Local Thermal Non-Equilibrium (LTNE) Model

Heat transfer within a porous material—like water flowing through sand or air through a metal foam—presents a fundamental choice of perspective. Do we track the thermal energy in every individual grain and pore, or can we step back and describe the system with a simplified, continuous model? While this "averaged" view is powerful, it often relies on the assumption of Local Thermal Equilibrium (LTE), where the solid and fluid phases are considered to be at the same temperature everywhere. This article addresses the crucial question: what happens when this assumption fails, and the solid and fluid exist at different local temperatures? It delves into the Local Thermal Non-Equilibrium (LTNE) model, a framework designed for this exact scenario.

This article will guide you through the two-temperature world of LTNE. In the first part, ​​Principles and Mechanisms​​, we will deconstruct the two-equation model that forms the heart of LTNE, exploring the physical meaning behind each term and the conditions that drive the system away from equilibrium. In the second part, ​​Applications and Interdisciplinary Connections​​, we will see the model in action, demonstrating its essential role in solving real-world challenges in chemical engineering, materials science, and even microscale physics.

Imagine you are looking at a beautiful pointillist painting by Georges Seurat. From a distance, your eyes blur the millions of tiny, distinct dots of color into a smooth, coherent image of a park scene. You perceive continuous shades of green, blue, and yellow. But step up close, and the illusion shatters. You see only individual, separate dots. The world of heat transfer in porous media—think of water flowing through sand, air through a sponge, or coolant through a high-tech metal foam—presents us with a similar choice of perspective. Do we track the heat in every single grain of sand and every microscopic pocket of water? Or can we step back and describe the system with a "blurry," averaged, continuous picture?

The answer, just like with the painting, is that the blurry view is incredibly useful, but only if we respect the rules of perspective. This is the first and most fundamental principle of our journey.

To treat a complex, heterogeneous material as a continuous medium, we need what physicists call a ​​Representative Elementary Volume (REV)​​. This is a small, imaginary cube of the material that is, on the one hand, much larger than the individual grains or pores (the "dots" of our painting), so it captures a fair, statistical average of the microstructure. On the other hand, this cube must be much smaller than the overall size of the system and the scale over which things like temperature are changing.

This leads to a crucial condition of ​​scale separation​​. The characteristic length of the microscopic structure, let's call it the correlation length $\ell_c$, must be much, much smaller than the macroscopic length scale, $L$, over which the temperature field is changing. There must exist a "window" for the size of our averaging volume, $\ell_{\text{REV}}$, such that $\ell_c \ll \ell_{\text{REV}} \ll L$. If this condition holds, our averaging process works beautifully. The properties we calculate, like thermal conductivity, become stable and independent of the exact size of our averaging box, a defining signature of a valid REV.

But what happens if this condition breaks down? What if we try to model heat flow in a material where the pore size is comparable to the size of the whole object ($L \sim \ell_c$)? The concept of an REV collapses. Our "blurry" picture becomes meaningless. The effective properties we measure would depend on the size of our sample, and the heat flux at one point would depend not just on the temperature gradient at that exact spot, but on the temperature field in a whole neighborhood around it. The physics becomes ​​nonlocal​​, and our simple, continuous models fail us. For the rest of our discussion, we will assume we are in a situation where this scale separation holds, and we are allowed to use the powerful tool of averaging.

Having decided we can use a "blurry" or averaged description, we face a new, profound question. Within our averaged viewpoint, at any given location $(x,y,z)$, does everything have the same temperature?

The simplest assumption is ​​Local Thermal Equilibrium (LTE)​​. This model posits that even though the material is made of distinct solid and fluid phases, heat transfer between them is so incredibly efficient that, at our macroscopic scale, they are always at the same temperature. We can describe the entire system with a single temperature field, $T(\mathbf{x}, t)$. This is a beautiful simplification, and it works remarkably well in many situations.

But nature is not always so simple. What if the heat transfer between the phases is not infinitely fast? What if one phase is heated internally, but the other is not? In these cases, the solid and the fluid, at the very same macroscopic location, can have different average temperatures. This is the world of ​​Local Thermal Non-Equilibrium (LTNE)​​. To describe it, we must abandon the simplicity of a single temperature and embrace the reality of two: a temperature for the fluid, $T_f(\mathbf{x}, t)$, and a separate temperature for the solid, $T_s(\mathbf{x}, t)$.

