Lankford Coefficient

From the body panels of a car to a simple aluminum soda can, modern manufacturing relies on the ability to shape flat sheets of metal into complex, durable forms. While we might assume these metal sheets behave uniformly, they often harbor a hidden directional preference, a property known as anisotropy. This directionality can be a blessing or a curse: it can make a material perfectly suited for a task or cause it to fail unexpectedly during production. The key to mastering metal forming lies in understanding and quantifying this behavior.

Simple isotropic models, which assume material properties are the same in all directions, fall short in predicting the behavior of these materials. This creates a critical knowledge gap between theoretical models and real-world manufacturing challenges, leading to costly defects and process limitations. To bridge this gap, engineers and scientists rely on a powerful metric: the Lankford coefficient.

This article provides a comprehensive exploration of the Lankford coefficient, or r-value. In the following chapters, we will first dissect the fundamental theory behind the r-value, exploring its physical origins in the material's microstructure and its mathematical formalization within plasticity theory. Subsequently, we will see this theory in action, examining how the Lankford coefficient is used to solve practical engineering problems, from preventing manufacturing defects like earing to its role as a diagnostic tool in the development of advanced materials.

Imagine you take a sheet of modeling clay, a wonderfully uniform and pliable material. If you stretch it in one direction, you'd expect it to shrink down equally in its width and thickness. Its properties are the same no matter which way you pull it; we call this ​​isotropic​​. But many materials we rely on, especially the high-strength metal sheets that form the body of a car or the shell of an airplane, don't behave this way. They have a secret directional preference, a kind of internal "grain" that makes them stronger or more stretchable in one direction than another. This property is called ​​anisotropy​​, and understanding it is not just an academic exercise—it is the key to shaping metal into complex, durable parts without it tearing apart.

Let's do a thought experiment. Take a rectangular strip of sheet metal and pull on its ends, stretching it along its length. It gets longer, of course. To conserve its volume (a very good approximation for metals under plastic deformation), it must get thinner and narrower. But does it shrink by the same proportion in width and in thickness? The surprising answer is, generally, no.

To quantify this, we define a simple number, an exquisite measure of this anisotropy, called the ​​Lankford coefficient​​, or ​​r-value​​. It's the ratio of the true plastic strain in the width direction, $\epsilon_w^p$, to the true plastic strain in the thickness direction, $\epsilon_t^p$:

For an isotropic material like our clay, which shrinks equally in the two lateral directions, we would have $r=1$. But for a typical sheet of steel or aluminum, this value might be $1.5$, or $0.8$, or even $2.0$. What does this number tell us? Let's use the principle of ​​plastic incompressibility​​, which states that the sum of the true plastic strains in three perpendicular directions must be zero: $\epsilon_l^p + \epsilon_w^p + \epsilon_t^p = 0$, where $\epsilon_l^p$ is the strain along the stretching direction.

With a little algebra, we can use the definition of $r$ to find out how much the sheet thins. Substituting $\epsilon_w^p = r \epsilon_t^p$ into the incompressibility equation, we get $\epsilon_l^p + r \epsilon_t^p + \epsilon_t^p = 0$, which gives us a beautiful and simple relationship for the thickness strain:

This equation is remarkably insightful. For a given amount of stretching ($\epsilon_l^p$), a material with a higher r-value will thin down less. This is tremendously important in manufacturing. Consider the process of ​​deep drawing​​, where a flat sheet of metal is pressed by a punch into a die to form a cup-like shape, such as a soda can or a kitchen sink. The main way this process fails is that the wall of the cup gets stretched too thin and tears. A material that naturally resists thinning—a material with a high r-value—is a godsend. It allows us to form deeper, more complex shapes, a property known as high ​​drawability​​. The simple r-value, a number born from a stretching experiment, is a direct predictor of a material's performance in a sophisticated industrial process.

So, where does this mysterious anisotropy come from? Why would a seemingly uniform sheet of metal have a directional character? The answer lies deep inside, at the level of the microscopic crystals, or ​​grains​​, that make up the metal.

Metals are not an amorphous continuum; they are a vast collection of tiny, individual crystals. Within each crystal, atoms are arranged in a highly ordered, repeating lattice. Plastic deformation doesn't happen by atoms just randomly squishing around. It occurs through a process called ​​slip​​, where planes of atoms slide over one another along specific crystallographic directions, much like sliding cards in a deck. For slip to happen, the shear stress resolved onto that specific plane and in that specific direction must reach a critical value—a rule known as the ​​Schmid law​​.

