Normal Stress Differences

When stirred, simple liquids like water or honey are pushed outwards by centrifugal force, creating a dip at the center. But other fluids, from shampoo to industrial polymers, behave in a startlingly different way, climbing up the rotating rod in defiance of gravity. This counter-intuitive behavior reveals the presence of hidden internal forces—​​normal stress differences​​—that are absent in ordinary liquids and challenge our everyday understanding of fluid flow. Our intuition, built upon the concept of viscosity alone, is insufficient to explain these phenomena, pointing to a gap in knowledge that can only be filled by exploring the combined liquid-like and solid-like nature of these complex fluids.

This article delves into these fascinating forces. In the first chapter, ​​"Principles and Mechanisms,"​​ we will uncover the physical origin of normal stress differences, exploring how the stretching of a fluid's microstructure generates forces perpendicular to the flow. We will contrast these viscoelastic materials with simpler fluids and use foundational models to understand why "memory" is the secret ingredient. In the second chapter, ​​"Applications and Interdisciplinary Connections,"​​ we will witness how these principles manifest in the real world, from manufacturing challenges in the plastics industry to the celestial mechanics governing Saturn's rings. To begin, we will investigate the fundamental principles that cause a fluid to defy gravity and climb a rotating rod.

Imagine dipping a rotating rod into a bucket of honey. What do you expect to happen? The honey, a typical viscous liquid, is spun outwards by centrifugal force, and the surface level dips down around the rod. Now, imagine we replace the honey with a solution of long-chain polymers—something like a shampoo or a polymer melt. As the rod spins, something extraordinary and counter-intuitive occurs: the fluid defies gravity and climbs up the rotating rod! This bizarre phenomenon, known as the ​​Weissenberg effect​​, is a dramatic visual clue that we are no longer dealing with simple liquids. These "complex fluids" harbor internal stresses that ordinary fluids like water or honey do not. To understand this strange new world, we must venture beyond viscosity and uncover the principles of elasticity in fluids.

So, what is the secret behind the climbing fluid? The answer lies in forces that are "normal" (perpendicular) to the surfaces on which they act, forces that are generated in addition to the familiar shear stresses that resist flow. To get a handle on this, let's consider the simplest possible shearing motion: ​​simple shear flow​​. Imagine a fluid sandwiched between two large parallel plates. The bottom plate is fixed, and the top plate moves at a constant speed, dragging the fluid along. We can set up a coordinate system: the flow is in the $x$-direction, the velocity changes along the $y$-direction (the gradient direction), and the third, neutral direction is the $z$-direction (the vorticity direction).

In a polymer solution under shear, the long, spaghetti-like polymer chains are stretched and aligned, on average, along the flow direction, $x$. Think of this like stretching a rubber band. A stretched rubber band pulls inwards on its ends. Similarly, the collection of stretched polymer chains generates a tensile stress—an extra "pull"—along the flow direction. This means the normal stress in the $x$-direction, $\sigma_{xx}$, becomes larger than the normal stress in the $y$-direction, $\sigma_{yy}$. This inequality gives rise to the ​​first normal stress difference​​, denoted $N_1$:

For nearly all polymeric fluids, experiments show that $N_1 > 0$ in shear flow. Now, let's return to our climbing rod. The fluid is swirling in a circle. The "flow direction" is now the azimuthal (hoop) direction. A positive $N_1$ manifests as a "hoop stress"—an elastic tension along the circular streamlines, just like the tension in a stretched rubber band wrapped around the rod. This hoop tension squeezes the fluid inwards, towards the rod. This inward force must be balanced by an increase in pressure as one moves towards the center. This pressure gradient pushes the fluid up against gravity, creating the spectacular rod-climbing effect. So, the "magic" is nothing more than the macroscopic consequence of molecules being stretched!

