Thermoelastic Constitutive Laws

In the realms of physics and engineering, the interplay between temperature and mechanical forces is a constant and critical consideration. Nearly all materials expand when heated and contract when cooled, a seemingly simple phenomenon that can generate immense internal forces known as thermal stress. Unmanaged, these stresses can lead to the buckling of bridges, the cracking of components, and the failure of complex systems. This article addresses the fundamental question of how to describe, predict, and engineer with these forces by exploring the thermoelastic constitutive laws. We will first uncover the foundational theories in the "Principles and Mechanisms" chapter, examining how total strain is decomposed and how stress arises from constrained thermal deformation. Following this, the "Applications and Interdisciplinary Connections" chapter will demonstrate how these laws govern everything from simple bimetallic strips to the design of advanced aerospace composites and the computational analysis of modern structures.

Now that we have a sense of what thermoelasticity is all about, let’s peel back the layers and look at the engine that drives it. Like any good piece of physics, it starts with a simple, beautiful idea and then blossoms into a rich and complex description of the world. Our journey will take us from a simple stretched rubber band to the subtle dance of heat and vibration inside a solid.

Imagine you are holding a metal rod. What can you do to make it longer? The most obvious way is to pull on it. You apply a force, the material stretches, and we say it is under ​​strain​​. The internal resistance to this strain is what we call ​​stress​​. For small stretches, Robert Hooke gave us a simple rule that has served us well for centuries: stress is proportional to strain. This is the world of pure elasticity.

But there is another way to make the rod longer: heat it up. As you increase its temperature, the atoms jiggle more vigorously and push each other further apart. The rod expands. This is also a type of strain, a ​​thermal strain​​. Here’s the crucial difference: if the rod is free to expand, it does so happily, without any internal stress. It’s a natural, stress-free change of shape.

So, what happens if you do both? What if you heat the rod and pull on it? The elegant answer, which forms the bedrock of our theory, is that you simply add the effects. The total change in shape you observe, the ​​total strain​​ ($\boldsymbol{\varepsilon}$), is the sum of the strain caused by mechanical forces—the ​​elastic strain​​ ($\boldsymbol{\varepsilon}^{e}$)—and the strain caused by the temperature change—the ​​thermal strain​​ ($\boldsymbol{\varepsilon}^{th}$).

This principle of superposition is the key. And it holds a deep physical insight: the material only generates stress in response to the elastic part of the strain. It’s the elastic strain that represents a true distortion of the atomic lattice from its preferred configuration at a given temperature. The thermal strain is just the material adjusting to its new preferred size. We can give this stress-free thermal strain a rather wonderful name: ​​eigenstrain​​. It is the "innate" or "characteristic" strain the material desires. Stress is simply the penalty the material pays for being denied its heart's desire.

Let’s translate this into the language of mathematics, which is how physics tells its most precise stories. We can rearrange our little equation to say that the stress-producing elastic strain is the total strain you measure, minus the strain the material wanted to have because of heat: $\boldsymbol{\varepsilon}^{e} = \boldsymbol{\varepsilon} - \boldsymbol{\varepsilon}^{th}$.

Now, we bring back Hooke's Law, but insist that it only applies to this elastic strain: $\boldsymbol{\sigma} = \mathbb{C} : \boldsymbol{\varepsilon}^{e}$, where $\mathbb{C}$ is the material's stiffness tensor. For a simple ​​isotropic​​ material (one that behaves the same in all directions), the thermal eigenstrain is just a uniform expansion, written as $\boldsymbol{\varepsilon}^{th} = \alpha \Delta T \mathbf{I}$. Here, $\alpha$ is the familiar coefficient of linear thermal expansion, $\Delta T$ is the change from a reference temperature, and $\mathbf{I}$ is the identity tensor, a neat mathematical way of saying "the same in all directions."

