<p>The dynamic equilibrium equation expressed in terms of virtual work can be written, see Remseth (1978)</p>
<p><a class="anchor" id="tm_eq_2_14"></a> </p>
<p class="formulaDsp">
<img class="formulaDsp" alt="\[ \int_{V_0} \boldsymbol {S} : \delta \boldsymbol{E} \mathrm{d} V_0 + \int_{V_0} \rho_0 \boldsymbol {\ddot{u}} \cdot \delta \boldsymbol{u} \mathrm{d} V_0 + \int_{V_0} \tilde{c} \boldsymbol {\dot{u}} \cdot \delta \boldsymbol{u} \mathrm{d} V_0 = \int_{A_0} \boldsymbol{t}_0 \cdot \delta \boldsymbol{u} \mathrm{d} A_0 + \int_{V_0} \boldsymbol{f}_0 \cdot \delta \boldsymbol{u} \mathrm{d} V_0 \tag{2.14} \]" src="form_58.png"/>
</p>
<p>where <img class="formulaInl" alt="$ \rho_0 $" src="form_59.png"/> denotes mass density and <img class="formulaInl" alt="$\tilde{c}$" src="form_60.png"/> is a viscous damping density function, i.e. damping forces are proportional to velocity. The inertia forces are proportional to the mass and acceleration of the structure. The additional terms in <a class="el" href="a00004.html#tm_eq_2_14">Eq. (2.14)</a> as compared to <a class="el" href="a00004.html#tm_eq_2_12">Eq. (2.12)</a>, give rise to the mass and the damping matrices.</p>
<p>The incremental form of the virtual work equation yields</p>
<p class="formulaDsp">
<img class="formulaDsp" alt="\[ \begin{array}{l} \displaystyle \int_{V_0} (\boldsymbol {S} : \delta \Delta \boldsymbol{E} + \Delta \boldsymbol {S} : \delta \boldsymbol{E})\mathrm{d} V_0 + \int_{V_0} \rho_0 \Delta \boldsymbol {\ddot{u}} \cdot \delta \boldsymbol{u} \mathrm{d} V_0 + \int_{V_0} \tilde{c} \Delta \boldsymbol {\dot{u}} \cdot \delta \boldsymbol{u} \mathrm{d} V_0 \\ \\ \displaystyle = \int_{A_0} \Delta \boldsymbol {t}_0 \cdot \delta \boldsymbol{u} \mathrm{d} A_0+ \int_{V_0} \Delta \boldsymbol {f}_0 \cdot \delta \boldsymbol{u} \mathrm{d} V_0 \end{array}\tag{2.15} \]" src="form_61.png"/>
</p>
<h1><a class="anchor" id="tm_finite_implementation"></a>
Implementation of the Finite Element Method</h1>
<h2><a class="anchor" id="tm_finite_implementation_large"></a>
Large Rotation in Space</h2>
<p>The finite element nodal points may have up to six degrees of freedom, i.e. three translations and three rotations. In the case of translational degrees of freedom only, the position of the displaced nodal point is uniquely defined by the initial position vector <img class="formulaInl" alt="$\boldsymbol{X}$" src="form_1.png"/> and the nodal displacement vector <img class="formulaInl" alt="$\boldsymbol{v}$" src="form_62.png"/>.</p>
<p><a class="anchor" id="tm_eq_2_16"></a> </p>
<p class="formulaDsp">
<img class="formulaDsp" alt="\[ \boldsymbol{x} = \boldsymbol{X} + \boldsymbol{v} \quad \mathrm{or} \quad x_i \boldsymbol{I}_i = (X_i + v_i) \boldsymbol{I}_i \tag{2.16} \]" src="form_63.png"/>
</p>
<p>The case of a node that is both translated and rotated must be treated more carefully. This is because large rotations in space are not true vectors that may be expressed by vectorial components in a base coordinate system. The current formulation uses a coordinate system with base vectors <img class="formulaInl" alt="$\boldsymbol{\bar{i}}_i$" src="form_64.png"/> that are frozen to the nodal point and that follow the movement of the node. This coordinate system is parallel to the global base vectors, <img class="formulaInl" alt="$\boldsymbol{\bar{I}}_i$" src="form_65.png"/> in the initial configuration as shown in <a class="el" href="a00004.html#Nodal_point_with_translational">Nodal point with translational and rotational degrees of freedom</a>. The orientation of the nodal point in space is uniquely defined by the base vector transformation.</p>
<p><a class="anchor" id="tm_eq_2_17"></a> </p>
<p class="formulaDsp">
<img class="formulaDsp" alt="\[ \boldsymbol{\bar{i}}_i^{(n)} = \boldsymbol{\bar{T}}_{ij}^{(n)} \boldsymbol{I}_j \tag{2.17} \]" src="form_66.png"/>
</p>
<p>The rotation of a node is thus not given by angles, but by the components of a rotation matrix which is orthogonal. Consequently, in the current formulation the general motion of a nodal point is described by the three displacement components <img class="formulaInl" alt="$ v_i $" src="form_67.png"/> and the nine elements of the rotation matrix <img class="formulaInl" alt="$\boldsymbol{\bar{T}}_{ij}$" src="form_68.png"/>.</p>
<p>During the incremental solution procedure <img class="formulaInl" alt="$\boldsymbol{\bar{T}}_{ij}$" src="form_68.png"/> is updated by treating the incremental rotational components about the global axes as rotations for small displacement problems. This will be described in more detail in <a class="el" href="a00004.html#tm_finite_implementation_updating">Updating of the Rotation Matrix</a>.</p>
<p><a class="anchor" id="Nodal_point_with_translational"></a> </p>
<div class="image">
<img src="RIFLEX_TheoryManual_42_v0_87.jpg" alt="RIFLEX_TheoryManual_42_v0_87.jpg"/>
<div class="caption">
Nodal point with translational and rotational degrees of freedom</div></div>
<h2><a class="anchor" id="tm_finite_implementation_finite"></a>
Finite Rigid Bodies at Nodes</h2>
<p>Finite rigid bodies may be introduced at the nodes. This feature makes it possible to accommodate for finite, rigid structural parts and to model eccentricities at nodes. This is illustrated in <a class="el" href="a00004.html#Motion_of_a_point_P">Motion of a point P at the surface of a rigid body</a>, where the element is thought to be attached to the surface of the rigid body. The initial location of the surface point is defined in terms of an eccentricity vector <img class="formulaInl" alt="$\boldsymbol{e}^0$" src="form_69.png"/>. After deformation has taken place, the position of the surface point is given by its position vector <img class="formulaInl" alt="$\boldsymbol{x}_p$" src="form_70.png"/>.</p>
<p><a class="anchor" id="tm_eq_2_18"></a> </p>
<p class="formulaDsp">
<img class="formulaDsp" alt="\[ \boldsymbol{x}_p =\boldsymbol{x} +\boldsymbol{e} = (X_i + v_i + e^0_j \boldsymbol{\bar{T}}_{ij})\boldsymbol{I}_i \tag{2.18} \]" src="form_71.png"/>
</p>
<p>where <img class="formulaInl" alt="$\boldsymbol{e}^0_j$" src="form_72.png"/> are the components of the initial eccentricity vector <img class="formulaInl" alt="$\boldsymbol{e}^0$" src="form_69.png"/> and <img class="formulaInl" alt="$\boldsymbol{\bar{T}}_{ij}$" src="form_68.png"/> is the rotation matrix components of the nodal point defined by <a class="el" href="a00004.html#tm_eq_2_17">Eq. (2.17)</a>.</p>
<p><a class="anchor" id="Motion_of_a_point_P"></a> </p>
<div class="image">
<img src="RIFLEX_TheoryManual_42_v0_94.jpg" alt="RIFLEX_TheoryManual_42_v0_94.jpg"/>
<div class="caption">
Motion of a point P at the surface of a rigid body</div></div>
<h2><a class="anchor" id="tm_finite_implementation_updating"></a>
Updating of the Rotation Matrix</h2>
<p>The transformation expressed by <a class="el" href="a00004.html#tm_eq_2_18">Eq. (2.18)</a> needs to be updated during the incremental solution procedure. The problem at hand is to establish the new transformation</p>
<p class="formulaDsp">
<img class="formulaDsp" alt="\[ \boldsymbol{\bar{i}}_i^{(n+1)} = \boldsymbol{\bar{T}}_{ij}^{(n+1)} \boldsymbol{I}_j \tag{2.19} \]" src="form_73.png"/>
</p>
<p>for configuration <img class="formulaInl" alt="$ C_{n+1} $" src="form_15.png"/> when the corresponding transformation for configuration <img class="formulaInl" alt="$ C_{n} $" src="form_74.png"/> is known. Provided that the incremental rotations from state <img class="formulaInl" alt="$ C_{n} $" src="form_74.png"/> to <img class="formulaInl" alt="$ C_{n+1} $" src="form_15.png"/> are small, this rotational motion ( <img class="formulaInl" alt="$\boldsymbol{\bar{i}}_i^{(n)} \rightarrow \boldsymbol{\bar{i}}_i^{(n+1)}$" src="form_75.png"/>) may be expressed as</p>
<p class="formulaDsp">
<img class="formulaDsp" alt="\[ \Delta\theta = \Delta\theta_i\boldsymbol{I}_i\tag{2.20} \]" src="form_76.png"/>
</p>
<p>where <img class="formulaInl" alt="$ \Delta\theta_i$" src="form_77.png"/> denotes the rotational components about the global axes <img class="formulaInl" alt="$\boldsymbol{I}_i$" src="form_0.png"/>. Bergan et al. (1985) and Mollestad (1983) show that the orientation of the nodal coordinate system in <img class="formulaInl" alt="$ C_{n+1} $" src="form_15.png"/> is given as</p>
<p class="formulaDsp">
<img class="formulaDsp" alt="\[ \boldsymbol{\bar{i}}_i^{(n+1)} = \boldsymbol{\bar{T}}_{ik}^{(n)} \boldsymbol{\tilde{T}}_{kj} \boldsymbol{I}_j \tag{2.21} \]" src="form_78.png"/>
</p>
<p>thus</p>
<p><a class="anchor" id="tm_eq_2_22"></a> </p>
<p class="formulaDsp">
<img class="formulaDsp" alt="\[ \boldsymbol{\bar{T}}_{ij}^{(n+1)} = \boldsymbol{\bar{T}}_{ik}^{(n)} \boldsymbol{\tilde{T}}_{kj} \tag{2.22} \]" src="form_79.png"/>
</p>
<p><img class="formulaInl" alt="$\boldsymbol{\tilde{T}}_{kj}$" src="form_80.png"/> denotes the components of the rotation matrix representing the incremental rotations from <img class="formulaInl" alt="$ C_n $" src="form_14.png"/> to <img class="formulaInl" alt="$ C_{n+1} $" src="form_15.png"/>, and given as</p>
<p><a class="anchor" id="tm_eq_2_23"></a> </p>
<p class="formulaDsp">
<img class="formulaDsp" alt="\[ \boldsymbol{\tilde{T}}_{kj} = \begin{bmatrix} C2 \cdot C3 &amp; \quad S3 \cdot C1 + S1 \cdot S2 \cdot C3 &amp; \quad S1 \cdot S3 - S2 \cdot C1 \cdot C3 \\ -S3 \cdot C2 &amp; \quad C1 \cdot C3 - S1 \cdot S2 \cdot S3 &amp; \quad S1 \cdot C3 + S2 \cdot S3 \cdot C1 \\ S2 &amp; \quad -S1 \cdot C2 &amp; \quad C1 \cdot C2 \end{bmatrix} \tag{2.23} \]" src="form_81.png"/>
</p>
<p>where</p>
<p><a class="anchor" id="tm_eq_2_24"></a> </p>
<p class="formulaDsp">
<img class="formulaDsp" alt="\[ \begin{array}{l} S1 = \sin (\Delta \theta_1) \quad C1 = \cos (\Delta \theta_1) \\ S2 = \sin (\Delta \theta_2) \quad C2 = \cos (\Delta \theta_2) \\ S3 = \sin (\Delta \theta_3) \quad C3 = \cos (\Delta \theta_3) \end{array}\tag{2.24} \]" src="form_82.png"/>
</p>
<p>The above expression for <img class="formulaInl" alt="$\boldsymbol{\tilde{T}}_{kj}$" src="form_80.png"/> is found by taking the rotations <img class="formulaInl" alt="$ \Delta\theta_i$" src="form_77.png"/> in the order <img class="formulaInl" alt="$ \Delta\theta_1 $" src="form_83.png"/>, <img class="formulaInl" alt="$ \Delta\theta_2 $" src="form_84.png"/>, <img class="formulaInl" alt="$ \Delta\theta_3$" src="form_85.png"/>. By taking these rotations in another order, the result in <a class="el" href="a00004.html#tm_eq_2_23">Eq. (2.23)</a> will become somewhat different. However, these differences become negligible when the rotations are small. It should also be kept in mind that <a class="el" href="a00004.html#tm_eq_2_23">Eq. (2.23)</a> is to be used in connection with a linearized incremental formulation which assumes small increments. The most important property of the rotation matrix is that it does not introduce orthogonality errors during the sequence of incremental rotations <img class="formulaInl" alt="$ \Delta\theta_1 $" src="form_83.png"/>, <img class="formulaInl" alt="$ \Delta\theta_2 $" src="form_84.png"/>, <img class="formulaInl" alt="$ \Delta\theta_3$" src="form_85.png"/>.</p>
<p>The updated rotation matrix in <a class="el" href="a00004.html#tm_eq_2_22">Eq. (2.22)</a> and the translations <a class="el" href="a00004.html#tm_eq_2_16">Eq. (2.16)</a> now defines the new orientation of the node in configuration <img class="formulaInl" alt="$ C_{n+1} $" src="form_15.png"/>.</p>
<p>Finally, it should be noted from <a class="el" href="a00004.html#tm_eq_2_23">Eq. (2.23)</a> that the following relations hold when the rotations are small</p>
<p><a class="anchor" id="tm_eq_2_25"></a> </p>
<p class="formulaDsp">
<img class="formulaDsp" alt="\[ \begin{array}{l} \Delta\theta_1 \approx \tilde{T}_{23} \\ \Delta\theta_2 \approx \tilde{T}_{31} \\ \Delta\theta_3 \approx \tilde{T}_{12} \end{array} \tag{2.25} \]" src="form_86.png"/>
</p>
<h2><a class="anchor" id="tm_finite_implementation_bar"></a>
Bar Element</h2>
<p>The spatial bar element is described in a total Lagrangian formulation. The description is based on Bergan (1984) and adjusted to a formulation based on integrated cross-section forces and small strain theory.</p>
<p><a class="anchor" id="Bar_element_in_initial"></a> </p>
<div class="image">
<img src="RIFLEX_TheoryManual_42_v0_125.jpg" alt="RIFLEX_TheoryManual_42_v0_125.jpg"/>
<div class="caption">
Bar element in initial and deformed configuration.</div></div>
<p>The element is assumed to be straight with an initial cross-sectional area <img class="formulaInl" alt="$ A_0 $" src="form_48.png"/> which is constant along the element length. Each of the two nodes has three translational degrees of freedom, which are expressed directly in the global coordinate system (see <a class="el" href="a00004.html#Bar_element_in_initial">Bar element in initial and deformed configuration</a>). The element length is denoted <img class="formulaInl" alt="$ L_0 $" src="form_87.png"/> and <img class="formulaInl" alt="$ L $" src="form_88.png"/> in the initial and deformed configuration, respectively. The deformed element length is given by</p>
<p class="formulaDsp">
<img class="formulaDsp" alt="\[ L = \sqrt{\Delta x^2 + \Delta y^2 + \Delta z^2}\tag{2.26} \]" src="form_89.png"/>
</p>
<p>where</p>
<p><img class="formulaInl" alt="$ \Delta x = x_2 - x_1 $" src="form_90.png"/></p>
<p><img class="formulaInl" alt="$ \Delta y = y_2 - y_1 $" src="form_91.png"/></p>
<p><img class="formulaInl" alt="$ \Delta z = z_2 - z_1 $" src="form_92.png"/></p>
<p>Based on a total Lagrangian formulation and linear displacement functions, the Green strain is expressed</p>
<p class="formulaDsp">
<img class="formulaDsp" alt="\[ E_f = \frac{1}{2}\frac{L^2-L_0^2}{L_0^2} = \frac{1}{2L_0^2} (\Delta x^2 + \Delta y^2 + \Delta z^2 - L_0^2)\tag{2.27} \]" src="form_93.png"/>
</p>
<p>The Piola-Kirchhoff stress <img class="formulaInl" alt="$ S_f $" src="form_94.png"/> is found through the constitutive law</p>
<p class="formulaDsp">
<img class="formulaDsp" alt="\[ S_f = S_f(E_f, E_0, S_0)\tag{2.28} \]" src="form_95.png"/>
