Strain Energy and FEM Principles

TL;DR

Strain energy is the internal energy stored in a deformable body due to its deformation, crucial for understanding how structures respond to loads. In Finite Element Method (FEM), we use strain energy principles to formulate stiffness matrices and solve complex engineering problems. Understanding these concepts helps predict structural behavior and design more efficient components.

1. The Mental Model

Imagine stretching a rubber band; the energy you put into stretching it isn't lost, it's stored inside the rubber band, ready to be released. That stored energy is strain energy, and FEM uses this idea to break down big problems into smaller, manageable pieces to figure out how a whole structure behaves.

2. The Core Material

Strain energy is the work done by external forces on a deformable body that's stored internally as potential energy due to elastic deformation. Think of it as the energy "absorbed" by the material when it's stressed or strained. When the load is removed, this energy can be recovered, assuming elastic deformation.

Understanding Strain Energy Density

Instead of thinking about the total energy in a whole object, it's often more useful to think about strain energy density, which is the strain energy per unit volume. It's like asking, "how much energy is stored in a tiny cube of this material?"

For a simple uniaxial stress state (like a rod being pulled), the strain energy density ($U_0$) can be expressed as:

$U_0 = \frac{1}{2} \sigma \epsilon$

Where:
* $\sigma$ (sigma) is the stress
* $\epsilon$ (epsilon) is the strain

Using Hooke's Law ($\sigma = E\epsilon$), we can also write it as:

$U_0 = \frac{1}{2} E \epsilon^2 = \frac{1}{2E} \sigma^2$

Where $E$ is Young's Modulus (a measure of stiffness).

To get the total strain energy (U) in the entire body, you'd integrate this density over the volume of the body:

$U = \int_V U_0 dV$

This might look a bit complex, but the core idea is simple: the more you deform a material (higher strain), or the stiffer it is (higher E), the more strain energy it can store.

Strain Energy and FEM

FEM is a numerical technique used to find approximate solutions to boundary value problems. When applied to structural mechanics, it relies heavily on energy principles, particularly the Principle of Minimum Potential Energy. This principle states that for an elastic body in equilibrium, the total potential energy of the system is at its minimum.

The total potential energy ($\Pi$) of a system is the sum of its internal strain energy ($U$) and the potential energy of the external loads ($V_e$):

$\Pi = U + V_e$

In FEM, we discretize (break down) a continuous structure into a finite number of discrete elements connected at nodes. For each element, we approximate the displacement field using shape functions. These shape functions relate the displacement at any point within the element to the nodal displacements.

The strain energy for a single element can be expressed in terms of its nodal displacements. When you assemble all the elements, you get a global system of equations that can be solved for the nodal displacements. The goal is to minimize the total potential energy of the entire structure.

Here's a simplified flow for how FEM uses these principles:

graph TD
    A["Structure & Loads"] --> B["Discretize into Elements"]
    B --> C["Assume Displacement Field (Shape Functions) for each Element"]
    C --> D["Calculate Element Strain & Stress from Displacements"]
    D --> E["Formulate Element Strain Energy"]
    E --> F["Formulate Element Stiffness Matrix ([k])"]
    F --> G["Assemble Global Stiffness Matrix ([K])"]
    G --> H["Apply Boundary Conditions & Solve for Nodal Displacements ({u})"]
    H --> I["Calculate Element Strains & Stresses"]
    I --> J["Evaluate Total Strain Energy"]

The Element Stiffness Matrix [k]

Photo by Markus Spiske on Pexels

The element stiffness matrix [k] is at the heart of FEM. It relates the forces at the nodes of an element to the displacements of those nodes. You can derive this matrix by considering the strain energy within an element.

For an element, the strain energy can be written in matrix form:

$U_e = \frac{1}{2} \{\delta\}^T [k] \{\delta\}$

Where:
* $\{\delta\}$ is the vector of nodal displacements for that element.
* $[k]$ is the element stiffness matrix.

This equation looks a lot like the 1/2 k x^2 for a spring, but generalized for a more complex element. The terms in [k] depend on the element's material properties (like Young's Modulus) and its geometry.

By minimizing the total potential energy with respect to each nodal displacement, we arrive at the fundamental FEM equation:

[K]{U} = {F}

Where:
* [K] is the global stiffness matrix (assembled from all [k] elements).
* {U} is the global vector of unknown nodal displacements.
* {F} is the global vector of external nodal forces.

Solving this system gives you the nodal displacements, from which you can then calculate strains, stresses, and the total strain energy throughout the structure.