This naturally raises the question: why doesn't this happen with water or honey? These are described by the ​​Newtonian fluid​​ model, where the extra stress ($\boldsymbol{\tau}$) is directly proportional to the rate of deformation ($\mathbf{D}$). The constitutive law is a simple, elegant relationship: $\boldsymbol{\tau} = 2\eta\mathbf{D}$, where $\eta$ is the viscosity.

If we calculate the rate-of-deformation tensor $\mathbf{D}$ for a simple shear flow, we find that its diagonal components ($D_{xx}$, $D_{yy}$, $D_{zz}$) are all zero. Consequently, the normal stresses $\tau_{xx}$, $\tau_{yy}$, and $\tau_{zz}$ are also zero. This means for any Newtonian fluid, $N_1 = \tau_{xx} - \tau_{yy} = 0 - 0 = 0$. No extra tension, no hoop stress, no rod climbing.

You might argue, "But many fluids aren't perfectly Newtonian! Their viscosity changes with the shear rate; they are 'shear-thinning' like paint." That's true, and these are described by ​​Generalized Newtonian Fluid (GNF)​​ models, where the stress is still proportional to the rate of deformation, but the viscosity $\eta$ is now a function of the shear rate $\dot{\gamma}$, i.e., $\boldsymbol{\tau} = 2\eta(\dot{\gamma})\mathbf{D}$. Does this added complexity change anything? As a careful analysis shows, it does not. Since the diagonal components of $\mathbf{D}$ are still zero, the normal stresses remain zero, and $N_1$ is still identically zero. Shear-thinning behavior alone is not enough to make a fluid climb a rod. We are missing a crucial ingredient.

The missing piece of the puzzle is ​​viscoelasticity​​—the combination of viscous (liquid-like) and elastic (solid-like) behavior. Unlike a simple Newtonian fluid, a viscoelastic fluid has a "memory" of its past shape. Its stress today depends not just on the rate of deformation right now, but on its entire history of deformation.

How can we build a mathematical model with memory? One of the simplest and most illuminating is the ​​Upper Convected Maxwell (UCM) model​​. You can think of it as representing the fluid with a combination of a spring (to store elastic energy) and a dashpot (to dissipate energy viscously). The model introduces a new parameter, the ​​relaxation time​​ $\lambda$, which quantifies how long the fluid "remembers" a deformation.

When we apply the UCM model to a simple shear flow, it predicts the shear stress, but it also predicts something remarkable about the normal stresses. Due to a term in the model that accounts for how the stresses are stretched and rotated by the flow (the upper-convected derivative), we find that the first normal stress difference is not zero! Instead, it is given by:

This beautiful result reveals the physics in a nutshell. $N_1$ is positive, just as we observed. It's proportional to the relaxation time $\lambda$—the longer the fluid's memory, the more elastic it is, and the larger the normal stress effect. And it's proportional to the shear rate squared, $\dot{\gamma}^2$, a characteristic signature of this elastic response at low shear rates. By adding the single, simple concept of memory, we have unlocked the origin of the Weissenberg effect. The ratio of this elastic force to the viscous shearing force, $N_1 / \tau_{yx}$, is found to be proportional to the product of the relaxation time and the shear rate, a quantity known as the Weissenberg number, which tells us the degree of elastic effects in the flow.

Knowing that elasticity is the key, we can dig even deeper and ask: what is the microscopic origin of this behavior? The answer depends dramatically on what the fluid is made of. Let's compare two different complex fluids: a polymer melt and a dense suspension of solid particles.

​​1. Polymer Chains and the Second Normal Stress Difference:​​

For polymers, we've seen that the positive $N_1$ comes from the stretching and alignment of chains in the flow. But there is a second, more subtle effect. Physicists also define the ​​second normal stress difference​​, $N_2 = \sigma_{yy} - \sigma_{zz}$. This compares the normal stress in the gradient direction ($y$) with the one in the neutral, vorticity direction ($z$).

