Listening to Stretch: How Wave Propagation Refines Soft Material Models

Author: Denis Avetisyan

New research shows that analyzing guided waves in stretched materials offers a powerful way to distinguish between different hyperelastic models, going beyond the limitations of traditional mechanical testing.

The study demonstrates how manipulating elastomer plate elongation-specifically stretching to <span class="katex-eq" data-katex-display="false">\lambda\in[1.03,2.27]</span> and <span class="katex-eq" data-katex-display="false">\lambda\in[1.11,2.22]</span>- fundamentally alters the propagation of in-plane (<span class="katex-eq" data-katex-display="false">S\!H_0</span> and <span class="katex-eq" data-katex-display="false">S_0</span>) and out-of-plane (<span class="katex-eq" data-katex-display="false">A_0</span>) modes, revealing a complex relationship between mechanical deformation and wave dynamics within soft materials. — The study demonstrates how manipulating elastomer plate elongation-specifically stretching to $\lambda\in[1.03,2.27]$ and $\lambda\in[1.11,2.22]$ – fundamentally alters the propagation of in-plane ( $S\!H_0$ and $S_0$ ) and out-of-plane ( $A_0$ ) modes, revealing a complex relationship between mechanical deformation and wave dynamics within soft materials.

Guided wave analysis provides enhanced insight into the strain energy density functions of soft, incompressible materials like rubber and elastomers.

Accurately characterizing the complex mechanical behavior of rubber-like materials remains a significant challenge in materials science. This is addressed in ‘Guided elastic waves informed material modelling of soft incompressible media’, which proposes a novel dynamic mechanical analysis exploiting the propagation of guided waves to refine hyperelastic material models. By monitoring wave dispersion during uniaxial extension, the research demonstrates an ability to differentiate between material models-specifically lifting the degeneracy in the $C_2$ term-that are indistinguishable from static tensile tests alone. Could this approach pave the way for more accurate and robust material characterization, ultimately improving the predictive capability of simulations for soft, incompressible media?

Beyond Linear Assumptions: The Whispers of Complex Materials

Many engineering designs rely on material models to predict how a component will behave under stress, but simplistic approaches like linear elasticity frequently fall short when dealing with modern materials. These models assume a direct proportionality between stress and strain, which works well for small deformations of metals but fails spectacularly with polymers and composites. These materials exhibit non-linear behavior-their response to force changes depending on how much they’ve already been stretched or deformed. Furthermore, polymers demonstrate behaviors like viscoelasticity – a combination of viscous and elastic characteristics – and rate-dependent effects, meaning their response varies with how quickly a load is applied. Composites, with their heterogeneous structure, introduce additional complexity, as different constituents respond differently to stress, leading to localized failures that simple models cannot predict. Consequently, designs based on inadequate material representations can lead to inaccurate simulations, unreliable predictions, and ultimately, structural failures.

Traditional material models often fall short when subjected to the stresses found in real-world applications, necessitating the use of hyperelastic models to accurately depict material behavior. These advanced frameworks move beyond the limitations of linear elasticity, allowing for the description of significant, non-reversible deformations – think of a rubber band stretching far beyond its initial length. Crucially, hyperelasticity also accounts for rate-dependent effects, meaning a material’s response changes based on how quickly it is loaded or deformed. This is particularly important for polymers and composites, where the speed of deformation can dramatically alter their behavior – a quick impact yielding a different result than a slow, sustained force. By incorporating these complex responses, hyperelastic models provide a more realistic and reliable foundation for simulations and predictive analyses in fields ranging from automotive safety to biomedical engineering.

The fidelity of engineering simulations and predictive models hinges directly on the accuracy with which material behaviors are characterized. Insufficiently defined material properties can lead to significant discrepancies between modeled outcomes and real-world performance, potentially compromising structural integrity, system efficiency, and overall product reliability. Consequently, substantial effort is dedicated to developing and validating sophisticated characterization techniques – encompassing mechanical testing, advanced imaging, and computational modeling – to precisely capture even subtle nuances in material response. This detailed understanding allows engineers to confidently predict how a component or system will behave under various loading conditions, optimize designs for enhanced performance, and ultimately minimize the risk of failure, driving innovation across diverse fields from aerospace and automotive engineering to biomedical device development.