To capture the physics of LTNE, we write down the law of energy conservation—the first law of thermodynamics—separately for each phase. This gives us a pair of coupled equations that form the heart of the LTNE model. Let's look at them, not as abstract mathematics, but as a story about the life of heat in each phase.

For the fluid phase, the story goes like this: $\varepsilon \rho_f c_{pf} \frac{\partial T_f}{\partial t} + \varepsilon \rho_f c_{pf} \mathbf{u} \cdot \nabla T_f = \nabla \cdot (k_{f,\text{eff}} \nabla T_f) + h_{sf} a_{sf} (T_s - T_f) + \varepsilon \dot{q}_f'''$

And for the solid phase: $(1-\varepsilon) \rho_s c_{ps} \frac{\partial T_s}{\partial t} = \nabla \cdot (k_{s,\text{eff}} \nabla T_s) + h_{sf} a_{sf} (T_f - T_s) + (1-\varepsilon) \dot{q}_s'''$

Let's break this down. Each term tells part of the story:

• ​​Storage:​​ The terms on the far left, like $\varepsilon \rho_f c_{pf} \frac{\partial T_f}{\partial t}$, describe how much heat is being stored or released over time in each phase within a unit volume. The porosity $\varepsilon$ (the fraction of volume occupied by the fluid) appears because these equations are written per unit of total bulk volume.

• ​​Advection:​​ The term $\varepsilon \rho_f c_{pf} \mathbf{u} \cdot \nabla T_f$ appears only in the fluid equation. It tells us about the heat that is physically carried along by the moving fluid. An interesting subtlety here is the velocity $\mathbf{u}$. Physicists use two types of averaged velocity: the ​​superficial velocity​​ $U$, which is the total flow rate divided by the total area (solid included), and the ​​interstitial velocity​​ $u = U/\varepsilon$, which is the actual average speed of fluid particles within the pores. The form of the advection term depends on which velocity you use, but the underlying physics is the same: faster flow means more heat transport.

• ​​Conduction:​​ The terms like $\nabla \cdot (k_{f,\text{eff}} \nabla T_f)$ describe the diffusion of heat through each phase, governed by their respective effective thermal conductivities. This is heat spreading out, like warmth from a hot poker.

• ​​Source:​​ The terms $\dot{q}'''$ represent any internal heat generation, perhaps from a chemical reaction or microwave heating.

• ​​The Crucial Link:​​ The final term, $h_{sf} a_{sf} (T_s - T_f)$, is the most important of all. This is the ​​interfacial heat exchange​​ term. It's the bridge that connects the two separate worlds of the fluid and the solid. Notice its form: it's proportional to the temperature difference $(T_s - T_f)$. If the solid is hotter, this term is positive in the fluid equation (a heat source for the fluid) and negative in the solid equation (a heat sink for the solid). Heat flows from hot to cold, and the total energy is perfectly conserved. This single term is the mathematical embodiment of the non-equilibrium interaction.

The interfacial exchange term is often written as $H(T_s - T_f)$, where $H = h_{sf} a_{sf}$ is the volumetric interfacial heat transfer coefficient. But what really is this coefficient $H$? It's not a fundamental constant of nature; it is a composite character, the product of two very different physical quantities.

• $a_{sf}$: The ​​specific surface area​​. This is a purely geometric property of the porous matrix. It is the total surface area of the solid-fluid interface packed into a unit of bulk volume. A fine-grained sand has a much larger $a_{sf}$ than a coarse gravel. A metal foam designed for heat exchange will have an enormous $a_{sf}$. Just like a radiator uses fins to increase its surface area, a porous medium with a high $a_{sf}$ has a large capacity for heat exchange. For a fixed geometry, this value is constant.

• $h_{sf}$: The ​​interfacial heat transfer coefficient​​. This is where the dynamics come in. It measures the efficiency of heat transfer per unit of interface area. How easily can a parcel of heat jump from the solid surface into the fluid? The answer lies in the physics of the thin ​​thermal boundary layer​​ in the fluid right next to the solid surface. The coefficient $h_{sf}$ is essentially the fluid's thermal conductivity, $k_f$, divided by the thickness of this boundary layer, $\delta_T$. Anything that makes this boundary layer thinner will increase $h_{sf}$ and improve heat transfer.

