The Volumetric-Isochoric Split in Continuum Mechanics

When you stretch a rubber band or knead dough, you are witnessing two fundamental types of deformation: a change in volume and a change in shape. In physics and engineering, nearly every material deformation is a complex mixture of these two actions. This raises a crucial question: how can we mathematically untangle this mixture to understand how a material separately resists being squeezed versus being sheared? The answer lies in a powerful and elegant concept known as the volumetric-isochoric split.

While the total deformation is captured by a single mathematical object, the deformation gradient, its raw form lumps these distinct physical responses together. This article unpacks this concept, providing a clear path from fundamental theory to practical application. We will first explore the "Principles and Mechanisms" behind the split, detailing the mathematical tools used to isolate volume change from shape distortion and how this decomposition allows us to model a material's internal energy and stress. Following this, the "Applications and Interdisciplinary Connections" section reveals why this is more than just a mathematical convenience. We will see how the principle is essential for modern computational simulations, connects mechanics to fields like thermodynamics and plasticity, and even provides a robust framework for building physics-informed artificial intelligence for materials science.

Imagine you are stretching a rubber band. It gets longer, of course, but it also gets thinner. You're changing both its shape and its size (its volume) simultaneously. Or think of kneading dough: you are constantly changing its shape, but its total volume stays more or less the same. It turns out that nearly every deformation you can imagine—from the inflation of a car tire to the gentle swelling of a soft biological tissue—is a mixture of these two fundamental effects: a change in volume (a ​​dilatation​​) and a change in shape (a ​​distortion​​ or ​​shear​​).

For a physicist or an engineer, this presents a wonderful puzzle. How can we mathematically describe a deformation in a way that cleanly separates these two actions? If we can do this, we can begin to understand how materials resist each action separately. How much energy does it take to just squeeze a material, versus just shearing it? The journey to answer this question reveals a beautiful piece of mathematical physics, the ​​volumetric-isochoric split​​.

Our main tool for describing any deformation is a mathematical object called the ​​deformation gradient​​, denoted by the symbol $\boldsymbol{F}$. You can think of $\boldsymbol{F}$ as a complete instruction manual for the deformation. It takes any infinitesimally small fiber in the original, undeformed body and tells you exactly how it has been stretched and rotated to arrive at its new state in the deformed body. A tiny line element $\mathrm{d}\boldsymbol{X}$ in the original body becomes $\mathrm{d}\boldsymbol{x} = \boldsymbol{F} \mathrm{d}\boldsymbol{X}$ in the new one.

While $\boldsymbol{F}$ contains all the information, it's all mixed together. Rotation, shearing, and volume change are all bundled up inside. Our first task is to find a way to isolate just the change in volume. You might first guess that a simple measure, like the average stretch of the coordinate axes (the trace of the matrix, $\mathrm{tr}\,\boldsymbol{F}$), could tell us about the volume change. It seems plausible, but it turns out to be wrong. It's entirely possible to cook up two different deformations that have the exact same trace but produce completely different volume changes. One might preserve volume perfectly, while the other causes significant expansion. This tells us we need a more sophisticated tool.

That tool is the ​​determinant​​ of the deformation gradient, a single number we call the ​​Jacobian​​, symbolized by $J = \det \boldsymbol{F}$. This number, it turns out, is the one true measure of how the local volume changes.

Imagine a tiny, tiny cube in the material before deformation, with a volume we'll call $\mathrm{d}V$. After the deformation, this cube will have been squashed and stretched into a little parallelepiped with a new volume, $\mathrm{d}v$. The fundamental relationship between these volumes is beautifully simple:

If a deformation causes a material to swell to twice its local volume, then $J=2$. If it's compressed to half its volume, $J=0.5$. And if the shape changes but the volume stays exactly the same—like our kneading dough—this is called an ​​isochoric​​ deformation, and it corresponds to the special case where $J=1$.

If we consider a simple stretching along the main axes by factors of $\lambda_1, \lambda_2,$ and $\lambda_3$, the new volume of a unit cube is simply $\lambda_1 \times \lambda_2 \times \lambda_3$. This product is precisely the determinant of the diagonal deformation gradient matrix, so we can see intuitively why $J = \lambda_1 \lambda_2 \lambda_3$ captures the volume change. The Jacobian $J$ has unlocked the secret of the squeeze.