Rheological measurements of Ecoflex OO-30 reveal its complex shear modulus <span class="katex-eq" data-katex-display="false">\mu^{\\prime}</span> and <span class="katex-eq" data-katex-display="false">\mu^{\\prime\\prime}</span>, accurately modeled by a fractional Kelvin-Voigt model, and subsequent dispersion relations for in-plane (<span class="katex-eq" data-katex-display="false">S\\!H\\_{0}</span>, <span class="katex-eq" data-katex-display="false">S\\_{0}</span>) and out-of-plane (<span class="katex-eq" data-katex-display="false">A\\_{0}</span>) modes at stretch ratios of <span class="katex-eq" data-katex-display="false">\lambda=1.03</span> and <span class="katex-eq" data-katex-display="false">\lambda=1.11</span> closely match both long-wavelength approximations and full semi-analytical predictions. — Rheological measurements of Ecoflex OO-30 reveal its complex shear modulus $\mu^{\\prime}$ and $\mu^{\\prime\\prime}$ , accurately modeled by a fractional Kelvin-Voigt model, and subsequent dispersion relations for in-plane ( $S\\!H\\_{0}$ , $S\\_{0}$ ) and out-of-plane ( $A\\_{0}$ ) modes at stretch ratios of $\lambda=1.03$ and $\lambda=1.11$ closely match both long-wavelength approximations and full semi-analytical predictions.

Unlocking Material Secrets: Evidence from Mechanical Testing

The uniaxial extension test is a primary method for determining the constitutive parameters used in hyperelastic material models. This test involves subjecting a specimen to a controlled tensile deformation and measuring the resulting force. The data obtained – specifically, the relationship between applied stress and resulting strain – forms the basis for calibrating model parameters such as $C_{10}, C_{01}, C_{11}$ in common models like the Mooney-Rivlin or Yeoh models. By fitting the model’s predicted stress-strain curve to the experimentally derived curve, accurate material properties can be established, enabling prediction of material behavior under more complex loading scenarios. The test is typically performed at varying strain rates and temperatures to fully characterize material response and ensure model validity across a range of operating conditions.

Hyperelastic material model parameters, such as those defining the Mooney-Rivlin or Ogden models, directly dictate the predicted stress response for a given strain. These parameters quantify material stiffness, anisotropy, and incompressibility; alterations to their values result in corresponding changes to the modeled material’s behavior under tensile, compressive, shear, or complex loading scenarios. Specifically, parameters influence the material’s resistance to deformation, its tendency to recover its original shape, and its overall load-bearing capacity. Consequently, the fidelity of simulations involving these materials – including finite element analysis – is intrinsically linked to the accurate determination of these parameters through mechanical testing and model calibration.

The predictive capability of hyperelastic material models is directly and significantly impacted by the fidelity and scope of mechanical testing data used during parameterization. Insufficient data, particularly within specific strain ranges or lacking representation of relevant loading conditions, can lead to inaccurate model predictions. A broader range of data, encompassing various strain rates, temperatures, and pre-stretch conditions, allows for a more robust calibration process and improves the model’s ability to accurately represent material behavior across a wider spectrum of applications. Furthermore, data quality, including minimization of experimental error and accurate measurement of both force and displacement, is paramount to ensuring reliable model parameters and minimizing discrepancies between simulation results and physical experiments.

Uniaxial extension testing of an Ecoflex OO-30 sample demonstrates the material's behavior in terms of engineering stress <span class="katex-eq" data-katex-display="false">\sigma^e</span> versus stretch ratio λ, and its representation in Mooney space using equation 6 reveals further material characteristics. — Uniaxial extension testing of an Ecoflex OO-30 sample demonstrates the material’s behavior in terms of engineering stress $\sigma^e$ versus stretch ratio λ, and its representation in Mooney space using equation 6 reveals further material characteristics.

Whispers Through the Structure: Guided Waves as a Diagnostic Tool

Guided wave propagation utilizes the principle that elastic waves can efficiently travel long distances within a plate or structural component due to repeated reflections at its boundaries. Unlike traditional ultrasonic testing which relies on short-wavelength waves and localized inspection, guided waves possess wavelengths significantly larger than the component thickness, enabling inspection of extended areas from a single excitation point. The wave modes-characterized by specific displacement patterns-are sensitive to changes in material properties, such as stiffness, density, and the presence of defects like cracks or corrosion. Analysis of the wave’s velocity, attenuation, and mode shapes provides quantitative information regarding these material characteristics, facilitating non-destructive evaluation without requiring access to multiple locations on the structure.