  • In slow, conduction-dominated flows, the boundary layer is thick, roughly the size of the pores themselves.
  • In faster, convection-dominated flows, the moving fluid shears the boundary layer and makes it much thinner, dramatically increasing $h_{sf}$.
  • Furthermore, the very shape of the pores matters! Curvy passages can induce secondary flows that scrub the surface, thinning the boundary layer. Microscopic roughness can trip up the flow, creating turbulence that enhances mixing. All of these pore-scale phenomena are bundled up into the macroscopic parameter $h_{sf}$.

So, the total coupling $H$ is a marriage of static geometry ($a_{sf}$) and fluid dynamics ($h_{sf}$). To truly understand heat transfer in a porous medium, you must appreciate both.

We have this elegant two-equation model, but when do we really need it? When is the simple LTE assumption not good enough? Non-equilibrium reigns when there is a mechanism actively driving the two temperatures apart, or when the system changes too fast for them to keep up.

A classic example is ​​non-coincident heating​​. Imagine a catalytic converter, where chemical reactions on the solid surfaces generate heat. Here, $q_s''' > 0$ but $q_f''' = 0$. The solid is being directly heated, while the fluid is not. To get rid of this heat, the solid's temperature must rise above the fluid's temperature to create the driving potential $(T_s - T_f)$ needed for heat to flow into the fluid. A simple calculation for a stationary system shows that the steady-state temperature difference is directly driven by the imbalance between heating and conductivity in the two phases: a mismatch in the ratio $q'''/k$ is the fundamental driver of LTNE.

A more general way to think about this is to compare timescales. There are two competing clocks in this system:

1. The ​​macroscopic diffusion time, $\tau_d$​​: This is the slow clock. It's the time it takes for heat to diffuse across the entire system, roughly $L^2 / \alpha_{\text{eff}}$, where $L$ is the system size and $\alpha_{\text{eff}}$ is the effective thermal diffusivity.
2. The ​​interfacial exchange time, $\tau_i$​​: This is the fast clock. It's the time it takes for the fluid and solid in a small local region to equilibrate their temperatures, proportional to $(\varepsilon \rho_f c_{pf}) / (h_{sf} a_{sf})$.

The ratio of these two times gives us a powerful dimensionless number, $\Pi = \tau_d / \tau_i$.

• If $\Pi \gg 1$, it means the local equilibration is lightning-fast compared to the overall process ($\tau_i \ll \tau_d$). The fluid and solid are always in sync locally. The LTE assumption is perfectly valid. For a typical water-saturated granular medium, this number can be on the order of 100 or more, justifying the use of the simpler LTE model.
• If $\Pi \sim 1$ or smaller, the time it takes for the system to change is comparable to or even faster than the time the phases have to talk to each other. Significant temperature differences will appear, and the full LTNE model is essential.

This explains a seeming paradox. Consider a metal foam saturated with air. The solid metal is an excellent conductor, while the air is an insulator ($k_s \gg k_f$). Does this promote or hinder equilibrium? The answer is: it depends!.

• If you heat the whole system slowly from the boundary, the highly conductive metal network creates a very smooth, slowly changing temperature field. The sluggish air has plenty of time to catch up with the local metal temperature. High $k_s$ promotes LTE.
• But now, imagine you zap the foam with a short, intense laser pulse that only heats the metal. The high $k_s$ now works against equilibrium! It whisks the heat away through the solid network so quickly that the poorly-conducting air, which also receives heat slowly across the interface, is left far behind. A massive temperature gap opens up. In this case, high $k_s$ actively drives the system into a state of profound non-equilibrium.

Ultimately, the question of equilibrium is not a property of the material alone, but of the material and the process it is undergoing. This beautiful interplay between material properties, geometry, and dynamic processes is what makes the study of heat transfer in porous media so rich and fascinating. And it all begins with the courage to ask: are we looking at one world, or two?

We have spent some time understanding the machinery of the Local Thermal Non-Equilibrium (LTNE) model, this idea that within a porous material, the fluid and the solid can exist at two different temperatures. It might seem like an abstract, almost academic, distinction. But the world is rarely as simple as our averages suggest. The real beauty of a physical law or model is not just in its elegant formulation, but in the vast and often surprising landscape of phenomena it allows us to understand, predict, and even control. Now, let us embark on a journey to see where this "two-temperature" idea truly comes to life, moving from the laboratory bench to the heart of industrial processes and the frontiers of modern science.