Now that we have a dial, $J$, that tunes the volume, how do we "factor out" its effect from the total deformation $\boldsymbol{F}$? The wonderfully elegant idea is to split $\boldsymbol{F}$ not by adding, but by multiplying. We propose that any deformation can be seen as a sequence of two simpler ones:

Here, $\boldsymbol{F}_{\text{vol}}$ is a deformation that only changes volume, and $\boldsymbol{F}_{\text{iso}}$ is one that only changes shape (isochoric). What should $\boldsymbol{F}_{\text{vol}}$ look like? A pure volume change, with no change in shape, is an equal stretch in all directions—an isotropic dilatation. We can represent this with a scalar multiple of the identity tensor $\boldsymbol{I}$. To get the total volume change right, this scaling factor must be $J^{1/3}$. The $1/3$ power comes from the fact that we are in three dimensions, where volume scales with the cube of length. So, $\boldsymbol{F}_{\text{vol}} = J^{1/3}\boldsymbol{I}$.

With this, we can now solve for the shape-changing part. We define a new, modified deformation gradient, usually written as $\bar{\boldsymbol{F}}$:

By its very construction, the determinant of $\bar{\boldsymbol{F}}$ is guaranteed to be 1, meaning it represents a purely isochoric (volume-preserving) transformation. We have successfully split the total deformation into a pure dilatation ($J^{1/3}$) and a pure distortion ($\bar{\boldsymbol{F}}$).

This split naturally extends to the strain tensors we use in practice. For instance, the ​​right Cauchy-Green tensor​​ $\boldsymbol{C} = \boldsymbol{F}^{\mathsf{T}}\boldsymbol{F}$, which measures squared-length changes and is cleverly immune to rigid body rotations, also gets a split. Its isochoric counterpart, $\bar{\boldsymbol{C}}$, is found to be:

The exponent $-2/3$ arises directly from squaring the $J^{-1/3}$ factor in the definition of $\bar{\boldsymbol{F}}$. Just as with $\bar{\boldsymbol{F}}$, the determinant of $\bar{\boldsymbol{C}}$ is always 1, signifying its purely distortional nature.

This mathematical decomposition might seem like a clever formal trick, but its real power is physical. It allows us to build models of materials that reflect a deep truth about how they behave.

Think of a water-filled balloon. It is incredibly difficult to squeeze the balloon to a smaller volume, because water is nearly incompressible. This is a high resistance to volume change (a high ​​bulk modulus​​). However, it is quite easy to change the balloon's shape by squishing it in one direction while letting it bulge in others. This is a low resistance to shape change (a low ​​shear modulus​​).

Most soft materials, from rubber to biological tissues, behave this way. To describe them accurately, we need our theory of stored energy to respect this split personality. We postulate that the total strain energy density, $W$, can be written as the sum of a purely volumetric part and a purely isochoric part:

Here, $U(J)$ is the ​​volumetric energy​​, which depends only on the volume ratio $J$. It acts like a powerful penalty, shooting up in value if $J$ dares to stray from 1, capturing the material's immense resistance to being compressed. $W_{\text{iso}}(\bar{\boldsymbol{C}})$ is the ​​isochoric energy​​, describing how energy is stored during shape-changing distortions at constant volume. This formulation is not just an elegant choice; it's essential for preventing our models from predicting unphysical behaviors, such as a pure shear deformation generating a pressure, or a pure volume expansion creating shear stresses.

This beautiful separation of energy leads directly to a separation of stress. The total stress $\boldsymbol{\sigma}$ inside the material splits neatly into two parts. One is the ​​hydrostatic pressure​​, $p$, which arises solely from the volumetric energy $U(J)$. The other is the ​​deviatoric stress​​, which accounts for shearing and arises solely from the isochoric energy $W_{\text{iso}}$. The pressure is the material's response to being squeezed, and the deviatoric stress is its response to being distorted.

It is worth noting that the volumetric-isochoric split is not the only way to factor the deformation gradient. Another famous method is the ​​polar decomposition​​, $\boldsymbol{F} = \boldsymbol{R}\boldsymbol{U}$. This split is different: it separates the local rigid-body ​​rotation​​ ($\boldsymbol{R}$) from the pure ​​stretch​​ ($\boldsymbol{U}$). In this case, the stretch tensor $\boldsymbol{U}$ still contains both shape change and volume change mixed together.

The two decompositions answer different questions:

• ​​Polar Decomposition ($\boldsymbol{F}=\boldsymbol{R}\boldsymbol{U}$)​​: "Ignoring rotation, how has the material been stretched?"
• ​​Volumetric-Isochoric Split ($\boldsymbol{F}=J^{1/3}\bar{\boldsymbol{F}}$)​​: "Ignoring volume change, how has the material been distorted and rotated?"