Putting it all together, we arrive at the master equation of linear thermoelasticity, often called the ​​Duhamel-Neumann relation​​:

In a more explicit form for isotropic materials using Lamé parameters $\lambda$ and $\mu$, it looks like this:

This equation might seem a bit intimidating, but its story is the one we've just told. The stress ($\boldsymbol{\sigma}$) depends on the total strain ($\boldsymbol{\varepsilon}$) you see, but corrected for the bit of strain ($\alpha \Delta T \mathbf{I}$) the material was going to have anyway because it got hotter. This law isn't just an ad-hoc guess; it can be rigorously derived from the principles of thermodynamics, specifically from a potential function called the Helmholtz free energy. This tells us the law is not just a good empirical fit, but a fundamental feature of nature's bookkeeping of energy.

So, we have a law. How does it actually produce stress? The first way is the most obvious: you get in the way.

Let's go back to our metal rod. This time, we wedge it between two immovable, unyielding walls, so its length cannot change. The total strain $\epsilon$ is forced to be zero. Now, we heat the rod by $\Delta T$. It wants to expand by a thermal strain of $\epsilon^{th} = \alpha \Delta T$. But it can't.

Our law tells us what must happen. The elastic strain becomes $\epsilon^e = \epsilon - \epsilon^{th} = 0 - \alpha\Delta T = -\alpha\Delta T$. The rod has been forced into a state of elastic compression relative to the size it wants to be. The resulting stress is $\sigma = E \epsilon^e = -E\alpha\Delta T$, a compressive stress. The rod is pushing furiously against the walls, which are pushing back. A state of stress has been created simply by preventing the free thermal expansion. This is the essence of a thermo-mechanical ​​boundary value problem​​: the interaction between a material's inherent behavior and the constraints the outside world places upon it.

This is where things get really beautiful. Can an object develop stress from heating even if there are no walls, clamps, or external forces of any kind? Absolutely. All you need to do is heat it unevenly.

Imagine a large, flat disc lying on a table, completely free to expand. Now, you use a magnifying glass to focus sunlight on its very center, making the center hot while the rim stays cool. The hot center wants to expand a great deal. The cool rim only wants to expand a little. But they are part of the same disc; they must stay connected.

What happens is a microscopic tug-of-war. The expanding center pushes outwards on the rim, trying to stretch it. The stiff, cool rim pulls inwards on the hot center, constraining its expansion. The result is a self-equilibrated field of ​​internal stress​​. The hot center is in compression (it's being squeezed by the rim), and the cool rim is in tension (it's being stretched by the center). The object is at war with itself.

The deep reason for this is a wonderful concept called ​​strain compatibility​​. For a strain field to correspond to a real, continuous deformation, its parts must fit together perfectly, like the pieces of a jigsaw puzzle. A non-uniform thermal eigenstrain, like our hot-center disc, is often ​​incompatible​​—it describes a shape that would have gaps or overlaps if it were allowed to happen. The material cannot tear or overlap, so it must deform elastically to force the pieces to fit. This "fixing" strain, the elastic strain, is precisely what generates the stress. The incompatibility of the thermal field acts as an internal source of stress, equivalent to having an invisible field of forces distributed throughout the body.

We've been talking about simple isotropic materials. But the world is filled with materials that have a "grain" or preferred direction. Wood is stronger along the grain. Advanced composites used in aircraft and satellites are made of strong fibers embedded in a matrix. These are ​​anisotropic​​ materials.

For them, the simple scalar coefficient of thermal expansion $\alpha$ becomes a tensor $\boldsymbol{\alpha}$. The material might expand a lot in one direction but very little in another. This opens a new Pandora's box for thermal stresses.

Consider a modern ​​laminate​​, made by gluing together layers of a composite material. Imagine one layer has its stiff fibers running along the x-direction ($0^\circ$), while the layer beneath it has fibers running along the y-direction ($90^\circ$). When you heat this laminate, the $0^\circ$ layer wants to expand a lot in x but little in y. The $90^\circ$ layer wants to do the exact opposite. Since they are bonded together, they fight each other. This creates tremendous internal stresses, which are notorious for causing problems at the free edges of the laminate where the layers try to pull apart.

