Problem 5 - Concepts and Applications of Finite Element Analysis

Question

Let loads \(P_{1}\) and \(P_{2}\) be functions of displacements \(D_{1}\) and \(D_{2}\); that is, \(P_{1}=\) \(f_{1}\left(D_{1}, D_{2}\right)\) and \(P_{2}=f_{2}\left(D_{1}, D_{2}\right)\). Let \(D_{A}\) and \(D_{B}\) be exact values of \(D_{1}\) and \(D_{2}\) produced by loads \(P_{A}\) and \(P_{B}\). Let \(D_{A}^{*}\) and \(D_{B}^{*}\) be approximations of \(D_{A}\) and \(D_{B}\). Assume that \(D_{A}=D_{A}^{*}+\triangle D_{A}\) and \(D_{B}=D_{B}^{2}+\Delta D_{B}\). Derive the following equa. tions of the \(\mathrm{N}-\mathrm{R}\) method: $$ \left[\begin{array}{ll} \partial P_{1} / \partial D_{1} & \partial P_{1} / \partial D_{2} \\ \partial P_{2} / \partial D_{1} & \partial P_{2} / \partial D_{2} \end{array}\right]_{D_{\lambda}^{*}, D ;}\left\\{\begin{array}{l} \Delta D_{A} \\ \Delta D_{B} \end{array}\right\\}=\left\\{\begin{array}{l} P_{A}-f_{1}\left(D_{A}^{*}, D_{B}^{*}\right) \\ P_{B}-f_{2}\left(D_{A}^{*}, D_{B}^{*}\right) \end{array}\right\\} $$

Step-by-Step Solution

Answer

The resulted equations represent the Newton-Raphson method; they are obtained by linearizing the system of equations, then rearranging and written in a matrix form. The matrix on the left side of the equation is the Jacobian matrix, which is the matrix of all first-order partial derivatives. The vector on the right side of the equation is calculated from the difference between the imposed loads and the corresponding reactions from the structure.

Step 1: Understand the Newton-Raphson Method Context

The Newton-Raphson method is an iterative scheme used to find roots of equations. In this case, we are applying it to a system of two equations \(P_1=f_1(D_1,D_2)\) and \(P_2=f_2(D_1,D_2)\). Iterating the scheme involves linearizing the functions around the current approximation (here \(D^*_A\) and \(D^*_B\)) and solving for the correction \(\Delta D_A\) and \(\Delta D_B\).

Step 2: Formulate the Linearization

The linearization of the system can be written as \(P_A -f_1(D_{A}^{*}, D_{B}^{*})= \partial P_1/ \partial D_1 * \Delta D_{A} + \partial P_1/ \partial D_2 * \Delta D_{B}\) and \(P_B -f_2(D_{A}^{*}, D_{B}^{*})= \partial P_2/ \partial D_1 * \Delta D_{A} + \partial P_2/ \partial D_2 * \Delta D_{B}\). These are our two equations in the form \(\Delta F = J \Delta D\), where \(\Delta F\) is the difference between our load vectors, \(J\) is the Jacobian matrix formed by the partial derivates, and \(\Delta D\) is our correction vector.

Step 3: Rearranging into Matrix form

The two newly derived equations can be reformulated into a matrix equation. In this case we have \(\left[\begin{array}{ll} \partial P_{1} / \partial D_{1} & \partial P_{1} / \partial D_{2} \ \partial P_{2} / \partial D_{1} & \partial P_{2} / \partial D_{2} \end{array}\right]_{D_{\lambda}^{*}, D }\left\{\begin{array}{l} \Delta D_{A} \ \Delta D_{B} \end{array}\right\}=\left\{\begin{array}{l} P_{A}-f_{1}\left(D_{A}^{*}, D_{B}^{*}\right) \ P_{B}-f_{2}\left(D_{A}^{*}, D_{B}^{*}\right) \end{array}\right\}\).

Key Concepts

Numerical Methods

Numerical methods are computational algorithms for approximating solutions to complex mathematical problems that are difficult or impossible to solve analytically. In engineering, science, and mathematics, these methods are vital for analyzing models and making predictions when closed-form expressions are not available.

Specifically, numerical methods enable us to solve problems such as finding the roots of nonlinear equations, integrating functions over complicated domains, and solving differential equations. These methods typically involve iterations or discretization techniques and are crucial when dealing with large-scale problems in Finite Element Analysis (FEA) and other computational frameworks.

For accurate and reliable outcomes, numerical methods must be carefully chosen and validated to ensure convergence and stability. The Newton-Raphson (N-R) method is one exemplary numerical technique that iteratively refines solutions to yield highly accurate results when applied correctly.

Finite Element Analysis

Finite Element Analysis (FEA) is an advanced computational technique used to predict how products will react to real-world forces, vibration, heat, fluid flow, and other physical effects. FEA works by breaking down a real object into a large number (finite) of elements, where equations are solved in each element through numerical methods such as the Newton-Raphson method.

In the context of FEA, the Newton-Raphson method helps solve complex nonlinear equations by approximating the distribution of forces and displacements within elements. This is crucial when the behavior of materials and systems under stress needs to be predicted with high precision. The method's iterative nature allows for updating of the model until the solution converges to an acceptable level of accuracy, enabling engineers and scientists to model scenarios and test designs virtually, saving time and resources.

Iterative Techniques

Iterative techniques are procedures in which successive approximations are used to approach the desired solution of a problem. The Newton-Raphson method is one such iterative technique widely used in numerical computations. It involves making an initial guess at the solution and then refining that guess with repeated iterations until the changes in the solution become negligibly small.

The key to the success of iterative techniques lies in their ability to hone in on a precise solution through repeated application of a specific algorithm. This method is particularly useful when direct solutions are complicated or impossible to derive. Iterative methods often incorporate stopping criteria based on error tolerances, ensuring that the computed solution is as accurate as necessary for practical purposes.

Crucial to many fields is the implementation of iteration that is computationally efficient – converging quickly to the correct result without excessive calculations that would slow down the process.

Linearization

Linearization is a mathematical strategy that simplifies a nonlinear function by approximating it as a linear function within a small region around a certain point. This process is fundamental in applying the Newton-Raphson method as it converts a set of nonlinear equations into a series of linear equations that can be more easily solved.

The linearization in Newton-Raphson involves computing the Jacobian matrix of partial derivatives, which provides local linear models of the original nonlinear functions. By repeatedly linearizing and solving at each iteration step, we create a pathway that ideally leads to the root of the original nonlinear system. Linearization is an essential step in many branches of engineering and science as it enables us to apply linear analysis methods to complex, nonlinear problems and arrive at an approximate solution that is sufficient for practical purposes.