Strain-Displacement Relations

When a solid object is subjected to forces, it moves and changes shape. But how do we precisely quantify this change in shape, separating it from simple rigid-body motion like translation or rotation? The answer lies in the strain-displacement relations, a cornerstone concept in solid mechanics that provides the mathematical language to describe deformation. These relations address the fundamental problem of connecting the observable displacement of points within a body to the internal strains—the stretching, shrinking, and shearing—that define its change in shape. This article provides a comprehensive overview of this critical topic, structured to build understanding from the ground up.

The following chapters will guide you through this essential concept. In "Principles and Mechanisms," we will explore the mathematical foundation of the strain-displacement relations, uncovering how local motion is decomposed into pure deformation and rigid rotation, and discussing the crucial assumptions and limitations of this framework. Subsequently, "Applications and Interdisciplinary Connections" will demonstrate how this purely geometric idea becomes a powerful tool, forming the basis for elasticity theory and modern computational methods like the Finite Element Method, with connections to fields ranging from thermodynamics to fracture mechanics.

Imagine you are looking at a large, soft block of gelatin. If you push the whole block from one side of the table to the other, every single point in it has moved, or been displaced. Yet, has the block deformed? Has its shape changed? Of course not. It has just undergone a rigid translation. Now, if you squeeze the block between your hands, it certainly deforms. Points that were once far apart are now closer, and the block bulges out at the sides.

This simple thought experiment gets to the heart of our story. To understand how a body changes shape, it's not enough to know how much each point has moved. The crucial information lies in how the relative positions of neighboring points have changed. The physics of deformation is a local affair.

To get a handle on this local change, we look at how the displacement vector, let's call it $\boldsymbol{u}$, varies from place to place. The mathematical tool for this is the ​​displacement gradient​​, a tensor written as $\nabla \boldsymbol{u}$. This object contains everything we need to know about the local motion. It tells us, for instance, how the $x$-component of displacement changes as we move a tiny step in the $y$-direction.

But as we saw with the gelatin block, a local change in displacement can be a mix of two very different things: a genuine change in shape and a simple rotation. A key insight of mechanics, beautiful in its simplicity, is that we can cleanly separate these two effects. Any square matrix—and our displacement gradient is just a matrix of numbers at each point—can be split into the sum of a symmetric matrix and a skew-symmetric (or anti-symmetric) matrix. This isn't just a mathematical parlor trick; it's a profound physical decomposition.

The symmetric part is what we call the ​​infinitesimal strain tensor​​, $\boldsymbol{\varepsilon}$. It is the true measure of deformation—stretching, shrinking, and shearing (changes in angle). It is defined as:

The skew-symmetric part is the ​​infinitesimal rotation tensor​​, $\boldsymbol{\omega}$. It tells us how the material at that point is locally spinning as a rigid body, without any change in shape. It is defined as:

So, the full picture of local motion is simply $\nabla \boldsymbol{u} = \boldsymbol{\varepsilon} + \boldsymbol{\omega}$.

Let's make this concrete. Suppose we measure the displacement gradient in a small region of a deforming metal plate and find it to be:

What is the real deformation? We apply our formulas. The strain is:

And the rotation is:

This tells us that the material is being stretched in the $x$ and $y$ directions, and also sheared, while simultaneously rotating. A motion of ​​pure strain​​ would occur only if the rotation tensor $\boldsymbol{\omega}$ were zero, which requires the displacement gradient to be symmetric ($b=c$). Conversely, a ​​pure rigid-body rotation​​ would have zero strain, $\boldsymbol{\varepsilon} = \mathbf{0}$, which requires the displacement gradient to be skew-symmetric ($a=d=0$ and $b=-c$).

You might be wondering about the word "infinitesimal" that keeps popping up. This is not just a fancy adjective; it's a crucial warning label. Our neat separation of strain and rotation works perfectly only when the displacements and their gradients are very, very small compared to the size of the object. This is the famous ​​small-strain assumption​​.

To see why, let's consider a dramatic counterexample. Imagine a cube that we rotate by a large angle, say $120^\circ$ ($\frac{2\pi}{3}$ radians), around the $z$-axis. Has the cube been strained? No. Every line drawn on it has the same length it did before. The true, physical strain is zero.

However, if we calculate the displacement gradient $\nabla \boldsymbol{u}$ for this large rotation, we find it is not zero. In fact, its components are numbers on the order of 1, which is not small at all! If we were to naively calculate the infinitesimal strain tensor $\boldsymbol{\varepsilon}$ from this large $\nabla \boldsymbol{u}$, we would get a non-zero result, incorrectly concluding that the cube has deformed.