For polymers, the tumbling chains are somewhat constrained in their motion by the velocity gradient in the $y$-direction. In the $z$-direction, they are much freer to fluctuate. This subtle difference in confinement makes the stress in the $y$-direction slightly smaller than in the $z$-direction. The result is that $N_2$ is typically negative and much smaller in magnitude than $N_1$ (e.g., $|N_2|/N_1 \approx 0.1 - 0.3$). In fact, some of the simplest molecular models, like the ​​Rouse model​​, fail to capture this subtlety and incorrectly predict $N_2=0$. More advanced models like the ​​Phan-Thien-Tanner (PTT) model​​ must be specifically constructed with parameters that allow one to tune the ratio $N_2/N_1$ to match experimental results, showing how rheologists build progressively better descriptions of reality.

It is worth noting that not all non-Newtonian models get the physics right. For instance, the purely viscous ​​Reiner-Rivlin fluid​​ model is known to be inadequate for polymers, as it leads to unphysical predictions like a positive second normal stress difference ($N_2 > 0$) and can allow for $N_1=0$ while $N_2 \neq 0$—contrary to the dominant behavior seen in experiments. This serves as a powerful reminder that a mathematical model must be rooted in correct physical mechanisms to be useful.

​​2. Dense Suspensions of Particles:​​

Now, let's switch from flexible polymer chains to a dense slurry of hard, solid particles, like sand in water. The physics completely changes. Here, the stresses are dominated by particles colliding and grinding past each other—a phenomenon driven by "jamming" and friction, not entropic elasticity.

In a shear flow, these particles are forced into layers. Motion in the gradient ($y$) direction is heavily constrained as particles from one layer push against the next. This creates a large compressive stress in the $y$-direction. In contrast, particles are much freer to move in the vorticity ($z$) direction. This leads to a stress state where the normal stress in the gradient direction is far more compressive than in the vorticity direction. The result? A large and negative second normal stress difference, $N_2 0$. In stark contrast to polymers, for dense suspensions, $N_2$ is often the dominant normal stress effect. The first normal stress difference, $N_1$, is typically much smaller and can even be negative. This beautiful comparison shows that simply saying a fluid is "viscoelastic" is not enough; the specific microscopic structure dictates the nature of the stresses it can support.

Is there a more fundamental way to think about these elastic stresses, one that doesn't rely on proposing specific models? Here, physics offers a truly profound and beautiful insight through the ​​Fluctuation-Dissipation Theorem​​. One of its consequences in rheology is a relationship called the ​​Lodge-Yamamoto relation​​.

It states that the elastic response of a fluid—specifically, the coefficient related to the first normal stress difference, $\Psi_1 = N_1 / \dot{\gamma}^2$—is directly proportional to the time integral of the random, spontaneous fluctuations of shear stress that occur in the fluid at complete rest.

Think about what this means. If you could somehow measure the fleeting, microscopic shear stresses that are constantly appearing and disappearing in a bucket of polymer solution due to the thermal jiggling of molecules, you could predict how high that fluid will climb a rotating rod! The way the system dissipates these natural fluctuations at equilibrium holds the complete blueprint for its non-equilibrium elastic response. It is a deep and powerful testament to the unity of physics, connecting the random world of statistical mechanics to the deterministic world of continuum mechanics. It's in these moments, where disparate-seeming ideas link up to reveal a simpler, underlying truth, that we can truly appreciate the inherent beauty of the science of flow.

In the previous chapter, we journeyed into the strange and wonderful world of viscoelasticity, uncovering the origin of normal stress differences. We saw that in a fluid with memory, the simple act of shearing creates forces perpendicular to the direction of flow—forces that our intuition, honed by experience with simple liquids like water, would never predict. You might be tempted to think of these as mere mathematical curiosities, confined to the arcane equations of rheologists. But that would be a tremendous mistake. These unseen forces are everywhere, shaping the world in profound and often startling ways. They are at play when we manufacture plastics, when we design advanced machinery, and, as we shall see, even in the majestic dance of planetary rings millions of miles away.