Now, imagine a single crystal. Its resistance to deformation will strongly depend on how it's oriented relative to the direction you're pulling it. For some orientations, it's easy to activate a slip system; for others, it's difficult.

In a freshly cast piece of metal, these millions of tiny grains are usually oriented randomly in every possible direction. On a large scale, these random orientations average out, and the material behaves isotropically. But then we process the metal. To make a sheet, we pass a thick slab through massive rollers, squeezing it thinner and longer. This ​​cold rolling​​ process is violent at the microstructural level. It forces the individual crystal grains to rotate and align themselves in a common, preferred orientation. The result is what we call a ​​crystallographic texture​​. The metal is no longer a random collection of crystals; it's more like a highly disciplined army of crystals, mostly facing the same way.

When you have a material with a strong texture, its macroscopic properties—like strength and stretchability—will naturally depend on direction. Pulling along the rolling direction might be different from pulling across it because you are interacting with the aligned "deck of cards" in different ways. This collective, direction-dependent response of the textured crystals is the physical origin of the Lankford coefficient and, more generally, of plastic anisotropy. We need a more sophisticated "rule book" than the simple one for isotropic materials.

To bring this phenomenon into the realm of predictive engineering, we need a mathematical framework. In plasticity theory, the "rule book" that governs when a material starts to deform permanently is called a ​​yield criterion​​, which defines a ​​yield surface​​ in the space of stresses. For a simple isotropic material that doesn't care about direction, the go-to model is the ​​von Mises yield criterion​​. In the space of principal stresses, its yield surface is a perfectly circular cylinder, reflecting the fact that the yielding rule is the same in all directions.

But we know our rolled sheet is not isotropic. Its yield surface cannot be a perfect cylinder. It must be distorted in a way that reflects the material's underlying orthotropic symmetry (the symmetry of a brick, with three perpendicular planes of mirror symmetry corresponding to the rolling, transverse, and normal directions). In 1948, the brilliant applied mathematician Rodney Hill proposed a wonderfully elegant generalization of the von Mises criterion for just this case. The ​​Hill 1948 yield criterion​​ is a quadratic equation in the stresses:

Look at this equation. It's built from the differences in normal stresses, which means that adding a uniform hydrostatic pressure (squeezing it equally from all sides) doesn't change the value. This correctly captures the fact that yielding in dense metals is driven by shape-changing shear, not by pressure. The magic lies in the coefficients: $F, G, H, L, M,$ and $N$. These are not universal constants; they are parameters we measure for a specific material. They are the "knobs" we can tune to stretch and shape our yield surface so that it accurately describes the anisotropic behavior. If we set these knobs to very specific values ($F=G=H=1/2$ and $L=M=N=3/2$), Hill's criterion beautifully simplifies and becomes the isotropic von Mises criterion. So, Hill's model contains the isotropic case as a special instance but gives us the power to describe a whole universe of anisotropic materials.

Here is where the story gets truly beautiful. You might think Hill's criterion only tells us when the material will yield. But it does more. It also tells us how the material will flow once yielding begins. This is thanks to a profound concept in plasticity called the ​​associated flow rule​​, or the ​​normality rule​​.

Imagine the yield surface not as a boundary, but as a hill in "stress space." The associated flow rule states that when the material is plastic, the direction of the plastic strain rate (how it's deforming) is always perpendicular (or "normal") to the yield surface at the current stress state.

This is the unifying principle. The very same function, $\Phi(\boldsymbol{\sigma})$, that defines the shape of the yield surface also serves as a "potential" from which the flow direction can be derived by taking its gradient. This means the shape of the yield surface dictates the direction of plastic flow. An elliptical yield surface will not produce the same flow direction as a circular one.

And now we can connect everything. The Lankford coefficient, $r$, is just a ratio of plastic strain rates. According to the normality rule, these rates are determined by the derivatives of the Hill potential function. When we carry out the mathematics for a tensile test along the rolling direction ($\theta=0^\circ$), we find an astonishingly simple result:

Similarly, for a test in the transverse direction ($\theta=90^\circ$), we find $r_{90} = H/F$. Suddenly, the Lankford coefficient, our experimental measure of anisotropy, is tied directly to the coefficients that define the shape of our theoretical yield surface! This closes the loop. We can perform simple tensile tests to measure $r_0$ and $r_{90}$, use these to determine the ratios of our Hill coefficients, and thereby construct a mathematical model that can predict the material's behavior under any complex, multiaxial stress state. It's a textbook example of how theory and experiment work together to build a powerful predictive tool.

