🧑🎓
What exactly is a "hyperelastic" material? Is it just a fancy word for rubber?
🎓
Basically, yes! Hyperelasticity is the theory we use to model materials like rubber, silicone, and some biological tissues. The key idea is that the material is perfectly elastic—it returns to its original shape after you remove the force—and its stress comes from a "strain energy density" function. In practice, this means we can model huge, nonlinear deformations that metals can't handle. Try moving the "Stretch (λ)" slider above from 1 to 3; you'll see the stress rise nonlinearly for all three models.
🧑🎓
Wait, really? So why are there three different models (Neo-Hookean, Mooney-Rivlin, Ogden) in the simulator? Which one is "correct"?
🎓
Great question! None is universally "correct"; they offer a trade-off between simplicity and accuracy. The Neo-Hookean model is the simplest, using just one material constant, $C_{10}$. It works well for moderate stretches. Mooney-Rivlin adds a second constant, $C_{01}$, which makes it more accurate for a wider range of deformations, like biaxial stretching. The Ogden model is the most flexible, using multiple terms (with parameters $\mu_i$ and $\alpha_i$) to fit complex experimental data perfectly. In the simulator, switch between them using the dropdown and see how their stress-stretch curves differ!
🧑🎓
So if I'm simulating a car tire, which model should I pick? And how do I know what values to put in for those C10 or μ parameters?
🎓
For a tire, which undergoes complex multiaxial loading, you'd likely start with a Mooney-Rivlin or Ogden model. The parameter values come from fitting the model to experimental data from a uniaxial tension test—exactly the kind of curve you're generating in this simulator! A common case is setting $C_{10}$ and $C_{01}$ so their sum relates to the initial shear modulus. Play with the $C_{10}$ and $C_{01}$ sliders: notice how increasing $C_{10}$ raises the entire curve, while adjusting $C_{01}$ changes the curvature, especially at higher stretches.
The core of hyperelasticity is the strain energy density function, $W$, which measures the energy stored per unit volume due to deformation. For an incompressible material (like rubber), the Cauchy stress for uniaxial tension is derived from the derivative of $W$ with respect to the stretch, $\lambda$.
$$ \sigma = \lambda \frac{\partial W}{\partial \lambda}- p $$
Here, $\sigma$ is the true (Cauchy) stress, $\lambda$ is the stretch ratio (deformed length / original length), and $p$ is a hydrostatic pressure eliminated using the incompressibility condition ($\lambda_1 \lambda_2 \lambda_3 = 1$). The different models define $W$ differently.
The simulator implements the following stress-stretch relations for a uniaxial test, which are the final working equations after applying the incompressibility condition.
$$ \text{Neo-Hookean: }\sigma = 2C_{10}(\lambda^2 - \lambda^{-1}) $$
$$ \text{Mooney-Rivlin: }\sigma = 2(\lambda^2 - \lambda^{-1})(C_{10}+ C_{01}/\lambda) $$
$$ \text{Ogden (1-term): }\sigma = \mu_1(\lambda^{\alpha_1 - 1}- \lambda^{-\alpha_1/2 - 1}) $$
Variable Definitions:
- $C_{10}, C_{01}$: Material constants (MPa) related to shear modulus.
- $\mu_i$: Ogden shear modulus parameter (MPa).
- $\alpha_i$: Ogden exponent (dimensionless) controlling nonlinearity.
The physical meaning is clear when you change them: $C_{10}$ and $\mu_i$ scale the stress, while $\alpha_i$ and the $C_{01}/C_{10}$ ratio define the curve's shape.
Common Misconceptions and Points to Note
First, understand that "material constants are not catalog values for the material." For example, when you hear that a certain rubber has a C10=0.5 MPa, you might tend to think it's a direct indicator of tensile strength or hardness. However, these constants are merely the result of fitting a specific mathematical model called a "strain energy density function." Even for the same material, the C10 value fitted for Neo-Hookean and the C10 fitted for Mooney-Rivlin are physically different values. Therefore, it's dangerous to use constants from literature or other companies' data as-is; you must always re-fit them based on your own test data.