</p>
<p>where <img class="formulaInl" alt="$ E_0 $" src="form_96.png"/> is initial strain and <img class="formulaInl" alt="$ S_0 $" src="form_97.png"/> is initial stress.</p>
<p>The linear displacement function (field function) for the three displacement components is given by</p>
<p><a class="anchor" id="tm_eq_2_29"></a> </p>
<p class="formulaDsp">
<img class="formulaDsp" alt="\[ \boldsymbol{N} = \boldsymbol{N}_u = \boldsymbol{N}_v = \boldsymbol{N}_w = \begin{bmatrix} 1-\xi &amp; \quad \xi \end{bmatrix} \tag{2.29} \]" src="form_98.png"/>
</p>
<p>where <img class="formulaInl" alt="$ \xi = \frac{x}{L} $" src="form_99.png"/> is normalized length coordinate of the element.</p>
<p>In the following small strain theory is used, and it is assumed that <img class="formulaInl" alt="$ L_0 $" src="form_87.png"/> is the initial stress free element length. Thus, the axial force of the element is given by</p>
<p class="formulaDsp">
<img class="formulaDsp" alt="\[ N = \frac{L-L_0}{L_0}(EA)\tag{2.30} \]" src="form_100.png"/>
</p>
<p>where <img class="formulaInl" alt="$ L_0 $" src="form_87.png"/> is initial, stress free element length and <img class="formulaInl" alt="$ EA$" src="form_101.png"/> is the axial stiffness. The strain, <img class="formulaInl" alt="$ \varepsilon $" src="form_102.png"/>, is given by</p>
<p class="formulaDsp">
<img class="formulaDsp" alt="\[ \varepsilon = \frac{L-L_0}{L_0}\tag{2.31} \]" src="form_103.png"/>
</p>
<p><a class="anchor" id="internal_loads"></a></p>
<h3>Internal loads</h3>
<p>Based on the virtual work equation, <a class="el" href="a00004.html#tm_eq_2_12">Eq. (2.12)</a>, the internal reaction force vector is expressed by</p>
<p class="formulaDsp">
<img class="formulaDsp" alt="\[ \boldsymbol{S}_{int} = \begin{bmatrix} S_{x1} \\ S_{y1} \\ S_{z1} \\ S_{x2} \\ S_{y2} \\ S_{z2} \\ \end{bmatrix} = \frac{N}{L_0}\cdot \begin{bmatrix} -\Delta x \\ -\Delta y \\ -\Delta z \\ \Delta x \\ \Delta y \\ \Delta z \end{bmatrix} \tag{2.32} \]" src="form_104.png"/>
</p>
<h3>Tangential stiffness</h3>
<p>The incremental form of the virtual work principle given by <a class="el" href="a00004.html#tm_eq_2_13">Eq. (2.13)</a> is used to derive the following tangential stiffness relation for the element</p>
<p class="formulaDsp">
<img class="formulaDsp" alt="\[ \Delta\boldsymbol{S}_{int} = \boldsymbol{k}\Delta\boldsymbol{v} = (\boldsymbol{k}_G + \boldsymbol{k}_M)\Delta \boldsymbol{v}\tag{2.33} \]" src="form_105.png"/>
</p>
<p>where <img class="formulaInl" alt="$ \Delta\boldsymbol{S}_{int} $" src="form_106.png"/> denotes increment in the internal load vector and <img class="formulaInl" alt="$ \boldsymbol{k}_G $" src="form_107.png"/> and <img class="formulaInl" alt="$ \boldsymbol{k}_M $" src="form_108.png"/> denote geometric and material stiffness matrices, respectively. <img class="formulaInl" alt="$ \Delta\boldsymbol{v} $" src="form_109.png"/> is the incremental displacement vector. It may be shown that <img class="formulaInl" alt="$ \boldsymbol{k}_G $" src="form_107.png"/> and <img class="formulaInl" alt="$ \boldsymbol{k}_M $" src="form_108.png"/> are given as</p>
<p class="formulaDsp">
<img class="formulaDsp" alt="\[ \boldsymbol{k}_G = \frac{N}{L_0} \begin{bmatrix} \boldsymbol{I}_3 &amp; -\boldsymbol{I}_3 \\ -\boldsymbol{I}_3 &amp; \boldsymbol{I}_3 \\ \end{bmatrix} \tag{2.34} \]" src="form_110.png"/>
</p>
<p class="formulaDsp">
<img class="formulaDsp" alt="\[ \boldsymbol{k}_M = \frac{(EA)_T}{L_0^3} \begin{bmatrix} \boldsymbol{D}^T\boldsymbol{D} &amp; -\boldsymbol{D}^T\boldsymbol{D} \\ -\boldsymbol{D}^T\boldsymbol{D} &amp; \boldsymbol{D}^T\boldsymbol{D} \end{bmatrix} \tag{2.35} \]" src="form_111.png"/>
</p>
<p>where</p>
<p class="formulaDsp">
<img class="formulaDsp" alt="\[ \boldsymbol{D} = \begin{bmatrix} \Delta x &amp; \Delta y &amp; \Delta z \end{bmatrix}\tag{2.36} \]" src="form_112.png"/>
</p>
<p>and <img class="formulaInl" alt="$ I_3 $" src="form_113.png"/> is the 3x3 unity matrix.</p>
<p><img class="formulaInl" alt="$ (EA)_T $" src="form_114.png"/> is the tangential axial stiffness giving the relation between incremental strain and incremental axial force</p>
<p class="formulaDsp">
<img class="formulaDsp" alt="\[ \Delta N = (EA)_T\Delta \varepsilon\tag{2.37} \]" src="form_115.png"/>
</p>
<h3>Distributed external loads</h3>
<p>The consistent external nodal load vector is generally written</p>
<p><a class="anchor" id="tm_eq_2_38"></a> </p>
<p class="formulaDsp">
<img class="formulaDsp" alt="\[ \begin{array}{l} \boldsymbol{S}_u^E = \displaystyle\int_L\boldsymbol{N}_u^Tp_x(x)\mathrm{d}x \\ \\ \boldsymbol{S}_v^E = \displaystyle\int_L\boldsymbol{N}_v^Tp_y(x)\mathrm{d}x \\ \\ \boldsymbol{S}_w^E = \displaystyle\int_L\boldsymbol{N}_w^Tp_z(x)\mathrm{d}x \end{array}\tag{2.38} \]" src="form_116.png"/>
</p>
<p>where <img class="formulaInl" alt="$ p(x) $" src="form_117.png"/> is distributed external loads per unit length in the deformed configuration. <img class="formulaInl" alt="$ \boldsymbol{N}_u $" src="form_118.png"/>, <img class="formulaInl" alt="$ \boldsymbol{N}_v $" src="form_119.png"/> and <img class="formulaInl" alt="$ \boldsymbol{N}_w $" src="form_120.png"/> are the displacement functions given by <a class="el" href="a00004.html#tm_eq_2_29">Eq. (2.29)</a>.</p>
<p>In most cases it is assumed that the load components are linearly distributed over the element. Using the relation in <a class="el" href="a00004.html#tm_eq_2_29">Eq. (2.29)</a>, the external load vector is given by</p>
<p><a class="anchor" id="tm_eq_2_39"></a> </p>
<p class="formulaDsp">
<img class="formulaDsp" alt="\[ \begin{array}{l} \boldsymbol{S}_u^E = \displaystyle\int_L\boldsymbol{N}^T\boldsymbol{N}\mathrm{d}x \cdot \boldsymbol{p}_u \\ \\ \boldsymbol{S}_v^E = \displaystyle\int_L\boldsymbol{N}^T\boldsymbol{N}\mathrm{d}x \cdot \boldsymbol{p}_v \\ \\ \boldsymbol{S}_w^E = \displaystyle\int_L\boldsymbol{N}^T\boldsymbol{N}\mathrm{d}x \cdot \boldsymbol{p}_w \\ \end{array}\tag{2.39} \]" src="form_121.png"/>
</p>
<p>where <img class="formulaInl" alt="$ \boldsymbol{p}_u $" src="form_122.png"/>, <img class="formulaInl" alt="$ \boldsymbol{p}_v $" src="form_123.png"/> and <img class="formulaInl" alt="$ \boldsymbol{p}_w $" src="form_124.png"/> are load intensity vectors at the element nodes.</p>
<p>By carrying out the integrals in <a class="el" href="a00004.html#tm_eq_2_39">Eq. (2.39)</a> the external nodal load vector is written</p>
<p class="formulaDsp">
<img class="formulaDsp" alt="\[ \boldsymbol{S}^E = \begin{bmatrix} S_{x1} \\ S_{y1} \\ S_{z1} \\ S_{x2} \\ S_{y2} \\ S_{z2} \end{bmatrix} = \frac{L}{6} \begin{bmatrix} 2p_{x1} + p_{x2}\\ 2p_{y1} + p_{y2}\\ 2p_{z1} + p_{z2}\\ p_{x1} + 2p_{x2}\\ p_{y1} + 2p_{y2}\\ p_{z1} + 2p_{z2} \end{bmatrix}\tag{2.40} \]" src="form_125.png"/>
</p>
<p>The external load vector can alternatively be calculated using a lumped load formulation, which gives</p>
<p class="formulaDsp">
<img class="formulaDsp" alt="\[ \boldsymbol{S}^E = \begin{bmatrix} S_{x1} \\ S_{y1} \\ S_{z1} \\ S_{x2} \\ S_{y2} \\ S_{z2} \end{bmatrix} = \frac{L}{2} \begin{bmatrix} p_{x1} \\ p_{y1} \\ p_{z1} \\ p_{x2} \\ p_{y2} \\ p_{z2} \end{bmatrix}\tag{2.41} \]" src="form_126.png"/>
</p>
<p>It should be noted that volume loads (weight and buoyancy) are calculated directly in the global reference system, while other load contributions are calculated in the local element system before transformation to global system.</p>
<p>For elements penetrating the sea surface, contributions to the external load vector from hydrodynamic loads and buoyancy are calculated considering the wetted element length. In this case, a linear distribution is assumed from the submerged element end to the intersection point between the element and the sea surface.</p>
<p>For floating partly submerged cross-sections (special option) consistent formulation is always used. The contributions from hydrodynamic loading are calculated based on numerical integration of <a class="el" href="a00004.html#tm_eq_2_38">Eq. (2.38)</a>.</p>
<p>Note that hydrodynamic load contributions proportional to structural accelerations and velocities, i.e. added mass and hydrodynamic damping forces, are included in the mass and damping matrices respectively.</p>
<h3>Mass matrix</h3>
<p>The mass matrix contains contributions from structural and hydrodynamic (added) mass according to</p>
<p class="formulaDsp">
<img class="formulaDsp" alt="\[ \boldsymbol{m} = \boldsymbol{m}^s + \boldsymbol{m}^h \tag{2.42a} \]" src="form_127.png"/>
</p>
<p>where <img class="formulaInl" alt="$ \boldsymbol{m}^s $" src="form_128.png"/> is the structural mass matrix and <img class="formulaInl" alt="$ \boldsymbol{m}^h $" src="form_129.png"/> is the added mass matrix.</p>
<p>The structural mass matrix is calculated according to</p>
<p class="formulaDsp">
<img class="formulaDsp" alt="\[ \boldsymbol{m}^s = m^s \int_{L_0}\boldsymbol{N}^T\boldsymbol{N}\mathrm{d}x \tag{2.42b} \]" src="form_130.png"/>
</p>
<p>where <img class="formulaInl" alt="$ \boldsymbol{m}^s $" src="form_128.png"/> is structural mass per unit length and <img class="formulaInl" alt="$ \boldsymbol{N} $" src="form_131.png"/> is the displacement function matrix given by</p>
<p><a class="anchor" id="tm_eq_2_43"></a> </p>
<p class="formulaDsp">
<img class="formulaDsp" alt="\[ \boldsymbol{N} = \begin{bmatrix} N_{u1} &amp; &amp; &amp; &amp; &amp; \\ &amp; N_{v1} &amp; &amp; &amp; &amp; \\ &amp; &amp; N_{w1} &amp; &amp; &amp; \\ &amp; &amp; &amp; N_{u2} &amp; &amp; \\ &amp; &amp; &amp; &amp; N_{v2} &amp; \\ &amp; &amp; &amp; &amp; &amp; N_{w2} \end{bmatrix}\tag{2.43} \]" src="form_132.png"/>
</p>
<p> All terms that are not indicated in <a class="el" href="a00004.html#tm_eq_2_43">Eq. (2.43)</a> are equal to zero.</p>
<p><img class="formulaInl" alt="$ \boldsymbol{N}_u $" src="form_118.png"/>, <img class="formulaInl" alt="$ \boldsymbol{N}_v $" src="form_119.png"/> and <img class="formulaInl" alt="$ \boldsymbol{N}_w $" src="form_120.png"/> are defined in <a class="el" href="a00004.html#tm_eq_2_29">Eq. (2.29)</a>.</p>
<p>The direction-dependent added mass matrix is calculated in the local system according to</p>
<p><a class="anchor" id="tm_eq_2_44"></a> </p>
<p class="formulaDsp">
<img class="formulaDsp" alt="\[ \begin{array}{l} \boldsymbol{\boldsymbol{m}}^h_{uu} = \displaystyle\int_{L_0} \boldsymbol{N}^Tm_x^h\boldsymbol{N}\mathrm{d}x \\ \\ \boldsymbol{m}^h_{vv} = \displaystyle\int_{L_0} \boldsymbol{N}^Tm_y^h\boldsymbol{N}\mathrm{d}x \\ \\ \boldsymbol{m}^h_{ww} = \displaystyle\int_{L_0} \boldsymbol{N}^Tm_z^h\boldsymbol{N}\mathrm{d}x \end{array}\tag{2.44} \]" src="form_133.png"/>
</p>
<p>where <img class="formulaInl" alt="$ m_x^h $" src="form_134.png"/> , <img class="formulaInl" alt="$ m_y^h $" src="form_135.png"/> and <img class="formulaInl" alt="$ m_z^h $" src="form_136.png"/> are added mass per unit length referring to local x, y and z-axes, respectively.</p>
<p>For elements penetrating the sea surface, only the submerged part contributes to the added mass.</p>
<p>The mass matrix can alternatively be calculated using a lumped mass formulation, which in the local system is given by</p>
<p><a class="anchor" id="tm_eq_2_45"></a> </p>
<p class="formulaDsp">
<img class="formulaDsp" alt="\[ \boldsymbol{m} = \frac{1}{2} L \cdot \begin{bmatrix} m^s + m^h_x &amp; &amp; &amp; &amp; &amp; \\ &amp; m^s + m_y^h &amp; &amp; &amp; &amp; \\ &amp; &amp; m^s + m_z^h &amp; &amp; &amp; \\ &amp; &amp; &amp; m^s + m^h_x &amp; &amp; \\ &amp; &amp; &amp; &amp; m^s + m_y^h &amp; \\ &amp; &amp; &amp; &amp; &amp; m^s + m_z^h \end{bmatrix}\tag{2.45} \]" src="form_137.png"/>
</p>
<p> All terms that are not indicated in <a class="el" href="a00004.html#tm_eq_2_45">Eq. (2.45)</a> are equal to zero.</p>
<p>Due to directional dependency the mass matrix contains off-diagonal terms after transformation to the global system. To obtain a diagonal mass matrix after transformation, the off-diagonal terms are simply added to the diagonal for the lumped formulation.</p>
<p>For floating partly submerged cross-sections (special option) the added mass may vary over the element length. Consequently, the added mass matrix is calculated based on numerical integration of <a class="el" href="a00004.html#tm_eq_2_44">Eq. (2.44)</a>. This option is restricted to the consistent formulation.</p>
<h3>Damping matrix</h3>
<p>The damping matrix consists of contributions from structural damping, <img class="formulaInl" alt="$ \boldsymbol{c}^s$" src="form_138.png"/>, and hydrodynamic damping, <img class="formulaInl" alt="$ \boldsymbol{c}^h$" src="form_139.png"/>.</p>
<p class="formulaDsp">
<img class="formulaDsp" alt="\[ \boldsymbol{c} = \boldsymbol{c}^s + \boldsymbol{c}^h\tag{2.46} \]" src="form_140.png"/>
</p>
<p>Hydrodynamic damping is only considered for partly submerged cross-sections (special option).</p>
<p>For description of structural damping, confer <a class="el" href="a00010.html#tm_dynamic_structural">Structural Damping Models</a> in <a class="el" href="a00010.html#dynamic_time">Dynamic Time Domain Analysis</a>.</p>
<p>The hydrodynamic damping matrix in local system is calculated according to</p>
<p class="formulaDsp">
<img class="formulaDsp" alt="\[ \begin{array}{l} c_{uu} = 0 \\ \\ c_{vv} = \displaystyle\int_{L_0} \boldsymbol{N}_v^T c_y^h(x)\boldsymbol{N}_v \mathrm{d}x \\ \\ c_{vv} = \displaystyle\int_{L_0} \boldsymbol{N}_w^T c_x^h(x)\boldsymbol{N}_w \mathrm{d}x \end{array}\tag{2.47} \]" src="form_141.png"/>
</p>
<p>where <img class="formulaInl" alt="$ c^h_y $" src="form_142.png"/> and <img class="formulaInl" alt="$ c^h_z $" src="form_143.png"/> are hydrodynamic damping per unit length.</p>
<p>As indicated, the damping may vary along the element. Thus, the damping matrix is calculated based on numerical integration.</p>
<h2><a class="anchor" id="tm_finite_implementation_beam"></a>
Beam Element</h2>
