Problem 7 - Concepts and Applications of Finite Element Analysis

Question

The sketch shows an axially loaded structure and its structural equations after boundary conditions have been imposed. (a) Use Gauss elimination to solve for \(u_{2}, u_{3}\), and \(u_{4}\). (b) After the first elimination, what physical interpretation can be given to \(K_{22} ?\). And what interpretation to \(K_{33}\) after the second elimination? (c) Structural equations can be written in the following form: $$ u_{2}=\left(24+6 u_{3}\right) / 12 \quad u_{3}=\left(24+6 u_{2}+6 u_{4}\right) / 12 \quad u_{4}=u_{3} $$ In Gauss-Seidel iteration, one solves the equations serially, using the most recently calculated values of the \(u_{i}\) in each equation. Thus, in the first iteration, starting with \(u_{2}=u_{3}=u_{4}=0\), one obtains \(u_{2}=2\) from the first equation, then \(u_{3}=3\) from the second equation, then \(u_{4}=3\) from the fourth equation. Camy out another three cycles of this process. (Note: There are ways to greatly increase the convergence rate.)

Step-by-Step Solution

Answer

Gauss elimination allows for simpler solving of the equations, but without the specifics of the matrix, it's not possible to perform. The components \(K_{22}\) and \(K_{33}\) potentially relate to the stiffness and deformation of structural elements after specific forces are applied. In Gauss-Seidel iteration, for given equations, three cycles of updating \(u_{2}\), \(u_{3}\), and \(u_{4}\) gives \(u_{2}=2\), \(u_{3}=3\), and \(u_{4}=3\) at each cycle.

Step 1: Solve using Gauss elimination

We do not have the specific equations or the matrix in the exercise, but generally, Gauss elimination involves turning an input matrix into row-echelon form through a series of row operations, leading to simplified equations and easier calculation of unknowns.

Step 2: Interpret physical meaning

Without specific context, it can be inferred that \(K_{22}\) and \(K_{33}\) are components of a stiffness matrix appearing after Gauss elimination. They relate to how much an element of the structure changes shape when a specific force is applied. Again, without specific physical context, it's hard to specify their exact physical interpretation.

Step 3: Applying Gauss-Seidel iteration

Given the equations and initial values \(u_{2}=u_{3}=u_{4}=0\), we iterate through the equations, first finding \(u_{2}=2\), then substituting this into the second equation to find \(u_{3}=3\), and hence \(u_{4}=3\), as \(u_{4}\) equals \(u_{3}\) from the structured equations. We repeat this process three times as per the exercise.

Key Concepts

Gauss Elimination

Gauss elimination is a method used in linear algebra to solve systems of linear equations. The main idea is to transform the coefficient matrix into an upper triangular form using a series of row operations, which include swapping rows, multiplying a row by a non-zero scalar, and adding or subtracting the multiple of one row from another. This transformation simplifies solving for the variables step-by-step, starting with the last row and working upwards.

The process of elimination makes it easier to identify and solve for unknown variables in a structural analysis, especially in the context of finite element analysis. In the given exercise, Gauss elimination would use these techniques to solve for the displacement values \( u_{2}, u_{3}, \) and \( u_{4} \) in a structural matrix equation. Understanding how each row corresponds to a balance of forces helps interpret why this method is efficient and reliable.

Additionally, during this process, specific elements like \( K_{22} \) and \( K_{33} \) in the stiffness matrix might emerge as key pivot positions. These values represent the effective stiffness of the structure during each stage of elimination. Thus, Gauss elimination not only provides a solution but also insight into the mechanical properties related to the system's response under load conditions.

Gauss-Seidel Iteration

The Gauss-Seidel iteration method is an approach for solving a system of linear equations iteratively. Unlike direct methods like Gauss elimination, Gauss-Seidel iterates to refine guesses until reaching an acceptable level of accuracy.

For the exercise in question, Gauss-Seidel iteration tackles each equation one by one. It updates the values of unknowns using the most recent values calculated. This serial process involves determining the value of one variable, substituting it into the next equation, and cycling through the system multiple times. Initially, as mentioned, starting efforts give: \( u_{2} = 2 \), \( u_{3} = 3 \), and consequently, \( u_{4} = 3 \) due to equivalence.

Here are some practical steps involved in Gauss-Seidel:

• Choose an initial guess for each variable (e.g., all zeros in this problem).
• Sequentially use each equation to update each variable, substituting back immediately any newly computed values into the subsequent equations.
• Repeat the above process for several cycles to enhance accuracy, noting do note how the choice of initial guesses and equation ordering impacts convergence rate.
• Stop once changes in variable values fall below a predetermined threshold, indicating stability.

By iteratively updating with feedback from previous iterations, Gauss-Seidel can be an efficient approach, particularly for large systems in finite element analysis. However, the method's success and speed depend heavily on matrix properties like diagonal dominance.

Stiffness Matrix

In the context of finite element analysis, the stiffness matrix plays a crucial role. It is a systematic representation that relates forces applied on a structural model to its displacements. Each element of the stiffness matrix is crucial because they describe how the structure resists deformation under load.

The stiffness matrix is structured such that each row and column corresponds to degrees of freedom in the model. It embodies the mechanical relationships between nodal points. It effectively acts as a map outlining how displacements respond to forces applied, often represented in the form \( [K][u] = [F] \), where \( [K] \) is the stiffness matrix, \( [u] \) the displacement vector, and \( [F] \) the force vector.

Under finite element analysis, constructing the stiffness matrix involves understanding its components:

• The diagonal elements (like \( K_{22} \) and \( K_{33} \)) often show self-stiffness. These represent resistance factors on corresponding degrees of freedom, often altered during matrix reduction techniques like Gauss elimination.
• Off-diagonal elements indicate how a displacement in one degree of freedom affects other directions. These terms are crucial for understanding coupling across different parts of the structure.
• It is typically symmetrical, simplifying computation in mechanical analysis.

Thus, mastering the nuances of stiffness matrices is central for designing stable and efficient structures, crucial in engineering applications such as bridges, buildings, and aerospace components.