Next, be aware of the pitfall that "not everything is determined by uniaxial tensile data alone." While uniaxial testing is simple, relying solely on it often leads to underestimating behavior in biaxial or shear modes. For instance, if you determine the C01 for Mooney-Rivlin from only uniaxial data, there are cases where you could underestimate the stress by more than 50% when an actual component undergoes biaxial deformation. In practice, ideally, you should acquire test data from compression or planar tension (close to biaxial) if possible, and adjust the parameters so the curve fits across multiple deformation modes. With this simulator, you can switch deformation modes and see at a glance how a curve fitted for uniaxial data deviates in other modes.
Finally, "a higher order N for the Ogden model is not always better." While the Ogden model can fit complex curves by increasing terms (N=1,2,3...), there's a risk of overfitting. With N=3 or higher, you might get a perfect fit within the error range of the test data, but the model can exhibit unrealistic behavior between data points (interpolation) or beyond them (extrapolation). Start with N=1 or 2, and judge the necessity based on the quality of fit. Computational cost is also a non-negligible factor in CAE.
Related Engineering Fields
The concept of hyperelasticity handled by this tool is deeply connected to biomechanics. Biological tissues like blood vessels, cartilage, and skin are also hyperelastic materials exhibiting large deformations. For example, Mooney-Rivlin or Fung-type models are used for stress analysis of aneurysms. In the design of medical devices, such as catheters or artificial valves, these material models are essential for simulating interactions with biological tissues.
Another important field is soft robotics and wearable devices. Soft grippers actuated by pneumatics or stretchable substrates for health monitors worn on the body are made of silicone rubber or TPU (thermoplastic polyurethane). When predicting the durability and operational accuracy of these products with CAE, the ability to accurately represent large stretches (stretch ratio λ > 5) using models like Ogden can make or break the design.
Furthermore, automotive tire engineering is a classic application of hyperelastic analysis. Tires are composite materials, but the behavior of the rubber components, which constitute the majority, dictates performance. Phenomena like rolling resistance, heat generation, wear, and grip on wet surfaces are closely related to the hysteresis (history-dependent properties) of rubber. While this simulator deals with the elastic part, learning about "viscoelasticity" as a next step will lead you to an understanding of the dynamic behavior of tires and heat generation mechanisms.
For Further Learning
As a recommended next step, consider incorporating the concept of "viscoelasticity". Real-world rubber is not a perfectly elastic body as calculated by this tool; it exhibits stress dependent on deformation rate (viscosity) and hysteresis (energy loss). For example, doesn't a vibration isolation rubber feel completely different in stiffness when pressed slowly versus when struck with a hammer? This is modeled using Maxwell models or Prony series. In CAE software, combining "hyperelasticity + viscoelasticity" allows you to represent more realistic hysteresis loops and frequency dependencies.
If you want to deepen the mathematical background, study the fundamentals of continuum mechanics, especially "descriptions of deformation" and the "principle of material frame-indifference." The formulas for hyperelasticity are not mere empirical equations; they are derived from physical requirements like material isotropy and the objectivity of deformation (independence from the coordinate system). Understanding the differences between strain tensors (Green-Lagrange, Cauchy-Green) and stress tensors (Piola-Kirchhoff, Cauchy) will help you grasp why you need to distinguish between nominal stress and true stress, and the meaning of CAE software output will click.
For a practical learning step, try your hand at "curve fitting with real data." Moving sliders in this simulator to match a curve is intuitive, but in reality, optimal material constants are found using algorithms like the least squares method. Using Excel's Solver or Python's SciPy library, you can even create your own fitting program. The experience of simultaneously fitting uniaxial, biaxial, and volumetric compression (bulk modulus) data is excellent training for understanding the essence of material parameter setting.