<p>The beam element is formulated using the concept of co-rotated ghost reference as outlined in <a class="el" href="a00004.html#tm_finite_fundamental">Fundamental Continuum Mechanics Theory</a>. The present formulation is, apart from a few modifications, identical to the beam element described by Bergan (1984). A detailed discussion of this element together with examples demonstrating its capabilities may be found in Mollestad (1983) and Engseth (1984).</p>
<p>As indicated in <a class="el" href="a00004.html#Nodal_degrees_of_freedom">Nodal degrees of freedom for beam element</a>, the beam has 3 translational and 3 rotational degrees of freedom at each node. They are defined in relation to the local x, y, z-system in the <img class="formulaInl" alt="$ C_{0n}$" src="form_144.png"/> configuration. The x-axis is oriented along the secant between the two end nodes 1 and 2 of the element and goes through the centroid of the cross section. The y-axis is taken as the average direction of the principal y-axes at the two beam ends, and the z-axis is finally taken to be perpendicular to the x-y plane. The <img class="formulaInl" alt="$ C_{0n}$" src="form_144.png"/> configuration is then oriented along the x-axis with cross-sectional principal axis in the y- and z-directions. It is important to note that the rotational degrees of freedom in <a class="el" href="a00004.html#Nodal_degrees_of_freedom">Nodal degrees of freedom for beam element</a> express deformational rotations in relation to the co-rotated straight element.</p>
<p><a class="anchor" id="Nodal_degrees_of_freedom"></a></p>
<div class="image">
<img src="RIFLEX_TheoryManual_42_v0_201.jpg" alt="RIFLEX_TheoryManual_42_v0_201.jpg"/>
<div class="caption">
Nodal degrees of freedom for beam element.</div></div>
<p>The beam theory is based on the following assumptions:</p>
<ul>
<li>A plane section of the beam initially normal to the x-axis, remains plane and normal to the x-axis during deformations.</li>
<li>Lateral contraction caused by axial elongation is neglected.</li>
<li>The strains are small.</li>
<li>Shear deformations due to lateral loading are accounted for by modifying the bending stiffness as described in Bell (1987).</li>
<li>St. Venant torsion model is applied.</li>
<li>Torsional warping resistance is neglected.</li>
<li>Coupling between torsion and bending is accounted for by a second order approximation of torsion and bending curvature, however, the coupling is not reflected in the stiffness matrix.</li>
</ul>
<p>As shown in the figure <a class="el" href="a00004.html#Prismatic_beam">Prismatic beam</a>, the displacement of an arbitrary point P with coordinates x, y and z may be expressed as</p>
<p class="formulaDsp">
<img class="formulaDsp" alt="\[ \begin{array}{l} \displaystyle {u}(x,y,z) = {u}_0(x)-y \frac{\mathrm{d}{v}_0}{\mathrm{d}x} -z\frac{\mathrm{d}{v}_0}{\mathrm{d}x} \\\\ {v}(x,y,z) = {v}_0(x)-z\theta \\\\ {w}(x,y,z) = {w}_0(x)+y\theta \end{array}\tag{2.48} \]" src="form_145.png"/>
</p>
<p>where <img class="formulaInl" alt="$ {u}_0 $" src="form_146.png"/>, <img class="formulaInl" alt="$ {v}_0 $" src="form_147.png"/> and <img class="formulaInl" alt="$ {w}_0 $" src="form_148.png"/> are the displacements of the corresponding point on the reference axis. <a class="el" href="a00004.html#tm_eq_2_8">Eq. (2.8)</a> may now be used to find the axial strain. If all quadratic strain terms that are zero on the x-axis are neglected, and if the quadratic axial strain term is assumed to be negligible, then the Green strain may be expressed as</p>
<p><a class="anchor" id="tm_eq_2_49"></a> </p>
<p class="formulaDsp">
<img class="formulaDsp" alt="\[ {E}_{xx} ={u}_{0,xx} -y {v}_{0,xx} -z{w}_{0,xx} +\frac{1}{2} ({v}^2_{0,x}+{w}^2_{0,x})\tag{2.49} \]" src="form_149.png"/>
</p>
<p>The torsional behaviour of the beam is based on the relationship</p>
<p class="formulaDsp">
<img class="formulaDsp" alt="\[ {M}_\theta = GI_t\theta_x \tag{2.50} \]" src="form_150.png"/>
</p>
<p>where <img class="formulaInl" alt="$ \boldsymbol{M}_\theta $" src="form_151.png"/> is the twist moment and <img class="formulaInl" alt="$ GI_t $" src="form_152.png"/> is the torsional stiffness.</p>
<p><a class="anchor" id="Prismatic_beam"></a> </p>
<div class="image">
<img src="RIFLEX_TheoryManual_42_v0_212.jpg" alt="RIFLEX_TheoryManual_42_v0_212.jpg"/>
<div class="caption">
Prismatic beam.</div></div>
<p>A standard element formulation gives</p>
<p><a class="anchor" id="tm_eq_2_51"></a> </p>
<p class="formulaDsp">
<img class="formulaDsp" alt="\[ \begin{array}{l} {u}_0 =\boldsymbol{N}_u \boldsymbol{v}_u \quad \mathrm{where} \quad\boldsymbol{v}^T_u = \begin{bmatrix} v_{x1} &amp; v_{x2}\end{bmatrix} \\ \\ {v}_0 =\boldsymbol{N}_v \boldsymbol{v}_v \quad \mathrm{where} \quad\boldsymbol{v}^T_v = \begin{bmatrix} v_{y1} &amp; \theta_{z1} &amp; v_{y2} &amp; \theta_{z2}\end{bmatrix} \\ \\ {w}_0 =\boldsymbol{N}_w \boldsymbol{v}_w \quad \mathrm{where} \quad\boldsymbol{v}^T_w = \begin{bmatrix} v_{z1} &amp; \theta_{y1} &amp; v_{z2} &amp; \theta_{y2}\end{bmatrix} \\ \\ {\theta}_0 =\boldsymbol{N}_\theta \boldsymbol{v}_\theta \quad \mathrm{where}\quad\boldsymbol{v}^T_\theta = \begin{bmatrix} \theta_{x1} &amp; \theta_{x2}\end{bmatrix} \end{array}\tag{2.51} \]" src="form_153.png"/>
</p>
<p>where <img class="formulaInl" alt="$ \boldsymbol{N}_u $" src="form_118.png"/> and <img class="formulaInl" alt="$ \boldsymbol{N}_\theta $" src="form_154.png"/> are linear interpolation functions, while <img class="formulaInl" alt="$ \boldsymbol{N}_v $" src="form_119.png"/> and <img class="formulaInl" alt="$ \boldsymbol{N}_w $" src="form_120.png"/> express cubic interpolation functions.</p>
<p>The interpolation functions are given by</p>
<p><a class="anchor" id="tm_eq_2_52"></a> </p>
<p class="formulaDsp">
<img class="formulaDsp" alt="\[ \begin{array}{l} \boldsymbol{N}_u =\boldsymbol{N}_\theta = \begin{bmatrix} 1 -\displaystyle\frac{x}{L} &amp; \quad \displaystyle\frac{x}{L} \end{bmatrix} \\ \\ \boldsymbol{N}^T_v = \begin{bmatrix} 1-3\displaystyle\frac{x^2}{L^2}+2 \frac{x^3}{L^3} \\ \\ x - 2\displaystyle\frac{x^2}{L^2} + \frac{x^3}{L^2} \\ \\ 1 - 3\displaystyle\frac{(L-x)^2}{L^2} + 2 \frac{(L-x)^3}{L^3} \\ \\ - (L-x) + 2\displaystyle\frac{(L-x)^2}{L} - \frac{(L-x)^3}{L^2} \end{bmatrix} \\ \\ \\ \boldsymbol{N}^T_w = \begin{bmatrix} 1-3\displaystyle\frac{x}{L^2}+2 \frac{x^3}{L^3} \\ \\ -x + 2\displaystyle\frac{x^2}{L} - \frac{x^3}{L^2} \\ \\ 1 - 3\displaystyle\frac{(L-x)^2}{L^2} + 2 \frac{(L-x)^3}{L^3} \\ \\ L-x - 2\displaystyle\frac{(L-x)^2}{L} + \frac{(L-x)^3}{L^2} \end{bmatrix} \end{array}\tag{2.52} \]" src="form_155.png"/>
</p>
<h3><a class="anchor" id="tm_internal_loads"></a>
Internal loads</h3>
<p>Based on the internal virtual work equation, <a class="el" href="a00004.html#tm_eq_2_12">Eq. (2.12)</a>, using the results expressed by <a class="el" href="a00004.html#tm_eq_2_49">Eq. (2.49)</a> and <a class="el" href="a00004.html#tm_eq_2_51">Eq. (2.51)</a>, it may be shown that the expressions for the internal reaction forces are given by</p>
<p><a class="anchor" id="tm_eq_2_53"></a> </p>
<p class="formulaDsp">
<img class="formulaDsp" alt="\[ \begin{array}{l} \boldsymbol{S}_u = \displaystyle\int_{L_0}\boldsymbol{N}^T_{u,x} N_{xx}\mathrm{d}x \\ \\ \boldsymbol{S}_v = \displaystyle\int_{L_0}\boldsymbol{N}^T_{v,xx} M_{y}\mathrm{d}x \\ \\ \boldsymbol{S}_w = \displaystyle\int_{L_0}\boldsymbol{N}^T_{w,xx} M_{z}\mathrm{d}x \\ \\ \boldsymbol{S}_\theta = \displaystyle\int_{L_0}\boldsymbol{N}^T_{\theta,xx} M_{\theta}\mathrm{d}x \end{array}\tag{2.53} \]" src="form_156.png"/>
</p>
<p>where <img class="formulaInl" alt="$ N_{xx}$" src="form_157.png"/>, <img class="formulaInl" alt="$ M_y$" src="form_158.png"/>, <img class="formulaInl" alt="$ M_z$" src="form_159.png"/> and <img class="formulaInl" alt="$ M_\theta$" src="form_160.png"/> denote the cross-sectional force resultants. <img class="formulaInl" alt="$ \boldsymbol{N}_u $" src="form_118.png"/>, <img class="formulaInl" alt="$ \boldsymbol{N}_v $" src="form_119.png"/>, <img class="formulaInl" alt="$ \boldsymbol{N}_w $" src="form_120.png"/> and <img class="formulaInl" alt="$ \boldsymbol{N}_\theta $" src="form_154.png"/> are taken from <a class="el" href="a00004.html#tm_eq_2_52">Eq. (2.52)</a> with the <img class="formulaInl" alt="$ _x $" src="form_161.png"/> and <img class="formulaInl" alt="$ _{xx} $" src="form_162.png"/> subscripts denoting first and second derivatives of <img class="formulaInl" alt="$ x $" src="form_163.png"/>. In the current formulation these resultants will be determined from predefined curves giving the moment-curvature relationship and the relation between axial force and elongation for the given cross section. This eliminates the numerical integration over the cross-section. The axial force <img class="formulaInl" alt="$ N_x$" src="form_164.png"/> is positive when in tension, and the sign conventions for the moments <img class="formulaInl" alt="$ M_y$" src="form_158.png"/> and <img class="formulaInl" alt="$ M_z$" src="form_159.png"/> are given in <a class="el" href="a00004.html#Sign_conventions_for_moments">Sign conventions for moments</a>. Note that <img class="formulaInl" alt="$ M_y$" src="form_158.png"/> and <img class="formulaInl" alt="$ M_z$" src="form_159.png"/> denote the moments giving transverse displacements in the y- and z-directions, respectively. Some higher order terms have been neglected in <a class="el" href="a00004.html#tm_eq_2_53">Eq. (2.53)</a>.</p>
<p><a class="anchor" id="Sign_conventions_for_moments"></a></p>
<div class="image">
<img src="RIFLEX_TheoryManual_42_v0_241.jpg" alt="RIFLEX_TheoryManual_42_v0_241.jpg"/>
<div class="caption">
Sign conventions for moments.</div></div>
<p>The increment in internal virtual work, <a class="el" href="a00004.html#tm_eq_2_13">Eq. (2.13)</a>, is used to calculate the element stiffness matrix. The expression for the geometric stiffness matrix becomes</p>
<p><a class="anchor" id="tm_eq_2_54"></a> </p>
<p class="formulaDsp">
<img class="formulaDsp" alt="\[ \boldsymbol{k}_G = \begin{bmatrix} 0 &amp; 0 &amp; 0 &amp; 0 \\ 0 &amp; \boldsymbol{k}_{Gvv} &amp; 0 &amp; 0 \\ 0 &amp; 0 &amp; \boldsymbol{k}_{Gww} &amp; 0 \\ 0 &amp; 0 &amp; 0 &amp; 0 \end{bmatrix}\tag{2.54} \]" src="form_165.png"/>
</p>
<p>where</p>
<p class="formulaDsp">
<img class="formulaDsp" alt="\[ \begin{array}{l} \boldsymbol{k}_{Gvv} = \displaystyle\int_{L_0} N_{xx} \boldsymbol{N}^T_{v,x} \mathrm{d}x \\ \\ \boldsymbol{k}_{Gww} = \displaystyle\int_{L_0} N_{xx} \boldsymbol{N}^T_{w,x} \mathrm{d}x \end{array}\tag{2.55} \]" src="form_166.png"/>
</p>
<p>From this result one may see that the only geometric effect accounted for by the current beam element, is the influence of axial force on the coupling between transverse forces and transverse displacements. <a class="el" href="a00004.html#tm_eq_2_54">Eq. (2.54)</a> shows no coupling between axial forces and twist, which signifies that torsional stability problems are excluded.</p>
<p>The material part of the tangent stiffness matrix takes the following form</p>
<p><a class="anchor" id="tm_eq_2_56"></a> </p>
<p class="formulaDsp">
<img class="formulaDsp" alt="\[ \boldsymbol{k}_M = \begin{bmatrix} \boldsymbol{k}_{Muu} &amp; 0 &amp; 0 &amp; 0 \\ 0 &amp; \boldsymbol{k}_{Mvv} &amp; 0 &amp; 0 \\ 0 &amp; 0 &amp; \boldsymbol{k}_{Mww} &amp; 0 \\ 0 &amp; 0 &amp; 0 &amp; \boldsymbol{k}_{M\theta\theta} \end{bmatrix}\tag{2.56} \]" src="form_167.png"/>
</p>
<p>where</p>
<p><a class="anchor" id="tm_eq_2_57"></a> </p>
<p class="formulaDsp">
<img class="formulaDsp" alt="\[ \begin{array}{l} \boldsymbol{k}_{Muu} = \displaystyle\int_{L_0} {N}_{T}\boldsymbol{N}^T_{u,x} \boldsymbol{N}_{u,x} \mathrm{d}x \\ \\ \boldsymbol{k}_{Mvv} = \displaystyle\int_{L_0} {B}_{yT}\boldsymbol{N}^T_{v,xx} \boldsymbol{N}_{v,xx} \mathrm{d}x \\ \\ \boldsymbol{k}_{Mww} = \displaystyle\int_{L_0} {B}_{zT}\boldsymbol{N}^T_{w,xx} \boldsymbol{N}_{w,xx} \mathrm{d}x \\ \\ \boldsymbol{k}_{M\theta\theta} = \displaystyle\int_{L_0} GI_{\theta} \boldsymbol{N}^T_{\theta,x} \boldsymbol{N}_{\theta,x} \mathrm{d}x \\ \end{array}\tag{2.57} \]" src="form_168.png"/>
</p>
<p>In <a class="el" href="a00004.html#tm_eq_2_57">Eq. (2.57)</a> <img class="formulaInl" alt="$ N_T$" src="form_169.png"/>, <img class="formulaInl" alt="$ {B}_{yT} $" src="form_170.png"/>, <img class="formulaInl" alt="$ {B}_{zT} $" src="form_171.png"/> denote the tangential cross-sectional moduli where <img class="formulaInl" alt="$ N_T$" src="form_169.png"/> is the axial stiffness and <img class="formulaInl" alt="$ {B}_{yT} $" src="form_170.png"/> and <img class="formulaInl" alt="$ {B}_{zT} $" src="form_171.png"/> are the bending stiffnesses. It is seen from <a class="el" href="a00004.html#tm_eq_2_56">Eq. (2.56)</a> that there is no coupling between axial terms and bending terms, and no coupling between bending about the y- and z-axes. This is because the x-axis is chosen to go through the centroid and y and z to follow the principal axes.</p>
<p>The tangential stiffness is expressed as</p>
<p class="formulaDsp">
<img class="formulaDsp" alt="\[ \boldsymbol{k} = \boldsymbol{k}_G + \boldsymbol{k}_M\tag{2.58} \]" src="form_172.png"/>
</p>
<p>Eccentricties for the area and shear centers are handled as described in Section <a class="el" href="a00004.html#element_local_trans">Element local transformations</a>.</p>
<h3>Distributed external loads</h3>
<p>The external nodal load vector due to distributed external loads is generally written</p>
<p><a class="anchor" id="tm_eq_2_59"></a> </p>
<p class="formulaDsp">
<img class="formulaDsp" alt="\[ \begin{array}{l} \boldsymbol{S}_u = \displaystyle\int_L\boldsymbol{N}^T_u p_x(x)\mathrm{d}x \\ \\ \boldsymbol{S}_v = \displaystyle\int_L\boldsymbol{N}^T_v p_y(x)\mathrm{d}x \\ \\ \boldsymbol{S}_w = \displaystyle\int_L\boldsymbol{N}^T_w p_z(x)\mathrm{d}x \\ \\ \boldsymbol{S}_\theta = \displaystyle\int_L\boldsymbol{N}^T_\theta p_\theta(x)\mathrm{d}x \end{array}\tag{2.59} \]" src="form_173.png"/>
</p>
<p>where <img class="formulaInl" alt="$ p(x) $" src="form_117.png"/> is distributed external loads per unit length.</p>