Let us now embark on a tour of the universe as seen through the lens of normal stress differences, and discover how this single concept provides a unifying thread through a vast tapestry of phenomena.

Perhaps the most dramatic and visually striking demonstrations of normal stresses occur in the flow of polymer solutions and melts. Imagine a rotating rod dipped into a vat of water. The water surface near the rod dips down, carved out by centrifugal force trying to fling the liquid outwards. Now, let's replace the water with a viscoelastic fluid, like a solution of long-chain polymers. What happens is astonishing: the fluid climbs the rod, seemingly defying gravity and centrifugal force in equal measure! This is the famed ​​Weissenberg effect​​, and it is a direct visualization of the first normal stress difference, $N_1$.

Think of the circular path of the fluid around the rod. The shearing motion stretches the long polymer chains along this path. This stretching creates tension, much like the tension in a stretched rubber band. If we imagine these tension-filled streamlines as a series of nested hoops, the "hoop stress" they generate squeezes the fluid inward and, with nowhere else to go, forces it up the rod. This effect is a potent one, capable of overcoming not only gravity but also the powerful forces of surface tension that dominate at the micro-scale.

This "elastic memory" also leads to another famous oddity in polymer processing: ​​extrudate swell​​. When you squeeze a Newtonian fluid like honey through a narrow tube, the stream that emerges is thinner than the tube's opening. But when a polymer melt is extruded from a die to make a plastic rod or fiber, the emerging stream swells to a diameter significantly larger than the die itself. Why? Inside the die, the polymer chains are stretched and aligned by the flow, storing elastic energy and generating large normal stresses. As the fluid exits the die, this constraint is released. The molecules recoil and relax, causing the fluid stream to expand in a collective sigh of relief. Understanding and predicting the amount of swell, which is directly related to the first normal stress difference, is absolutely critical for designing accurate molds and dies for manufacturing everything from plastic bottles to synthetic fibers. This release of stored elastic energy even results in a measurable "elastic thrust" on the die itself.

While these effects can be useful, they can also be a source of trouble. Consider a viscoelastic fluid being pumped at high speed around a sharp bend. The tension along the curved streamlines, a manifestation of $N_1$, can become immense. This tension can literally pull the liquid apart, causing the local pressure to plummet below the fluid's vapor pressure. When this happens, bubbles of vapor spontaneously form—a phenomenon known as cavitation. Unlike the cavitation in water caused by high speed flow (the Bernoulli effect), this is a form of elastic cavitation, born from the material's own internal tensions. Preventing this is a major design challenge in polymer processing equipment.

The concepts we've developed are not limited to liquids. The line between a solid and a fluid is, after all, a matter of timescale. And in the world of elasticity, we find the same principles at work. Consider a block of soft rubber. If you subject it to simple shear—by, for example, gluing it between two plates and sliding one past the other—it doesn't just resist the shear. It also pushes the plates apart. This is the ​​Poynting effect​​, the solid-state analog of normal stresses in fluids.

The microscopic picture is nearly identical: the network of cross-linked polymer chains in the rubber is distorted by the shear, leading to an anisotropic stress state that includes a stress component normal to the plane of shear. Fundamental models of rubber elasticity, such as the Neo-Hookean and Mooney–Rivlin models, explicitly predict this positive first normal stress difference, showing that it is an inherent property of these materials. This reveals a beautiful unity: the same underlying physics governs the bizarre behavior of polymer goo and the elasticity of a simple rubber block.

A good engineer sees a problem not as an obstacle, but as an opportunity. The same normal stress effects that can cause instabilities can also be harnessed for clever design. A wonderful example is the ​​viscoelastic thrust bearing​​. A standard thrust bearing, which must support a load between a rotating and a stationary surface, relies on a thin film of lubricant. For a simple Newtonian lubricant, the surfaces must be slightly angled to form a converging "wedge" that generates hydrodynamic pressure and lift.