Analysis of guided wave modes-specifically the Pseudo-Longitudinal (S0), Shear Horizontal (SH0), and Flexural (A0) modes-provides quantitative data regarding material properties. The S0 mode, characterized by both longitudinal and shear components, is highly sensitive to changes in stiffness and is commonly used for flaw detection. SH0 waves, purely shear, are less dispersive and propagate efficiently in thicker structures, offering robust assessment of shear properties. The A0 mode, an out-of-plane flexural wave, is sensitive to surface and near-surface defects, and its velocity is strongly influenced by plate thickness and material damping. By measuring the phase velocity and attenuation of these modes, material stiffness (Young’s modulus and shear modulus) and damping characteristics can be determined, enabling non-destructive evaluation of structural integrity and material degradation.

The Long-Wavelength Approximation, utilized in guided wave analysis, is predicated on the assumption that the wavelength of the guided wave is significantly larger than the thickness of the plate or structure being evaluated. This simplification enables the decoupling of wave propagation behavior in the thickness direction, effectively reducing a three-dimensional problem to a two-dimensional one. Consequently, computational demands are substantially reduced, facilitating rapid and efficient assessment of material properties such as stiffness and damping over extended areas. The approximation is most accurate for low-frequency waves and relatively thin structures, allowing for practical, large-scale non-destructive evaluation.

The Gent and Dobrynin & Carrillo models, enhanced with a Gent-Thomas term, accurately fit simple tension test data for Ecoflex OO-30, and predictions based on equations (13) and (14) successfully model the normalized wave velocities for <span class="katex-eq" data-katex-display="false">SH_0</span> and <span class="katex-eq" data-katex-display="false">S_0</span> modes propagating parallel and perpendicular to elongation at <span class="katex-eq" data-katex-display="false">170 Hz</span>, as well as flexural modes at <span class="katex-eq" data-katex-display="false">50 Hz</span> in the <span class="katex-eq" data-katex-display="false">m{e}_1</span> and <span class="katex-eq" data-katex-display="false">m{e}_3</span> directions. — The Gent and Dobrynin & Carrillo models, enhanced with a Gent-Thomas term, accurately fit simple tension test data for Ecoflex OO-30, and predictions based on equations (13) and (14) successfully model the normalized wave velocities for $SH_0$ and $S_0$ modes propagating parallel and perpendicular to elongation at $170 Hz$ , as well as flexural modes at $50 Hz$ in the $m{e}_1$ and $m{e}_3$ directions.

Acoustoelasticity: Listening to the Stress Within

Acoustoelasticity investigates how alterations in stress within a material directly impact the speed at which waves travel through it, offering a powerful means to characterize material properties and internal stress states. This field rests on the principle that stress modifies a material’s stiffness, and since wave velocity is fundamentally linked to stiffness, measurable changes in wave speed correlate with applied or residual stress. By carefully analyzing these velocity shifts – detected through techniques like ultrasound – researchers can not only determine the magnitude of stress but also map its distribution within a component. This is particularly valuable for non-destructive evaluation, allowing assessment of structural integrity without compromising the material itself, and has applications ranging from monitoring stress in aerospace components to detecting defects in composite materials and even characterizing subsurface geological formations.

The propagation of acoustic waves through a material isn’t simply a function of its inherent properties, but is significantly altered by the application of stress, particularly when that stress exists in multiple directions simultaneously. Research into biaxial stress states – where a material is stretched or compressed in two perpendicular axes – reveals a nuanced interplay between stress and wave velocity. Unlike uniaxial stress, which simplifies the analysis, biaxial loading introduces complexities arising from the material’s anisotropic response and the coupling of stress components. This means that a wave traveling through a stressed material will experience variations in speed and direction dependent not only on the magnitude of the stress, but also on its orientation relative to the wave’s propagation path. Consequently, accurate modeling of wave behavior under these conditions requires a sophisticated understanding of the material’s constitutive laws and the ability to account for these multi-axial interactions, demonstrating that materials respond in fundamentally different ways depending on the stress environment.

A comprehensive understanding of viscoelasticity-a material’s tendency to exhibit both viscous and elastic characteristics-is crucial when interpreting acoustoelastic data, as frequency-dependent properties significantly influence wave propagation. Recent research confirms the ability of guided wave measurements to effectively distinguish between various constitutive models-mathematical descriptions of a material’s behavior-demonstrating a high degree of accuracy. Specifically, this work achieved relative errors in measured wave velocity of less than 7.0%, highlighting the potential for precise material characterization through the analysis of how stress affects wave speed and attenuation at different frequencies. This level of precision offers improved capabilities for non-destructive evaluation and advanced material modeling.