<p>In most cases it is assumed that the load components are linearly distributed over the element</p>
<p><a class="anchor" id="tm_eq_2_60"></a> </p>
<p class="formulaDsp">
<img class="formulaDsp" alt="\[ p(x) =\boldsymbol{Np} \tag{2.60} \]" src="form_174.png"/>
</p>
<p>where <img class="formulaInl" alt="$ \boldsymbol{p} $" src="form_175.png"/> is the load intensity vector at the element nodes and <img class="formulaInl" alt="$ \boldsymbol{N} $" src="form_131.png"/> is linear displacement function given by</p>
<p class="formulaDsp">
<img class="formulaDsp" alt="\[ \boldsymbol{N} = \begin{bmatrix} \displaystyle 1-\frac{x}{L} &amp; \displaystyle\quad \frac{x}{L} \end{bmatrix} \tag{2.61} \]" src="form_176.png"/>
</p>
<p>By use of the relation in <a class="el" href="a00004.html#tm_eq_2_60">Eq. (2.60)</a> the external load vector is calculated according to</p>
<p><a class="anchor" id="tm_eq_2_62"></a> </p>
<p class="formulaDsp">
<img class="formulaDsp" alt="\[ \begin{array}{l} \boldsymbol{S}_u = \displaystyle\int_L\boldsymbol{N}^T_u \boldsymbol{N} \mathrm{d}x \cdot \boldsymbol{p}_x \\ \\ \boldsymbol{S}_v = \displaystyle\int_L\boldsymbol{N}^T_v \boldsymbol{N} \mathrm{d}x \cdot \boldsymbol{p}_y \\ \\ \boldsymbol{S}_w = \displaystyle\int_L\boldsymbol{N}^T_w \boldsymbol{N} \mathrm{d}x \cdot \boldsymbol{p}_z \\ \\ \boldsymbol{S}_\theta = \displaystyle\int_L\boldsymbol{N}^T_\theta \boldsymbol{N} \mathrm{d}x \cdot \boldsymbol{p}_\theta \end{array}\tag{2.62} \]" src="form_177.png"/>
</p>
<p>where <img class="formulaInl" alt="$ \boldsymbol{p}_x $" src="form_178.png"/>, <img class="formulaInl" alt="$ \boldsymbol{p}_y $" src="form_179.png"/>, <img class="formulaInl" alt="$ \boldsymbol{p}_z $" src="form_180.png"/> and <img class="formulaInl" alt="$ \boldsymbol{p}_\theta $" src="form_181.png"/> are load intensities at the element nodes.</p>
<p>It should be noted that contributions to fixed end moments in <img class="formulaInl" alt="$ \boldsymbol{S}_v $" src="form_182.png"/> and <img class="formulaInl" alt="$ \boldsymbol{S}_w $" src="form_183.png"/> are disregarded. This is not a fully consistent formulation, but it avoids superposition of the particular solution in the nonlinear stress calculation for the element. Further, the principle of superposition is really not valid in nonlinear analysis. On the other hand, this introduces an error strongly dependent on the actual discretization and consequently one should avoid using adjacent elements with large deviations in length.</p>
<p>By carrying out the integral in <a class="el" href="a00004.html#tm_eq_2_62">Eq. (2.62)</a> disregarding fixed end moments the external load vector is written</p>
<p class="formulaDsp">
<img class="formulaDsp" alt="\[ \boldsymbol{S}^E = \begin{bmatrix} S_{x1} \\ S_{y1} \\ S_{z1} \\ S_{\theta x1} \\ S_{\theta y1} \\ S_{\theta z1} \\ S_{x2} \\ S_{y2} \\ S_{z2} \\ S_{\theta x2} \\ S_{\theta y2} \\ S_{\theta z2} \end{bmatrix} =\begin{bmatrix} \frac{L}{6}(2p_{x1}+ p_{x2}) \\ \frac{L}{60}(21p_{y1} + 9p_{y2}) \\ \frac{L}{60}(21p_{z1} + 9p_{z2}) \\ \frac{L}{6}(2p_{\theta1} + p_{\theta2}) \\ 0 \\ 0 \\ \frac{L}{6}(p_{x1} + 2p_{x2}) \\ \frac{L}{60}(9p_{y1} + 21p_{y2}) \\ \frac{L}{60}(9p_{z1} + 21p_{z2}) \\ \frac{L}{6}(p_{\theta1} + 2p_{\theta2}) \\ 0 \\ 0 \end{bmatrix}\tag{2.63} \]" src="form_184.png"/>
</p>
<p>The external load vector can alternatively be calculated using a lumped load formulation, which gives</p>
<p class="formulaDsp">
<img class="formulaDsp" alt="\[ \boldsymbol{S}^E = \begin{bmatrix} S_{x1} \\ S_{y1} \\ S_{z1} \\ S_{\theta x1} \\ S_{\theta y1} \\ S_{\theta z1} \\ S_{x2} \\ S_{y2} \\ S_{z2} \\ S_{\theta x2} \\ S_{\theta y2} \\ S_{\theta z2} \end{bmatrix} =\frac{L}{2}\begin{bmatrix} p_{x1} \\ p_{y1} \\ p_{z1} \\ p_{\theta1} \\ 0 \\ 0 \\ p_{x2} \\ p_{y2} \\ p_{z2} \\ p_{\theta2} \\ 0 \\ 0 \end{bmatrix}\tag{2.64} \]" src="form_185.png"/>
</p>
<p>For elements penetrating the sea surface, contributions to the external load vector from hydrodynamic loads and buoyancy are calculated considering the submerged element length. In this case, linear distribution is assumed from submerged element end to the intersection point between the element and the sea surface.</p>
<p>For partly submerged cross-sections, special option, the contributions from hydrodynamic loading are calculated based on numerical integration of <a class="el" href="a00004.html#tm_eq_2_59">Eq. (2.59)</a>. In this case linear displacement functions are used for all degrees of freedom.</p>
<p>Note that hydrodynamic load contributions proportional to structural acceleration and velocity, i.e. added mass and hydrodynamic damping, are included in the mass and the damping matrices respectively.</p>
<p>Eccentricities for the gravity and buoyancy centers are handled as described in Section <a class="el" href="a00004.html#element_local_trans">Element local transformations</a>.</p>
<h3>Mass matrix</h3>
<p>The mass matrix contains contributions from the structural and hydrodynamical (added) mass according to</p>
<p class="formulaDsp">
<img class="formulaDsp" alt="\[ \boldsymbol{m} = \boldsymbol{m}^s +\boldsymbol{m}^h \tag{2.65} \]" src="form_186.png"/>
</p>
<p>where <img class="formulaInl" alt="$ \boldsymbol{m}^s $" src="form_128.png"/> is the structural mass matrix and <img class="formulaInl" alt="$ \boldsymbol{m}^h $" src="form_129.png"/> is the added mass matrix.</p>
<p>The structural mass matrix is calculated according to</p>
<p class="formulaDsp">
<img class="formulaDsp" alt="\[ \begin{array}{l} \boldsymbol{m}^s_{uu} = m^s \displaystyle\int_{L_0} \boldsymbol{N}^T_u \boldsymbol{N}_u \mathrm{d}x \\ \\ \boldsymbol{m}^s_{vv} = m^s \displaystyle\int_{L_0} \boldsymbol{N}^T_v \boldsymbol{N}_v \mathrm{d}x \\ \\ \boldsymbol{m}^s_{ww} = m^s \displaystyle\int_{L_0} \boldsymbol{N}^T_w \boldsymbol{N}_w \mathrm{d}x \\ \\ \boldsymbol{m}^s_{\theta\theta} = m^s_\theta \displaystyle\int_{L_0} \boldsymbol{N}^T_\theta \boldsymbol{N}_\theta \mathrm{d}x \\ \end{array}\tag{2.66} \]" src="form_187.png"/>
</p>
<p>where <img class="formulaInl" alt="$ m^s$" src="form_188.png"/> is the structural mass per unit length and <img class="formulaInl" alt="$ m^s_\theta$" src="form_189.png"/> is the mass moment of inertia for rotation about the element x axis per unit length.</p>
<p>The added mass matrix is calculated according to</p>
<p><a class="anchor" id="tm_eq_2_67"></a> </p>
<p class="formulaDsp">
<img class="formulaDsp" alt="\[ \begin{array}{l} \boldsymbol{m}^h_{uu} = \displaystyle\int_{L_0} \boldsymbol{N}^T_u m^h_x \boldsymbol{N}_u \mathrm{d}x \\ \\ \boldsymbol{m}^h_{vv} = \displaystyle\int_{L_0} \boldsymbol{N}^T_v m^h_y \boldsymbol{N}_v \mathrm{d}x \\ \\ \boldsymbol{m}^h_{ww} = \displaystyle\int_{L_0} \boldsymbol{N}^T_w m^h_z \boldsymbol{N}_w \mathrm{d}x \\ \\ \boldsymbol{m}^h_{\theta\theta} = \displaystyle\int_{L_0} \boldsymbol{N}^T_\theta m^h_\theta \boldsymbol{N}_\theta \mathrm{d}x \end{array}\tag{2.67} \]" src="form_190.png"/>
</p>
<p>where <img class="formulaInl" alt="$ m_x^h$" src="form_191.png"/>, <img class="formulaInl" alt="$ m_y^h$" src="form_192.png"/> and <img class="formulaInl" alt="$ m_z^h$" src="form_193.png"/>, are added mass per unit length and <img class="formulaInl" alt="$ m^h_\theta$" src="form_194.png"/> is added mass due to rotation around the local x-axis.</p>
<p>For elements penetrating the sea surface only the submerged part gives contribution to the added mass matrix.</p>
<p>The mass matrix can alternatively be calculated using a lumped mass formulation. Due to direction dependency the mass matrix will contain off-diagonal terms after transformation to global system. To obtain a diagonal global mass matrix, off-diagonal terms are then simply added to the diagonal.</p>
<p>For floating partly submerged cross-section (special option), the added mass may vary over the element length. Consequently, the added mass matrix is calculated by numerical integration of <a class="el" href="a00004.html#tm_eq_2_67">Eq. (2.67)</a>. This option is restricted to consistent formulation.</p>
<p>Mass center eccentricities are handled as described in Section <a class="el" href="a00004.html#element_local_trans">Element local transformations</a>.</p>
<h3>Damping matrix</h3>
<p>The damping matrix consists of contributions from structural damping, <img class="formulaInl" alt="$ \boldsymbol{c}^s $" src="form_195.png"/>, and hydrodynamic damping, <img class="formulaInl" alt="$\boldsymbol{c}^h $" src="form_196.png"/>.</p>
<p class="formulaDsp">
<img class="formulaDsp" alt="\[ \boldsymbol{c} =\boldsymbol{c}^s + \boldsymbol{c}^h \tag{2.68} \]" src="form_197.png"/>
</p>
<p>For description of structural damping, confer <a class="el" href="a00010.html#tm_dynamic_structural">Structural Damping Models</a> in <a class="el" href="a00010.html#dynamic_time">Dynamic Time Domain Analysis</a>.</p>
<p>Hydrodynamic damping is only considered for floating partly submerged cross-sections (special option). The hydrodynamic damping matrix is calculated according to</p>
<p class="formulaDsp">
<img class="formulaDsp" alt="\[ \begin{array}{l} \boldsymbol{c}_{uu} = 0 \\ \\ \boldsymbol{c}_{vv} = \displaystyle\int_{L_0} \boldsymbol{N}^T_vc^h_y(x) \boldsymbol{N}_v \mathrm{d}x \\ \\ \boldsymbol{c}_{ww} = \displaystyle\int_{L_0} \boldsymbol{N}^T_wc^h_z(x) \boldsymbol{N}_w \mathrm{d}x \\ \\ \boldsymbol{c}_{\theta\theta} = \displaystyle\int_{L_0} \boldsymbol{N}^T_\theta c^h_\theta(x) \boldsymbol{N}_\theta \mathrm{d}x \\ \\ \end{array}\tag{2.69} \]" src="form_198.png"/>
</p>
<p>where <img class="formulaInl" alt="$ c_y^h$" src="form_199.png"/> and <img class="formulaInl" alt="$ c_z^h$" src="form_200.png"/> are hydrodynamic damping per unit length due to displacements in y- and z-directions and <img class="formulaInl" alt="$ c_\theta^h$" src="form_201.png"/> is hydrodynamic damping due to rotation around local x-axis.</p>
<p>As indicated the damping may vary along the element. Thus, the damping matrix is calculated based on numerical integration.</p>
<h3>Coupled bending and torsion model</h3>
<p>The governing equations for the coupled bending and torsion model is developed by representing the curvature in terms of a second order approximation of the deformational motion. The same second order approximation of the deformational motion as used by Battini (2002) is applied, which when introduced into the longitudinal Green strain definition gives the following curvature components,</p>
<p><a class="anchor" id="tm_eq_2_bt1"></a> </p>
<p class="formulaDsp">
<img class="formulaDsp" alt="\[ \begin{array}{l} \kappa_x = \theta_{x,x} + \frac{1}{2} [ \theta_{y,x}\theta_z - \theta_y \theta_{z,x}] \\ \\ \kappa_y = \theta_{y,x} + \frac{1}{2} [ \theta_{z,x}\theta_x - \theta_z \theta_{x,x}] \\ \\ \kappa_z = \theta_{z,x} + \frac{1}{2} \left[ \theta_{x,x}\theta_y - \theta_x \theta_{y,x} \right] \\ \\ \end{array} \]" src="form_202.png"/>
</p>
<p>which can be expressed on matrix format as,</p>
<p><a class="anchor" id="tm_eq_2_bt2"></a> </p>
<p class="formulaDsp">
<img class="formulaDsp" alt="\[ \boldsymbol{\kappa} = \left[ \boldsymbol{I}_3 - \frac{1}{2} \boldsymbol{W}\left(\boldsymbol{\theta}\right) \right] \boldsymbol{\theta}_{,x} \qquad \boldsymbol{\theta} = \begin{bmatrix} \theta_{x} \\ \theta_{y} \\ \theta_{z} \end{bmatrix} \quad \boldsymbol{\kappa} = \begin{bmatrix} \kappa_{x} \\ \kappa_{y} \\ \kappa_{z} \end{bmatrix} \quad \boldsymbol{W}\left(\boldsymbol{\theta}\right) = \begin{bmatrix} 0 &amp; -\theta_z &amp; \theta_y \\ \theta_z &amp; 0 &amp; -\theta_x \\ -\theta_y &amp; \theta_x &amp; 0 \\ \end{bmatrix} \]" src="form_203.png"/>
</p>
<p>where the bending curvature components about the co-rotated <img class="formulaInl" alt="$y$" src="form_204.png"/> and <img class="formulaInl" alt="$z$" src="form_205.png"/> axes are denoted <img class="formulaInl" alt="$\kappa_y$" src="form_206.png"/> and <img class="formulaInl" alt="$\kappa_x$" src="form_207.png"/>, respectively, and <img class="formulaInl" alt="$\kappa_x$" src="form_207.png"/> is the torsion. The interpolations used for the cross-section rotation vector <img class="formulaInl" alt="$\boldsymbol{\theta}$" src="form_208.png"/> and its longitudinal gradient <img class="formulaInl" alt="$\boldsymbol{\theta}_{,x}$" src="form_209.png"/> are based on <a class="el" href="a00004.html#tm_eq_2_51">Eq. (2.51)</a> and <a class="el" href="a00004.html#tm_eq_2_52">Eq. (2.52)</a>,</p>
<p><a class="anchor" id="tm_eq_2_bt3"></a> </p>
<p class="formulaDsp">
<img class="formulaDsp" alt="\[ \boldsymbol{\theta} = \boldsymbol{H}_{\theta} \boldsymbol{v} = \begin{bmatrix} 0 &amp; 0 &amp; 0 &amp; 0 &amp; 0 &amp; 0 &amp; 0 &amp; 0 &amp; 0 &amp; 0 &amp; \boldsymbol{N}_{\theta} &amp; \\ 0 &amp; 0 &amp; 0 &amp; 0 &amp; 0 &amp; 0 &amp; -\boldsymbol{N}_{w,x} &amp; &amp; &amp; &amp; 0 &amp; 0 \\ 0 &amp; 0 &amp; \boldsymbol{N}_{v,x} &amp; &amp; &amp; &amp; 0 &amp; 0 &amp; 0 &amp; 0 &amp; 0 &amp; 0 \end{bmatrix} \begin{bmatrix} v_{x1} \\ v_{x2} \\ v_{y1} \\ \theta_{z1} \\ v_{y2} \\ \theta_{z2}\\ v_{z1} \\ \theta_{y1} \\ v_{z2} \\ \theta_{y2} \\ \theta_{x1} \\ \theta_{x2} \end{bmatrix} \]" src="form_210.png"/>
</p>
<p><a class="anchor" id="tm_eq_2_bt3"></a> </p>
<p class="formulaDsp">
<img class="formulaDsp" alt="\[ \boldsymbol{\theta}_{,x} = \boldsymbol{H}_{\theta,x} \boldsymbol{v} \]" src="form_211.png"/>
</p>
<p>where the cross-section rotation interpolation matrix is denoted <img class="formulaInl" alt="$\boldsymbol{H}_{\theta}$" src="form_212.png"/> to avoid confusion with <img class="formulaInl" alt="$\boldsymbol{N}_{\theta}$" src="form_213.png"/>. The virtual curvature is selected as,</p>
<p><a class="anchor" id="tm_eq_2_bt4"></a> </p>
<p class="formulaDsp">
<img class="formulaDsp" alt="\[ \delta \boldsymbol{\kappa} = \left[ \boldsymbol{I}_3 - \frac{1}{2} \boldsymbol{W}\left(\boldsymbol{\theta}\right) \right] \delta \boldsymbol{\theta}_{,x} \]" src="form_214.png"/>
</p>