Before we can apply a model with confidence, we must first ask ourselves: is it real? How could we ever prove that a tiny parcel of fluid is hotter than the solid scaffold surrounding it? Imagine a simple, yet profound, experiment. We take a long tube packed with cool ceramic beads and suddenly start pumping hot water in at one end. A thermal "wave" begins to travel down the tube. The water, of course, heats up the beads it touches. But this takes time. The water is a nimble courier of heat, but the solid beads are more sluggish, possessing a greater thermal inertia.

If we could place ourselves at some point downstream, we would witness a race. The hot fluid arrives first, its temperature jumping upwards. The beads, however, lag behind. For a few precious moments, the water is measurably hotter than the solid it surrounds. To capture this fleeting temperature gap, we would need incredibly small and fast thermometers—perhaps a microscopic thermistor suspended in the fluid-filled void, and another tiny thermocouple embedded just beneath the surface of an adjacent bead. By measuring the two temperatures simultaneously, we could directly observe the non-equilibrium state, $T_f \neq T_s$. This isn't just a confirmation of the model; it's a direct window into the dynamic interplay of heat transfer at the pore scale.

This ability to measure the temperature difference is more than just a party trick; it's the foundation of a powerful engineering tool. By analyzing the precise way the fluid and solid temperatures evolve in such experiments, engineers can work backwards. They can use the LTNE model to deduce crucial, otherwise invisible, properties of the porous material, such as the effective interfacial heat transfer coefficient, $h_{sf}$, and the specific interfacial area, $a_{sf}$. This is akin to a form of "thermal tomography," where we use controlled heat waves to map out the intricate internal thermal landscape of a material, revealing how efficiently heat can pass between the phases. This characterization is essential for designing everything from advanced insulation to chemical reactors.

In many cases, thermal non-equilibrium is a subtle effect that we must account for to achieve accuracy. But what if we could design a device where non-equilibrium isn't just a feature, but the entire point of its operation? Enter the ​​regenerator​​, a clever type of heat exchanger.

Imagine you have a hot exhaust stream from a factory and you want to use its heat to pre-warm a cold incoming air stream, saving energy. Instead of having the two streams flow simultaneously on opposite sides of a metal plate (which would be a recuperator), a regenerator uses a single packed bed of ceramic spheres and a set of valves. First, the hot exhaust gas flows through the bed for a period of time, "charging" it with thermal energy. The ceramic spheres heat up, storing the energy like a thermal battery. During this phase, the solid is playing catch-up to the hot gas, and the essential fact is that its temperature is changing—the storage term in the energy balance, $(1-\varepsilon)\rho_s c_s \frac{\partial T_s}{\partial t}$, is large and positive.

Then, the valves switch. The hot exhaust is diverted, and the cold incoming air is sent through the now-hot bed. The stored heat flows from the hot ceramic spheres into the cold air, pre-warming it. During this "discharging" phase, the solid's temperature drops, and the storage term is large and negative. The heat is transferred from the hot stream to the cold stream indirectly, using the solid matrix as a temporary storage medium. The entire process fundamentally relies on the cyclic storage and release of energy in the solid—a process that only exists because the solid and fluid are not in equilibrium. Here, $T_f \neq T_s$ is not a bug, it's the feature that makes the whole machine work.

The true power of the LTNE framework is its flexibility. It provides a scaffold upon which we can build models for an astonishing variety of complex, interdisciplinary problems.

Chemistry and Safety: The Runaway Reactor

Porous materials are the heart of the chemical industry, often acting as catalysts that speed up reactions. Consider a packed bed reactor where an exothermic (heat-releasing) reaction occurs on the surface of the solid catalyst pellets. The heat is generated directly in the solid, causing its temperature, $T_s$, to rise. This heat is then transferred to the fluid flowing through the bed and is also conducted through the solid skeleton itself. Because the heat source is in the solid, it's natural for the solid to become hotter than the fluid, $T_s > T_f$.

This temperature difference is not merely a detail; it can be a matter of life and death. The rate of most chemical reactions increases exponentially with temperature (the Arrhenius law). If the solid catalyst gets too hot, the reaction speeds up, which generates more heat, which makes the solid even hotter. This can lead to a vicious cycle known as ​​thermal runaway​​, potentially causing an explosion. An LTNE model provides a far more realistic picture of safety than a simple one-temperature model. It correctly identifies the two primary pathways for cooling the hot solid catalyst: conduction through the solid network (governed by $k_s$) and, crucially, transfer to the fluid (governed by the interphase coefficient $H$). By understanding this dual-pathway cooling, engineers can better predict the critical conditions for runaway and design safer reactors. In the limit of very strong interphase coupling ($H \to \infty$), the two phases are locked together thermally, and the system behaves like a single material with an effective thermal conductivity of $k_{\text{eff}} = k_{s,\text{eff}} + k_{f,\text{eff}}$, providing an enhanced pathway for heat removal.