This paradox reveals that for large rotations, the displacement gradient itself becomes entangled with rotation in a complex, non-linear way. The infinitesimal strain tensor $\boldsymbol{\varepsilon}$ is a linearized approximation of a more complicated, fully non-linear strain measure (like the ​​Green-Lagrange strain tensor​​) which correctly reports zero strain for any rigid rotation, large or small. Our simple formula for $\boldsymbol{\varepsilon}$ is the first term in a Taylor series expansion, and it is fantastically accurate as long as gradients are small, but misleading otherwise.

This isn't just an academic curiosity. It has profound practical implications. The classic Lamé solution for the stresses in a thick-walled cylinder under pressure—a cornerstone of mechanical engineering—is built entirely upon this small-strain assumption. The elegant formulas engineers use for pressure vessels and pipes are only valid because they assume the radial displacement and its derivatives are tiny compared to the radius of the pipe. If you tried to use them to analyze the inflation of a rubber balloon, where strains are enormous, the results would be completely wrong.

We've seen that if you give me a valid displacement field $\boldsymbol{u}$, I can compute the corresponding strain field $\boldsymbol{\varepsilon}$. Now let's turn the question around. Suppose I'm a designer, and I just write down a strain field on a piece of paper. Maybe I want the top of a beam to be compressed by $0.01$ and the bottom to be stretched by $0.01$. Can I be sure that this imagined strain field corresponds to a real, possible deformation of a continuous body?

The answer is, not necessarily! For a strain field to be physically possible, its components must obey a set of mathematical consistency conditions known as the ​​Saint-Venant compatibility equations​​.

The intuition is this: the strains must "fit together". The change in length of a tiny horizontal fiber ($\varepsilon_{xx}$) must be consistent with the changes in length of the vertical fibers above and below it ($\varepsilon_{yy}$) and the shearing of the angles between them ($\varepsilon_{xy}$). If they aren't compatible, trying to integrate the strain field to find the underlying displacement would lead you to a contradiction—you'd find that a point would need to be in two places at once, or that gaps and overlaps would have to appear in your supposedly continuous body.

What's truly beautiful about these compatibility conditions is that they are purely kinematic. They are a statement about the geometry of continuous space and have nothing to do with the material's properties. Whether the object is made of steel, rubber, or jelly, the rules for how strains must fit together remain exactly the same.

So where does our strain-displacement relation fit into the grand symphony of solid mechanics? It serves as the crucial bridge connecting the geometry of motion to the forces that cause it. The entire theory of linear elasticity rests on three pillars:

1. ​​Kinematics (Geometry of Motion):​​ The strain-displacement relations, $\boldsymbol{\varepsilon} = \frac{1}{2}(\nabla \boldsymbol{u} + (\nabla \boldsymbol{u})^T)$, which we have been exploring.

2. ​​Constitutive Law (Material Behavior):​​ This describes the "personality" of the material. For many materials, this is Hooke's Law, which states that stress is proportional to strain: $\boldsymbol{\sigma} = \boldsymbol{C} : \boldsymbol{\varepsilon}$, where $\boldsymbol{C}$ is the stiffness tensor.

3. ​​Kinetics (Balance of Forces):​​ This is simply Newton's Second Law applied to a continuum, which in static equilibrium says that the divergence of the stress must be balanced by any body forces (like gravity): $\nabla \cdot \boldsymbol{\sigma} + \boldsymbol{b} = 0$.

The magic happens when we put them all together. We can take the strain-displacement relation (Pillar 1) and plug it into the constitutive law (Pillar 2). Then we take the resulting expression for stress and plug it into the equilibrium equation (Pillar 3). The result is a single, powerful set of partial differential equations for the displacement field $\boldsymbol{u}$, known as the ​​Navier-Cauchy equations​​.

This is the master equation of linear elastostatics. It shows that if you know the material properties ($\lambda$ and $\mu$) and the forces ($\boldsymbol{b}$), you can, in principle, solve for the displacement everywhere in the body. And once you have the displacement, our strain-displacement relation gives you the strain, and Hooke's Law gives you the stress.

In practice, solving these equations is incredibly difficult for anything but the simplest geometries. This is where numerical techniques like the ​​Finite Element Method (FEM)​​ come in. In FEM, we chop a complex body into a mesh of simple elements (like tiny triangles or bricks). Within each element, we approximate the displacement field with a simple function, like a linear one. A linear displacement field, as we can derive from $\varepsilon_{xx} = \partial u_x / \partial x$, corresponds to a constant strain within the element. By ensuring the displacements match up at the corners of the elements and satisfying the governing equations in an average sense, we can build an approximate solution for the entire body.