<p>where the term <img class="formulaInl" alt="$-\frac{1}{2}\boldsymbol{W}\left(\delta\boldsymbol{\theta}\right)\boldsymbol{\theta}_{,x}$" src="form_215.png"/> is discarded because it introduces spurious internal loads for the standard co-rotational beam formulation. The internal virtual work is given by the left-hand side of <a class="el" href="a00004.html#tm_eq_2_12">Eq. (2.12)</a>, which for the virtual work contributions due to bending and torsion can be expressed as follows for a single beam element,</p>
<p><a class="anchor" id="tm_eq_2_bt5"></a> </p>
<p class="formulaDsp">
<img class="formulaDsp" alt="\[ \delta W_{bt} = \int_{L_0} \delta \boldsymbol{\kappa}_{,x}^T \boldsymbol{C} \boldsymbol{\kappa}_{,x} \mathrm{d}x \qquad \boldsymbol{C} = \begin{bmatrix} GI_{\theta} &amp; 0 &amp; 0 \\ 0 &amp; B_{zT} &amp; 0 \\ 0 &amp; 0 &amp; B_{yT} \end{bmatrix} \]" src="form_216.png"/>
</p>
<p>in which the co-rotated <img class="formulaInl" alt="$y$" src="form_204.png"/> and <img class="formulaInl" alt="$z$" src="form_205.png"/> axes are assumed to coincide with the principal axes for the second area moments of the cross-section. Inserting the curvature vector <img class="formulaInl" alt="$\boldsymbol{\kappa}$" src="form_217.png"/> and the virtual curvature vector <img class="formulaInl" alt="$\delta\boldsymbol{\kappa}$" src="form_218.png"/>, and expanding the matrix products yield,</p>
<p><a class="anchor" id="tm_eq_2_bt6"></a> </p>
<p class="formulaDsp">
<img class="formulaDsp" alt="\[ \delta W_{bt} = \int_{L_0} \delta \boldsymbol{\theta}_{,x}^T \left[ \boldsymbol{C} + \frac{1}{2} \boldsymbol{W}\left(\boldsymbol{\theta}\right)\boldsymbol{C} - \frac{1}{2} \boldsymbol{C}\boldsymbol{W}\left(\boldsymbol{\theta}\right) - \frac{1}{4} \boldsymbol{W}\left(\boldsymbol{\theta}\right) \boldsymbol{C} \boldsymbol{W}\left(\boldsymbol{\theta}\right) \right] \boldsymbol{\theta}_{,x} \mathrm{d}x \]" src="form_219.png"/>
</p>
<p>where the first term inside the bracket represents the standard bending and torsion contributions, see <a class="el" href="a00004.html#tm_eq_2_53">Eq. (2.53)</a>. The fourth term is neglected because it is one order less than the second and third terms. The internal load vector representing coupling between bending and torsion is therefore expressed as follows,</p>
<p><a class="anchor" id="tm_eq_2_bt7"></a> </p>
<p class="formulaDsp">
<img class="formulaDsp" alt="\[ \boldsymbol{S}_{vw\theta} = \frac{1}{2} \int_{L_0} \boldsymbol{H}_{\theta,x}^T \left[ \boldsymbol{W}\left(\boldsymbol{\theta}\right)\boldsymbol{C} - \boldsymbol{C}\boldsymbol{W}\left(\boldsymbol{\theta}\right) \right] \boldsymbol{H}_{\theta,x} \boldsymbol{v} \mathrm{d}x \]" src="form_220.png"/>
</p>
<p>The stiffness matrix contribution due to coupling between bending and torsion is neglected because it does not significantly affect the convergence rate of the solution procedure.</p>
<p><a class="anchor" id="deformational_displacements"></a></p>
<h3>Deformational displacements</h3>
<p>In the following the procedure that is used to determine the deformational displacements of the element referred to the <img class="formulaInl" alt="$ C_{0n}$" src="form_144.png"/> configuration is described. Before doing so, a few additional coordinate systems need to be defined.</p>
<p>The figure <a class="el" href="a00004.html#Deformed_beam_element">Deformed beam element</a> below shows the beam element in deformed configuration with the co-rotated reference configuration <img class="formulaInl" alt="$ C_{0n}$" src="form_144.png"/> close to <img class="formulaInl" alt="$ C_{n}$" src="form_221.png"/>. In addition to the two coordinate systems with base vectors <img class="formulaInl" alt="$ \boldsymbol{I}_i $" src="form_222.png"/> and <img class="formulaInl" alt="$ \boldsymbol{\bar{i}}_i $" src="form_223.png"/>, two more coordinate systems are used. These are shown in <a class="el" href="a00004.html#Deformed_beam_element">Deformed beam element</a> and are represented by the base vectors <img class="formulaInl" alt="$ \boldsymbol{i}_i^0 $" src="form_224.png"/> and <img class="formulaInl" alt="$ \boldsymbol{i}_i $" src="form_225.png"/> . The first one is a coordinate system that follows the deformation of the beam end. There is one such system for each beam end, however, only the coordinate system at end "b" is shown in <a class="el" href="a00004.html#Deformed_beam_element">Deformed beam element</a>. The coordinate system with base vectors <img class="formulaInl" alt="$ \boldsymbol{\bar{i}}_i $" src="form_223.png"/> gives the orientation of the ghost reference configuration <img class="formulaInl" alt="$ C_{0n}$" src="form_144.png"/>. The figure <a class="el" href="a00004.html#Deformed_beam_element">Deformed beam element</a> also illustrates that the nodes may be eccentrically connected to the beam ends. Note that the transformation between the beam end coordinate system and the corresponding nodal point system <img class="formulaInl" alt="$ \boldsymbol{\bar{i}}_i $" src="form_223.png"/> does not change during deformation.</p>
<p><a class="anchor" id="Deformed_beam_element"></a></p>
<div class="image">
<img src="RIFLEX_TheoryManual_42_v0_316.jpg" alt="RIFLEX_TheoryManual_42_v0_316.jpg"/>
<div class="caption">
Deformed beam element.</div></div>
<p>The translations are easily found as</p>
<p><a class="anchor" id="tm_eq_2_70"></a> </p>
<p class="formulaDsp">
<img class="formulaDsp" alt="\[ \begin{array}{l} v_{x1} = v_{y1} = v_{z1} = v_{y2} = v_{z2} = 0 \\ \\ v_{x2} = L - L_0 \end{array}\tag{2.70} \]" src="form_226.png"/>
</p>
<p>The orientation of the beam end coordinate system may be expressed by the following transformation</p>
<p><a class="anchor" id="tm_eq_2_71"></a> </p>
<p class="formulaDsp">
<img class="formulaDsp" alt="\[ \boldsymbol{i}^0_i = \boldsymbol{\bar{T}}^0_{ik} \boldsymbol{I}_k \tag{2.71} \]" src="form_227.png"/>
</p>
<p>where <img class="formulaInl" alt="$ \boldsymbol{\bar{T}}^0_{ik} $" src="form_228.png"/> is different at the two beam ends. The transformation between the element coordinate system ( <img class="formulaInl" alt="$ C_{0n}$" src="form_144.png"/>-configuration) and the global system may stated as</p>
<p><a class="anchor" id="tm_eq_2_72"></a> </p>
<p class="formulaDsp">
<img class="formulaDsp" alt="\[ \boldsymbol{i}_i = \boldsymbol{T}_{ij} \boldsymbol{I}_j\tag{2.72} \]" src="form_229.png"/>
</p>
<p>The deformational rotation at the beam ends now appears through the rotation matrix between the <img class="formulaInl" alt="$ \boldsymbol{i}^0_i $" src="form_230.png"/>-system in configuration <img class="formulaInl" alt="$ C_n$" src="form_231.png"/> and the <img class="formulaInl" alt="$ \boldsymbol{i}_i$" src="form_232.png"/>-system in the <img class="formulaInl" alt="$ C_{0n}$" src="form_144.png"/>-configuration. By inverting <a class="el" href="a00004.html#tm_eq_2_72">Eq. (2.72)</a> and inserting the result into <a class="el" href="a00004.html#tm_eq_2_71">Eq. (2.71)</a> one gets</p>
<p><a class="anchor" id="tm_eq_2_73"></a> </p>
<p class="formulaDsp">
<img class="formulaDsp" alt="\[ \boldsymbol{i}^0_i = \boldsymbol{\bar{T}}^0_{ij} \boldsymbol{T}_{kj} \boldsymbol{i}_k \tag{2.73} \]" src="form_233.png"/>
</p>
<p>This result is analogous to the transformation expressed by <img class="formulaInl" alt="$ \boldsymbol{\tilde{T}}_{kj} $" src="form_234.png"/> in <a class="el" href="a00004.html#tm_eq_2_23">Eq. (2.23)</a>. By using the result from <a class="el" href="a00004.html#tm_eq_2_25">Eq. (2.25)</a> and comparing this to <a class="el" href="a00004.html#tm_eq_2_73">Eq. (2.73)</a>, one finds the following expressions for the beam end rotations</p>
<p><a class="anchor" id="tm_eq_2_74"></a> </p>
<p class="formulaDsp">
<img class="formulaDsp" alt="\[ \begin{array}{l} \theta_x = \bar{T}^0_{21}T_{31} + \bar{T}^0_{22}T_{32} + \bar{T}^0_{23}T_{33} \\ \\ \theta_y = \bar{T}^0_{31}T_{11} + \bar{T}^0_{32}T_{12} + \bar{T}^0_{33}T_{13} \\ \\ \theta_z = \bar{T}^0_{11}T_{21} + \bar{T}^0_{12}T_{22} + \bar{T}^0_{13}T_{23} \end{array}\tag{2.74} \]" src="form_235.png"/>
</p>
<p>There is one such relation for each beam end. The element displacement vector referred to the local base vectors <img class="formulaInl" alt="$ \boldsymbol{\bar{i}}_i $" src="form_223.png"/> is now determined by the results in <a class="el" href="a00004.html#tm_eq_2_70">Eq. (2.70)</a> and <a class="el" href="a00004.html#tm_eq_2_74">Eq. (2.74)</a>.</p>
<p><a class="anchor" id="tm_eq_2_74b"></a> </p>
<p class="formulaDsp">
<img class="formulaDsp" alt="\[ \boldsymbol{v} = \begin{bmatrix} v_{x1} &amp; v_{x2} &amp; v_{y1} &amp; \theta_{z1} &amp; v_{y2}&amp; \theta_{z2}&amp; v_{z1} &amp; \theta_{y1} &amp; v_{z2}&amp; \theta_{y2} &amp; \theta_{x1} &amp; \theta_{x2} \end{bmatrix}^T \]" src="form_236.png"/>
</p>
<p>where it is emphasized that the components of <img class="formulaInl" alt="$ \boldsymbol{v} $" src="form_237.png"/> refer to the nodal points.</p>
<p><a class="anchor" id="element_local_trans"></a></p>
<h3>Element local transformations</h3>
<p>The element displacement vector is re-organized as follows prior to system equation assembly</p>
<p><a class="anchor" id="tm_eq_2_bt3"></a> </p>
<p class="formulaDsp">
<img class="formulaDsp" alt="\[ \boldsymbol{\bar{v}} = \begin{bmatrix} v_{x1} &amp; v_{y1} &amp; v_{z1} &amp; \theta_{x1} &amp; \theta_{y1} &amp; \theta_{z1}&amp; v_{x2} &amp; v_{y2} &amp; v_{z2} &amp; \theta_{x2} &amp; \theta_{y2} &amp; \theta_{z2} &amp; \end{bmatrix}^T \]" src="form_238.png"/>
</p>
<p>which formally is expressed by</p>
<p><a class="anchor" id="tm_eq_2_74d"></a> </p>
<p class="formulaDsp">
<img class="formulaDsp" alt="\[ \boldsymbol{{A}} = \begin{bmatrix} 1 &amp; 0 &amp; 0 &amp; 0 &amp; 0 &amp; 0 &amp; 0 &amp; 0 &amp; 0 &amp; 0 &amp; 0 &amp; 0 \\ 0 &amp; 0 &amp; 1 &amp; 0 &amp; 0 &amp; 0 &amp; 0 &amp; 0 &amp; 0 &amp; 0 &amp; 0 &amp; 0 \\ 0 &amp; 0 &amp; 0 &amp; 0 &amp; 0 &amp; 0 &amp; 1 &amp; 0 &amp; 0 &amp; 0 &amp; 0 &amp; 0 \\ 0 &amp; 0 &amp; 0 &amp; 0 &amp; 0 &amp; 0 &amp; 0 &amp; 0 &amp; 0 &amp; 0 &amp; 1 &amp; 0 \\ 0 &amp; 0 &amp; 0 &amp; 0 &amp; 0 &amp; 0 &amp; 0 &amp; 1 &amp; 0 &amp; 0 &amp; 0 &amp; 0 \\ 0 &amp; 0 &amp; 0 &amp; 1 &amp; 0 &amp; 0 &amp; 0 &amp; 0 &amp; 0 &amp; 0 &amp; 0 &amp; 0 \\ 0 &amp; 1 &amp; 0 &amp; 0 &amp; 0 &amp; 0 &amp; 0 &amp; 0 &amp; 0 &amp; 0 &amp; 0 &amp; 0 \\ 0 &amp; 0 &amp; 0 &amp; 0 &amp; 1 &amp; 0 &amp; 0 &amp; 0 &amp; 0 &amp; 0 &amp; 0 &amp; 0 \\ 0 &amp; 0 &amp; 0 &amp; 0 &amp; 0 &amp; 0 &amp; 0 &amp; 0 &amp; 1 &amp; 0 &amp; 0 &amp; 0 \\ 0 &amp; 0 &amp; 0 &amp; 0 &amp; 0 &amp; 0 &amp; 0 &amp; 0 &amp; 0 &amp; 0 &amp; 0 &amp; 1 \\ 0 &amp; 0 &amp; 0 &amp; 0 &amp; 0 &amp; 0 &amp; 0 &amp; 0 &amp; 0 &amp; 1 &amp; 0 &amp; 0 \\ 0 &amp; 0 &amp; 0 &amp; 0 &amp; 0 &amp; 1 &amp; 0 &amp; 0 &amp; 0 &amp; 0 &amp; 0 &amp; 0 \\ \end{bmatrix} \]" src="form_239.png"/>
</p>
<p>applied to the element load vectors</p>
<p><a class="anchor" id="tm_eq_2_74e"></a> </p>
<p class="formulaDsp">
<img class="formulaDsp" alt="\[ \boldsymbol{\check{S}} = \boldsymbol{{A}}^T \!\boldsymbol{S} \qquad \boldsymbol{S} = \begin{bmatrix} \boldsymbol{S}_u &amp; \boldsymbol{S}_v &amp; \boldsymbol{S}_w &amp; \boldsymbol{S}_{\theta} \end{bmatrix}^T \]" src="form_240.png"/>
</p>
<p>and the element matrices</p>
<p><a class="anchor" id="tm_eq_2_74f"></a> </p>
<p class="formulaDsp">
<img class="formulaDsp" alt="\[ \begin{array}{l} \boldsymbol{\check{k}} = \boldsymbol{{A}}^T \!\boldsymbol{k} \,\boldsymbol{{A}} \\ \\ \boldsymbol{\check{c}} = \boldsymbol{{A}}^T \!\boldsymbol{c} \,\boldsymbol{{A}} \\ \\ \boldsymbol{\check{m}} = \boldsymbol{{A}}^T \!\boldsymbol{m} \,\boldsymbol{{A}} \\ \\ \end{array} \]" src="form_241.png"/>
</p>
<p><a class="anchor" id="gen_nonsym_crossec"></a> </p>
<div class="image">
<img src="CRS7_allecc.svg" alt="CRS7_allecc.svg"/>
<div class="caption">
General cross-section</div></div>
<p>The constitutive response of general cross-sections is expressed in the principal axes coordinates to allow for direct use of the shear stiffness model in Bell (1987). The principal axes V and W for the second area moment are formally determined by the requirement <img class="formulaInl" alt="$ \int_A V \, W \,\,\mathrm{d}A = 0 $" src="form_242.png"/> where <img class="formulaInl" alt="$ A $" src="form_243.png"/> is the cross-section area. The angle of the principal <img class="formulaInl" alt="$V$" src="form_244.png"/>-axis measured relative to the beam element <img class="formulaInl" alt="$Y$" src="form_245.png"/>-axis is denoted <img class="formulaInl" alt="$ \theta $" src="form_246.png"/>, see figure <a class="el" href="a00004.html#gen_nonsym_crossec">General cross-section</a>. The transformation between the principal axes and the beam element <img class="formulaInl" alt="$Y$" src="form_245.png"/>- and <img class="formulaInl" alt="$Z$" src="form_247.png"/>-axis is defined by</p>
<p><a class="anchor" id="tm_eq_2_74g"></a> </p>
<p class="formulaDsp">
<img class="formulaDsp" alt="\[ \boldsymbol{\hat{T}}_{\theta} = \begin{bmatrix} \boldsymbol{T}_{\theta} &amp; \boldsymbol{0}_{3\times 3} &amp; \boldsymbol{0}_{3\times 3} &amp; \boldsymbol{0}_{3\times 3} \\ \boldsymbol{0}_{3\times 3} &amp; \boldsymbol{T}_{\theta} &amp; \boldsymbol{0}_{3\times 3} &amp; \boldsymbol{0}_{3\times 3} \\ \boldsymbol{0}_{3\times 3} &amp; \boldsymbol{0}_{3\times 3} &amp; \boldsymbol{T}_{\theta} &amp; \boldsymbol{0}_{3\times 3} \\ \boldsymbol{0}_{3\times 3} &amp; \boldsymbol{0}_{3\times 3} &amp; \boldsymbol{0}_{3\times 3} &amp; \boldsymbol{T}_{\theta} \\ \end{bmatrix} \qquad \boldsymbol{T}_{\theta} = \begin{bmatrix} 1 &amp; 0 &amp; 0 \\ 0 &amp; \cos(\theta) &amp; \sin(\theta) \\ 0 &amp; -\sin(\theta) &amp; \cos(\theta) \\ \end{bmatrix} \]" src="form_248.png"/>
</p>
<p>where <img class="formulaInl" alt="$\boldsymbol{I}_{3\times 3}$" src="form_249.png"/> and <img class="formulaInl" alt="$\boldsymbol{0}_{3\times 3}$" src="form_250.png"/> denote the identity and zero matrix of dimension 3 times 3. The transformation of the element internal load vector reads as follows</p>
<p><a class="anchor" id="tm_eq_2_74h"></a> </p>
<p class="formulaDsp">