A powerful theory shouldn't just explain what we already know; it should predict things we might not have expected. Our anisotropic plasticity framework does exactly that, beautifully explaining a common and costly manufacturing defect known as ​​earing​​.

When you deep draw a cup from a perfectly circular blank of anisotropic metal, the top rim of the finished cup is often not a perfect circle. Instead, it has a wavy, scalloped edge with several high points, or "ears," and low points, or "valleys." This is undesirable; the ears must be trimmed off, wasting material and adding an extra manufacturing step.

Why does this happen? It's the Lankford coefficient, varying with direction in the plane of the sheet. The material's resistance to thinning isn't the same all the way around the cup. To characterize this, we define two very useful average parameters based on tensile tests in the rolling direction ($0^\circ$), diagonal direction ($45^\circ$), and transverse direction ($90^\circ$):

• ​​Normal Anisotropy ($\bar{r}$):​​ $\bar{r} = \frac{r_0 + 2r_{45} + r_{90}}{4}$. This is a measure of the sheet's average resistance to thinning. As we saw, a high $\bar{r}$ is good for overall drawability.

• ​​Planar Anisotropy ($\Delta r$):​​ $\Delta r = \frac{r_0 - 2r_{45} + r_{90}}{2}$. This measures the variation of the r-value within the plane. It compares the properties along the axes to those along the diagonal.

If a material were perfectly uniform in the plane, we'd have $r_0=r_{45}=r_{90}$, and $\Delta r$ would be zero. But if $\Delta r \neq 0$, it's a sign that earing will occur. The theory is even more powerful than that. The sign of $\Delta r$ predicts where the ears will form!

• If $\Delta r > 0$, it means the r-values are highest along the $0^\circ$ and $90^\circ$ directions. Material flowing from these directions will resist thinning the most, so it gets "piled up" into ears. We get four ears, aligned with the rolling and transverse directions.

• If $\Delta r 0$, the r-value is highest at $45^\circ$. We again get four ears, but this time they are located at $45^\circ$ to the primary axes.

This is a spectacular success of the theory. The simple Hill model, calibrated with a few tensile tests, can predict a complex, three-dimensional defect in a major industrial process. An engineering headache is transformed into a solved physics problem.

The story, of course, does not end there. Science is a continuous journey of refinement. The beautiful picture we painted with Hill's 1948 model is a fantastic first approximation, but more precise measurements reveal further subtleties.

For instance, careful experiments show that the Lankford coefficients are not constant; they can change as the material is stretched. An $r_0$ value that starts at $1.0$ might evolve to $1.4$ after $15\%$ strain. This implies that the very shape of the yield surface is changing during deformation. This phenomenon, known as ​​distortional hardening​​, requires more advanced models where the yield surface is a "living" object whose shape evolves with the plastic history, often linked to the complex interactions between different slip systems at the crystal level.

Furthermore, in some materials, the beautiful unity of the associated flow rule breaks down slightly. The "rule book" for yielding (the yield function $f$) and the "rule book" for flow (the plastic potential $g$) might be two different, albeit similar, functions. This is the realm of ​​non-associated plasticity​​, a necessary complication to explain data where, for instance, equal r-values in two directions do not correspond to equal yield strengths.

These frontiers do not diminish the beauty of the classical theory. On the contrary, they show its power as a foundation. The simple, elegant ideas of the r-value, the anisotropic yield surface, and the normality rule provide the essential language and concepts we need to explore and understand the richer, more complex symphony of plastic deformation in the real world.

Have you ever looked closely at a simple aluminum soda can? It starts its life as a flat, circular disk of metal. In a series of violent, yet incredibly precise, steps, it's forced through dies to become the familiar cylinder we all know. This process is called "deep drawing," and it's a cornerstone of modern manufacturing, used to make everything from kitchen sinks to car doors. Now, what if I told you there's a single number, a simple ratio, that holds the secret to whether that process succeeds brilliantly or fails spectacularly?