Multiphysics I: The Physics of Drying

Think about the process of drying, whether it's paper in a mill, food being preserved, or soil after a rainstorm. As water evaporates from the pores of the material, it requires a tremendous amount of energy—the latent heat of vaporization. This energy is drawn from the immediate surroundings, causing a strong local cooling effect.

In an LTNE context, this cooling primarily affects the fluid phase (the mixture of air and water vapor) right at the interface where evaporation occurs. The solid matrix, with its typically larger thermal mass, cannot cool down as quickly. This creates a sharp local temperature difference, with the fluid phase becoming significantly colder than the solid. The LTNE model can be extended to include this powerful latent heat sink in the fluid's energy equation, often appearing as a term like $- a_s \dot{m}''_e L$, where $\dot{m}''_e$ is the evaporation rate and $L$ is the latent heat. Accurately modeling this intense, localized non-equilibrium is vital for optimizing drying processes to save energy and improve product quality.

Multiphysics II: Heat in the Furnace and on the Fin

At very high temperatures, such as in a solar thermal receiver or industrial furnace, heat doesn't just conduct and convect—it flies through space as thermal radiation. In a porous medium, this adds another layer of complexity that the LTNE framework is uniquely suited to handle. The solid skeleton, being opaque, interacts with radiation at its surface, emitting and absorbing according to its surface emissivity, $\varepsilon_s$. The gas filling the pores might be transparent, or it could be a "participating" medium (like CO$_2$ or water vapor) that absorbs and emits radiation throughout its volume, governed by an absorption coefficient, $\kappa_f$.

The LTNE model allows us to write separate radiative source terms for each phase. The solid's energy balance gains a term proportional to $a_s \varepsilon_s \sigma (T_r^4 - T_s^4)$, reflecting surface-based exchange, while the fluid's balance might include a term like $4 \kappa_f \sigma (T_r^4 - T_f^4)$ for volumetric exchange. This detailed accounting is crucial for designing systems that can withstand and efficiently manage extreme heat. A related application is in heat transfer enhancement using modern materials like porous metal foams for fins. While often simplified to an LTE model with an effective conductivity, the underlying principles of heat flowing through a complex, two-phase structure are the same.

The Very Small: When Molecules Misbehave

Finally, the LTNE model provides a bridge connecting the macroscopic engineering world to the microscopic world of kinetic theory. Consider a gas flowing through a porous material at very low pressure, or through a material with extremely small pores, like in some microelectronic cooling applications. In these situations, the ​​mean free path​​ of the gas molecules—the average distance a molecule travels before colliding with another—can become comparable to the size of the pores. This is the realm of rarefied gas dynamics, characterized by the Knudsen number, $Kn = \lambda/d_p$.

When the Knudsen number is not negligible, the very concept of a continuous fluid with a single temperature at a wall breaks down. A gas molecule hitting the solid surface may not have enough subsequent collisions with other gas molecules to fully "accommodate" to the solid's temperature. It can bounce off carrying away only partial information about the wall's thermal state. The result is a finite "temperature jump" right at the gas-solid interface. This microscopic phenomenon manifests as an additional thermal resistance. Within the LTNE framework, this isn't a crisis; it's an opportunity. We can modify our trusted interfacial heat transfer coefficient, $h_{sf}$, making it a function of the Knudsen number. A common correction takes the form $\mathrm{Nu}_{sf} = \frac{\mathrm{Nu}_0}{1 + C_j Kn \mathrm{Nu}_0}$, where $\mathrm{Nu}_0$ is the Nusselt number in the continuum regime. This beautiful synthesis shows how a macroscopic engineering parameter can be corrected using insights from the molecular world, extending the model's reach into nanotechnology and vacuum systems.

From industrial furnaces to microchips, the story is the same. The Local Thermal Non-Equilibrium model is far more than a mathematical refinement. It is a lens that provides a sharper, more truthful image of how energy moves through the complex, multiphase world around us. It reminds us that to truly understand nature, we must often look beyond the simple average and appreciate the rich, dynamic, and beautiful interplay happening just beneath the surface.