This highlights the richness of continuum mechanics. We can even combine these ideas in a three-part decomposition, $\boldsymbol{F} = (J^{1/3}\boldsymbol{I})\bar{\boldsymbol{R}}\bar{\boldsymbol{U}}$, which isolates all three fundamental components of deformation: pure volume change ($J^{1/3}\boldsymbol{I}$), pure rotation ($\bar{\boldsymbol{R}}$), and pure, volume-preserving shape change ($\bar{\boldsymbol{U}}$).

For most problems in the mechanics of soft, nearly incompressible materials, however, it is the volumetric-isochoric split that takes center stage. It provides the essential conceptual and mathematical bridge between the geometry of deformation we can see and the energetic and stress response of the material we seek to understand and predict. It is a perfect example of how choosing the right mathematical description can reveal the inherent beauty and unity of the physical world. For anisotropic materials, this clean separation is typically a modeling assumption, a choice to build a material model that behaves in this decoupled way.

You might be thinking that this whole business of splitting deformation into a part that changes volume and a part that changes shape is a rather clever mathematical trick. A fine bit of bookkeeping for the theoretically inclined, perhaps, but what is it good for? Well, it turns out that this isn't just a trick; it's a deep physical insight. Nature herself seems to have adopted this division of labor. One mechanism is responsible for resisting compression, like the force between atoms getting too close, and a completely different mechanism is responsible for resisting a change in shape, like the uncoiling of long polymer chains. By separating these two effects in our theory, we haven't just simplified the math; we have started to think like the material itself. And when you do that, a whole world of applications and connections suddenly snaps into focus, from building better tires to designing new materials with artificial intelligence. Let's take a tour.

The first and most direct use of our split is in the craft of constitutive modeling—the art of writing down the mathematical laws that describe a specific material. When an engineer looks at a deforming body, they want to know two things above all else: what is the pressure inside it, and what are the shear stresses? Pressure, a spherical stress, tells you if a balloon will pop or a submarine will implode. Shear, or deviatoric stress, tells you if a beam will tear or an axle will twist apart. The beauty of the volumetric-isochoric split of the energy is that it elegantly separates the resulting stress into these very two parts. The volumetric part of the energy gives rise only to the pressure-like spherical stress, while the isochoric part gives rise only to the shape-distorting deviatoric stress. This gives us a powerful toolkit. We can now build a model for a material piece by piece.

First, how does it resist being squeezed? We can perform a hydrostatic compression test, submerging the material in a fluid and increasing the pressure. By measuring how its volume $J$ changes with pressure $p$, we can work backward to deduce the exact mathematical form of the volumetric energy, $U(J)$. This part of the model isn't guesswork; it's a direct reflection of experimental data.

Then comes the more subtle part: how does it resist changing shape? This is where the physics of different materials really shines. For a soft rubber, for example, the resistance to shape change comes from the entropy of its tangled network of long polymer chains. When you stretch the rubber, you are uncoiling these chains, making them more ordered and thus decreasing their entropy. This creates a restoring force. More advanced models, like the Gent model, even account for the fact that these chains have a finite length and cannot be stretched indefinitely. As the material approaches its maximum extension, it becomes incredibly stiff. This "locking" behavior is captured beautifully and entirely within the isochoric part of the strain energy, linking a macroscopic stiffening effect to the microscopic physics of polymer chains. For other materials, we might not have such a neat microscopic picture. In these cases, we can use more general, phenomenological forms like the Ogden model. This model uses a series of terms with different parameters that can be adjusted to match the material's measured response in shear. The key point is that we can tune these shear parameters, which live in the isochoric energy function, without messing up the carefully calibrated volumetric response, and vice-versa. We are building a composite sketch of the material, one fundamental behavior at a time.

Having a good material model is one thing; using it to predict the behavior of a complex object, like a car tire hitting a pothole or a stent expanding in an artery, is another. This is the domain of computational mechanics, particularly the Finite Element Method (FEM). Here, a complex object is broken down into a mesh of simpler "elements," and the laws of physics are solved on this mesh. And it is here that our humble split saves engineers from a notorious numerical trap: ​​volumetric locking​​.

Nearly incompressible materials, like rubber and biological tissues, are everywhere. Their volume is very hard to change, but their shape is easy to change. When you try to simulate such materials with standard, low-order finite elements, you often find that the model becomes pathologically, artificially stiff. It "locks up" and refuses to deform, even in ways that should be easy, like simple bending. It's a disaster! The simulation gives completely wrong answers.