In the previous chapter, we delved into the principles behind the Lankford coefficient, or r-value, a measure of plastic anisotropy. We saw that for many materials, a pull in one direction doesn't produce uniform thinning; the material has a "preference" for how it deforms. Now, we're going to see that number in action. We'll embark on a journey to see how this seemingly abstract concept is not just a curiosity, but a powerful tool that connects a factory floor to the frontiers of materials science, and the atomic lattice to the supercomputers that design our world. It's a beautiful example of how a single, carefully observed piece of nature’s puzzle can illuminate a vast landscape of science and technology.

Let’s return to our soda can. When you draw a flat circular blank into a cup, you want a cup with a nice, even rim. But often, the metal doesn't cooperate. Instead of a flat rim, you get a wavy, scalloped edge with several high points, or "ears." This isn't just a cosmetic problem; those ears represent wasted material that must be trimmed off, adding cost and complexity to manufacturing. This phenomenon, aptly named ​​earing​​, is a direct manifestation of planar anisotropy.

Imagine the material flowing into the die from the flat flange. If the material resists thinning more in certain directions (i.e., has a higher r-value), it will tend to draw in more from the sides, pushing material upwards to form an ear in that direction. Conversely, in directions with a lower r-value, the material thins more easily and doesn't flow as far, creating a valley. By measuring the Lankford coefficients at $0^\circ$ (the rolling direction), $45^\circ$, and $90^\circ$ to the rolling direction, we can define a quantity called the ​​planar anisotropy​​, $\Delta r = \frac{r_0 - 2r_{45} + r_{90}}{2}$. The sign and magnitude of $\Delta r$ give us a remarkably accurate prediction of the earing pattern. A positive $\Delta r$, for instance, typically leads to ears forming along the $0^\circ$ and $90^\circ$ directions, with valleys in between at $45^\circ$. This simple measurement allows engineers to anticipate and mitigate this costly problem, perhaps by altering the rolling process to create a more balanced texture or by designing the tooling to account for the uneven flow.

Anisotropy, however, governs more than just the final shape; it can dictate the very moment of failure. When you stretch a sheet of metal, it doesn't just get thinner and thinner until it vanishes. At a certain point, the deformation decides to "localize" into a narrow band. All subsequent stretching occurs in this band, which thins rapidly and leads to a tear. This is called ​​localized necking​​, and it represents the forming limit of the material. Where does this neck form? Naively, one might guess it forms perpendicular to the stretching direction, but the reality is more subtle and beautiful.

For a perfectly isotropic material under uniaxial tension, this localization typically occurs in a band perpendicular to the stretching direction. But for an anisotropic sheet, the r-value changes the game. A deep analysis based on the principles of mechanical instability reveals that the angle of the necking band depends directly on the r-value in the direction of stretching. For a typical rolled steel sheet where $r_0 > 1$, the necking band actually rotates closer to the rolling direction. The higher the anisotropy (the larger the $r_0$ value), the smaller the angle becomes. This is a profound insight: the same property that causes cosmetic ears also controls the geometry of catastrophic failure. By understanding this, engineers can design forming processes that avoid stress states likely to trigger such instabilities.

The ability to predict earing and failure is powerful, but modern engineering demands more. We want to build virtual prototypes in a computer, to simulate the entire manufacturing process and test a thousand designs before a single piece of metal is cut. To do this, we need to translate our physical understanding of anisotropy into the language of mathematics, into so-called ​​constitutive models​​.

The heart of such a model is a ​​yield criterion​​, a mathematical function $f(\boldsymbol{\sigma})$ that defines a surface in the multi-dimensional space of stress. Inside this surface, the material behaves elastically; when the stress state reaches the surface, the material yields and flows plastically. For an isotropic material, this surface is simple and symmetric, like the one described by von Mises. But for an anisotropic material, this surface is warped and distorted. How do we describe this distorted shape? You guessed it: with the Lankford coefficients.

In one of the earliest and most influential models, Hill's 1948 quadratic criterion, the r-values measured in simple tensile tests are used directly to calculate the coefficients that define the shape of the yield surface. This is a remarkable leap: a measurement of strain ratios (kinematics) is used to define a criterion for stress (kinetics). Once the yield surface is calibrated, the model can predict how the material will behave under any complex combination of stresses, far beyond the simple pulling used to measure the r-values in the first place. This predictive power is the engine of modern Finite Element Analysis (FEA) software used across the automotive and aerospace industries.