Uniaxial elongation testing of Ecoflex OO-30 in Mooney space, combined with analysis of normalized wave velocities for <span class="katex-eq" data-katex-display="false">SH_0</span>, <span class="katex-eq" data-katex-display="false">S_0</span>, and <span class="katex-eq" data-katex-display="false">A_0</span> modes, demonstrates the material's anisotropic behavior as predicted by long-wavelength theory and semi-analytical models for varying stretch λ and frequency <span class="katex-eq" data-katex-display="false">f=170</span> Hz. — Uniaxial elongation testing of Ecoflex OO-30 in Mooney space, combined with analysis of normalized wave velocities for $SH_0$ , $S_0$ , and $A_0$ modes, demonstrates the material’s anisotropic behavior as predicted by long-wavelength theory and semi-analytical models for varying stretch λ and frequency $f=170$ Hz.

Towards an Effective Material Model: The Art of Persuasion

A truly robust understanding of any material’s behavior necessitates a multifaceted characterization strategy. Rather than relying on a single technique, researchers are increasingly integrating mechanical testing – which reveals how a material deforms under applied forces – with guided wave propagation, a method that explores material properties through the way waves travel within it. This experimental data is then synthesized with computational modeling, allowing for the creation of virtual representations that can predict performance under conditions difficult or impossible to replicate physically. This combined approach not only validates model accuracy against real-world observations, but also allows scientists to extrapolate material responses to novel scenarios, ultimately leading to more informed design and engineering practices.

To accurately capture the intricacies of material behavior under realistic, complex loading, researchers are increasingly turning to Genetic Algorithms for model parameter optimization. These algorithms, inspired by natural selection, iteratively refine a material model’s parameters by ‘evolving’ a population of potential solutions. Each iteration assesses how well the model, with a given set of parameters, predicts experimental results, favoring those that demonstrate higher accuracy. Through processes analogous to crossover and mutation, the algorithm explores a vast parameter space, identifying combinations that minimize the discrepancy between predicted and observed material response. This approach circumvents the limitations of traditional optimization techniques, particularly when dealing with non-linear material models and noisy experimental data, ultimately leading to a more robust and reliable representation of material characteristics.

A streamlined material model, leveraging the generalized neo-Hookean formulation augmented with a Gent-Thomas term, demonstrates a compelling balance between accuracy and simplicity. This approach successfully captures material behavior using only three parameters, a significant reduction compared to more complex models, while maintaining strong agreement with experimental data gathered across a frequency spectrum of 170 Hz to 50 Hz. The development of such an ‘Effective Model’ isn’t merely an academic exercise; it directly addresses a crucial need in engineering design – the ability to reliably predict material response under varied and often complex conditions. By minimizing the number of required parameters, engineers gain a powerful tool for simulations and analyses, ultimately leading to more efficient designs and improved product reliability.

The pursuit of accurate material modeling feels less like physics and more like divination. This research, focused on guided waves in stretched plates, exemplifies the challenge. It’s not about finding the true material behavior, but discerning the model that best persuades the data. The study highlights how acoustoelasticity offers additional constraints, a finer granularity of observation beyond simple tensile tests. As Blaise Pascal observed, “The eloquence of the tongue deceives, but the eloquence of the eyes does not.” Similarly, these guided waves aren’t revealing inherent truth, but offering a clearer signal – a more compelling illusion – within the inherent noise of material properties. It’s a testament to the idea that noise isn’t necessarily error, but simply truth lacking sufficient resolution.

Where the Waves Lead

The pursuit of material truth, as this work suggests, is less about capturing ‘the’ correct model and more about elegantly deceiving oneself with a useful one. Traditional tensile tests offer a singular narrative, a pull and a measure. But the propagation of guided waves… that’s a chorus of whispers, a diffraction of possibility. It reveals that even seemingly homogeneous rubber holds hidden geometries, internal inconsistencies that traditional methods smooth away. There’s truth, hiding from aggregates.

The immediate challenge isn’t refining existing hyperelastic models-they will all continue to lie, some with more grace than others-but embracing the inherent non-uniqueness. Perhaps the future lies not in a single strain energy density function, but in a probabilistic field, a distribution of possibilities weighted by the observed wave behavior. Each measurement doesn’t converge on ‘the’ answer, but nudges the probability landscape.

Ultimately, this work is a reminder that materials don’t ‘have’ properties; they respond to interrogation. And the more cleverly one asks the question – employing guided waves, dynamic analyses, or whatever subtle probe comes to hand – the more nuanced, and delightfully ambiguous, the answer becomes. The illusion of control is compelling, but the elegance lies in acknowledging the beautiful chaos.

Original article: https://arxiv.org/pdf/2603.18839.pdf

Contact the author: https://www.linkedin.com/in/avetisyan/