This story becomes even more dramatic when we consider that material properties, like stiffness, themselves change with temperature. Typically, materials get "softer" when they get hot. If a laminate is heated from one side, the hot, soft top layer may not be able to "fight" as hard, forcing the cooler, stiffer layers below to carry more of the internal load. This causes a complex, evolving dance of stress redistribution during the heating process, posing a major challenge for engineers.

So far, we have only considered things sitting still. What happens when we shake a thermoelastic solid? We send a sound wave through it. A sound wave is an alternating series of compressions and rarefactions.

Here's the magic: when you rapidly compress a piece of the material, you do work on it, and its temperature rises. When it expands, it does work on its surroundings and its temperature drops. This is the ​​thermoelastic effect​​. So, a sound wave is also a temperature wave!

Now, think about the heat generated in the compressed (hot) regions. It wants to flow to the rarefied (cold) regions. Two extreme scenarios are possible:

1. ​​Isothermal Limit (Slow Waves)​​: If the wave vibrates very slowly, there is plenty of time for heat to flow and equalize the temperature. The process occurs at a constant temperature.
2. ​​Adiabatic Limit (Fast Waves)​​: If the wave vibrates very fast, there is no time for any significant heat to flow during a cycle. Hot spots stay hot, and cold spots stay cold.

The interesting part is what happens in between. In this case, heat starts to flow from hot to cold, but it doesn't have time to fully equalize. The flow of heat down a temperature gradient is a one-way street in nature. It is an ​​irreversible​​ process, a fundamental expression of the Second Law of Thermodynamics. This irreversible flow generates entropy and dissipates energy. Where does this energy come from? It's drained from the sound wave itself.

This means the wave will lose amplitude as it travels—it is ​​damped​​ or ​​attenuated​​. Thermoelastic coupling is a universal and fundamental mechanism of energy loss for mechanical vibrations in solids. It’s one of the reasons why you can’t hear an echo forever. The strength of this damping is a delicate function of the material's properties (proportional to $\alpha^2$) and the frequency of the wave. By studying this damping, we can learn a great deal about a material's internal structure. In some exotic theories, heat itself doesn't just diffuse but can travel as a wave ("second sound"), which further modifies this story at extremely high frequencies, showing that there is always more to discover in the intricate dance between heat and motion.

"A uniform temperature is necessary for the repose of a body," a physicist might say. And in principle, it's true. But in our world, uniform temperatures are a rare luxury. The sun shines on one side of a bridge but not the other. A hot liquid is poured into a cool ceramic bowl. An engine block heats up rapidly, while the air around it remains cool. In all these moments, we are witnessing a subtle but powerful drama unfold within the material itself. Heat is not merely "energy"; it is a command to expand or contract. And when that command is met with resistance—from an unyielding support, from a neighboring part made of a different material, or even from a cooler layer within the same object—an internal battle of forces ensues. These are thermal stresses.

Understanding these forces is not just an academic exercise. It is fundamental to the art of engineering. The thermoelastic laws we have explored are the grammar of a language spoken by materials under the influence of heat. Learning this language allows us to prevent catastrophic failures, like a railway track buckling on a summer day, but it also empowers us to create remarkable technologies, from the simple, reliable thermostat that controls our home's temperature to advanced composite materials that maintain their shape in the extreme environments of outer space. Let us now take a journey through some of these applications, to see how these elegant principles shape the world we have built.

Imagine a simple steel beam. Heat it up, and it grows. Let it cool, and it shrinks. As long as it is free to do so, it feels no stress. It is in a state of happy, unhindered thermal expansion. But what if we trap it? What if we wedge it perfectly between two immovable stone walls and then raise its temperature?