<img class="formulaDsp" alt="\[ \boldsymbol{\hat{S}} = \boldsymbol{\hat{T}}_{\theta}^T \boldsymbol{\check{S}} \]" src="form_251.png"/>
</p>
<p>while the material stiffness matrix is transformed according to</p>
<p><a class="anchor" id="tm_eq_2_74i"></a> </p>
<p class="formulaDsp">
<img class="formulaDsp" alt="\[ \boldsymbol{\hat{k}} = \boldsymbol{\hat{T}}_{\theta}^T \boldsymbol{\check{k}} \, \boldsymbol{\hat{T}}_{\theta} \]" src="form_252.png"/>
</p>
<p>The shear forces and torsion moments in <img class="formulaInl" alt="$\boldsymbol{\hat{S}}$" src="form_253.png"/> refer to the shear center, while the axial force and the bending moments in <img class="formulaInl" alt="$\boldsymbol{\hat{S}}$" src="form_253.png"/> refer to the area center. For general cross-sections the shear and area centers may be eccentrically located. This eccentricity is accounted for by the transformation matrix</p>
<p><a class="anchor" id="tm_eq_2_74k"></a> </p>
<p class="formulaDsp">
<img class="formulaDsp" alt="\[ \boldsymbol{T}_{sa} = \begin{bmatrix} \boldsymbol{I}_{3\times 3} &amp; \boldsymbol{e}_{sa} &amp; \boldsymbol{0}_{3\times 3} &amp; \boldsymbol{0}_{3\times 3} \\ \boldsymbol{0}_{3\times 3} &amp; \boldsymbol{I}_{3\times 3} &amp; \boldsymbol{0}_{3\times 3} &amp; \boldsymbol{0}_{3\times 3} \\ \boldsymbol{0}_{3\times 3} &amp; \boldsymbol{0}_{3\times 3} &amp; \boldsymbol{I}_{3\times 3} &amp; \boldsymbol{e}_{sa} \\ \boldsymbol{0}_{3\times 3} &amp; \boldsymbol{0}_{3\times 3} &amp; \boldsymbol{0}_{3\times 3} &amp; \boldsymbol{I}_{3\times 3} \\ \end{bmatrix} \qquad \boldsymbol{e}_{sa} = \begin{bmatrix} 0 &amp; Z_a &amp; -Y_a \\ -Z_s &amp; 0 &amp; 0 \\ Y_s &amp; 0 &amp; 0 \\ \end{bmatrix} \]" src="form_254.png"/>
</p>
<p>where <img class="formulaInl" alt="$ (Y_a,Z_a) $" src="form_255.png"/> and <img class="formulaInl" alt="$ (Y_s,Z_s) $" src="form_256.png"/> are respectively the coordinates of the area and shear centers expressed in the beam element system, see figure <a class="el" href="a00004.html#gen_nonsym_crossec">General cross-section</a>. The eccentricity transformation applies to the element internal load vector and material stiffness matrix according to</p>
<p><a class="anchor" id="tm_eq_2_74m"></a> </p>
<p class="formulaDsp">
<img class="formulaDsp" alt="\[ \boldsymbol{\bar{S}} = \boldsymbol{T}_{sa}^T \boldsymbol{\hat{S}} \]" src="form_257.png"/>
</p>
<p><a class="anchor" id="tm_eq_2_74n"></a> </p>
<p class="formulaDsp">
<img class="formulaDsp" alt="\[ \boldsymbol{\bar{k}} = \boldsymbol{T}_{sa}^T \boldsymbol{\hat{k}} \, \boldsymbol{T}_{sa} \]" src="form_258.png"/>
</p>
<p>It is emphasized that <img class="formulaInl" alt="$ \boldsymbol{\bar{S}} $" src="form_259.png"/> and <img class="formulaInl" alt="$ \boldsymbol{\bar{k}} $" src="form_260.png"/> are energy-conjugate to the element displacement vector <img class="formulaInl" alt="$ \boldsymbol{\bar{v}} $" src="form_261.png"/> which refers to the nodal points of the beam element.</p>
<p>Similarly, general cross-sections may have an eccentric mass center that must be accounted for by the following transformation of the element mass matrix</p>
<p><a class="anchor" id="tm_eq_2_74o"></a> </p>
<p class="formulaDsp">
<img class="formulaDsp" alt="\[ \boldsymbol{T}_{m} = \begin{bmatrix} \boldsymbol{I}_{3\times 3} &amp; \boldsymbol{e}_{m} &amp; \boldsymbol{0}_{3\times 3} &amp; \boldsymbol{0}_{3\times 3} \\ \boldsymbol{0}_{3\times 3} &amp; \boldsymbol{I}_{3\times 3} &amp; \boldsymbol{0}_{3\times 3} &amp; \boldsymbol{0}_{3\times 3} \\ \boldsymbol{0}_{3\times 3} &amp; \boldsymbol{0}_{3\times 3} &amp; \boldsymbol{I}_{3\times 3} &amp; \boldsymbol{e}_{m} \\ \boldsymbol{0}_{3\times 3} &amp; \boldsymbol{0}_{3\times 3} &amp; \boldsymbol{0}_{3\times 3} &amp; \boldsymbol{I}_{3\times 3} \\ \end{bmatrix} \qquad \boldsymbol{e}_{m} = \begin{bmatrix} 0 &amp; Z_m &amp; -Y_m \\ -Z_m &amp; 0 &amp; 0 \\ Y_m &amp; 0 &amp; 0 \\ \end{bmatrix} \]" src="form_262.png"/>
</p>
<p><a class="anchor" id="tm_eq_2_74q"></a> </p>
<p class="formulaDsp">
<img class="formulaDsp" alt="\[ \boldsymbol{\bar{m}} = \boldsymbol{T}_{m}^T \boldsymbol{\check{m}} \, \boldsymbol{T}_{m} \]" src="form_263.png"/>
</p>
<p>where <img class="formulaInl" alt="$ (Y_m,Z_m) $" src="form_264.png"/> is the mass center expressed in the beam element coordinates. The eccentricity transformation applies also to the load vector containing the gravity and buoyancy force</p>
<p><a class="anchor" id="tm_eq_2_74q"></a> </p>
<p class="formulaDsp">
<img class="formulaDsp" alt="\[ \boldsymbol{\bar{S}} = \boldsymbol{T}_{m}^T \boldsymbol{\check{S}} \]" src="form_265.png"/>
</p>
<p>in which the buoyancy and mass centers are assumed to be coincident.</p>
<h3>Equilibrium projection of load vector</h3>
<p>The internal load vector in Section <a class="el" href="a00004.html#internal_loads">Internal loads</a> is in equilibrium with respect to the undeformed element. In nonlinear analysis the load vector should be in equilibrium with respect to deformed element, which is particularly important if the deformed length <img class="formulaInl" alt="$L$" src="form_266.png"/> is significantly different from the initial length <img class="formulaInl" alt="$L_0$" src="form_267.png"/>. The equilibrium projection of the internal load vector is expressed as follows</p>
<p><a class="anchor" id="tm_eq_2_74q"></a> </p>
<p class="formulaDsp">
<img class="formulaDsp" alt="\[ \boldsymbol{\tilde{S}} = \boldsymbol{{P}}^T \boldsymbol{\bar{S}} \]" src="form_268.png"/>
</p>
<p>where the projection matrix can be shown to be given by</p>
<p><a class="anchor" id="tm_eq_2_74p"></a> </p>
<p class="formulaDsp">
<img class="formulaDsp" alt="\[ \boldsymbol{{P}}^T = \begin{bmatrix} \frac{1}{2} &amp; 0 &amp; 0 &amp; 0 &amp; 0 &amp; 0 &amp; -\frac{1}{2} &amp; 0 &amp; 0 &amp; 0 &amp; 0 &amp; 0 \\ 0 &amp; 0 &amp; 0 &amp; 0 &amp; 0 &amp;\frac{1}{L}&amp; 0 &amp; 0 &amp; 0 &amp; 0 &amp; 0 &amp; \frac{1}{L} \\ 0 &amp; 0 &amp; 0 &amp; 0 &amp;-\frac{1}{L}&amp; 0 &amp; 0 &amp; 0 &amp; 0 &amp; 0 &amp;-\frac{1}{L}&amp; 0 \\ 0 &amp; 0 &amp; 0 &amp; \frac{1}{2} &amp; 0 &amp; 0 &amp; 0 &amp; 0 &amp; 0 &amp; -\frac{1}{2} &amp; 0 &amp; 0 \\ 0 &amp; 0 &amp; 0 &amp; 0 &amp; 1 &amp; 0 &amp; 0 &amp; 0 &amp; 0 &amp; 0 &amp; 0 &amp; 0 \\ 0 &amp; 0 &amp; 0 &amp; 0 &amp; 0 &amp; 1 &amp; 0 &amp; 0 &amp; 0 &amp; 0 &amp; 0 &amp; 0 \\ -\frac{1}{2} &amp; 0 &amp; 0 &amp; 0 &amp; 0 &amp; 0 &amp; \frac{1}{2} &amp; 0 &amp; 0 &amp; 0 &amp; 0 &amp; 0 \\ 0 &amp; 0 &amp; 0 &amp; 0 &amp; 0 &amp;-\frac{1}{L}&amp; 0 &amp; 0 &amp; 0 &amp; 0 &amp; 0 &amp;-\frac{1}{L} \\ 0 &amp; 0 &amp; 0 &amp; 0 &amp;\frac{1}{L}&amp; 0 &amp; 0 &amp; 0 &amp; 0 &amp; 0 &amp;\frac{1}{L}&amp; 0 \\ 0 &amp; 0 &amp; 0 &amp; -\frac{1}{2} &amp; 0 &amp; 0 &amp; 0 &amp; 0 &amp; 0 &amp; \frac{1}{2} &amp; 0 &amp; 0 \\ 0 &amp; 0 &amp; 0 &amp; 0 &amp; 0 &amp; 0 &amp; 0 &amp; 0 &amp; 0 &amp; 0 &amp; 1 &amp; 0 \\ 0 &amp; 0 &amp; 0 &amp; 0 &amp; 0 &amp; 0 &amp; 0 &amp; 0 &amp; 0 &amp; 0 &amp; 0 &amp; 1 \\ \end{bmatrix} \]" src="form_269.png"/>
</p>
<h3>Transformation of load vector to global system</h3>
<p>Prior to assembly of the system equations, the element internal load vector <img class="formulaInl" alt="$\boldsymbol{S}^g$" src="form_270.png"/> is transformed to global coordinates according to</p>
<p><a class="anchor" id="tm_eq_2_74r"></a> </p>
<p class="formulaDsp">
<img class="formulaDsp" alt="\[ \boldsymbol{\hat{T}} = \begin{bmatrix} \boldsymbol{T} &amp; \boldsymbol{0}_{3\times 3} &amp; \boldsymbol{0}_{3\times 3} &amp; \boldsymbol{0}_{3\times 3} \\ \boldsymbol{0}_{3\times 3} &amp; \boldsymbol{T} &amp; \boldsymbol{0}_{3\times 3} &amp; \boldsymbol{0}_{3\times 3} \\ \boldsymbol{0}_{3\times 3} &amp; \boldsymbol{0}_{3\times 3} &amp; \boldsymbol{T} &amp; \boldsymbol{0}_{3\times 3} \\ \boldsymbol{0}_{3\times 3} &amp; \boldsymbol{0}_{3\times 3} &amp; \boldsymbol{0}_{3\times 3} &amp; \boldsymbol{T} \\ \end{bmatrix} \qquad \boldsymbol{T} = \begin{bmatrix} \boldsymbol{i}_1^T \\ \boldsymbol{i}_2^T \\ \boldsymbol{i}_3^T \\ \end{bmatrix} \]" src="form_271.png"/>
</p>
<p><a class="anchor" id="tm_eq_2_74s"></a> </p>
<p class="formulaDsp">
<img class="formulaDsp" alt="\[ \boldsymbol{S}^g = \boldsymbol{\hat{T}}^T \boldsymbol{{P}}^T \boldsymbol{\bar{S}} \]" src="form_272.png"/>
</p>
<p>where superscript <img class="formulaInl" alt="$ g $" src="form_273.png"/> signifies global coordinates and <img class="formulaInl" alt="$ \boldsymbol{i}_i $" src="form_225.png"/> is the element system unit vectors in global coordinates, see Section <a class="el" href="a00004.html#deformational_displacements">Deformational displacements</a>.</p>
<h3>Tangent stiffness matrix in the co-rotational framework</h3>
<p>The element tangent stiffness matrix presented in this section is at present implemented only for the CRS7 cross-section.</p>
<p>A consistent tangent stiffness matrix requires that increments of the internal load vector are related to increments of the total displacements, which enables quadratic convergence rate of Newton's method in the vicinity of the solution point. The tangent stiffness relation is derived from</p>
<p><a class="anchor" id="tm_eq_2_74t"></a> </p>
<p class="formulaDsp">
<img class="formulaDsp" alt="\[ \mathrm{d} \boldsymbol{S}^g = \mathrm{d} \boldsymbol{\hat{T}}^T \boldsymbol{P}^T \boldsymbol{\bar{S}} + \boldsymbol{\hat{T}}^T \mathrm{d} \boldsymbol{P}^T \boldsymbol{\bar{S}} + \boldsymbol{\hat{T}}^T \boldsymbol{P}^T \mathrm{d} \boldsymbol{\bar{S}} \]" src="form_274.png"/>
</p>
<p>where the semi-tangential moment contribution to stiffness from the internal moments are neglected, see Krenk (2009). The linearization of the projection matrix <img class="formulaInl" alt="$ \mathrm{d} \mathrm{\boldsymbol{P}} $" src="form_275.png"/> is also neglected as the change of element length within a step of the Newton solution procedure is assumed negligible. The tangent stiffness relation can therefore be expressed as</p>
<p><a class="anchor" id="tm_eq_2_74u"></a> </p>
<p class="formulaDsp">
<img class="formulaDsp" alt="\[ \mathrm{d} \boldsymbol{S}^g = \mathrm{d} \boldsymbol{\hat{T}}^T \boldsymbol{\tilde{S}} + \boldsymbol{\hat{T}}^T \boldsymbol{P}^T \mathrm{d} \boldsymbol{\bar{S}} \]" src="form_276.png"/>
</p>
<p>where the linearization in the first term provides the geometric stiffness matrix and is associated with rigid body rotation of the element. The second term represents change of internal loads due to deformation and can hence be expressed by the material stiffness matrix <img class="formulaInl" alt="$ \boldsymbol{\bar{k}} $" src="form_260.png"/> from Section <a class="el" href="a00004.html#element_local_trans">Element local transformations</a> as follows</p>
<p><a class="anchor" id="tm_eq_2_74v"></a> </p>
<p class="formulaDsp">
<img class="formulaDsp" alt="\[ \mathrm{d} \boldsymbol{\bar{S}} = \boldsymbol{\bar{k}} \boldsymbol{P} \mathrm{d} \boldsymbol{\bar{v}} \]" src="form_277.png"/>
</p>
<p>which gives the following material tangent stiffness matrix in global coordinates</p>
<p><a class="anchor" id="tm_eq_2_74w"></a> </p>
<p class="formulaDsp">
<img class="formulaDsp" alt="\[ \boldsymbol{k}_M^g = \boldsymbol{\hat{T}}^T \! \boldsymbol{P}^T \boldsymbol{\bar{k}} \boldsymbol{P} \boldsymbol{\hat{T}} \]" src="form_278.png"/>
</p>
<p>The linearization of the transformation matrix <img class="formulaInl" alt="$ \boldsymbol{{T}} $" src="form_279.png"/> is expressed as follows</p>
<p><a class="anchor" id="tm_eq_2_74y"></a> </p>
<p class="formulaDsp">
<img class="formulaDsp" alt="\[ \mathrm{d}\boldsymbol{T} = \frac{\partial{\boldsymbol{T}}}{\partial{\boldsymbol{\omega}}} \mathrm{d}{\boldsymbol{\omega}} = \begin{bmatrix} \boldsymbol{0}_{1 \times 3} \\ {\boldsymbol{i}_3}^T \\ -{\boldsymbol{i}_2}^T \\ \end{bmatrix} \mathrm{d}\omega_x + \begin{bmatrix} -{\boldsymbol{i}_3}^T \\ \boldsymbol{0}_{1 \times 3} \\ {\boldsymbol{i}_1}^T \\ \end{bmatrix} \mathrm{d}\omega_y + \begin{bmatrix} {\boldsymbol{i}_2}^T \\ -{\boldsymbol{i}_1}^T \\ \boldsymbol{0}_{1 \times 3} \\ \end{bmatrix} \mathrm{d}\omega_z \]" src="form_280.png"/>
</p>
<p>where <img class="formulaInl" alt="$\mathrm{d} \boldsymbol{\omega} $" src="form_281.png"/> is the incremental element rigid body rotation. The relation can be expressed in terms of the skew-symmetric matrix <img class="formulaInl" alt="$ \boldsymbol{W} $" src="form_282.png"/> and the element transformation matrix <img class="formulaInl" alt="$ \boldsymbol{T} $" src="form_283.png"/> as follows,</p>
<p><a class="anchor" id="tm_eq_2_74z"></a> </p>
<p class="formulaDsp">
<img class="formulaDsp" alt="\[ \mathrm{d}\boldsymbol{T} = \begin{bmatrix} 0 &amp; \mathrm{d}\omega_z &amp; -\mathrm{d}\omega_y \\ -\mathrm{d}\omega_z &amp; 0 &amp; \mathrm{d}\omega_x \\ \mathrm{d}\omega_y &amp; -\mathrm{d}\omega_x &amp; 0 \\ \end{bmatrix} \begin{bmatrix} {{\boldsymbol{i}}_1}^T \\ {{\boldsymbol{i}}_2}^T \\ {{\boldsymbol{i}}_3}^T \\ \end{bmatrix} = -\boldsymbol{W}(\mathrm{d}{\boldsymbol{\omega}} ) \boldsymbol{T} \]" src="form_284.png"/>
</p>
<p>and</p>
<p><a class="anchor" id="tm_eq_2_74aa"></a> </p>
<p class="formulaDsp">
<img class="formulaDsp" alt="\[ \mathrm{d}\boldsymbol{T}^T = \boldsymbol{T}^T \boldsymbol{W} (\mathrm{d}{\boldsymbol{\omega}} ) \]" src="form_285.png"/>
</p>
<p>The geometric stiffness matrix can now be established</p>
<p><a class="anchor" id="tm_eq_2_74ab"></a> </p>
<p class="formulaDsp">
<img class="formulaDsp" alt="\[ \mathrm{d}{\boldsymbol{\hat{T}}}^T \boldsymbol{\tilde{S}} = \begin{bmatrix} \mathrm{d}{\boldsymbol{T}}^T \tilde{\boldsymbol{f}}_1 \\ \mathrm{d}{\boldsymbol{T}}^T \tilde{\boldsymbol{m}}_1 \\ \mathrm{d}{\boldsymbol{T}}^T \tilde{\boldsymbol{f}}_2 \\ \mathrm{d}{\boldsymbol{T}}^T \tilde{\boldsymbol{m}}_2 \\ \end{bmatrix} \]" src="form_286.png"/>
</p>
<p>where the components of the equilibrium projected internal load vector are decomposed into contributions representing forces <img class="formulaInl" alt="$ \tilde{\boldsymbol{f}}_i $" src="form_287.png"/> and moments <img class="formulaInl" alt="$ \tilde{\boldsymbol{m}}_i $" src="form_288.png"/> at element node <img class="formulaInl" alt="$ i $" src="form_289.png"/></p>