But what if you use a viscoelastic lubricant? It turns out you can support a load with two perfectly parallel plates! As one plate rotates, the shear flow generates normal stresses in the lubricating fluid. The interplay between the first and second normal stress differences ($N_1$ and $N_2$) creates a remarkable pressure profile: the pressure is lowest at the edge and rises to a peak in the center. This centrally-peaked pressure pushes the plates apart, creating a load-bearing capacity out of thin air—or rather, out of the fluid's elastic memory. An effect that seems to defy logic becomes the very principle of operation.

Thus far, our examples have been earth-bound. But the principles of physics are universal. Let us now travel hundreds of millions of miles to the majestic rings of Saturn. These are not solid structures, but a vast collection of trillions of icy particles, from dust specks to house-sized boulders, all in orbit around the planet. On a large scale, this "sea" of particles behaves like a fluid—a ​​granular fluid​​.

The rings are in differential rotation: particles in inner orbits travel faster than those in outer orbits. This creates a constant, massive shear flow. What happens when you shear a collection of particles? The collisions between them, which are the source of pressure in this "gas," are no longer random. Particles are more likely to collide with higher relative velocity in the direction of flow. This means the pressure is anisotropic: the effective pressure in the azimuthal (flow) direction, $P_{\theta\theta}$, is greater than the pressure in the radial (gradient) direction, $P_{rr}$. This difference, $P_{\theta\theta} - P_{rr}$, is precisely the first normal stress difference, $N_1$!

Astrophysicists who model the dynamics and stability of planetary rings must therefore become rheologists. They use kinetic theory to derive the relationship between the shear rate, the particle density, and the resulting normal stress differences. These stresses play a crucial role in preventing the rings from collapsing into thin strands and in transporting momentum and energy throughout the ring system. It is a stunning realization that the same essential physics that makes a polymer solution climb a spinning rod also governs the structure of one of the most beautiful objects in our solar system.

In the modern era, our understanding is often advanced in the digital world of computer simulation. Molecular dynamics allows us to build a "virtual laboratory" and watch the behavior of atoms and molecules directly. But to make these simulations realistic, we must correctly tell the computer the laws of physics.

Suppose we want to simulate a polymer melt under shear at a constant ambient pressure. A standard barostat algorithm, designed for equilibrium fluids, would try to make the pressure equal in all directions. But we know a sheared viscoelastic fluid has an anisotropic pressure tensor with large normal stress differences! Applying a naive barostat would wage war against the fluid's intrinsic physics, leading to a completely wrong result. Advanced non-equilibrium simulation techniques require barostats that are "aware" of normal stresses. These algorithms don't force the pressure to be isotropic; instead, they control the average pressure while allowing the normal stress differences to develop naturally, matching the physics of the real world.

Furthermore, researchers have developed sophisticated experimental techniques to probe these effects with incredible precision. In methods like Large Amplitude Oscillatory Shear (LAOS), the material is subjected to a sinusoidal shearing motion. For a non-linear material like a polymer melt, the resulting normal stresses oscillate not just at the driving frequency, but at its harmonics, and most curiously, they develop a non-zero time-average, a DC offset that pushes outwards even though the average shear is zero. For material scientists, characterizing this behavior, and understanding how it changes with temperature using principles like time-temperature superposition, is essential for designing the next generation of advanced materials.

From the kitchen to the cosmos, the concept of normal stress differences is a profound and unifying one. It reminds us that our simple, everyday intuition about how things should behave can be beautifully, wonderfully wrong. It is by embracing these counter-intuitive phenomena, by asking why the fluid climbs the rod, that we uncover deeper truths about the nature of matter—truths that connect the mundane task of molding a plastic cup to the magnificent, silent orbiting of ice in the rings of Saturn.