From the simple idea of measuring relative motion, we have journeyed through the subtle interplay of rotation and deformation, uncovered the critical assumptions that underpin engineering practice, and arrived at the master equations that govern the response of solids to forces. The strain-displacement relation is the golden thread that ties the geometry of motion to the reality of stress and force, forming the very foundation of solid mechanics.

Having established the intimate kinematic relationship between strain and displacement, we might now ask: what is it good for? The answer, it turns out, is almost everything in the world of deformable bodies. This purely geometric rule, devoid of any mention of forces or material properties, is the universal language of deformation. It is the unseen architect that provides the geometric blueprint for the entire edifice of solid mechanics, connecting abstract theory to tangible reality. Let us embark on a journey to see how this simple idea blossoms into a rich and powerful set of tools for understanding and engineering our world.

The first and most profound application of the strain-displacement relation is in building the very laws that govern how materials behave under load. By itself, it is just one piece of a three-part puzzle. The other two are the law of equilibrium (a statement of Newton's laws for a continuum, $\nabla \cdot \boldsymbol{\sigma} + \mathbf{b} = \mathbf{0}$) and a material's constitutive law (which describes its unique personality, e.g., Hooke's Law, $\boldsymbol{\sigma} = \boldsymbol{C} : \boldsymbol{\varepsilon}$).

When you put these three ingredients together—kinematics, equilibrium, and material law—you create a complete mathematical description of the problem. For instance, if you substitute the strain-displacement relation into Hooke's Law, and then substitute the resulting stress into the equilibrium equation, you eliminate stress and strain entirely, arriving at a single, powerful equation for the displacement field itself. This is the celebrated Navier-Cauchy equation, the cornerstone of elasticity theory. It’s a beautiful demonstration of synthesis: three distinct physical principles are woven together by mathematics into a single governing equation that we can then solve to predict how a structure will deform.

There is another, equally elegant, path one can take. Instead of focusing on displacement, we can focus on stress. Engineers in the 19th century, like George Biddell Airy, realized that for 2D problems, one could define a clever mathematical construct—the Airy stress function, $\Phi$—from which all stress components could be derived such that the equilibrium equations are automatically satisfied. This is a brilliant trick, but it raises a crucial question: are these stress fields physically possible? Can the material actually deform in a way that produces them? The answer lies, once again, with the strain-displacement relations. For a strain field to be integrable into a valid, continuous displacement field, it must satisfy a "compatibility condition." When this condition is translated into terms of stress, it leads to the famous biharmonic equation, $\nabla^4 \Phi = 0$. This ensures that the beautiful stress patterns we can dream up with the Airy function correspond to a real-world deformation. So, whether we start with displacements or with stresses, the strain-displacement relation stands as the fundamental arbiter of geometric possibility.

While the analytical theories of Navier and Airy provide profound insight, they are often limited to objects with simple, idealized geometries. What about the intricate shape of an engine component, an artificial hip joint, or an entire airplane? For these, we must teach the laws of physics to a computer. This is the realm of the Finite Element Method (FEM), a revolutionary technique that has transformed modern engineering, and the strain-displacement relation is its absolute heart.

The core idea of FEM is "divide and conquer." A complex object is broken down into a mesh of simple, manageable pieces called "elements"—typically triangles or quadrilaterals in 2D. Within each small element, we approximate the continuous displacement field as a simple function (e.g., a linear one) interpolated from the displacements at the element's corners, or "nodes."

Once we have this simple approximation for displacement, we can apply the strain-displacement differentiation rule. The magic is that this turns a calculus problem into an algebra problem. The continuous relation $\boldsymbol{\varepsilon} = \mathcal{L}(\mathbf{u})$ becomes a simple matrix equation for the element: $\boldsymbol{\varepsilon} = \mathbf{B} \mathbf{d}$, where $\mathbf{d}$ is the vector of nodal displacements and $\mathbf{B}$ is the all-important strain-displacement matrix. For more complex elements used to model curved geometries, this process involves a beautiful mathematical mapping from a perfect "parent" shape to the real-world distorted element, requiring the use of the Jacobian to correctly transform the derivatives.