<p><a class="anchor" id="tm_eq_2_74ac"></a> </p>
<p class="formulaDsp">
<img class="formulaDsp" alt="\[ \boldsymbol{\tilde{S}} = \begin{bmatrix} \tilde{\boldsymbol{f}}_1 \\ \tilde{\boldsymbol{m}}_1 \\ \tilde{\boldsymbol{f}}_2 \\ \tilde{\boldsymbol{m}}_2 \\ \end{bmatrix} \]" src="form_290.png"/>
</p>
<p>Introducing the expression for <img class="formulaInl" alt="$ \mathrm{d} \boldsymbol{T}^T $" src="form_291.png"/> gives</p>
<p><a class="anchor" id="tm_eq_2_74ad"></a> </p>
<p class="formulaDsp">
<img class="formulaDsp" alt="\[ \mathrm{d} \boldsymbol{\hat{T}}^T \boldsymbol{\tilde{S}} = \begin{bmatrix} \boldsymbol{T}^T \boldsymbol{W} ( \mathrm{d} \boldsymbol{\omega} ) \tilde{\boldsymbol{f}}_1 \\ \boldsymbol{T}^T \boldsymbol{W} ( \mathrm{d} \boldsymbol{\omega} ) \tilde{\boldsymbol{m}}_1 \\ \boldsymbol{T}^T \boldsymbol{W} ( \mathrm{d} \boldsymbol{\omega} ) \tilde{\boldsymbol{f}}_2 \\ \boldsymbol{T}^T \boldsymbol{W} ( \mathrm{d} \boldsymbol{\omega} ) \tilde{\boldsymbol{m}}_2 \\ \end{bmatrix} = - \boldsymbol{\hat{T}}^T \begin{bmatrix} \boldsymbol{W} ( \tilde{\boldsymbol{f}}_1 ) \\ \boldsymbol{W} ( \tilde{\boldsymbol{m}}_1 ) \\ \boldsymbol{W} ( \tilde{\boldsymbol{f}}_2 ) \\ \boldsymbol{W} ( \tilde{\boldsymbol{m}}_2 ) \\ \end{bmatrix} \mathrm{d} \boldsymbol{\omega} \]" src="form_292.png"/>
</p>
<p>The rigid body rotation of the element can be shown to be given by the following matrix relation for small increments</p>
<p><a class="anchor" id="tm_eq_2_74ae"></a> </p>
<p class="formulaDsp">
<img class="formulaDsp" alt="\[ \mathrm{d} \boldsymbol{\omega} = \boldsymbol{G}\, \boldsymbol{\hat{T}}\, \mathrm{d} \boldsymbol{\bar{v}} \qquad \boldsymbol{G} = \begin{bmatrix} 0 &amp; 0 &amp; 0 &amp; \frac{1}{2} &amp; 0 &amp; 0 &amp; 0 &amp; 0 &amp; 0 &amp; \frac{1}{2} &amp; 0 &amp; 0 \\ 0&amp; 0&amp; \frac{1}{L}&amp; 0&amp; 0&amp; 0&amp; 0&amp; 0&amp; -\frac{1}{L}&amp; 0&amp; 0&amp; 0\\ 0&amp; -\frac{1}{L}&amp; 0&amp; 0&amp; 0&amp; 0&amp; 0&amp; \frac{1}{L}&amp; 0&amp; 0&amp; 0&amp; 0\\ \end{bmatrix} \]" src="form_293.png"/>
</p>
<p>and thus</p>
<p><a class="anchor" id="tm_eq_2_74af"></a> </p>
<p class="formulaDsp">
<img class="formulaDsp" alt="\[ \mathrm{d}{\boldsymbol{\hat{T}}}^T \boldsymbol{\tilde{S}} = - \boldsymbol{\hat{T}}^T \! \begin{bmatrix} \boldsymbol{W} ( \tilde{\boldsymbol{f}}_1 ) \\ \boldsymbol{W} ( \tilde{\boldsymbol{m}}_1 ) \\ \boldsymbol{W} ( \tilde{\boldsymbol{f}}_2 ) \\ \boldsymbol{W} ( \tilde{\boldsymbol{m}}_2 ) \\ \end{bmatrix} \! \boldsymbol{G} \, \boldsymbol{\hat{T}} \mathrm{d} \boldsymbol{\bar{v}} \]" src="form_294.png"/>
</p>
<p>in which the geometric stiffness matrix in global coordinates reads,</p>
<p><a class="anchor" id="tm_eq_2_74ag"></a> </p>
<p class="formulaDsp">
<img class="formulaDsp" alt="\[ \boldsymbol{k}_G^g = - \boldsymbol{\hat{T}}^T \! \begin{bmatrix} \boldsymbol{W} ( \tilde{\boldsymbol{f}}_1 ) \\ \boldsymbol{W} ( \tilde{\boldsymbol{m}}_1 ) \\ \boldsymbol{W} ( \tilde{\boldsymbol{f}}_2 ) \\ \boldsymbol{W} ( \tilde{\boldsymbol{m}}_2 ) \\ \end{bmatrix}\! \boldsymbol{G} \, \boldsymbol{\hat{T}} \]" src="form_295.png"/>
</p>
<p><img class="formulaInl" alt="$ \boldsymbol{k}_G^g $" src="form_296.png"/> is a rank 3 matrix which adds stiffness only to the 3 rigid body rotations of the element. It is important to note that the derived geometric stiffness matrix does not add stiffness for the deformation modes of the element. The RIFLEX equation solver handles only symmetric matrices while the derived geometric stiffness matrix is non-symmetric. An artificial symmetrization is therefore applied according to</p>
<p><a class="anchor" id="tm_eq_2_74ah"></a> </p>
<p class="formulaDsp">
<img class="formulaDsp" alt="\[ \boldsymbol{k}_G^g = \frac{1}{2} (\boldsymbol{k}_G^g + {\boldsymbol{k}_M^g}^T) \]" src="form_297.png"/>
</p>
<h2><a class="anchor" id="tm_finite_implementation_connector"></a>
Connector Element</h2>
<p>The primary purpose of the connector element is to enable modelling of structural joints. The connector element description here refers to ball joints in RIFLEX. Newer flex joint elements use a master-slave formulation to eliminate fixed degrees of freedom, and allow the user to specify stiffnesses in the free degrees of freedom. The connector element is associated with 2 separate nodes which occupy exactly the same initial position in space. Forces in the connector element are introduced by specified spring stiffness related to translation and rotational degrees of freedom, see <a class="el" href="a00004.html#Connector_element">Connector element</a> below.</p>
<p><a class="anchor" id="Connector_element"></a></p>
<div class="image">
<img src="RIFLEX_TheoryManual_42_v0_333.jpg" alt="RIFLEX_TheoryManual_42_v0_333.jpg"/>
<div class="caption">
Connector element.</div></div>
<p>For application the connector element is exclusively connected to a specified adjacent beam element and used to model internal moment-free joints in the structure only. The procedure used to determine local rotations is similar to the method used for the beam element except for choice of reference system.</p>
<p>As for the beam element, a base vector system for each of the link element ends is introduced, denoted <img class="formulaInl" alt="$ \boldsymbol{i}^{0a} $" src="form_298.png"/> and <img class="formulaInl" alt="$ \boldsymbol{i}^{0b} $" src="form_299.png"/>. In <a class="el" href="a00004.html#Base_vector_systems_for_the_connector">Base vector systems for the connector element</a> only <img class="formulaInl" alt="$ \boldsymbol{i}^{0a} $" src="form_298.png"/> is presented for readability. The transformations between these systems and the corresponding nodal point systems <img class="formulaInl" alt="$ \boldsymbol{\bar{i}}^{(a)} $" src="form_300.png"/> and <img class="formulaInl" alt="$ \boldsymbol{\bar{i}}^{(b)} $" src="form_301.png"/> (only <img class="formulaInl" alt="$ \boldsymbol{\bar{i}}^{(b)} $" src="form_301.png"/> is shown) are given by</p>
<p><a class="anchor" id="tm_eq_2_75"></a> </p>
<p class="formulaDsp">
<img class="formulaDsp" alt="\[ \begin{array}{l} \boldsymbol{i}^{0a}_i =\boldsymbol{T}^{0a}_{ij} \boldsymbol{\bar{i}}^{(a)}_j = \boldsymbol{T}^{0a}_{ij} \boldsymbol{\bar{T}}^{(a)}_{jk} \boldsymbol{I}_k = \boldsymbol{\bar{T}}^{0a}_{ik} \boldsymbol{I}_k \\ \\ \boldsymbol{i}^{0b}_i =\boldsymbol{T}^{0b}_{ij} \boldsymbol{\bar{i}}^{(b)}_j = \boldsymbol{T}^{0b}_{ij} \boldsymbol{\bar{T}}^{(b)}_{jk} \boldsymbol{I}_k = \boldsymbol{\bar{T}}^{0b}_{ik} \boldsymbol{I}_k \end{array}\tag{2.75} \]" src="form_302.png"/>
</p>
<p>where</p>
<p class="formulaDsp">
<img class="formulaDsp" alt="\[ \boldsymbol{T}^{0a}_{ij} = \boldsymbol{T}^{0b}_{ij} = \boldsymbol{T}^{0}_{ij} \tag{2.76} \]" src="form_303.png"/>
</p>
<p>Due to the definitions of the local coordinate systems used, <img class="formulaInl" alt="$ \boldsymbol{T}^0_{ij} $" src="form_304.png"/> also describes the transformation between the adjacent beam end base vector system and the nodal system. Keep in mind that <img class="formulaInl" alt="$ \boldsymbol{T}^0_{ij} $" src="form_304.png"/> remains constant through the deformations.</p>
<p><a class="anchor" id="Base_vector_systems_for_the_connector"></a></p>
<div class="image">
<img src="RIFLEX_TheoryManual_42_v0_345.jpg" alt="RIFLEX_TheoryManual_42_v0_345.jpg"/>
<div class="caption">
Base vector systems for the connector element.</div></div>
<p>Due to the relation with the adjacent beam element, the "ghost" reference for this beam element is adopted as reference system for the connector element. Employing <a class="el" href="a00004.html#tm_eq_2_75">Eq. (2.75)</a> and the inverse of <a class="el" href="a00004.html#tm_eq_2_72">Eq. (2.72)</a> gives the following transformation between the connector element end coordinate systems and the "ghost" reference system of the adjacent beam element</p>
<p class="formulaDsp">
<img class="formulaDsp" alt="\[ \begin{array}{l} \boldsymbol{i}^{0a}_i = \boldsymbol{\bar{T}}^{0a}_{ij} \boldsymbol{T}_{kj} \boldsymbol{i}_k \\ \\ \boldsymbol{i}^{0b}_i = \boldsymbol{\bar{T}}^{0b}_{ij} \boldsymbol{T}_{kj} \boldsymbol{i}_k \\ \end{array}\tag{2.77} \]" src="form_305.png"/>
</p>
<p>The local rotations of the ends of the connector element related to the "ghost" reference system are now found by inspecting the differences in the coordinate systems <img class="formulaInl" alt="$ \boldsymbol{i}^{0a}_c $" src="form_306.png"/>, <img class="formulaInl" alt="$ \boldsymbol{i}_i $" src="form_225.png"/> and <img class="formulaInl" alt="$ \boldsymbol{i}^{0b}_c $" src="form_307.png"/>, <img class="formulaInl" alt="$ \boldsymbol{i}_i $" src="form_225.png"/>.</p>
<p>The rotations in the connector element should be small, or primarily take place in one plane if they are large.</p>
<h3>Internal structural reaction forces and stiffness matrix</h3>
<p>The internal structural reaction forces are given by</p>
<p class="formulaDsp">
<img class="formulaDsp" alt="\[ \boldsymbol{S}^E = \begin{bmatrix} S_{x1} \\ S_{y1} \\ S_{z1} \\ S_{\theta x1} \\ S_{\theta y1} \\ S_{\theta z1} \\ S_{x2} \\ S_{y2} \\ S_{z2} \\ S_{\theta x2} \\ S_{\theta y2} \\ S_{\theta z2} \end{bmatrix} = \begin{bmatrix} - k_x \cdot (v_{x2}- v_{x1}) \\ - k_y \cdot (v_{y2}- v_{y1}) \\ - k_z \cdot (v_{z2}- v_{z1}) \\ - k_{\theta x} \cdot (\theta_{x2}- \theta_{x1}) \\ - k_{\theta y} \cdot (\theta_{y2}- \theta_{y1}) \\ - k_{\theta z} \cdot (\theta_{z2}- \theta_{z1}) \\ k_x \cdot (v_{x2}- v_{x1}) \\ k_y \cdot (v_{y2}- v_{y1}) \\ k_z \cdot (v_{z2}- v_{z1}) \\ k_{\theta x} \cdot (\theta_{x2}- \theta_{x1}) \\ k_{\theta y} \cdot (\theta_{y2}- \theta_{y1}) \\ k_{\theta z} \cdot (\theta_{z2}- \theta_{z1}) \\ \end{bmatrix}\tag{2.78} \]" src="form_308.png"/>
</p>
<p>where <img class="formulaInl" alt="$ k_x$" src="form_309.png"/>, <img class="formulaInl" alt="$ k_y$" src="form_310.png"/>, <img class="formulaInl" alt="$ k_z$" src="form_311.png"/>, <img class="formulaInl" alt="$ k_{\theta x}$" src="form_312.png"/>, <img class="formulaInl" alt="$ k_{\theta y}$" src="form_313.png"/> and <img class="formulaInl" alt="$ k_{\theta z}$" src="form_314.png"/> are spring stiffnesses associated with the different degrees of freedom.</p>
<p>In the current version of the program the spring stiffnesses are defined by the stiffness of the related beam element as</p>
<p class="formulaDsp">
<img class="formulaDsp" alt="\[ \begin{array}{l} \displaystyle k_x = k_y = k_z =\frac{(EA)}{L_0} \\ \\ \displaystyle k_{\theta x} = k_{\theta y} = k_{\theta z} =\frac{(EA)}{100 L_0} \end{array}\tag{2.79} \]" src="form_315.png"/>
</p>
<p>where <img class="formulaInl" alt="$ (EA)$" src="form_316.png"/> is the stiffness in axial direction and <img class="formulaInl" alt="$ L_0$" src="form_317.png"/> is the initial length of the related beam element. However, any of the rotational stiffnesses can be specified equal to zero.</p>
<p>The tangential stiffness matrix is computed using the tangential material moduli of the related element</p>
<p><a class="anchor" id="tm_eq_2_80"></a> </p>
<p class="formulaDsp">
<img class="formulaDsp" alt="\[ \boldsymbol{k} = \boldsymbol{k}_M = \left[ \begin{array}{cccccccccccc} k_x &amp; &amp; &amp; &amp; &amp; &amp; -k_x &amp; &amp; &amp; &amp; &amp; \\ &amp; k_y &amp; &amp; &amp; &amp; &amp; &amp; -k_y &amp; &amp; &amp; &amp; \\ &amp; &amp; k_z &amp; &amp; &amp; &amp; &amp; &amp; -k_z &amp; &amp; &amp; \\ &amp; &amp; &amp; k_{\theta x} &amp; &amp; &amp; &amp; &amp; &amp; -k_{\theta x} &amp; &amp; \\ &amp; &amp; &amp; &amp; k_{\theta y} &amp; &amp; &amp; &amp; &amp; &amp; -k_{\theta y} &amp; \\ &amp; &amp; &amp; &amp; &amp; k_{\theta z} &amp; &amp; &amp; &amp; &amp; &amp; -k_{\theta z} \\ -k_x &amp; &amp; &amp; &amp; &amp; &amp; k_x &amp; &amp; &amp; &amp; &amp; \\ &amp; -k_y &amp; &amp; &amp; &amp; &amp; &amp; k_y &amp; &amp; &amp; &amp; \\ &amp; &amp; -k_z &amp; &amp; &amp; &amp; &amp; &amp; k_z &amp; &amp; &amp; \\ &amp; &amp; &amp; -k_{\theta x} &amp; &amp; &amp; &amp; &amp; &amp; k_{\theta x} &amp; &amp; \\ &amp; &amp; &amp; &amp; -k_{\theta y} &amp; &amp; &amp; &amp; &amp; &amp; k_{\theta y} &amp; \\ &amp; &amp; &amp; &amp; &amp; -k_{\theta z} &amp; &amp; &amp; &amp; &amp; &amp; k_{\theta z} \\ \end{array} \right] \tag{2.80} \]" src="form_318.png"/>
</p>
<p>All terms that are not indicated in <a class="el" href="a00004.html#tm_eq_2_80">Eq. (2.80)</a> are equal to zero.</p>
<h1><a class="anchor" id="tm_finite_cross"></a>
Cross Sectional Modelling</h1>
<h2>Hyper-elastic constitutive models</h2>
<p>The detailed cross-sectional structure of many slender marine structures is normally very complex, e.g. flexible pipes, umbilicals and wires. In a global analysis program, it is therefore convenient to model the global cross-sectional behaviour in terms of force displacement relations. Such relations are axial force versus axial elongation, bending moment versus curvature and torsional moment versus torsional twist:</p>
<p><a class="anchor" id="tm_eq_2_81"></a> </p>
<p class="formulaDsp">
<img class="formulaDsp" alt="\[ \begin{array}{l} A = A(\Delta x) \\ M = M(R,\Delta p) \\ M_T = M_T (\Delta \theta) \end{array}\tag{2.81} \]" src="form_319.png"/>
</p>
<p>where</p>
<ul>
<li><img class="formulaInl" alt="$ A\:\:\:\:: \quad $" src="form_320.png"/> axial force</li>
<li><img class="formulaInl" alt="$ \Delta x\:: \quad $" src="form_321.png"/> axial elongation per unit length</li>
<li><img class="formulaInl" alt="$ M\:\:\:: \quad $" src="form_322.png"/> bending moment</li>
<li><img class="formulaInl" alt="$ R\:\:\:\:: \quad $" src="form_323.png"/> radius of curvature</li>
<li><img class="formulaInl" alt="$ \Delta p\:\:: \quad $" src="form_324.png"/> pressure difference</li>
<li><img class="formulaInl" alt="$ M_T: \quad $" src="form_325.png"/> torsional moment</li>
<li><img class="formulaInl" alt="$ \Delta \theta\: : \quad $" src="form_326.png"/> axial rotation per unit length</li>