The beam wants to expand by a certain fractional amount, let's say $\alpha \Delta T$, where $\alpha$ is its coefficient of thermal expansion and $\Delta T$ is the temperature change. But the walls say, "No, you cannot." To prevent this expansion, the walls must push on the ends of the beam, compressing it back to its original length. How much stress does this create? It's a simple and beautiful calculation. To counteract the thermal strain, a mechanical strain of $-\alpha \Delta T$ must be imposed. From Hooke's Law, the stress required to do this is simply Young's modulus times the strain: $\sigma = -E \alpha \Delta T$. A simple formula, yet it holds a tremendous lesson. This compressive stress can be immense, easily enough to buckle a slender beam or crush a brittle one. This single idea explains why engineers must leave expansion gaps in bridges and large structures; they are giving the material "room to breathe" thermally.

Nature becomes even more interesting when we join two different materials together. Imagine a composite bar made of an aluminum segment and a steel segment, bonded end-to-end and then locked between two rigid walls. Aluminum, for the same temperature increase, wants to expand much more than steel. When the whole assembly is heated, the aluminum pushes on the steel, and the steel pulls back on the aluminum. Neither is happy. The entire bar is forced to settle on a single, shared expansion that is too small for the aluminum and too large for the steel. The result? The aluminum segment finds itself in a state of compression, frustrated that it couldn't expand more, while the steel segment is in a state of tension, having been stretched by its overeager neighbor. This principle of "thermal mismatch" is a constant concern in engineering, from microchips where silicon is bonded to other materials, to large aircraft frames built from both aluminum and titanium alloys.

So far, we have imagined uniform heating. But what if the temperature is not uniform? Suppose we take a beam and heat only its top surface, leaving the bottom surface cool. The top fibers of the beam want to expand, while the bottom fibers are content to stay as they are. What can the beam do? The only way to accommodate this is to bend. The top, longer surface becomes the outside of a curve, and the bottom, shorter surface becomes the inside.

This bending is not arbitrary. If the temperature varies linearly from the bottom to the top, the beam will bend into a perfect circular arc. The curvature, $\kappa$, is found to be directly proportional to the thermal expansion coefficient $\alpha$ and the temperature gradient $\gamma$: $\kappa = \alpha \gamma$. This is the principle behind the bimetallic strip, the heart of many old thermostats and thermal switches. Two strips of different metals, say brass and steel, are bonded together. When heated, the brass ($\alpha_{\text{brass}} \gt \alpha_{\text{steel}}$) tries to expand more than the steel, forcing the strip to curl. This curling motion can make or break an electrical contact, elegantly turning a physical law into a control mechanism. This effect, while useful in thermostats, can be a nuisance elsewhere. For example, the precise athermalization of optical systems—ensuring that lenses in a telescope or camera do not go out of focus as the temperature changes—is a high-stakes engineering game played against the laws of thermoelastic bending.

Beams are simple, but much of our world relies on containing things under pressure and at high temperatures. Think of boilers, nuclear reactor pressure vessels, chemical pipes, and even soda bottles. These are often cylindrical or spherical shapes. Here, the interplay of geometry, pressure, and temperature gradients creates beautifully complex stress patterns.

Consider a thick-walled pipe, hot on the inside and cool on the outside. The inner layers want to expand more than the outer layers. This creates a kind of internal "fight": the hot inner surface is compressed by the cooler, more rigid outer layers, while the outer layers are pulled into tension. Now, add internal pressure, which wants to stretch the pipe and create tension. The genius of linear thermoelasticity is that we can often use superposition: we calculate the stresses from the pressure (the classic Lamé problem) and the stresses from the temperature gradient separately, and then simply add them together to get the total stress. This allows engineers to design vessels where the compressive thermal stress on the inside surface actually helps to counteract the tensile stress from the internal pressure, making the vessel safer and stronger.

The analysis also depends on the geometry of the part. In a very long pipe, the material can't easily contract or expand in the axial direction, a condition we call "plane strain." In a thin plate, however, the material is free to expand or contract through its thickness, leading to a state of "plane stress." The resulting stress fields can be quite different. For instance, if you were to completely encase a small block of material and heat it, the hydrostatic (inward pressure) stress it develops would be much higher under plane strain conditions than under plane stress conditions, because in plane strain, expansion is frustrated in all three directions.