But here is the real punchline. The $\mathbf{B}$ matrix is not just for finding strain. It is the key ingredient in computing the element's stiffness. The stiffness matrix, $\mathbf{K}$, which relates the nodal forces to the nodal displacements, is assembled by integrating the product $\mathbf{B}^T \mathbf{D} \mathbf{B}$ over the element's volume, where $\mathbf{D}$ is the material's property matrix from Hooke's law. Think about what this integral, $\int \mathbf{B}^T \mathbf{D} \mathbf{B} \, dV$, represents. The matrix $\mathbf{B}$ converts nodal motion into internal strain. The matrix $\mathbf{D}$ converts that strain into internal stress. And the matrix $\mathbf{B}^T$ (arising from the principle of virtual work) converts that stress back into the equivalent forces at the nodes. This chain directly links nodal displacements to nodal forces—which is precisely what stiffness is! The strain-displacement relation, embodied in $\mathbf{B}$, is the linchpin that holds this entire computational framework together.

The real world is rarely as simple as a block of steel under a pure mechanical load. The beauty of the strain-displacement framework is its ability to incorporate a vast range of physical phenomena.

Imagine heating a railway track on a summer day. It expands, but not because of a mechanical force. This is thermal strain. In the language of mechanics, the total strain—the actual geometric change we observe and measure—is the sum of the mechanical strain (which causes stress) and the thermal strain. The strain-displacement relation always describes this total, observable geometric strain. This allows us to neatly partition the effects. Using the principle of superposition, which rests on the linearity of the governing equations, we can solve a complex thermo-mechanical problem by solving two simpler ones: a pure mechanical problem and a pure thermal problem, and then adding the results. This elegant connection bridges solid mechanics with thermodynamics.

The framework also scales up to handle highly complex and non-linear behaviors. Consider the design of a gear tooth or a bolt connection. These involve contact: surfaces that can press against each other but cannot pass through one another. Solving such a problem requires adding logical, inequality-based conditions to our model. We might assume there is no contact, solve the problem, and check if the parts have interpenetrated. If they have, our assumption was wrong, and we must re-solve with the condition that the parts are touching and exerting a reaction force on each other. Throughout this iterative, logical dance, the core of the problem remains the same: integrating the strain-displacement relation to find the deformation under a given set of loads and constraints.

What about the most dramatic event in a material's life: fracture? When a crack runs through a material, the displacement field is no longer continuous; it has a jump across the crack faces. Our standard FEM, which assumes smooth displacement fields, fails. Here, modern methods like the Extended Finite Element Method (XFEM) come to the rescue. The idea is wonderfully intuitive: we take our standard approximation for displacement and "enrich" it by adding a special function—like a Heaviside step function—that explicitly introduces a jump. We then apply the same old strain-displacement rule to this new, more sophisticated displacement field. This allows us to model the crack without having the mesh conform to the crack geometry, a massive computational advantage for tracking how cracks grow and, ultimately, for predicting when a structure might fail.

The universality of the strain-displacement relation means it applies regardless of the material's "personality" or the scale at which we view it.

Some materials, like polymers or biological tissues, are viscoelastic. Their response depends on how fast you load them; they exhibit both fluid-like viscous behavior and solid-like elastic behavior. Their constitutive law is not a simple algebraic relation but a differential equation in time. Yet, the kinematic rules do not change. The strain is still the spatial derivative of displacement, and the equation of motion still relates the divergence of stress to acceleration. The purely geometric nature of kinematics provides a stable foundation upon which we can build models for these much more complex, time-dependent materials and study fascinating phenomena like wave propagation and damping.

Finally, the strain-displacement relation is a powerful tool for the art of engineering modeling. For thin structures like airplane wings or car bodies (plates and shells), solving the full 3D elasticity equations is computationally prohibitive. Instead, we create simplified 2D theories by making an intelligent guess—an ansatz—for how the displacement varies through the thin dimension. For example, First-Order Shear Deformation Theory (FSDT) assumes a linear variation, while more refined Third-Order (TSDT) theories assume a cubic variation. By applying the full 3D strain-displacement relations to these assumed kinematic fields, we can derive a consistent and powerful 2D plate theory from first principles. The choice of displacement assumption dictates the accuracy of the model, particularly in how well it captures shear deformation—a critical effect in modern composite laminates.

From the fundamental equations of physics to the most advanced computer simulations, from thermal expansion to fracture, the strain-displacement relation is the constant, reliable thread that unifies the mechanics of deformable solids. It is a testament to the power of separating a problem into its fundamental components—in this case, pure geometry—and a beautiful example of how a simple idea can give rise to a universe of applications.