</ul>
<p>The force/displacement relations can be specified as linear by giving the cross-section stiffness, or as nonlinear by specification of force/displacement relations in tabular form, see <a class="el" href="a00004.html#Nonlinear_cross_section_parameters">Nonlinear cross section parameters</a>.</p>
<p>It is also possible to specify a hysteretic bending moment/curvature relationship.</p>
<p>As indicated in <a class="el" href="a00004.html#tm_eq_2_81">Eq. (2.81)</a> it can also be proposed to include a pressure dependent bending stiffness to account for cross-sectional ovalization during bending. This effect is, however, not considered in the present program version.</p>
<p><a class="anchor" id="Nonlinear_cross_section_parameters"></a></p>
<div class="image">
<img src="RIFLEX_TheoryManual_42_v0_374.jpg" alt="RIFLEX_TheoryManual_42_v0_374.jpg"/>
<div class="caption">
Nonlinear cross section parameters.</div></div>
<h2>Elasto-plastic bi-linear hysteresis bending model</h2>
<p>The bending response of the tensile armour in a flexible pipe may be explained by considering an initially straight pipe section that is gradually bent with constant interlayer radial pressures. For small curvatures, the helix tendons are kept in place by friction, and a linear relationship thus exists between the curvature and the change of tendon stress. In this situation, the bending moment is related to the curvature by a factor equal to the tensile armour bending stiffness contribution in an equivalent cross-section not allowing for interlayer slippage. As the curvature increases, the change of tendon stress exceeds the available friction stress and slippage along the helices starts developing at the neutral axis of bending. When slippage is present around the whole circumference, the tendon stress and the resulting bending moment remain unchanged if the pipe is subjected to a further increase of the curvature. This response can be modelled by a tri-linear moment-curvature relation as proposed in work by Savik (2011). However, the transition between start of slippage to full slippage is small in terms of curvature. This motivates use of a bi-linear bending model which allows for a simpler algorithmic treatment that may achieve better numerical performance than more sophisticated models. Further, the bi-linear bending model can be linearized in a straightforward way leading to a consistent tangent stiffness relation.</p>
<p>The implemented bi-linear model assumes an additive split of the bending moment into elastic and hysteresis parts,</p>
<p><a class="anchor" id="tm_eq_2_82"></a> </p>
<p class="formulaDsp">
<img class="formulaDsp" alt="\[ \boldsymbol{M}_{tot} = \boldsymbol{M} + \boldsymbol{M}_e \qquad \boldsymbol{M} = \begin{bmatrix} M_{tot,y} &amp; M_{tot,z} \end{bmatrix}^{\boldsymbol{\top}} \tag{2.82} \]" src="form_327.png"/>
</p>
<p>where the constitutive relation for the elastic part reads,</p>
<p><a class="anchor" id="tm_eq_2_83"></a> </p>
<p class="formulaDsp">
<img class="formulaDsp" alt="\[ \boldsymbol{M}_e = EJ \boldsymbol{\kappa} \tag{2.83} \]" src="form_328.png"/>
</p>
<p>in which <img class="formulaInl" alt="$ \boldsymbol{\kappa} $" src="form_329.png"/> is the bending curvature vector and the bending stiffness <img class="formulaInl" alt="$ EJ $" src="form_330.png"/> is illustrated in the figure <a class="el" href="a00004.html#Bi_linear_hysteresis_bending_moment_constitutive_parameters">Bi-linear hysteresis bending moment constitutive parameters</a>.</p>
<p><a class="anchor" id="Bi_linear_hysteresis_bending_moment_constitutive_parameters"></a></p>
<div class="image">
<img src="bilinearhystmoment.svg" alt="bilinearhystmoment.svg"/>
<div class="caption">
Bi-linear hysteresis bending moment constitutive parameters</div></div>
<p>The constitutive model for the hysteresis part is formulated in the framework of computational elasto-plasticity. In order to determine whether slippage is present or not, a slip function <img class="formulaInl" alt="$ f $" src="form_331.png"/> is defined according to,</p>
<p><a class="anchor" id="tm_eq_2_84"></a> </p>
<p class="formulaDsp">
<img class="formulaDsp" alt="\[ f \left( \boldsymbol{M} \right) = |\boldsymbol{M}| - M_f \leq 0 \qquad \boldsymbol{M} = \begin{bmatrix} M_y &amp; M_z \end{bmatrix}^{\boldsymbol{\top}} \tag{2.84} \]" src="form_332.png"/>
</p>
<p>where <img class="formulaInl" alt="$f=0$" src="form_333.png"/> during slip, <img class="formulaInl" alt="$f<0$" src="form_334.png"/> indicates no slippage and <img class="formulaInl" alt="$f >0$" src="form_335.png"/> is an inadmissible state. <img class="formulaInl" alt="$M_f$" src="form_336.png"/> is the friction moment limit, corresponding to the bending moment where slip takes place.</p>
<p>The rate of change of bending moment with respect to time is given by,</p>
<p><a class="anchor" id="tm_eq_2_85"></a> </p>
<p class="formulaDsp">
<img class="formulaDsp" alt="\[ \dot{\boldsymbol{M}} = K_E \left( \dot{\boldsymbol{\kappa}} - \dot{\boldsymbol{\kappa}_p} \right) \tag{2.85} \]" src="form_337.png"/>
</p>
<p>where <img class="formulaInl" alt="$K_E$" src="form_338.png"/> is the elastic modulus corresponding to the tensile armour bending stiffness contribution in an equivalent cross-section not allowing for slippage, which in terms of the stiffness data in the figure <a class="el" href="a00004.html#Bi_linear_hysteresis_bending_moment_constitutive_parameters">Bi-linear hysteresis bending moment constitutive parameters</a> is expressed by,</p>
<p><a class="anchor" id="tm_eq_2_86"></a> </p>
<p class="formulaDsp">
<img class="formulaDsp" alt="\[ K_E = SF \cdot EJ - EJ \qquad SF \geq1.0 \tag{2.86} \]" src="form_339.png"/>
</p>
<p>In <a class="el" href="a00004.html#tm_eq_2_85">Eq. (2.85)</a>, <img class="formulaInl" alt="$\dot{\boldsymbol{\kappa}}$" src="form_340.png"/> is the rate of total bending curvature and <img class="formulaInl" alt="$\dot{\boldsymbol{\kappa}}_p$" src="form_341.png"/> is the rate of bending curvature associated with slip. The latter rate is given by an associative slip rule,</p>
<p><a class="anchor" id="tm_eq_2_87"></a> </p>
<p class="formulaDsp">
<img class="formulaDsp" alt="\[ \dot{\boldsymbol{\kappa}}_p = \dot{\lambda} \frac{ \partial }{ \partial \boldsymbol{M} } f\left( \boldsymbol{M} \right) = \dot{\lambda} \frac{ \boldsymbol{M} }{ |\boldsymbol{M}| } \qquad \dot{\lambda} \geq 0 \tag{2.87} \]" src="form_342.png"/>
</p>
<p> in which <img class="formulaInl" alt="$\dot{\lambda}$" src="form_343.png"/> is the plastic rate parameter, determined by the consistency criterion <img class="formulaInl" alt="$\dot{f}=0$" src="form_344.png"/> if slippage occurs and set to zero if <img class="formulaInl" alt="$f<0$" src="form_334.png"/>. In case of no slip, the incremental constitutive relation is obtained by straightforward integration of <a class="el" href="a00004.html#tm_eq_2_85">Eq. (2.85)</a> with <img class="formulaInl" alt="$\dot{\boldsymbol{\kappa}}_p=\boldsymbol{0}$" src="form_345.png"/>. When slippage occurs, the incremental relation valid for finite increments is established by linearizing the moment update scheme. A fully implicit backward Euler scheme is employed to integrate <a class="el" href="a00004.html#tm_eq_2_85">Eq. (2.85)</a> and <a class="el" href="a00004.html#tm_eq_2_87">Eq. (2.87)</a> and to satisfy <a class="el" href="a00004.html#tm_eq_2_85">Eq. (2.84)</a>, which when linearized results in the following consistent tangent stiffness relation,</p>
<p class="formulaDsp">
<img class="formulaDsp" alt="\[ \begin{bmatrix} dM_y\\ dM_z \end{bmatrix} = \frac{ K_E M_f }{ |\boldsymbol{M}^{tr}|^3 } \begin{bmatrix} |\boldsymbol{M}^{tr} |^2-M_y^2 &amp; -M_y M_z \\ -M_y Mz &amp; |\boldsymbol{M}^{tr}|^2-M_z^2 \\ \end{bmatrix} \begin{bmatrix} d \kappa_y \\ d \kappa_z \end{bmatrix} \tag{2.88} \]" src="form_346.png"/>
</p>
<p>where <img class="formulaInl" alt="$\mathbf{M}^{tr}$" src="form_347.png"/> is the trial moment obtained by assuming no slippage during the finite curvature increment, i.e. <img class="formulaInl" alt="$\dot{\boldsymbol{\kappa}}_p$" src="form_341.png"/> is set to zero in <a class="el" href="a00004.html#tm_eq_2_85">Eq. (2.85)</a> according to,</p>
<p><a class="anchor" id="tm_eq_2_89"></a> </p>
<p class="formulaDsp">
<img class="formulaDsp" alt="\[ \boldsymbol{M}^{tr} = K_E \,\dot{\boldsymbol{\kappa}} \Delta t + \boldsymbol{M}^p \tag{2.89} \]" src="form_348.png"/>
</p>
<p>where <img class="formulaInl" alt="$ \Delta t$" src="form_349.png"/> is the time increment and <img class="formulaInl" alt="$ \boldsymbol{M}^p $" src="form_350.png"/> is the bending moment in the previous equilibrium state.</p>
<p>The moment-curvature response obtained when subjecting the cross-section to a sinusoidal moment loading of 4 kNm about either the y- or z-axis is shown in the figure <a class="el" href="a00004.html#Moment_curvature_response_for_in_plane_loading">Moment-curvature response for uni-axial bending load</a>. The cross-section was modelled by <img class="formulaInl" alt="$EJ=20 kNm^2$" src="form_351.png"/>, <img class="formulaInl" alt="$M_f=1 kNm$" src="form_352.png"/> and <img class="formulaInl" alt="$SF=50$" src="form_353.png"/>.</p>
<p>The moment response for the same cross-section is shown in the figure <a class="el" href="a00004.html#Moment_response_for_3D_loading">Moment response for bi-axial moment loading</a> when subjected to 2 kNm sinusoidal bending moments about the y- and z-axis which are 90 degrees out of phase. Here, the y-moment load is first ramped to its full value before the z-moment is applied. Observe that the bending moment trajectory is circular, whereas the curvature in the figure <a class="el" href="a00004.html#Curvature_response_for_3D_loading">Curvature response for bi-axial moment loading</a> undergoes a gradual increase before it stabilizes at a circular trajectory.</p>
<p><a class="anchor" id="Moment_curvature_response_for_in_plane_loading"></a></p>
<div class="image">
<img src="hysteresis-BL-2Dmomentcurvature.png" alt="hysteresis-BL-2Dmomentcurvature.png"/>
<div class="caption">
Moment-curvature response for uni-axial bending load</div></div>
<p><a class="anchor" id="Moment_response_for_3D_loading"></a></p>
<div class="image">
<img src="hysteresis-BL-3Dmoment.png" alt="hysteresis-BL-3Dmoment.png"/>
<div class="caption">
Moment response for bi-axial moment loading</div></div>
<p><a class="anchor" id="Curvature_response_for_3D_loading"></a></p>
<div class="image">
<img src="hysteresis-BL-3Dcurvature.png" alt="hysteresis-BL-3Dcurvature.png"/>
<div class="caption">
Curvature response for bi-axial moment loading</div></div>
<h1><a class="anchor" id="tm_finite_guideline"></a>
Guideline to System Modelling</h1>
<p>Advice with respect to element length and bending stiffness for systems that may have zero tension or compression loads.</p>
<h3>Bar elements:</h3>
<p>Element are deformed only in the axial directions. Compressive loads may be limited by two mechanisms:</p>
<p>a) Lateral motion of nodal points, which imply rigid-body rotation of elements. The shorter the elements, the more rapid the lateral displacement may occur. Too short elements may, however, cause numerical instabilities. Hence, use of short elements to limit axial compressive force in a bar model, will require very short time step.</p>
<p>b) Use of nonlinear tension-strain relationship, specifying a very low stiffness for negative strain. This will allow the element to "shrink", and does not involve element rotations.</p>
<h3>Beam elements:</h3>
<p>Compressive load on a beam element may lead to stability problems if the element is long compared with its bending stiffness. In regions where zero tension or compressive loads are expected, element lengths should be limited according to</p>
<p>a) <img class="formulaInl" alt="$ 1 \leq 2 (EI/q)^{1/3}$" src="form_354.png"/></p>
<p>b) <img class="formulaInl" alt="$ 1 \leq \pi (EI/P_c)^{1/3}$" src="form_355.png"/></p>
<p>where</p>
<ul>
<li><img class="formulaInl" alt="$ EI \quad $" src="form_356.png"/> = cross-section stiffness, weakest axis if not rotation symmetric</li>
<li><img class="formulaInl" alt="$ q \quad \:\:\:\, $" src="form_357.png"/> = largest lateral load intensity, e.g. element weight/length</li>
<li><img class="formulaInl" alt="$ P_c \quad \: $" src="form_358.png"/> = largest compressive loads</li>
</ul>
<h1><a class="anchor" id="tm_finite_references"></a>
References</h1>
<p>Battini, J-M. (2002): "Co-rotational beam elements in instability problems" PhD Thesis, Department of Mechanics, Royal Institute of Technology, Stockholm.</p>
<p>Bell, K. (1987): "Matrisestatistikk" Tapir Forlag, Trondheim, Norway.</p>
<p>Bergan, P.G. et al. (1985): "FENRIS System Manual, Theory-Program Outline-Data Input" Report 83-6064, A.S. Veritec, Høvik, Norway.</p>
<p>Bergan, P.G. et al. (1984): "FENRIS Satellite 1 Manual, Theory-Program Outline-Data Input" Report 83-6062, A.S. Veritec, Høvik, Norway.</p>
<p>Mollestad, E. (1983): "Techniques for Static and Dynamic Solution of Non-linear Finite Element
Problems" Report No. 83-1, Div. of Structural Mechanics, Norwegian Institute of Technology, Trondheim.</p>
<p>Engseth, A. (1984): "Finite Element Collapse Analysis of Tubular Steel Offshore Structures" Dr.ing. Thesis, Div. of Marine Structures, The Norwegian Institute of Technology, Trondheim.</p>
<p>Krenk, S. (2009): "Non-linear Modeling and Analysis of Solids and Structures" Cambridge University Press, Cambridge, UK.</p>
<p>Malvern, L.E. (1969): "Introduction to the Mechanics of a Continuous Medium" Prentice Hall, Englewood Cliffs, New Jersey.</p>
<p>Remseth, S.N. (1978): "Nonlinear Static and Dynamic Analysis of Space Structures" Dr.ing. Thesis, Div. of Structural Mechanics, The Norwegian Institute of Technology, Trondheim.</p>
<p>Savik, S. (2011): "Theoretical and experimental studies of stresses in flexible pipes" Computers &amp; Structures, vol. 89, pp. 2273-91. </p>
</div></div><!-- contents -->
	</body>
</html>