Perhaps the most elegant illustration is a hollow sphere undergoing a uniform temperature change. If it is free to expand, it simply grows larger, with every point moving radially outward. The shape remains a perfect sphere. In this ideal case, no stress develops at all. The geometry perfectly accommodates the thermal expansion. But now, imagine constraining this expansion, for instance by building a rigid scaffold inside that prevents any radial motion. When heated, the shell desperately wants to expand, but it cannot. The result is a massive, uniform compressive stress within the shell material. This stark contrast between the free and constrained cases beautifully illustrates a central theme of thermoelasticity: it's not the temperature change itself that creates stress, but the constraint of the resulting deformation.

Does an object "remember" the thermal stresses it has been through? If the material is perfectly elastic, the answer is a surprising "no". Imagine taking a solid cylinder, plunging it into an ice bath so it cools rapidly from the outside in. During this process, complex transient stresses will arise. The cold outer shell contracts and squeezes the hot core, while the hot core pushes back. But, as the entire cylinder eventually reaches a uniform, cool temperature, and as long as it has remained perfectly elastic throughout, all these stresses will completely vanish. The final state is stress-free.

But what if the transient stresses are so large that they cause the material to yield, to deform plastically? This is where true "residual stress" is born. While the outer shell is cooling and contracting, it might be stretched plastically by the still-hot and expanded core. Later, as the core cools and contracts, it pulls on the already-cold and rigid outer shell. The end result, after everything has cooled, is not a stress-free state. The outer layers are left in a state of compression, and to balance this, the inner core is left in a state of tension. The material now has a "memory" of the violent cooling process locked into its structure. This is not a defect; it is a powerful engineering tool! The compressive residual stress on the surface of "tempered" or "toughened" glass is what makes it so much stronger than normal glass. The same principle, known as autofrettage, is used to strengthen cannon barrels and high-pressure vessels.

We can take this idea of controlling properties even further, down to the microscopic level. What if we don't like the thermal expansion coefficient of a material? We can design a new one. Consider a composite made of high-stiffness fibers embedded in a softer matrix. The fibers might have a very low (or even negative) thermal expansion, while the matrix has a high one. When heated, the matrix tries to expand, but it is held back by the stiff, unmoving fibers. The resulting effective thermal expansion of the composite is not a simple average of its components; it is a stiffness-weighted average. By cleverly choosing the materials and their volume fractions, engineers can design composites with near-zero thermal expansion, which are essential for precision telescopes, satellite structures, and sensitive electronics that must maintain their dimensions perfectly over wide temperature ranges.

The principles we've discussed are beautiful, but applying them to a real-world object like a car engine or a turbine blade, with all its complex curves and holes, is impossible to do with pen and paper. This is where the final, crucial connection is made: the bridge to computational mechanics.

The finite element method (FEM) translates the continuous physical laws of thermoelasticity into a discrete problem that a computer can solve. The complex part is broken down into a mesh of small, simple elements (like tiny bars or bricks). Within each element, the laws we've studied are applied. The principle of virtual work provides a powerful framework to assemble the equations for the entire structure. A fascinating and elegant trick used in these programs is the concept of "equivalent nodal loads." The internal urge of an element to expand due to a temperature change can be mathematically represented as a set of forces acting on its corners, or nodes. The computer then solves for the displacements and stresses under the action of these thermal "ghost forces" and any real external forces. This allows engineers to visualize the invisible world of thermal stress, to predict hot spots, to prevent failure, and to optimize their designs before a single piece of metal is ever cut.

What we find is that the core ideas—the simple constitutive law relating stress to strain and temperature, the principle of superposition, and the conditions of equilibrium and compatibility—remain the unshakable foundation upon which these powerful modern tools are built. They are a testament to the enduring power and beauty of fundamental physics.