But science never stands still. As metallurgists developed new alloys with ever more complex behaviors, engineers found that the elegant simplicity of the Hill 1948 model had its limits. For many modern steels and aluminum alloys, it became impossible to find a single set of parameters that could simultaneously match the experimental yield stresses in all directions and the experimental r-values. The quadratic nature of the model also meant it was inherently symmetric, predicting the same yield strength in tension and compression, a feature not always observed in real materials.

This challenge sparked a creative burst in the mechanics community, leading to a fascinating evolution of more sophisticated models. The non-quadratic Hill 1979 criterion introduced an exponent, $m$, to give the yield surface more shape flexibility. Later, models like Barlat's Yld2000-2d took a giant leap forward. By using a much larger set of anisotropy parameters and a more complex mathematical form, these models effectively "decouple" the fitting of yield stresses from the fitting of r-values. This gives the modeler the flexibility to precisely tailor the yield surface shape and its gradient at every point, allowing for a much more faithful "portrait" of the material's true behavior. This story of evolving models, from Hill to Barlat, is a perfect microcosm of the scientific process: a simple, powerful idea is tested to its limits, its shortcomings inspire new theories, and our predictive power grows ever stronger.

So far, we have treated the r-value as a given property of a material sheet. But the deepest and most beautiful questions in science are often the simplest: Why? Why does a material have a particular r-value? The answer takes us from the world of engineering to the world of physics and materials science, down to the level of individual crystals.

A piece of metal is not a uniform jelly; it is a vast, polycrystalline aggregate, a jumble of trillions of tiny, ordered crystals. Each crystal has its own orientation. The manufacturing process, particularly rolling, doesn't just flatten the metal; it forces these crystals to rotate into preferred orientations, creating what is called a ​​crystallographic texture​​. It is this collective alignment of crystals that is the ultimate source of macroscopic anisotropy. Using the framework of ​​crystal plasticity​​, we can build a model from the ground up. We start with the behavior of a single crystal, which deforms by slip on specific crystallographic planes. Then, using a homogenization scheme like the Taylor model which assumes all crystals deform together, we can mathematically average the response of all the differently oriented crystals in our texture. From this, we can derive, from first principles, the macroscopic Lankford coefficients. This is a breathtaking connection, linking the quantum mechanical rules that govern atomic planes to the formability of a car fender.

This microstructural connection opens up thrilling new possibilities. The r-value is no longer just a static parameter to be measured, but a dynamic quantity that can be engineered and used as a diagnostic tool.

Consider forming a component at high temperatures, a process called hot stamping. Temperature changes everything. It allows dislocations to move more easily and can trigger microstructural changes like recovery and recrystallization, which in turn alter the crystallographic texture. As the texture evolves, so does the anisotropy. A complete thermoplasticity model must therefore treat the anisotropy coefficients, and thus the r-values, as functions of temperature, $r(T)$.

Perhaps the most exciting application is in the study of advanced "smart" materials like ​​Transformation-Induced Plasticity (TRIP) steels​​. These materials have a remarkable ability: when they are stretched, their internal crystal structure transforms from a phase called austenite to a much harder phase called martensite. This transformation makes the material stronger and more ductile, a dream combination for making lightweight, crash-safe vehicles. This transformation is a dynamic process that happens during deformation. How can we track it? By watching the r-value! Experiments and models show that as the steel is strained, its r-value evolves. Initially, it may increase due to the texture evolution of the austenite. But then, as the TRIP effect kicks in and martensite begins to form rapidly, the trend reverses. By simultaneously measuring the r-values and the amount of martensite (using techniques like X-ray diffraction), scientists have found that a sharp change in the evolution of the r-value serves as a clear, real-time signature of the onset of the TRIP effect. The Lankford coefficient is transformed from a simple parameter into a dynamic probe of phase transformations happening deep within the material.

From the factory floor to the research laboratory, from ensuring the quality of a simple can to unlocking the secrets of advanced alloys, the Lankford coefficient proves its worth. It is a testament to the fact that in science, there are no "unimportant" details. A careful measurement of how a metal sheet thins is a thread that, when pulled, unravels a rich tapestry connecting engineering, physics, and materials discovery. It is a simple number that tells a wonderfully complex and beautiful story about the nature of the materials that build our world.