3. Worked Example

Let's consider a simple 1D bar element with length L, cross-sectional area A, and Young's Modulus E. It has two nodes, 1 and 2, with axial displacements u1 and u2 respectively.

1. Displacement Field: We assume a linear displacement variation within the element: $u(x) = N_1 u_1 + N_2 u_2$, where $N_1 = 1 - x/L$ and $N_2 = x/L$ are shape functions.

2. Strain: The axial strain $\epsilon_x = du/dx = \frac{d}{dx} (N_1 u_1 + N_2 u_2) = (-1/L)u_1 + (1/L)u_2 = \frac{1}{L}(u_2 - u_1)$.

3. Stress: Using Hooke's Law, $\sigma_x = E \epsilon_x = \frac{E}{L}(u_2 - u_1)$.

4. Strain Energy for the Element:
   $U_e = \int_V \frac{1}{2} \sigma_x \epsilon_x dV = \int_0^L \int_A \frac{1}{2} \left( \frac{E}{L}(u_2 - u_1) \right) \left( \frac{1}{L}(u_2 - u_1) \right) dA dx$
   Since $\sigma_x$ and $\epsilon_x$ are constant over the cross-section and length for this simple element, we can simplify:
   $U_e = \frac{1}{2} \left( \frac{E}{L}(u_2 - u_1) \right) \left( \frac{1}{L}(u_2 - u_1) \right) (AL)$
   $U_e = \frac{1}{2} \frac{AE}{L} (u_2 - u_1)^2$
   This can be written in terms of nodal displacements:
   $U_e = \frac{1}{2} \frac{AE}{L} (u_1^2 - 2u_1u_2 + u_2^2)$

5. Deriving Element Stiffness Matrix [k]:
   Recall $U_e = \frac{1}{2} \{\delta\}^T [k] \{\delta\} = \frac{1}{2} [u_1 \ u_2] \begin{bmatrix} k_{11} & k_{12} \\ k_{21} & k_{22} \end{bmatrix} \begin{bmatrix} u_1 \\ u_2 \end{bmatrix}$
   $= \frac{1}{2} (k_{11}u_1^2 + k_{12}u_1u_2 + k_{21}u_2u_1 + k_{22}u_2^2)$

   By comparing coefficients from step 4, we find:
   $k_{11} = \frac{AE}{L}$
   $k_{22} = \frac{AE}{L}$
   $k_{12} = k_{21} = -\frac{AE}{L}$

   So the element stiffness matrix for a 1D bar is:
   $[k] = \frac{AE}{L} \begin{bmatrix} 1 & -1 \\ -1 & 1 \end{bmatrix}$

This simple example shows how strain energy directly leads to the formulation of the element stiffness matrix, which is then used to build the full FEM problem.

4. Key Takeaways

• Strain energy is the internal energy stored in a body due to elastic deformation, representing the work done by external forces.
• Strain energy density is the energy stored per unit volume, often more useful for local analysis.
• The Principle of Minimum Potential Energy is a foundational concept in FEM, stating that a system's total potential energy is minimized at equilibrium.
• FEM uses strain energy principles to derive element stiffness matrices, which relate nodal forces to nodal displacements.
• The element stiffness matrix [k] encapsulates an element's resistance to deformation based on its material and geometry.
• Solving the global system [K]{U} = {F} yields nodal displacements, crucial for subsequent stress and strain calculations.

Common Mistakes to Avoid:
- Confusing Strain Energy with Stress/Strain: Strain energy is energy, while stress and strain are measures of internal force intensity and deformation, respectively. They are related, but not the same.
- Ignoring Elastic Limit: Strain energy principles, especially as applied in basic FEM, assume elastic deformation. If the material yields, these linear relationships no longer hold directly.
- Incorrectly Applying Boundary Conditions: Improperly constraining a model will lead to incorrect or unstable solutions, even if the strain energy formulation is correct.
- Misinterpreting [K]{U}={F}: Remember [K] is a property of the structure (stiffness), {U} is the response (displacement), and {F} is the input (force). Don't mix them up.

5. Now Try It

Imagine a cantilever beam fixed at one end and subjected to a point load at the free end. Sketch this beam. Now, mentally discretize it into two beam elements. Without calculating numbers, think about:
1. How would the strain energy change if you doubled the load?
2. How would the stiffness matrix for one of these beam elements be different from the 1D bar element example, considering it also needs to resist bending?
3. What global degrees of freedom (displacements/rotations) would you need to consider at eachnode to fully capture its behavior? What would success look like for this exercise? Success means you can articulate these differences and considerations qualitatively, showing you understand the concepts of load, stiffness, and element degrees of freedom in the context of strain energy and FEM.