The volumetric-isochoric split is the lantern that illuminates the cause of this pathology. The numerical procedure, when naively applied, inadvertently enforces the incompressibility constraint at too many points inside each small element. For a simple element that only has enough flexibility to produce a linear change in volume, the simulation tries to force this volume change to be zero at four different points. The only way to satisfy this is for the volume change to be zero everywhere, which over-constrains the element and prevents it from deforming even in pure shear.

Once you see the problem in this light, the solutions become clear. The error is in how we handle the volumetric part, not the isochoric part. So, we treat them differently! Techniques like ​​Selective Reduced Integration (SRI)​​ use a fine, accurate numerical scheme for the isochoric (shear) energy but a less stringent, "reduced" scheme (e.g., evaluating it at only one point) for the volumetric energy. This effectively enforces the incompressibility constraint in an averaged sense over the element, rather than pointwise, giving the element the freedom it needs to deform correctly. This idea, born directly from the energy split, is a cornerstone of modern software for simulating soft materials and structures.

The influence of the volumetric-isochoric split extends far beyond rubber and finite elements. It provides a unifying thread connecting mechanics to other branches of physics.

Consider ​​thermodynamics​​. What happens when you rapidly stretch a rubber band? If you touch it to your lip, you'll feel it get warmer. When you let it contract, it cools down. This is thermoelasticity in action. Where does this coupling between temperature and deformation come from? Our split points the way. Thermal expansion is a change in volume, not shape. So, it stands to reason that the primary connection to temperature should live in the volumetric part of the energy, $\Phi(J, \theta)$. By making the material's bulk modulus a function of temperature, for instance, we can model how the material expands when heated. But the beauty is deeper. The laws of thermodynamics then allow us to derive a Maxwell relation that directly links the change in the material's entropy to its change in volume. This explains the rubber band effect: stretching it changes its shape (isochoric), but the underlying micro-structural rearrangement also causes a tiny volume change, which through the thermo-volumetric coupling, changes its entropy and temperature.

Now, think about a different class of materials: ​​metals​​. When you bend a paperclip, it stays bent. This permanent deformation is called plasticity. At the microscopic level, plasticity in crystalline metals involves layers of atoms slipping past one another. Critically, this slipping process conserves volume almost perfectly. Plasticity is an isochoric phenomenon! This makes the split a natural framework for elastoplasticity. The material's total deformation is split into an elastic part and a plastic part. The elastic part can then be further analyzed using our volumetric-isochoric split. The plastic flow—the permanent, irreversible deformation—is driven only by the deviatoric (isochoric) part of the stress. The material's resistance to volumetric compression remains purely elastic and is completely unaffected by whether the material is yielding in shear. This not only makes perfect physical sense but also vastly simplifies the mathematical structure of advanced computational models for plasticity.

We now stand at the threshold of a new era in materials science, one powered by data and machine learning. Imagine performing complex experiments and feeding the results to a neural network, letting it learn the material's behavior directly. This is a tantalizing prospect, but it comes with a peril: a "black-box" AI that simply memorizes data without understanding physics is brittle and untrustworthy. It may violate fundamental laws like the conservation of energy or the principle that a material's properties shouldn't depend on how you're looking at it (objectivity).

Here again, the volumetric-isochoric split provides a guiding principle for building smarter, more reliable AI. Instead of asking the neural network to learn everything from scratch, we can build the physical principles right into its architecture. This is the idea behind ​​Physics-Informed Machine Learning​​. We can design a neural network that is tasked with learning only the complex, nonlinear, and hard-to-model isochoric response of a material. Meanwhile, the volumetric response can be handled by a simple, classical equation that we know is physically correct. By constructing the total energy as an explicit sum of the prescribed volumetric part and the learned isochoric part, we guarantee—by construction—that the resulting model is hyperelastic, objective, and respects the fundamental decoupling of volume and shape change.

The split acts as a set of guardrails for the AI, allowing it to explore the vast space of possible material behaviors without ever wandering off the path of physical reality. It is a perfect marriage of the deep insights of classical mechanics and the powerful capabilities of modern artificial intelligence.

From the simple act of stretching a rubber band to the frontiers of AI-driven science, the principle of separating volume change from shape change proves to be an astonishingly powerful and unifying concept. It is a beautiful reminder that in science, the most profound ideas are often the simplest ones—those that provide a new and clearer lens through which to view the world.